The Pennsylvania State University

The Graduate School

Eberly College of Science

MECHANISTIC, SPECTROSCOPIC, AND STRUCTURAL CHARACTERIZATION OF

TWO NOVEL REACTIONS CATALYZED BY RADICAL S-ADENOSYLMETHIONINE

DEPENDENT ENZYMES

A Dissertation in

Chemistry

Tyler L. Grove

 2013 Tyler L. Grove

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

May 2013

The dissertation of Tyler L. Grove was reviewed and approved* by the following:

Squire J. Booker Associate Professor of Chemistry and of Biochemistry and Molecular Biology Dissertation Advisor Co-Chair of Committee

Joseph Martin Bollinger, Jr. Professor of Chemistry and of Biochemistry and Molecular Biology Co-Chair of Committee

Carsten Krebs Associate Professor of Chemistry and of Biochemistry and Molecular Biology

Christopher House Associate Professor of Geosciences

Barbara J. Garrison Professor of Chemistry Head of the Department of Chemistry

*Signatures are on file in the Graduate School

iii

ABSTRACT

Part I: Characterization of Radical SAM-Dependent Methyl Synthases.

The bacterial ribosome is the target of about half of all antibiotics currently in use.

Antibiotics bind to this huge macromolecular machine, which is composed of both proteins and RNA, and disrupt its function, which is to synthesize proteins required for bacterial survival. Frequently, bacteria develop mechanisms to prevent antibiotics from binding to their ribosomes, which most often involve modifying or changing specific amino acid residues or nucleotides that line the antibiotic binding site. One such modification is the addition of a methyl (–CH3) group to adenosine 2503 (A2503) of 23S ribosomal RNA (rRNA). This nucleotide is absolutely conserved, and is located within domain V of the ribosome which is the catalytic center where peptide bonds are formed.

Methylation of carbon 2 of A2503 is catalyzed by the enzyme RlmN. This modification is ubiquitous among prokaryotes, and like most modifications to the ribosome, is believed to be important in optimal functioning of the ribosome. Interestingly, the addition of a methyl group to carbon 8 of A2503 confers resistance to over seven classes of antibiotics, some of which are antibiotics of last resort for patients infected with methicillin resistant

Staphylococcus aureus (MRSA) or vancomycin-resistant Enterococcus. This activity is encoded by the product of the cfr gene (Cfr), which has been found in MRSA isolated from hospitalized patients from numerous countries, portending a new and dangerous mechanism for the spread of antibiotic resistance.

RlmN and Cfr have been predicted to be evolutionarily related, as well as belong to the radical SAM (RS) superfamily of enzymes. RS enzymes all utilize S-

iv adenosylmethionine (SAM), as a precursor to a potently oxidizing 5’-deoxyadenosyl 5’- radical (5’-dA•). The commonality among the reactions catalyzed by these enzymes is that the 5’-dA• initiates catalysis by abstracting a key hydrogen atom from the substrate.

Our studies on RlmN and Cfr have elucidated the chemical logic by which these enzymes are capable of installing a methyl group onto an unreactive, sp2-hybridized carbon, revealing a completely new strategy that expands the repertoire of enzyme-catalyzed reactions.

To elucidate the mechanistic strategy of RlmN and Cfr, we used high-resolution mass spectrometry in concert with isotopically labeled substrates to show that, unlike most RS enzymes, RlmN and Cfr are capable of activating SAM toward two distinct types of reactivities within a single active site. Interestingly, the first step in each reaction was shown to be the SAM-dependent methylation of a conserved cysteinyl residue, in the process making a methylcysteinyl residue that is subsequently activated by the 5’-dA•.

The methylcysteinyl radical then adds to the nucleotide substrate, generating a radical enzyme–substrate crosslink, which we were able to trap in Cfr and characterize by electron paramagnetic resonance and electron-nuclear double resonance spectroscopies.

In addition we were able to show that cleave age of this radical enzyme–substrate crosslink is driven by formation of a radical disulfide anion bond between the methyl- carrying Cys residue and another strictly conserved Cys residue, with concomitant release of the product from the protein.

In addition, through a collaboration with Professor Amy Rosenzweig at

Northwestern University, we were able to determine the crystal structure of RlmN to 2.05

Å, which allowed the visualization of the methylated cysteinyl residue predicted in our earlier biochemical studies.

Part II: Characterization of Radical SAM-Dependent Dehydrogenases

The second part of this dissertation was the study of a different and unique class of RS enzymes, the RS dehydrogenases. These enzymes use radical chemistry to effect two-electron oxidations of organic substrates via intermediates containing single unpaired electrons. This unique reaction was predicted to contain multiple iron–sulfur (Fe/S) cluster. All RS enzymes require 1 [4Fe–4S] cluster, which is ligated by cysteines found in the RS motif, CxxxCxxC. The RS cluster interacts with SAM to facilitate its cleavage to the 5’-dA•. The existence and the role of the additional cluster(s) were not as clear. This details the characterization of three enzymes from this class: AtsB, anSMEcpe, and BtrN.

BtrN was previously shown to catalyze a key step in the biosynthesis of the antibiotic butirosin B, which is the oxidation of a secondary alcohol located on a deoxysugar to a ketone. AtsB and anSMEcpe, on the other hand, catalyze the oxidation of a seryl or cysteinyl residue on a cognate protein to a formylglycyl (FGly) residue. The FGly residue is found in arylsulfatases, where it plays a unique role as a cofactor catalyzing the hydrolysis of sulfate monoesters. We were able to shown that, in contrast to the majority of radical SAM enzymes, each of these proteins contains two or more [4Fe–4S] clusters that are absolutely required for turnover. In addition, the stoichiometry of the AtsB and anSMEcpe reactions was determined as well as the stereochemical course of their reactions; in both cases the pro-S hydrogen of the substrate is abstracted by the 5’-dA•.

In addition, several radical species have been trapped in reactions catalyzed by anSAMcpe which will allow a detail spectroscopic study of the active site.

To truly understand the role of the auxiliary cluster(s) in these reactions, we developed conditions for crystallizing both BtrN and anSMEcpe in the presence of substrates. In collaboration with Professor Cathy Drennan (MIT), we solved the structure of both anSMEcpe and BtrN to a resolutions of ~1.6 Å.

vii

TABLE OF CONTENTS

List of Abbreviations ...... v

List of Figures...... v

List of Tables ...... vi

Acknowledgements ...... vii

Chapter 1 ...... 1

Introduction ...... 1 1.1 Radical SAM enzymes and Human Health...... 3 1.2 Radical SAM enzymes with multiple iron–sulfur clusters...... 8 1.3 Maturation of complex metallocofactors...... 19 1.4 Radical SAM enzymes lacking the canonical CxxxCxxC motif...... 23 1.5 Advances in understanding the reductive cleavage of S- adenosylmethionine...... 28 1.6 Future directions...... 30 1.7 References ...... 33

Part I

Characterization of Radical SAM-Dependent Methyl Synthases...... 49

Chapter 2 A Radically Different Mechanism for S-adenosylmethionine-dependent Methyltransferases Exhibited by the Antibiotic Resistance Protein Cfr and its Homologue RlmN ...... 50

2.1 Abstract ...... 52 2.2 Introduction...... 54 2.3 Materials and Methods ...... 57 2.4 Results ...... 69 2.5 Discussion...... 88 2.6 Acknowledgements ...... 90 2.7 References...... 91

Chapter 3 Cfr and RlmN Contain Only One Binding Site for S-Adenosylmethionine, which Supports Two Distinct Reactivities — Methyl Transfer by SN2 Displacement and Radical Generation ...... 97

3.1 Abstract ...... 98 3.2 Introduction ...... 99 3.3 Materials and Methods...... 104 3.4 Results and Discussion ...... 111 3.5 Acknowledgements ...... 124 3.6 References ...... 125

viii

Chapter 4 A Kinetically Competent Substrate Radical Intermediate in Catalysis by the Antibiotic Resistance Protein Cfr ...... 131

4.1 Abstract ...... 132 4.2 Introduction ...... 134 4.3 Results and Discussion ...... 147 4.4 Conclusions ...... 161 4.5 Coordinates from Density Functional Theory ...... 162 4.6 References ...... 166

Chapter 5 Generation and Characterization of the Proposed Disulfide Intermediate in the Reactions Catalyzed by the Radical SAM Methylsynthases, RlmN and Cfr ...... 174

5.1 Abstract ...... 175 5.2 Introduction ...... 177 5.3 Materials and Methods ...... 181 5.4 Results ...... 190 5.5 Discussion...... 209 5.6 Conclusions ...... 219 5.7 References ...... 220

Part II

Characterization of Radical SAM-Dependent Dehydrogenases ...... 227

Chapter 6 In vitro Characterization of AtsB, a Radical SAM Formylglycine Generating Enzyme that Contains Three [4Fe–4S] Clusters† ...... 228

6.1 Abstract ...... 229 6.2 Introduction ...... 231 6.3 Materials and Methods ...... 236 6.4 Results ...... 244 6.5 Discussion...... 263 6.6 Acknowledgements ...... 276 6.7 References ...... 277

Chapter 7 A General Cofactor Requirement for Radical S-adenoyslmethionine-dependent Dehydrogenases? BtrN contains two [4Fe–4S] Clusters ...... 290

7.1 Abstract ...... 291 7.2 Introduction ...... 292 7.3 Results and Discussion ...... 298 7.4 Conclusions ...... 306 7.5 Acknowledgments ...... 307 7.6 References ...... 308

Chapter 8 Further Characterization of Cys-Type and Ser-Type Anaerobic Sulfatase Maturating Enzymes Suggests a Commonality in Mechanism of Catalysis† ...... 312

8.1 Abstract ...... 313 8.2 Introduction ...... 315 8.3 Materials and Methods ...... 319 8.4 Results ...... 329 8.5 Discussion...... 358 8.6 Acknowledgement ...... 365 8.7 References ...... 366

Appendix A EPR Characterization of Radical Speices Trapped in the Radical SAM dehydrogenases anSMEcpe and AtsB ...... 375

A.1 Abstract...... 376 A.2 Introduction ...... 377 A.3 Materials and Methods ...... 378 A.4 Results ...... 379 A.5 Conclusions...... 385 A.6 References...... 386

LIST OF ABBREVIATIONS

5’-dA 5’-deoxyadenosine

5’-dA• 5’-deoxyadenosyl 5’-radical aa amino acid

AI as-isolated

AIR 5-aminoimidazole ribonucleotide amino-DOI amino-2-deoxy-scyllo-inosose anSME anaerobic sulfatase maturating enzyme apo RlmN RlmN lacking its [4Fe–4S] cluster

AT allo-threonine

BDE bond-dissociation energy

BME 2-mercaptoethanol

BS biotin synthase

BSA bovine serum albumin

Cfrwt→RCN Cfr produced in its apo form and then reconstituted d3-SAM S-adenosyl-[methyl-d3]methionine ddH2O distilled and deionized water

DOIA 2-deoxy-scyllo-inosamine

DOS 2-deoxystreptamine

DT dithionite

DTT dithiothreitol

EDTA ethylenediaminetetraacetic acid

EPR electron paramagnetic resonance

F flavodoxin

Fe/S iron–sulfur cluster

FGE formylglycine generating enzyme

FGly formylglycine

Flv flavodoxin

Flx flavodoxin reductase

FR flavodoxin reductase

HAT histone acetyl transferase

HEPES N-(2-hydroxyethyl)piperazine-N’-(2-ethanesulfonic acid)

HNMP 4-amino-5-hydroxymethyl-2-methylpyrimidine phosphate

HPLC high-performance liquid chromatography

IMAC immobilized metal affinity chromatography

IPTG isopropyl-α-D-thiogalactopyranoside

IS internal standard

LAM lysine 2,3-aminomutase

LC liquid chromatography

LC/MS HPLC with detection by QQQ mass spectrometry

LS lipoyl synthase m/z mass-to-charge-ratio

xii

MALDI matrix-assisted laser desorption/ionization

MALDI–TOF matrix assisted laser desorption ionization time-of-flight mass spectrometry

MRM multiple reaction monitoring

MS mass spectrometry

MSD multiple sulfatase deficiency

MW molecular weight

Ni-NTA nickel nitrilotriacetic acid

OD600 optical density (i.e. absorbance) at 600 nm

PCR polymerase chain reaction

PFL-AE pyruvate formate–lyase activating enzyme

PLP pyridoxal 5’-phosphate

PQQ pyrryloquinone

RCN reconstituted

RlmNwt→RCN RlmN produced in its apo form and then reconstituted rRNA ribosomal RNA

RS radical SAM

SAM S-adenosyl-L-methionine

SDS-PAGE sodium dodecylsulfate-polyacrylamide gel electrophoresis

SeCys selenocysteine siRNA single interfering RNA

TDP thiamine diphosphate

xiii

TFA trifluoroacetic acid

UV-vis UV-visible

Ve elution volume

Vo void volume wt wild-type

xiv

LIST OF FIGURES

Figure 1-1. Binding mode of SAM in RS proteins. Binding of SAM to the [4Fe–4S] cluster of biotin synthase. Color scheme: Brown, Fe; blue, N; yellow, S; red, O; grey, C. Structure prepared using Pymol Molecular Graphics System (http://www.pymol.org) from PDB 3RFA ...... 2

Figure 1-2. Binding mode of SAM in RS pr.Binding mode of SAM in RS proteins The reactions catalyzed by RlmN and Cfr ...... 5

Figure 1-3. Structure of RlmN+SAM. RlmN with SAM bound showing the flewxibe loop dip into the active site (red region). Color scheme: brown, Fe; blue, N; yellow, S; red, O; grey, C. Structure prepared using Pymol Molecular Graphics System (http://www.pymol.org) from PDB 3RFA ...... 7

Figure 1-4. Reactions catalyzed by MoaA/MoaC. Numbers highlight changes in positioning of atoms during the rearrangement...... 9

Figure 1-5. Active site of MoaA. Structure of MoaA with both GTP and SAM bound. Color scheme: brown, Fe; blue, N; yellow, S; red, O; grey, C. Structure prepared using Pymol Molecular Graphics System (http://www.pymol.org) from PDB 2FB3. .... 10

Figure 1-6. Reactions catalyzed by Radical SAM dehydrogenases. A) BtrN oxidizes 2- DOIA to 2-DOI. B) anSMEcpe and AtsB (C) both oxidize an active site cys or ser residue within their cognate sulfatase, respectively. anSMEcpe and AtsB both form the formylglycine residue...... 13

Figure 1-7. Structure of anSMEcpe with both 2-DOIA and SAM bound. Color scheme: brown, Fe; blue, N; yellow, S; red, O; pink, C. Structure prepared using Pymol Molecular Graphics System (http://www.pymol.org) unpublished...... 14

Figure 1-8. Structure of BtrN with SAM and 2-DOIA bound. Color scheme: brown, Fe; blue, N; yellow, S; red, O; grey, C. Structure prepared using Pymol Molecular Graphics System (http://www.pymol.org) unpublished...... 16

Figure 1-9. Maturation of the H-cluster of the [FeFe]-hydrogenase. Structure on left represents HydA, the hydrogenase from Desulfovibrio desulfuricans, with a [4Fe– 4S] cluster bound. In the presence of HydE, HydF, HydG, and appropriate small molecules, the H-cluster is formed on HydA. Color scheme: Red, iron; yellow, sulfur; grey, carbon; blue, nitrogen; black, unidentified atom (X). Structure prepared using Pymol Molecular Graphics System (http://www.pymol.org) from PDB 1HFE. ... 19

Figure 1-10. The reaction catalyzed by ThiC. Color-coding depicts the change in positioning of certain atoms during the rearrangement as determined by labelling experiments...... 23

Figure 1-11. Model for the reductive cleavage of SAM to generate a 5’-deoxyadenosyl radical...... 29

Figure 2-1. Reactions catalyzed by RlmN and Cfr. RlmN catalyzes uniquely methylation at C2, whereas Cfr catalyzes methylation at C8 and C2, although C8 is the preferred target...... 55

Figure 2-2 Elution profile of standards using the LC method described above with monitoring by A260 nm. pU, pseudouridine; SAM, S-adenosylmethionine; C, cytidine; U, uridine; G, guanosine; SAH, S-adenosylhomocysteine; A, adenosine; Trp, L-tryptophan; 5’-dA, 5’-deoxyadenosine; m8A, 8-methyladenosine...... 68

Figure 2-3 LC-MS analyses of the methylated products of the RlmN and Cfr reactions. Single turnover experiment in the presence of 250 µM 7-mer RNA substrate using 500 µM RlmN and 1 mM SAM (red solid trace); 500 µM RlmN and 1 mM d3-SAM (black solid trace); or 500 µM Cfr and 1 mM d3-SAM (black dotted trace). The peak at 6.8 min (black dotted trace) corresponds to m8A, with its associated mass spectrum in the above inset (black dotted trace). The peak at 6.85 min corresponds to m2A generated in the presence of SAM or d3-SAM with their associated mass spectra in the above inset (red and black trace, respectively). The green trace is a control corresponding to the above reactions in the absence of Cfr or RlmN...... 70

Figure 2-4. Control to analyze for SAM tightly bound to RlmNWT subsequent to purification. RlmNWT (200µM) was denatured with an equal volume of 100 mM H2SO4 containing 100 µM tryptophan (IS). As can be observed from the UV-vis trace above, no SAM is detected...... 71

Figure 2-5. Overlay of m8A standard (black trace) and product of Cfr (red trace)...... 72

Figure 2-6. Single turnover experiments. 250 µM 7-mer RNA substrate with (A) 500 µM RlmN and 1 mM d3-SAM (black trace), 500 µM RlmN (d3-met) with 1 mM SAM (red trace), or (B) 500 µM Cfr and 1 mM d3-SAM (black trace), 500 µM Cfr (d3- met) with 1 mM SAM (red trace). Peaks at m/z 282.1 correspond to m2A (RlmN) or m8A (Cfr) with no deuterium enrichment, while peaks at m/z 284.1 correspond to m2A (RlmN) or m8A (Cfr) with two deuterium atoms. No significant peaks at m/z 285.1 are observed, which would correspond to m2A (RlmN) or m8A (Cfr) with three deuterium atoms...... 73

Figure 2-7. Abortive cleavage of SAM (1 mM) in presence of 2 mM dithionite (closed markers) or Flv/Flx/NADPH (open triangle marker) by 200 µM RlmNWt (triangles), RlmNC118A (circles), RlmNC355A (squares)...... 75

Figure 2-8. Domain V of the 23S rRNA. Position A2503 is highlighted in red...... 76

xvi

reverse phase chromatography and fragmented by CID in the ion trap of a LTQ- Orbitrap XL-ETD mass spectrometer (Thermo-Fisher) with a Nano-Electrospray ionization. The precursor peptide had an m/z of 909.42017 (doubly charged) that eluted at 37.05 min. and was identified with an Xcorr score of 5.02 by Sequest. Y ions are indicated in blue and b ions in red as well as the charge state of each. The y ions that indentify the methylation is at C355 are y10 and y11. Y10 has a m/z = 1043.5496 corresponding to the sequence GQLAGDVIDR and y11 has a mz = 1160.5206 corresponding to the sequence C355GQLAGDVIDR, an increase of 116.971 mass units indentifying the C355 as the site of methylation...... 79

Figure 2-11. Postulated mechanism of A2502 methylation. RlmN (top) and Cfr (bottom), mechanism for the methylation of C2 and C8 of A2503, respectively. In contrast to catalysis by RlmN, significant solvent hydron exchange takes place during turnover as shown in step 5 of the mechanism. See text on RlmN for description of the analogous mechanism...... 80

Figure 2-12. UV-visible spectra of A) RlmN Wt (16.8 μM), B) RlmN C118A (12.6 μM), C) RlmN C355A (12 μM), and D) Cfr Wt (27 μM)...... 82

Figure 2-16. Single turnover experiment with 300 µM d3-met Cfr in the presence of 1.5 mM SAM, and 10 µM 771-mer RNA substrate. The peak at 6.7 min corresponds to 8 8 a mixture of m A (1) and d2-m A with the associated mass spectrum in the above left inset. The peak at 7.7 min corresponds to a mixture of 2,8-dimethyladenosine and 2,8-d2-dimethyladenosine with the associated mass spectrum in the above right inset. The lack of complete transfer of deuterium from the enzyme to the product indicates a step in which substrate deuteria are exchanged with solvent hydrons...... 86

xvii

Figure 3-1. Reactions catalyzed by RlmN and Cfr in vivo...... 100

Figure 3-2. Mechanism proposed by Grove et al for the RlmN reaction.(7) ...... 101

Figure 3-3. Active site structure of RlmN with SAM present...... 102

Figure 3-4. UV-visible spectra of RlmNwt and Cfrwt . A) AI RlmNwt (11 μM, black trace, left Y-axis) and RCN RlmNwt (6 μM, red trace, right Y-axis). B) AI Cfrwt (7 μM, black trace, left Y-axis) and RCN Cfrwt (8μM, red trace, right Y-axis) The A278/A400 ratios are given in Table 3.2...... 112

Figure 3-5. 4.2-K/53-mT Mössbauer spectra of AI RlmNwt (A), RCN RlmNwt (B), AI Cfrwt (C), and RCN Cfrwt (D). Experimental spectra are shown as vertical bars. The solid lines are quadrupole doublet simulations with parameters quoted in the text, accounting for 93% (A), 95% (B), 86% (C), and 98% (D) of the total intensity...... 113

Figure 3-6. EPR spectra of RlmN and Cfr Mössbauer samples...... 114

Figure 3-7. UV-visible spectra of A) AI RlmNC118A (10 μM, black trace, left Y-axis) and AI RlmNC355A (6 μM, red trace, right Y-axis); and B) RCN RlmNC118A (12 μM, black trace, left Y-axis) and RCN RlmNC355A (7 μM, red trace, right Y-axis). The A280/A400 ratios for RlmNC355A are given in Table S2...... 115

EQ = 1.14 mm/s) accounting for 89% (A), 94% (B), 95% (C), and 97% (D) of the total intensity, indicating 0.9 (A), 1.0 (B), 1.1 (C), and 1.0 (D) [4Fe–4S] clusters per polypeptide (see Table S2)...... 116

Figure 3-9. Extracted ion chromatograms for the C355-containing peptide from trypsin digests of RlmN. Black traces (A, B, C, D, E, F) are EIC of m/z 931.0 [M+2H], indicative of carbamidomethyl modification to C355. Red traces (A, B, C, D, E, F) are EIC of m/z 909.6 [M+2H], indicative of a methyl modification of C355. Traces correspond to A) AI RlmNwt; B) AI RlmNC125A-C129A-C132A; C) AI RlmNC118A; D) apo RlmNwt, inclusion bodies; E) apo RlmNwt, soluble fraction; F) apo RlmNwtRCN after addition of 1.5 mM SAM. Spectral pair intensities are normalized to the most abundant EIC...... 118

xviii

circles) in the presence of 1.5 mM SAM; or apo RlmNwtRCN (150 µM) in the presence of 1.5 mM d3-SAM (black triangles). Error bars indicate one standard deviation from the average of three assays. C) Q-Tof MS analysis of tryptic peptides

from apo RlmNwtRCN after incubation in the presence of SAM (black trace) or d3- SAM (red trace). Indicated m/z values correspond to the +2 charge state...... 119

Figure 3-11. Time-dependent formation of m2A from an assay containing apo

RlmNwtRCN. The enzyme was incubated with the 771-mer RNA substrate and 1.5 mM A) SAM or B) d3-SAM. Time points are 0, 0.5, 1, 10, and 30 min...... 120

Figure 3-12. MS/MS of apo RlmNwt tryptic peptides after reconstitution and subsequent + incubation of the protein in the presence of SAM or d3-SAM. Selected y ions from collision induced dissociation (CID) of either the m/z = 909.6 (mCys peptide, M+2H) or the m/z = 911.2 (d3-mCys peptide, M+2H) species, indicating a difference 1+ of m/z = +3 for the sample treated with d3-SAM. Ion y11 corresponds to the 355 1+ sequence C GQLAGDVIDR, while ion y10 (not observed in this spectrum) 1+ corresponds to the sequence GQLAGDVIDR. The y11 ion displays m/z = 1160.59 (1163.6 for deuterated methyl group), corresponding to a m/z increase of 116.97 1+ over that of the y10 ion, the mass of a methylCys residue...... 121

Figure 3-13. Methyl transfer catalyzed by apo RlmNwtRCN (150 µM) in the presence of

the in vivo reducing system. Production of SAH by apo RlmNwtRCN in the presence of 1.5 mM SAM and the F/FR/NADPH reducing system (black triangles); or apo

RlmNwtRCN in the presence of 1.5 mM d3-SAM and the F/FR/NADPH reducing system (green diamonds)...... 122

Figure 4-1. Mechanistic proposal for catalysis by Cfr. The numbering of the carbons in the adenosine ring is displayed in 1...... 135

Figure 4-2. Mims and Davies pulse sequences used in this work (upper left), and illustration of distortions induced by the respective sequences relative to a

hypothetical axial ENDOR spectrum (S=1/2, I=1/2, Az=AII> Ax=Ay=A) . The lower part of the figure illustrates peak positions for two cases: weak (2νL> [A > AII]) and strong (2νL< [A > AII]) hyperfine coupling. This distinction arises from the fact that in standard ENDOR experiments no information about the sign of the corresponding nuclear transition frequency is obtained. Thus, in the strong coupling case, negative frequency signals are "folded" to the positive axis...... 142

Figure 4-3. UV-visible spectrum of FSQ• (left) and the corresponding CW-EPR spectrum recorded at 100 K (right, blue trace). The arrow above 580 nm indicates the peak at which the concentration of the semiquinone standard was determined. CW-EPR spectra were acquired as described in the Materials and Methods...... 145

Figure 4-4. CW EPR spectra (blue) in the presence of the 155mer substrate (A), deu155mer substrate, [2H] (B), doubly labeled sample with 2H and 13C isotopes [2H,13C] (C), and 13C-labeled mCys, [13C] (D). Inserts show corresponding isotope labeling positions of A2503 in the radical species. Experimental conditions (similar for all data): temperature, 100 K; MW frequency, 9.379 GHz; modulation amplitude,

xix

0.5 mT; MW power, 12.8 μW. Simulations accounting for C8-1H/2H, N7(14N) and 13C HF coupling constants from Table 1 are shown in red...... 148

Figure 4-5. Characterization of wt CfrapoRCN. A) UV–vis traces of wt Cfrapo (5.1 µM, solid black line) and wt CfrapoRCN (7.1 µM, solid red line). B) Methyl transfer catalyzed by wt CfrapoRCN (150 µM) in the presence of 2 mM SAM. The data were fitted to a pseudo-first order exponential equation (red line)...... 151

Figure 4-6. Comparative ENDOR measurements of the unlabeled (blue) and doubly labeled [2H, 13C] samples (red), which allow identification of a total of 5 HF coupling constants. Experimental conditions: A. Davies ENDOR sequence; Tinv, 200 ns; TRF=12 μs; MW frequency, 9.7119 GHz; Magnetic Field, 346.0 mT. B. Mims ENDOR sequence; τ, 200 ns; TRF=15 μs; MW frequency, 9.730 GHz; Magnetic Field, 3470 mT C. Mims ENDOR sequence; τ, 200 ns; TRF=15 μs; MW frequency, 9.710 GHz; Magnetic Field, 3462 mT. In C, the green trace is the difference of the experimental data for the unlabeled and [2H,13C] samples. All spectra were accumulated over a period of 40-60 hours...... 153

Figure 4-8. Simulation of the CW EPR spectra, presented in Figure 1, accounting for a strongly coupled proton at N7 (see Table S2 for the corresponding HF coupling constants), according to the protonated model...... 159

Figure 4-9. Kinetic Competence of Radical Intermediate: A. Representative CW-EPR spectra from the rapid freeze-quench time course (green trace, 180 s; red trace, 45 s; blue trace, 15 s). B. Time-dependent formation and decay of the A2503 radical species (solid triangles) and time-dependent formation of m8A (open triangles). The EPR data were fitted to an ABC kinetic model, while the m8A data were fitted to a pseudo-first order exponential equation. The extracted rate constants are reported in the text. Error bars, where indicated, denote the standard deviation from two independent reactions. CW-EPR spectra were collected as described in the Materials and Methods...... 160

Figure 5-1. Proposed mechanism of methylation by RlmN (A) and Cfr (B)...... 178

Figure 5-2. non-reducing SDS-PAGE gel of wt RlmNapoRCN and wt CfrapoRCN in the

absence of reductants: lane M, molecular weight marker; lane 1, CfrapoRCN; lane 2,

RlmNapoRCN. Standards on the left side indicate protein size in kDa...... 191

Figure 5-3. Active site of RlmN in the presence of SAM demonstrating the proximity of Cys118 and mCys355...... 193

Figure 5-4. Reaction catalyzed by diamide in the presence of free thiols...... 194

Figure 5-5. Effect of oxidized DsbA on intramolecular disulfide bond formation. Time

dependent production of SAH by A) 70 µM wt RlmNapoRCN and B) 70 µM wt CfrapoRCN in the presence of 300 µM oxDsbA. C) non-reducing SDS-PAGE gel of wt RlmNapoRCN and D) wt CfrapoRCN after a 60 min incubation with 300µM oxDsbA followed by reaction with 1, 2, 3, 6, 12, 25 equivalents of TrxC35A. lane M, molecular weight marker; lane 1, mixture in the presence of 1 mM DTT; lane 2, 1 eq. TrxC35A; lane 3, 2 eq. TrxC35A; lane 4, 3 eq. TrxC35A; lane 5, 6 eq. TrxC35A; lane 6, 12 eq. TrxC35A; lane 7, 25 eq. TrxC35A. Arrows on the left side indicate bands of importance. R-R, RlmN disulfide dimer; R-TC35A, RlmN- TrxC35A mixed disulfide; R, RlmN monomer; DsbA, oxidized DsbA; C-C, Cfr disulfide dimer; C-TC35A, Cfr-TrxC35A mixed disulfide; C, Cfr monomer. Standards on the left side indicate protein size in kDa...... 195

Figure 5-6. non-reducing SDS-PAGE gels of reaction between diamide and wt Cfr (A) or

wt RlmN (B). A) Time dependent formation of a disulfide bond on wt CfrapoRCN via 1 mM of diamide. The reactions were quenched with non-reducing SDS-PAGE buffer: lane M, molecular weight marker; lane 1, before diamide addition; lane 2, 1 min; lane 3, 5 min; lane 4, 10 min; lane 30, 1 min; lane 6, 60 min. B) Reaction of wt

RlmN apoRCN with 1, 2, 6, 10 equivalents of diamide. The reactions were quenched with non-reducing SDS-PAGE buffer: lane M, molecular weight marker; lane 1, 1 eq.; lane 2, 2 eq.; lane 3, 6 eq.; lane 4, 10 eq. Standards on the left side indicate protein size in kDa...... 196

Figure 5-7. Reversiblity of disulfide bond formed on wt RlmN apoRCN. 70 µM RlmN was reacted with 1 eq. of diamde for 10 min before 2 mM SAM was added (green trace). Addition of 10 mM DTT restores SAH productions (blue trace). Peak at 3.6 min corresponds to SAH and the peak at 5.2 min corresponds to tryptophan (IS)...... 197

Figure 5-8. Generation of a disulfide in wt RlmNapoRCN and wt CfrapoRCN via diamide oxidation for in-gel digestion and HRMS/MS analysis. The bands corresponding to

RlmNapoRCN (~47 kDa), oxidized Cfr apoRCN (~43 kDa, bottom band), and Cfr apoRCN (~45 kDa, top band) were excised, labeled, and digested as described in the Materials and Methods. Note the a small amount of dimer formation in the RlmN and Cfr samples (~ 95 kDa and ~ 80 kDa, respectively)...... 198

Figure 5-9. Selective labeling results of peptides analyzed by HRMS from the top band

(A) and bottom band (B) from the CfrapoRCN reaction with diamide. Bands were excised from the gel, reduced with DTT, labeled with d2-iodoacetamide, and digested with trypsin/gluC proteases. Letters above amino acids denote the type of modification. If two labels are present, the label closest to the sequence is in a higher concentration. C, CAM; O, oxidation; d, d2-CAM. Green highlighted areas indicated sequence coverage...... 199

xxi

patterns show the link is between Cys118 and Cys355. The table to the right is a list of m/z values for b and y ions labeled in this spectrum. D) MS/MS analysis of the Cfr disulfide linked peptide with m/z of 1230.2710. Fragmentation patterns show the link is between Cys105 and Cys338. The table to the right is a list of m/z values for b and y ions labeled in this spectrum...... 202

Figure 5-11. SAH production by of wt RlmNapoRCN, C118S RlmNapoRCN, and wt CfrapoRCN in a redox titration with the GSH:GSSG ratios found in Table 2. A) 75 µM wt RlmNapoRCN first oxidized with diamide, then mixed with the redox buffer. B) 75 µM wt CfrapoRCN first oxidized with diamide, then mixed with the redox buffer. C) 75 µM wt RlmNapoRCN (open circles) or 10 µM C118S RlmNapoRCN mixed with only the redox buffer. D) 75 µM wt CffapoRCN mixed with only the redox buffer. It should be noted that the concentration of SAH produced by C118S

RlmNapoRCN and wt CffapoRCN is low because of the poor stability of these proteins. Dashed lines represent the fit of the data using Equation 1...... 204

Figure 5-12. Reduction of disulfide bond by F/FR/NADPH reducing system. wt

RlmNapoRCN was mixed with diamide and then 2 mM SAM (black dotted line), or diamide then F/FR/NADPH, then 2 mM SAM (blue line). wt CfrapoRCN was mixed with diamide (red trace), or diamide then F/FR/NADPH, then 2 mM SAM (black

trace). For comparsin wt RlmNapoRCN was mixed with diamide then 10 mM GSH, then 2 mM SAM (green trace). The peak at 3.6 min corresponds to SAH, while the peak at 5.2 min corresponds to tryptophan (IS)...... 206

Figure 5-13. Multiple turnover of AI wt RlmN or AI wt Cfr in the absence and presence of reductants. A) 10 µM AI wt RlmN was reacted with 300 µM 155mer RNA substrate and 2 mM SAM in the presence of 10 mM GSH (black triangles), 10 mM GSSG (blue squares), or 10 mM DTT (red circles). B) 10 µM AI wt Cfr was reacted with 300 µM 155mer RNA substrate and 2 mM SAM in the presence of 10 mM GSH (black triangles) or10 mM GSSG (blue squares). Only the linear portion of m2A or m8A production is shown...... 208

Figure 5-14. Methylation of C2 in A2503 by RlmN with radical disulfide anion formation driving release of product...... 216

Figure 6-1. AtsB dependent oxidation of Ser72 of AtsA from K. pneumoniae to afford the FGly cofactor. Also shown is the subsequent hydration of the aldehyde to give the geminal diol, which is the active form of the cofactor (A). The proposed mechanism of action of the FGly cofactor. The first step involves a nucleophilic attack on the sulfate group by the geminal diol form of the FGly cofactor with concomitant modification of the FGly cofactor and release of the corresponding alcohol. The second step is an internal elimination, which results in regeneration of the FGly cofactor (B)...... 232

Figure 6-2. SDS-PAGE analysis of purified AtsB and AtsA (aa 21-578). Lane 1, molecular mass markers; lane 2, AtsB (MW: 46,431.9 Da) purified by IMAC using cobalt Talon affinity resin; lane 3, AtsA (MW: 66,318.9 Da) purified by IMAC using nickel-NTA resin. The gel was stained with coomassie brilliant blue...... 245

xxii

Figure 6-3. UV-visible spectra of (A) AI WT AtsB (9.88 µM, solid line, left Y-axis) and RCN WT AtsB (14.4 µM, dotted line, right Y-axis); and (B) AI C35A–C39A–C42A AtsB triple variant (11.9 µM). The A280/A395 of AI WT and RCN AtsB were 2.69 and 2.31, respectively. The A280/A395 of the AI C35A-C39A-C42A AtsB triple variant was 3.12...... 247

Figure 6-4. Mössbauer spectra of (A) AI WT AtsB; (B) RCN WT AtsB; (C) AI C35A– C39A–C42A AtsB triple variant; and (D) RCN C35A–C39A–C42A AtsB triple variant. All spectra were collected at 4.2 K in an external 53-mT magnetic field. The solid lines in (A) – (D) represent a quadrupole doublet with the following parameters: δ = 0.44 mm/s and ΔEQ = 1.17 mm/s...... 249

Figure 6-5. Quantification of FGly formation by MALDI MS in the presence of phenylhydrazine. The reaction contained 100 µM AI WT AtsB, 1 mM SAM, 2 mM peptide 1 (m/z: 1961.5), and 2 mM dithionite. (A) t = 0; (B) t = 60 min. Samples were quenched in 1 M HCl containing 100 mM phenylhydrazine, and spotted on a MALDI target plate as described in the Materials and Methods. The phenylhydrazine adduct displays m/z 2049.5. (C) t = 60 min for the reaction using the Flv/Flx/NADPH reducing system...... 252

Figure 6-8. HPLC detection of formylglycine formation. Reactions contained 50 µM AtsB WT or C35A–C39A–C42A triple variant, 1mM SAM, 2 mM dithionite, 1 mM tryptophan (IS), and 1 mM peptide 2, and were quenched in 100 mM H2SO4 containing 1 M hydroxylamine. (A) HPLC trace of reaction in the presence of AI WT AtsB. t = 0 (solid black line); t = 5 min (dotted line); t = 10 min (dashed line); and t = 20 min (solid gray line). Directional arrows correspond to increases or decreases in peaks that undergo turnover-dependent changes. Retention times: 5’-dA (5.8 min); L-tryptophan (Trp, 9.8 min) ; peak 1 (10.6 min); peak 2 (12.1 min); peak 3 (12.7 min); peptide 2 (20.4 min) (B) HPLC trace of reaction in the presence of AtsB C35A–C39A–C42A triple variant. t = 0 (solid black line); t = 20 (dashed line).... 257

xxiii

F Figure 6-9. Time dependent formation of 5’-dA (closed triangles, left Y-axis) and FGly (open squares, left Y-axis), and depletion of peptide 2 (closed squares, right Y-axis). Reactions contained 50 µM AtsB, 1 mM SAM, 1 mM peptide 2, 2 mM 1 mM tryptophan (IS) and dithionite (A) or 100 µM AtsB, 25 µm Flv, 5 µM Flx, and 2 mM NADPH (B). The data are the average of two independent trials, and error bars denote one standard deviation...... 258

Figure 6-10. Time dependent formation of 5’-dA (closed triangles, left Y-axis) and FGly (open squares, left Y-axis), and depletion of peptide 3 (closed squares, right Y-axis). Reactions contained 50 µM AtsB, 1 mM SAM, 1 mM peptide 3, 2 mM 1 mM tryptophan (IS) and dithionite (A) or 100 µM AtsB, 25 µm Flv, 5 µM Flx, and 2 mM NADPH (B). The data are the average of two independent trials, and error bars denote one standard deviation...... 262

Figure 6-11. Working hypothesis for AtsB-catalyzed FGly formation (Numbers in bold represent individual steps). The reaction is initiated by coordination of the hydroxyl group of the target seryl residue to an empty coordination site on one of the three Fe/S clusters. Subsequent to deprotonation (Step 1) and generation of a 5’-dA• (Step 2), the 5’-dA• abstracts a hydrogen atom from C-3 of the target seryl residue (Step 3), which is followed by inner-sphere electron transfer and radical recombination to afford the FGly product and a one-electron reduced Fe/S cluster (Step 4). Last, the electron is transferred to an unidentified acceptor (Step 5)...... 268

Figure 7-1. Reaction catalyzed by BtrN ...... 292

Figure 7-2. UV-visible spectra of AI BtrN WT (21.3 μM, solid line, left Y-axis) and RC BtrN WT (24.5μM, red dotted line, rightY-axis) The A280/A387 ratios of AI and RC were 2.28 and 2.1, respectively...... 299

Figure 7-3. 4.2-K/53-mT Mössbauer spectra of wt BtrN (A and B) and ...... 300

Figure 7-4. UV-visible spectra of AI BtrN C-16,20,23-A (46.8 μM, solid line, left Y-axis) and RC BtrN C-16,20,23-A (36.1μM, red dotted line, rightY-axis) The A280/A387 ratios of AI and RC were 3.0 and 2.5, respectively...... 302

Figure 7-5. UV-visible spectra of AI BtrN C-69-A (35.1 μM, solid line, left Y-axis) and RC BtrN C-69-A (15.2μM, red dotted line, rightY-axis) The A280/A387 ratios of AI and RC were 2.3 and 1.8, respectively...... 303

Figure 7-6. UV-visible spectra of AI BtrN C-235-A (24.1 μM, solid line, left Y-axis) and RC BtrN C-235A (14.1μM, red dotted line, rightY-axis) The A280/A387 ratios of AI and RC were 3.3 and 1.8, respectively...... 304

Figure 7-7. Proposed mechanism for BtrN. Binding of (co)substrate and dissociation of (co)products are labeled with ① and ②, respectively. Electron flow is indicated by red arrows...... 305

Figure 8-1. Reactions catalyzed by BtrN (A) and anSMEs (B)...... 316

xxiv

Figure 8-2. Equation describing a burst phase followed by a steady-state phase...... 329

Figure 8-3. SDS–PAGE analysis of anSMEcpe. Lane 1, molecular mass markers. Lane 2, purified anSMEcpe (45,740 Da). The gel was stained with Coomassie Brilliant Blue. .. 330

Figure 8-4. UV-visible spectra of A) AI WT anSMEcpe (5 μM, solid line, left Y-axis) and RC WT anSMEcpe (10 μM, dotted line, right Y-axis). The A279/A387 ratios of AI and RCN proteins were 2.8 and 2.1, respectively. B) UV-visible spectra of AI anSMEcpeC15A/C19A/C22A (9.4 µM, solid line, left Y-axis) and RCN anSMEcpeC15A/C19A/C22A (5.6 µM, dashed line, right Y-axis). The A279/A387 ratios of AI and RCN proteins were 4.2 and 2.4, respectively...... 331

Figure 8-5. 4.2-K/53-mT Mössbauer spectra of WT anSMEcpe (A and B) and anSMEcpeC15A/C19A/C22A (C and D). A and C are the AI forms, and (B) and (D) are the RCN forms. The solid lines are quadrupole doublet simulations with parameters quoted in the text...... 333

Figure 8-6. EPR of Mӧssbauer samples. anSMEcpe AI (red trace), anSMEcpe RCN (black trace), anSMEcpeC15A/C19A/C22A AI (green trace), and anSMEcpeC15A/C19A/C22A RCN (blue trace). Spectra were collected on a Bruker ESP-300 X-Band EPR spectrometer with the following parameters: frequency, 9.51 GHz; temperature, 13 K; power, 0.101 mW; and modulation amplitude, 10 Gauss. Spin quantification was performed by comparing the double integral of the obtained signal to that of a 1 mM Cu(II)-EDTA standard collected under identical conditions...... 337

Figure 8-7. Molecular-sieve analysis of WT anSMEcpe and AtsB. A) WT RCN anSMEcpe; B) WT RCN AtsB in the absence of substrate (blue trace), in the presence of 2 mM Kp18Ser (black trace), or in the presence of WT AtsA from Kp (red trace). Molecular-sieve chromatography was conducted under anaerobic conditions as described in Materials and Methods. WT RCN anSMEcpe eluted at 64.7 mL and WT RCN AtsB eluted at 65.4 mL, yielding calculated molecular masses of 37.5 kDa and 33.5 kDa, respectively. Molecular masses were determined from a plot of known standards that were analyzed under identical conditions (Figure 4A, inset)...... 339

Figure 8-8. LC-MS analysis of anSMEcpe assay. The assay was conducted as described in Materials and Methods using Kp18Cys as the substrate, Kp9Ser as an internal standard, and dithionite (4.5 mM) as the required reductant. Green (t=0); Red (t=30 min); Black (t=5, 10, and 20 min)...... 340

xxv

Flv/Flx/NADPH reducing system. Reaction mixtures contained 40 μM anSMEcpe, 1 mM SAM, 1 mM Kp18Cys, 50 μM Flv, 15 μM Flx, and 2 mM NADPH. The data are the averages of three independent trials, and error bars denote one standard -1 deviation. Vmax/[ET] values for 5’-dA and Kp18Cys are 0.28 ± 0.022 min and 0.26 ± 0.022 min-1, respectively...... 342

Figure 8-10. Turnover of WT RCN anSMEcpe with Cp18Cys. A) Time-dependent formation of 5’-dA (black triangles) and depletion of Cp18Cys (red squares) in the presence of dithionite. Reaction mixtures contained 4 μM anSMEcpe, 1 mM SAM, 1 mM Cp18Cys, and 3 mM dithionite. The data are the averages of two independent trials, and error bars denote one standard deviation. Vmax/[ET] values for 5’-dA formation and peptide consumption are 4.50 ± 0.052 min-1 and 1.91 ± 0.259 min-1, respectively. B) Time-dependent formation of 5’-dA (black triangles) and depletion of Cp18Cys (red squares) in the presence of the Flv/Flx/NADPH reducing system. Reaction mixtures contained 40 μM anSMEcpe, 1 mM SAM, 1 mM Cp18Cys, 50 μM Flv, 15 μM Flx, and 2 mM NADPH. The data are the averages of two independent trials, and error bars denote one standard deviation. Vmax/[ET] values for 5’-dA formation and peptide consumption are 0.22 ± 0.003 min-1 and 0.21 ± 0.032 min-1, respectively...... 343

Figure 8-11. Turnover of WT RCN anSMEcpe with Kp18Ser. A) Time-dependent formation of 5′-dA (closed triangles), and Kp18FGly (open squares) in the presence of DT. Reaction mixtures contained 4 μM anSMEcpe, 1 mM SAM, 1 mM Kp18Ser, and 3 mM DT. The data are the averages of two independent trials, and error bars denote one standard deviation. Vmax/[ET] values for 5’-dA and Kp18FGly are 1.00 ± 0.029 min-1 and 0.85 ± 0.001 min-1 , respectively. B) Time-dependent formation of 5′-dA (closed diamonds) and Kp18FGly (open squares) in the presence of the Flv/Flx/NADPH reducing system. Reaction mixtures contained 40 μM anSMEcpe, 1 mM SAM, 1 mM Kp18Ser, 50 μM Flv, 15 μM Flx, and 2 mM NADPH. The data are the averages of three independent trials, and error bars denote one standard -1 deviation. Vmax/[ET] values for 5’-dA and Kp18FGly are 0.074 ± 0.009 min and 0.073 ± 0.004 min-1, respectively...... 344

Figure 8-13. Correlation of spectral changes and product formation during anSMEcpe turnover. A) X-Band EPR (77 K) spectra of a reaction mixture containing 100 μM anSMEcpe, 2 mM SAM, 2 mM Kp18Cys and 204 μM Flv• at 1 min (red), 15 min (green), and 30 min (black). Spectra were recorded as described in Materials and Methods. B) Time-dependent quantification of Flv• (open circles), 5′-dA (closed triangles), and Kp18FGly (open squares). Lines (black line, 5’-dA; red line, Kp18FGly) represent fits to a burst phase followed by steady-state phase...... 347

Figure 8-14. Low-temperature X-Band EPR of Flv• and anSMEcpe during turnover. Reaction mixtures contained 100 μM anSMEcpe, 2 mM SAM, 2 mM Kp18Cys and

xxvi

-1 -1 204 μM Flv• (580=4.57 mM cm ) at 13 K at 1 min (green), 15 min (red), and 30 min (black). Spectra were collected on a Bruker ESP-300 X-Band EPR spectrometer under the following conditions: frequency, 9.51 GHz; temperature, 13 K; power, 0.101 mW; and modulation amplitude, 10 Gauss...... 348

Figure 8-15. Correlation of spectral changes and product formation during AtsB turnover with Kp18Cys. A) X-Band EPR (77 K) spectra of a reaction mixture containing 150 μM AtsB, 1 mM SAM, 1 mM Kp18Ser, 75 µM Flvox, and 75 μM Flv• at 1 min (red), 15 min (green), and 30 min (black). Spectra were recorded as described in Materials and Methods. B) Time-dependent quantification of Flv• (open circles) and 5′-dA (closed triangles). The black line is a fit of the 5’-dA data to an equation describing a burst phase followed by a steady-state linear phase...... 349

Figure 8-16. Stereochemical designation of threonine and allo-threonine...... 351

Figure 8-17. Turnover of WT RCN anSMEcpe with Kp18Thr or Kp18alloThr. A) Time-dependent formation of 5′-dA (closed black triangles) and disappearance of Kp18Thr (closed red squares). Reaction mixtures contained 100 μM anSMEcpe, 1 mM SAM, 0.5 mM Kp18Thr, and 3 mM DT. Vmax/[ET] values for 5’-dA formation and Kp18Thr disappearance are 0.29 min-1 and 0.11 min-1 , respectively. B) Time- dependent formation of 5′-dA (closed triangles) and disappearance of Kp18alloThr (closed red squares). Reaction mixtures contained 100 μM anSMEcpe, 1 mM SAM, 0.5 mM Kp18alloThr, and 3 mM DT. Vmax/[ET] values for 5’-dA formation and Kp18alloThr disappearance are 0.07 min-1 and 0.007 min-1, respectively...... 352

Figure 8-18. MALDI MS analysis of a WT RCN anSMEcpe reaction with Kp18Thr (A), or Kp18alloThr (B). Aliquots removed from the reaction at 0 min (black trace) and 10 min (red trace) were derivatized with DNPH as described in Materials and Methods. Spectra were recorded as previously described (2)...... 353

Figure 8-19. MALDI MS analysis of a WT RCN AtsB reaction with Kp18Thr (A), or Kp18alloThr (B). Aliquots removed from the reaction at 0 min (black trace) and 10 min (red trace) were derivatized with DNPH as described in Materials and Methods. Spectra were recorded as previously described (2)...... 353

Figure 8-20. UV-vis spectra of AI AtsB C127A and C245A. AI AtsB C127A (11.3 µM; solid black trace, left axis) contained 9.8 ± 0.1 irons and 9.6 ± 0.5 sulfides per polypeptide. AI AtsB C245A variant (6.2 µM; dashed red trace, right axis) contained 12.0 ± 1.1 irons and 15.0 ± 0.3 sulfides per polypeptide. The A395/A280 ratio for both is 0.38...... 354

Figure 8-21. UV-vis spectrum of AI AtsB C291A. The protein (6.4 µM), contained 6.7 ± 0.1 irons and 5.6 ± 0.6 sulfides per polypeptide. The A405/A280 ratio was 0.39...... 355

Figure 8-22. UV-vis spectrum of AI (solid line) and RCN (dashed line) anSMEcpe C276A. AI anSMEcpe C276A (4 μM, solid line, left Y-axis) and RCnN anSMEcpe C276A (8.4 μM, dotted line, right Y-axis). The A410/A280 ratios of AI and RCN proteins were 0.36 and 0.45, respectively...... 356

xxvii

Figure 8-23. Time-dependent formation of 5′-dA (closed triangles) and Kp18FGly (open squares) with RCN anSMEcpe C276A and Kp18Cys. Reaction mixtures contained 200 μM anSMEcpe C276A, 1 mM SAM, 1 mM Kp18Cys, and 3 mM DT. The data are the averages of two independent trials, and error bars denote one standard -1 deviation. Vmax/[ET] values for 5’-dA and Kp18FGly are 0.14 ± 0.001 min and 0.001 ± 0.0001 min-1, respectively...... 358

Figure A-1. Peptides substrates used to trap radical species in reaction by AtsB and anSMEcpe. The residue that is acted upon by the anSMEs is shown in stick format. The β-position is believed to be the target hydrogen that is removed in the reactions. ... 378

Figure A-2. SDS-PAGE anlaysis of AtsB reaction with Kp18Allylgly before (lane 1) after 20 min (lane 2). The band appears to shift upward indicating a cross-linked intermediate may form (red arrows)...... 379

Figure A-3. EPR of AtsB and anSMEcpe with Kp18Allyl peptide after 1 min. The spectra were recorded a 77K...... 380

Figure A-4. Peptides EPR of anSMEcpe labeled with 56Fe or 57Fe after reaction with Kp18Allyl peptide for 1 min. The spectra were recorded a 77K...... 381

Figure A-5. EPR of anSMEcpe with Kp18MeCys or Cp18Cys peptide after 30 s. The spectra were recorded a 77K...... 382

Figure A-6. Product analysis of anSMEcpe reaction with Kp18MeCys after 30 min. Kp18MeCys has an m/z of 2032 Da before the reaction. The product has an m/z of 2078 Da after the reaction. The increase, interestingly, is equivalent to the addition of a methyl and sulfur...... 383

Figure A-7. EPR of anSMEcpe with Kp18Thr peptide after 1 min. The spectra were recorded at 80K...... 384

xxviii

LIST OF TABLES

Table 2-1. Primers for RlmN variants C118A and C355A ...... 58

Table 3-1. Primers for constructing RlmN variants ...... 104

Table 3-2. UV–vis characteristics and Fe and S2- content of wt and variant RlmN or Cfr proteins ...... 106

Table 3-3. Retention time and fragmentation products monitored by LC-MS ...... 109

Table 4-1. HF coupling constants (all in MHz) extracted from the experimental data as compared with calculated values from spin-unrestricted DFT methods...... 147

Table 4-2. Comparison of HF coupling constants (in MHz) obtained experimentally or calculated for two models that differ by the protonation state of the N7 atom...... 149

Table 5-1.Custom peptide sequence used in Proteome Discoverer 1.3...... 186

Table 5-2. Ratios of GSH and GSSG used for redox titrations...... 188

Table 7-1. Retention times and monitored m/z values for products and substrates ...... 295

Table 7-2. Primers for constructing BtrN variants ...... 297

Table 7-3. Activity, Fe, and S of Wt and Variants of BtrN ...... 301

Table 8-1. List of primers for site-directed mutagenesis of AtsB ...... 322

Table 8-2. Retention times and monitored m/z values for Detection Method 1 ...... 326

Table 8-3. Retention times and monitored m/z values for Detection Method 2 ...... 327

xxix

ACKNOWLEDGEMENTS

As I sit here with my dissertation complete, I can’t help but reminisce about the journey to this point. I feel this section of the thesis should be dedicated to the people who have both inspired and motivated me to push the limits of my capabilities.

Going back in my memories, I find I have always had an innate curiosity of the natural world. Science was always my favorite subject spanning my entire education. I had numerous teachers in those early stages, but none stand out in my mind like Mrs.

Karen Ash. She was the taught every science class I had for three years in high school and laid the ground work for my love of science. Mrs. Ash approached every day and every class I was a part of with vigor and enthusiasm that instilled in me a deep passion for the pursuit of understanding. I am not sure I would have made it to this point without her persistent, yet gentle push towards a science major in college. I thank you for this part of my education.

When I arrived at Shepherd University, my original and almost thoughtless plan to pursue a degree in Environmental science—what I would have done with such a degree escapes me now—was derailed by my General Chemistry professor, Dr. Jack

Schmidt. His lighthearted and thorough approach to chemistry education was so impressive that I decided he would be an excellent academic advisor. One day after class

I asked him for this favor and he excitedly agreed. During our first scheduled meeting, I expressed second thoughts about my choice in major. I then suggested that I should maybe switch to chemistry or biology as a major. Dr. Schmidt, being an astute observer of character, immediately hinted that maybe chemistry would be a little too difficult for

xxx me knowing that this comment would instantly make me chose chemistry as a major. This was a quality of Dr. Schmidt that always stood out; how to say the right thing in subtle fashion that would have lasting impact. As he predicted, I immediately went to the registrar’s office and changed my major to chemistry to prove to Dr. Schmidt that I could accomplish this task! In the subsequent years, I had four more chemistry courses and one Russian course under Dr. Schmidt’s tutelage. Ultimately, Dr. Schmidt was responsible for not only my attendance of graduate school, but also the institution (Dr.

Schmidt also received his PhD from PSU).

In addition to the excellent chemistry education I received at Shepherd, I also had the pleasure of being taught by an additional three excellent professors of chemistry. Dr.

Robert Warburton got me started in Biochemistry and gave me great advice about graduate school that to this day has served me well. He also taught me how to run an assay on a purified enzyme; a skill that I utilize today. The remaining two professors in the Shepherd Chemistry department are chiefly responsible for my early indoctrination with research. Drs. Dan Dilella and Eugene Volker instructed me both in the classroom as well as the lab. Dr. Dilella and Dr. Volker directed my undergraduate research and both were never too busy to take time and talk science or results. Both men professed the benefits of independent research and enthusiastically pushed me to apply to conferences.

Indeed, when I had publishable results, Dr. Dilella and Dr. Volker took time out of their schedules to escort me to two meetings in consecutive years at the University of

Maryland, Baltimore County for undergraduate research symposia. I will never forget this dedication to my research education. In addition to supporting my research activities, both were also stanch supports of my applications for research fellowships,

xxxi which greatly aided in my education financially. I was sad to leave Shepherd mostly because I thought I would never again find an environment where all the professors in the department care so deeply about the successes of their students and were always excited to teach and discuss science.

It turns out the scientific environment I encountered at Shepherd is rare, but not unique. Luck have it, the “Bioinorganic Chemistry department” at Penn State felt a lot like home. This was the main attraction to this area of research. Four professors essentially comprise this cluster of intellect, passion, and ambition; Professor Mike

Green, Professor Carsten Krebs, Professor Marty Bollinger, and, most important to me,

Professor Squire Booker. The professors have a unique chemistry with each other that allows science to move fluidly from lab to lab, enhancing all involved. I will start with

Mike, who I would like to personally thank for all fun I had at his house. The man truly knows how to host a party! Thank you also for allowing me to constantly infiltrate your lab and chit-chat with your students as well as yourself. It was something that enriched my experience at PSU and often brightened my day.

My first real experience with the comradery among these professors was when I joined Dr. Booker’s lab. Back then our lab and Carsten’s lab were adjoining in the

South Frear laboratory, which meant I often crept out of my space and into Carsten’s. I was quick to note that Carsten possessed a lot of knowledge and enthusiasm that was contagious. He was always extremely excited to hear about results or discuss a future experiment. Our lab at that time was also bursting at the seams with people and glovebox space was a premium. In light of this situation, Carsten agreed to let me assemble and use his Coy glovebox, which at the time I did not know how much that

xxxii gesture would impact my future. Every anaerobic protein purification and assay I have ever done originated in that very Coy glovebox. So thank you Carsten for giving me some space and encouragement to start my research. I still remember how pleasant it was, though brief, to have my own glovebox!

My first interaction with Marty and Dr. Booker was a Bioorganic Mechanisms class I took my first semester in grad school. This class was team-taught by both professors, and I loved their enthusiasm and attitude about the subject. It was this interaction that ultimately led me to join Dr. Booker’s lab. At that time, Marty and Dr.

Booker would hold joint group meetings and the few I attended were unlike any I had witnessed. The students and professors discussed and commented on results in a casual repartee. I really enjoyed these interactions and knew this was the place I needed to start my graduate career; it just fit my personality. Marty I want to thank you for introducing me to mechanistic enzymology and also sharing your enthusiasm for science with both members of your lab as well as me. I also want to make note that you are partly responsible for the girth of this thesis. You may or may not remember a comment you made during my oral examination, but it stuck with me at the time. You stated that what proposed to do in my orals was going to be more than I could accomplish. I remember thinking, I will do all of that plus another project worth of data. Thank you for that subtle and effective push. Thank you also for your endless capacity for excitement about science, especially the science done by your students (and surrogate students). Every time I hear you give a talk, I just want to go back to the lab, pick up some pipettors and do great science. Truly inspiring!

xxxiii

Last, but not least of the Fantastic Four is my professor Squire J. Booker. When I joined his lab I was one of four new graduate students. This was a tall order for him, as his only other two graduate students were defending, which meant he would be solely responsible for educating us in the ways of the Booker lab. This was a tall order indeed, particularly in my case as I had come to graduate school with the intent of doing organic chemistry. Knowing this he took me in his lab and was patient as I learned the skills needed to succeed. He also has a unique ability to say what is needed to motivate people, which allows him to cultivate his student’s talents. He always knew when to intervene with experiments and when not. I am a hard headed person and I like fail and succeed on my own. Dr. Booker was always there to cheer at my success and commiserate in my failures. His stories of his own tribulations in grad school at MIT always gave me perspective and a smile on my face. I have met a lot of professors over my years in grad school, but I don’t think I have ever encountered a professor more dedicated to his students and his work than Dr. Booker. He truly takes pride in his student’s success and wants nothing more than to produce fundamentally sound scientists. I owe all of my success to his backing and education. When you graduate from his lab, there are three things you know how to do; 1) construct a coherent and well thought-out talk, 2) take pride in the appearance, style, and layout of a submitted manuscript, and 3) design and execute beautiful experiments. I owe everything to your patience and diligence with my education.

A PhD in chemistry is not obtained without a lot of help of many peers. I would like to first thank Megan Matthews for all of her support and friendship during my graduate career. One of the ways I have always motivated myself to become better is by

xxxiv pitting myself against the peer that I perceive the strongest. You may not have been aware of this, but Megan, you were this person for me when I arrived at PSU. I almost immediately recognized that you were smart and driven, and I realized that if I wanted to succeed, I would need to do as well as you. Your successes in grad school fueled my ambition. Your friendship also meant a lot to me during this time. I always loved discussing results with you as well as personal stories. You are one of the most open and inviting people I have ever met. In addition I would like to thank the three musketeers from Mike Green’s group; Courtney Krest, Tim Yosca, and John Riddle. When we moved next door in the chemistry building, I quickly recognized that you guys knew how to work very hard as well as enjoy some good lab antics that would be called “safety hazards” now. Every day I looked forward to crossing through the portal into the loud and rambunctious environment that was your lab. Thank you all.

As I mentioned above, when I joined Dr. Booker’s lab two graduate students were preparing to defend. Drs. Dave Iwig and Rob Cicchillo where in their last year and really didn’t have time to babysit new graduate students, but they did. They also taught us what it meant to be a member of the Booker Lab. You had to work hard and smart. I started out on the CFA synthase project, which meant Dave was my mentor by default.

Dave is the best analytical chemist I have ever met and his ability to run assays is unparalleled. He gave me the best advice I ever received; I asked “what happens if you mess up?” he answered “Don’t mess up!” This advice has lived with me since and makes me think hard and long about how I should construct and perform an experiment to avoid the dreaded “mess up.” Rob at the time was the most successful graduate student I had come across. I looked up to Rob and wanted to succeed as much as he had.

xxxv

This motivated me to work as much as he did and, as it turns out, I did well. Thanks Rob and Dave for setting the bar for what a Booker Lab member was. In addition I want to acknowledge Liz Bilgrin, Amy Griffiths, and Allison Saunders. They all joined the lab the same time as I did. It was fun and exciting all being new graduate students in the lab.

We were all learning and moving forward together. I want to especially thank Allison.

Allison is the type of person who can command respect without saying anything. She just has a presence. I admired that about her, which made her reprimands sink in. This was no easy task, but Allison made me a better scientist by making me more considerate and conscientious. I also appreciated her brilliance as a researcher, especially when she brought critical insight into problems with my projects during group meeting. You made group meeting a lot of fun. Thanks Allison. I also want to recognize Kyung Hoon Lee for his role in this thesis. He was a good friend and an excellent scientist that taught me everything I know about molecular biology, which admittedly is not much.

Lastly, I need to acknowledge the role of my family in this thesis. I come from a large, close knit family, and having their support means a lot to me. I always love going home for the weekend or the holidays and just hanging out with them. It was a way for me to get away from my troubles and relax. It meant the world to me that you all supported me during my graduate career. Mom and Dad, I want to especially thank you so much for all that you have done for me and especially welcoming me home with open arms when I needed to get away from grad school. This dissertation is dedicated to you both.

Chapter 1

Introduction

A seminal bioinformatics paper published in 2001 by Sofia et al identified and outlined a superfamily of metalloenzymes that had previously been shown to catalyze a rich assortment of reactions involved in numerous important biological pathways, such as the biosynthesis of a large number of enzyme cofactors, antibiotics and other natural products, the biosynthesis and repair of DNA, and general bacterial metabolism (1).

Although these reactions were diverse, they all shared the property of being initiated via removal of a target hydrogen atom (H•) from their relevant substrate by a 5’-dA• generated from a reductive cleavage of S-adenosyl-L-methionine (SAM). The authors coined the title “radical SAM” for this superfamily of enzymes to distinguish them from the classical SAM-dependent reactions that proceed via polar (e.g. SN2) mechanisms.

The thrust of that paper was the identification of telltale features within the primary structures of these proteins, most notably a CX3CXΦC motif (where Φ is an aromatic residue), which has facilitated the rapid discovery of radical SAM (RS) proteins by sequence gazing. Early spectroscopic and biochemical studies on canonical members of the RS superfamily showed that each contained a redox active [4Fe–4S]2+/+ cluster in which three of the four irons of the cubane structure are ligated by single cysteinyl residues lying in a CX3CXΦC motif. The fourth iron is chelated to the α-amino and α- carboxylate groups of SAM in a bidentate fashion, which presumably facilitates the

2 electron transfer step and ensuing cleavage reaction (Figure 1-1) (2, 3). However, as detailed below, this sequence, though overwhelmingly common, is not strictly conserved among all RS proteins, suggesting that this superfamily may be more diverse than previously imagined. At the time of the study by Sofia et al, the RS superfamily was predicted to contain over 600 members.

A recent analysis of the protein and DNA sequence database by the Enzyme

Function Imitative indicates that there are at least 48,100 proteins with at least 61 distinct reactions; all containing the CX3CXΦC motif sequence

(http://sfld.rbvi.ucsf.edu/django/superfamily/29/).

1.1 Radical SAM enzymes and Human Health.

Very recent findings portend that a number of exciting RS-dependent transformations that have compelling medical implications are on the horizon, including reactions involving bacterial defense against antibiotics, a predictive risk factor for type 2 diabetes, and host defence against invading viruses (4-9). The most interesting is Viperin

(Virus Inhibitory Protein, Endoplasmic Reticulum-associated, INterferon-inducible), a protein induced upon interferon stimulation, is involved in the antiviral defence against

DNA viruses such as cytomegalovirus, RNA viruses such as hepatitis C and influenza, and retroviruses such as human immunodeficiency virus (4, 10). The protein is highly conserved across high eukaryotic species, sharing significant sequence homology with similar proteins from trout and mouse, as well as a protein from rat, best5, which is expressed during osteoblast differentiation and bone formation (1, 4). Viperin is composed of three distinct domains, a variable N-terminal domain, a radical SAM domain, and a C-terminal domain, the last two of which are highly conserved. Mutations in the Viperin gene that give rise to CysAla substitutions at the protein level resulted in loss of antiviral effects against hepatitis C virus, demonstrating the importance of RS chemistry in antiviral activity (11). At present, the exact mechanism of action of Viperin is unknown, as is its direct target. It has been suggested that Viperin-dependent inhibition of influenza A virus involves perturbing its release from the plasma membrane during its budding cycle by affecting the formation of lipid rafts. This activity is believed to derive from an inhibition of farnesyl diphosphate synthase (FPPS) via an unknown mechanism.

4 Also unknown is the exact pathway downstream of FPPS inhibition that gives rise to viral inhibition (9). Recently it was shown that Viperin localizes to intracellular lipid-storage organelles called lipid droplets via an N-terminal amphipathic -helix, which may mediate its effect against hepatitis C virus; however, again, the exact mechanism of inhibition is unknown (12).

Two reports describing the first in vitro characterization of Viperin have recently appeared (13, 14). In the report by Shaveta et al, expression analysis of 12 fragments of the Viperin gene showed that a Viperin construct lacking the first 44 amino acids (i.e.

45–361) was a predominantly soluble protein that could be purified under native conditions by immobilized metal affinity chromatography. Analysis of the reconstituted protein by UV-visible spectroscopy supported the presence of an Fe/S species (14). In the report by Duschene et al, a Viperin construct spanning residues 43–360 was generated.

The purified protein had low amounts of iron, but was reconstituted to contain 3.7 irons per polypeptide. Both UV-vis and EPR analysis of the protein supported the presence of

[4Fe–4S] clusters. In addition, the protein was cable of catalyzing reduction of SAM to

5’-dA and methionine (13).

Cfr is another recently characterized RS protein with clear medical implications.

It confers resistance to five classes of antibiotics (phenicols, lincosamides, oxazolidinones, pleuromutilins, and streptogramin A)—all of which bind to the peptidyl transferase center of bacterial ribosomes—as well as the 16-membered macrolides josamycin and spiramycin (15). Its mode of action involves methylation of C8 of adenosine 2503 of 23S ribosomal RNA (rRNA), which sits in the center of the peptidyl transferase site (16, 17). This methylation has a negligible effect on peptidyl transferase

5 activity, but sterically impedes the binding of antibiotics that target the site. A more troubling aspect of the acquisition of the cfr gene does not confer a competitive disadvantage in growth competition, meaning the cfr gene requires little energy to maintain (18). A highly homologous protein, RlmN, targets C2 of the exact nucleotide

(Figure 1-2).

NH3 NH3 NH3 N N N N N N H3C N N N N CH N N CH O O 3 O 3 O RlmN O Cfr O 2 SAM 2 SAM O OH O OH O OH

Figure 1-2. Binding mode of SAM in RS pr.Binding mode of SAM in RS proteins The reactions catalyzed by RlmN and Cfr RlmN is endogenous to a wide number of bacteria and other organisms, believe to function in the fine-tuning of translation as well as interact with the nascent peptide in the exit tunnel of the ribosome (19, 20). By contrast, the cfr gene is acquired, and appears to be an evolutionary spin-off of the rlmN gene, arising from gene duplication and horizontal transfer (21). Recently a structural model of Cfr was generated using the

MoaA structure as a template for its central RS domain. The model included the expected

[4Fe–4S] cluster ligated by the RS motif as well as two molecules of bound SAM: one as the precursor to the 5’-dA• and one as the donor of the methyl group (21). The RNA substrate was not modeled into the structure. A more important aspect of the work was a study of amino acid substitutions at proposed sites for binding the [4Fe–4S] cluster and the two SAM molecules, as well as in the N-terminal and C-terminal domains of the protein. Among other observations, five Cys residues appeared critical for Cfr activity.

These included residues 112, 116, and 119 found in the RS signature sequence; Cys338,

6 found at the C-terminus of the protein; and Cys105, found in the binding pocket for the second SAM molecule. Interestingly, although the Cys105Ala substitution did not support methylation, it appears that some type of reaction took place that caused a stop in primer extension assays similar to the observed effect of C8 methylation (21).

A subsequent study by Yan et al expressed the genes for E. coli RlmN and S. aureus Cfr in E. coli, and isolated the corresponding hexahistidine-tagged proteins (22).

They found that the reconstituted protein contained 3.98 and 6.79 irons, respectively, which— along with mutagenesis of the RS cluster ligand—they concluded was support for formation of at least one [4Fe–4S] cluster. The salient feature of the study was in vitro activity determinations conducted with a series of potential substrates, including the whole ribosome, the naked 23S rRNA, and various truncations of Domain V of 23S rRNA. It was found that neither the intact 70S ribosome, nor the isolated 50S or 30S subunits served as substrates for either Cfr or RlmN. Only protein-free rRNA containing an adenosine at position 2503 was capable of being methylated, which suggests that these proteins catalyze their reactions before the ribosome is assembled. Further studies provided evidence for the formation of both 5’-dA and SAH. More importantly, when the reaction was conducted in the presence of S-adenosyl-L-[methyl-3H]methionine, radioactivity was found to be transferred to the rRNA substrate, indicating that SAM is the source of the appended methyl group (22).

Recently, a 2.05 Å resolution X-ray crystal structure of RlmN was solved in both absence and presence of SAM (23). The structure in the absence of SAM contain all the hallmarks of a RS enzyme, including the partial TIM barrel as well as the [4Fe–4S] cluster. This structure also showed the C118 (C105 in Cfr) was located within the active

7 site. The second cysteine, C355, was not visible in this structure and was proposed to be part of a flexible loop region. Indeed, the structure of RlmN+SAM not only has all of the features of the SAM absent structure but also shows C355 dipping into the active site of

RlmN (Figure 1-3). This structure shows that C355 resides in a highly flexible region within the C-terminus of the protein. Based on the proposed role of C355 being a mediator for the transferred methyl, it was proposed the this flexible can move in and out of the active site to accommodate substrate binding and product release (24).

8 1.2 Radical SAM enzymes with multiple iron–sulfur clusters.

The discovery that biotin synthase from E. coli contains two distinct Fe/S clusters per polypeptide, a [4Fe–4S] cluster and a [2Fe–2S] cluster, ushered in a new chapter in

RS enzymology, which highlighted the versatility of these enzymes as catalysts (25, 26).

With the exception of MoaA, all early RS members containing multiple Fe/S clusters catalyzed the insertion of sulfur deriving from the second cluster into unactivated C–H bonds: the [2Fe–2S] cluster on biotin synthase, and the [4Fe–4S] clusters on lipoyl synthase and MiaB (27-30). In addition to sulfur insertion into the hypermodified tRNA nucleoside N6-(isopentenyl)adenosine-37, MiaB transfers the methyl group from another molecule of SAM onto the inserted sulfur atom. This reaction takes place on the hypermodified tRNA nucleoside N6-(isopentenyl)adenosine-37, and involves a net methylthiolation at C2 of the adenine ring. Therefore, it appears that, as in the case of

RlmN and Cfr, a single polypeptide catalyzes both radical and polar SAM-dependent reactions (31).

Recently, Anton et al showed that the yliG gene in E. coli, the product of which was designated RimO, catalyzes a similar methylthiolation reaction on a universally conserved Asp residue (Asp88 in E. coli) of the S12 subunit of certain bacterial ribosomes (32). RimO proteins from E. coli (33) and Thermotoga maritima (34) were subsequently purified by two different groups, and in both instances shown to bind two

[4Fe–4S] clusters per polypeptide and to catalyze methylthiolation of a peptide substrate containing an Asp residue in the appropriate sequence context. RimO, like MiaB, contains six conserved Cys residues, all of which reside in the N-terminal region of the

9 protein, suggesting that this second cluster is also ligated by only three Cys residues. In analogy with the previously mentioned RS enzymes that catalyze sulfur insertion, it is believed that the second cluster provides an activated form of sulfide to be inserted into the substrate.

Figure 1-4. Reactions catalyzed by MoaA/MoaC. Numbers highlight changes in positioning of atoms during the rearrangement.

The RS enzyme MoaA has been characterized structurally and shown to bind an additional [4Fe–4S] cluster via three Cys residues located in the C-terminal region of its primary structure (35). Unlike the enzymes discussed above, its net reaction does not involve sulfur insertion, but is a cryptic rearrangement of GTP to yield precursor Z, an intermediate in the biosynthesis of the cofactor molybdopterin (Figure 1-4). The MoaA reaction was shown to be dependent on the second cluster and to require the accessory protein MoaC, which participates in some undefined role (36). The X-ray crystal structure of MoaA containing both clusters and in complex with both SAM and GTP provided valuable insight into the architecture of the active site (Figure 1-5) (37).

The C-terminal cluster appeared to interact with either the N1 or N3 nitrogen atoms of GTP; however, the poorly defined electron density of the substrate did not allow an exact determination of its binding mode. Recently, electron nuclear double resonance

(ENDOR) spectroscopy was used to show that the mode of binding involved coordination of the N1 nitrogen atom to the unique iron atom of the cluster at a distance of 1.94 Å. The authors suggested that this interaction should favor guanine binding to the unique iron atom as the enol rather than keto tautomer, which they stated may have mechanistic implications (38). Currently, little is known about the detailed mechanism of catalysis by

MoaA/MoaC. This C-terminal Fe/S cluster has been shown to be redox-active, and it has been speculated that it could play a role in electron transfer (37).

Another RS enzyme purported to be in the MoaA family is PqqE, which is one of six proteins required for the biogenesis of pyrroloquinoline quinone (PQQ). Unlike the other quino-cofactors, which are generated via posttranslational modifications of the core

11 catalytic proteins, PQQ is synthesized as a small-molecule cofactor that subsequently associates with the relevant catalyst via noncovalent interactions (39). The biogenesis of

PQQ is quite complex, involving the crosslinking of the side chains of glutamyl and tyrosyl residues from a core peptide composed of 23 amino acids, which serves as the skeleton of the cofactor (40). It has been suggested that PqqE might catalyze this crosslinking, generally believed to be the first step in the pathway. Similar to MoaA,

PqqE displays two highly conserved cysteine-containing motifs at the N-and C-termini of the protein, CX3CX2C and CX2CX27C, respectively. A recent study by Wecksler et al provided spectroscopic and analytical evidence for the presence of two Fe/S clusters on the enzyme from Klebsiella pneumoniae, and showed that the protein can catalyze cleavage of SAM to yield 5’-dA and methionine. However, no evidence for in vitro formation of PQQ was forthcoming despite a determined effort to provide it (41). As described for MoaA, no distinct mechanistic role for this second cluster has been assigned.

In addition to MoaA and PqqE, an emerging subclass of RS enzymes, that contains multiple Fe/S clusters that catalyze a seemingly simple two-electron oxidation of an alcohol or thiol group to the corresponding aldehyde or ketone. This class of enzymes has been termed RS dehydrogenases because the reaction they catalyze is formally a hydride transfer, but actually proceeds by single electron steps.

Three of these enzymes, spanning two distinct reaction types, have been characterized in vitro. The first, BtrN, catalyzes a key step in the biosynthesis of the aminoglycoside antibiotic butirosin B, which is the oxidation of the C3 alcohol of 2- deoxy-scyllo-inosamine (DOIA) to amino-2-deoxy-scyllo-inosose (amino-DOI) (42). The

12 second two, anSMEcpe and AtsB, are anaerobic sulfatase modifying enzymes from

Clostridium perfringens and Klebsiella pneumoniae, which catalyze the oxidation of a

Cys or Ser residue on a cognate protein to generate a formylglycyl cofactor (Figure 1-6)

(43, 44). The reactions proceed via abstraction of a hydrogen atom from the carbon to be oxidized by the 5’-dA•, followed by the uptake of an electron by an undetermined acceptor (42, 44, 45). Detailed analytical and spectroscopic analysis of AtsB showed that it contained three [4Fe–4S] clusters per polypeptide. It was postulated that one of the clusters binds in contact with the substrate to facilitate loss of an electron from the substrate-radical intermediate via an inner-sphere mechanism (44). Although the stoichiometry of Fe/S clusters on anSMEcpe has not been determined, the protein shares

48% sequence similarity with AtsB, including eleven conserved Cys residues.

Recently, a bioinformatics study by Haft and Basu groups all anaerobic formylglycine generation enzymes (anSMEs)—including anSMEcpe and AtsB— into a family a family of enzymes termed SPASM. The SPASM acronym derived from the most characterized enzymes included in this family; namely Subtilisin A (AlbA),

Pyrroloquinoline quinone (PqqE), Anaerobic Sulfatase maturating enzymes (AtsB, anSMEcpe), and Mycofactocin maturation (46, 47). The primary feature of the ~ 1400

Figure 1-6. Reactions catalyzed by Radical SAM dehydrogenases. A) BtrN oxidizes 2-DOIA to 2-DOI. B) anSMEcpe and AtsB (C) both oxidize an active site cys or ser residue within their cognate sulfatase, respectively. anSMEcpe and AtsB both form the formylglycine residue.

of this subfamily of RS enzymes is the presence of a 7-cysteine motif—CX9-15GX4C– gap–CX2CX5CX3C–gap–C—found in the C-terminus of the amino acid sequence of these proteins. Haft and Basu proposed that this domain bound two [4Fe–4S] clusters, with one of the clusters having an open coordination site. Members of this family were predicted to modify small, ribosomally produced peptides that upon additional modification become bacteriocins (46, 47). As only seven cysteine are found in this motif, one of the two clusters was proposed to have an open coordination site that could function to anchor substrate in the active site or directly bind the modified amino acid allowing facile electron transfer. Only four members of this family, to date, have been characterized and already a divide in reactivity is found. AtsB and anSMEcpe perform a dehydrogenation, while AlbA and SkfB both in insert a sulfur from a cysteine residue residing in the peptide substrate into the backbone carbon of an adjacent amino acid (48,

14 49). The reactions catalyzed by AlbA and SkfB are, in essence, a dehydrogenation reaction, but formally are sulfur insertions, which is a very different reaction. It will be interesting to see in the future what structural features the members possess.

Importantly, a recent X-ray crystal structure of anSMEcpe—in the presence of both SAM and peptide substrate—has allowed the first details into the exact nature of the

SPASM domain (ref PETER). Like all RS enzymes structurally characterized to date, anSMEcpe contains an N-terminal TIM barrel composed of parallel β6/α6 secondary

15 structural elements. (Figure 1-7) which ligate the RS cluster. The remaining cluster are housed entirely in the SPASM domain of the protein, with auxiliary cluster I (Aux I) at the interface between the RS domain and the SPASM making up half of the active site.

Auxiliary cluster II resides about 13 Å away from Aux I, which is predicted to be the exit point for the electron generated during turnover. In addition to providing detail about the auxiliary cluster distribution within the polypeptide, the structure also revealed that two basic residues lie with 4 Å of the substrate’s cysteinyl proton. Deprotonation of this residue is a key element of the reaction, as the reduction potentials of radical anion are severely depressed compared in relation to the radical—which would facilitate the reduction of the auxiliary I cluster (50). Tyrosine 24 and Aspartate 277 were selected for mutagenesis, which revealed the most likely base was the Asp277. Variants of Y24F had a small effect on activity, whereas the D277N variant was essentially inactive (PETER ref).

Interestingly, when the structure of MoaA is superimposed onto anSMEcpe, the second cluster in MoaA and the Aux I cluster of anSMEcpe align well. A major difference in the cluster from MoaA and Aux I is Aux I is full coordinate meaning substrate cannot access this cluster. As state earlier, MoaA uses the coordination site to ligate substrate. Since the substrate was not found to directly coordinate—as was predicted—in anSMEcpe, this leaves the role for the auxiliary clusters in question. It will be interesting to see if AlbA and SkfB fall within this structural motif or whether one of the auxiliary clusters will have an open coordination site.

BtrN was originally characterized to contain only one [4Fe–4S] cluster, suggesting that the presence of multiple Fe/S clusters is not a prerequisite for RS

16 dehydrogenation (51). Recently Mössbauer spectroscopy was used in concert with analytical determinations of iron content to re-evaluate the stoichiometry of Fe/S clusters rigorously, showed that indeed the protein contains two [4Fe–4S] clusters (52). In addition, it was proposed that the second cluster of BtrN fills a similar role as to anSMEs.

In support of the proposal, a recent structure of BtrN bound to both SAM and 2-

DOIA shows that—like MoaA and anSMEcpe—the auxiliary cluster is in close proximity to the active site (Figure 1-8) and resides in a partial SPASM domain termed a

“twitch” domain. An additional interesting aspect of the BtrN structure was the lack of a

17 traditional base in the active site. Unlike anSMEcpe, the only base in proximity to the 3’- hydroxyl group of 2-DOIA was Arginine152 (R152). Though uncommon, arginine’s can fill the role of a base in the active site of protein when ion pairs with a glutamate

(53). Indeed, in hydrogen bonding contact is E189. A glutamine variant of this residue should diminish the ability of R152 to act as a base.

DesII is a radical SAM enzyme responsible for the deamination of TDP-4- amino-4,6-dideoxy-D-glucose to give 3-TDP-keto-4,6-dideoxy-D-glucose (54). 3-TDP- keto-4,6-dideoxy-D-glucose is an intermediate in the pathway to D-desosamine, which is incorporated, by some Streptomyces species, into macrolide antibiotics (55, 56). An intriguing aspect of this redox neutral deamination reaction is DesII expends one equivalent of SAM for every product made (57). This result suggests that an electron must be donated to the radical intermediate for catalysis to be complete, which is an aspect of the chemistry that currently eludes characterization (58). A suggestion was made that the radical is quenched by the reduced Fe/S cluster, which in this case is the RS cluster, as DesII only contains three cysteines in the amino acid sequence—all required to bind the RS cluster. If this mechanism holds true, this will be the first example in which the RS cluster is involved two, chemically different one electron reductions (58). In addition, if DesII is reacted with a substrate analogue—TDP-D-quinovose which is lacking the amino group—the reaction catalyzed is instead a dehydrogenation similar to

BtrN. The interesting aspect of this non-physiological reaction stems from lack of external electron acceptors (58). Once the enzyme is primer for reaction with dithionite, with subsequent removal of excess dithionite, the reaction catalyzes multiple turnovers without the requirement of addition electrons. This indicates that the radical intermediate

18 generated during turnover is reducing the RS cluster to release product. How this interesting reaction fits in with the radical SAM dependent dehydrogenases remains to be seen.

19 1.3 Maturation of complex metallocofactors.

RS enzymes are involved in the maturation of at least three classes of complex metallocofactors, the iron-molybdenum cofactor (FeMo-co) of nitrogenase, the H-cluster of the [FeFe]-hydrogenase, and the mononuclear cluster of the [Fe]-hydrogenase

Significant gains have been made recently in understanding the biosynthesis of the H-cluster of the [FeFe]-hydrogenase, one of the enzymes responsible for the reversible reduction of protons to H2 (59). This cluster consists of a 2Fe subcluster coordinated by cyanide and carbon monoxide ligands, as well as a dithiolate moiety (–

SCH2–X–CH2S–), which is then bridged to a [4Fe–4S] cluster via a protein cysteinate ligand (Figure 1-9) (60, 61). The exact identity of X in the dithiolate moiety is still controversial, but most evidence suggests it is nitrogen (61, 62).

Figure 1-9. Maturation of the H-cluster of the [FeFe]-hydrogenase. Structure on left represents HydA, the hydrogenase from Desulfovibrio desulfuricans, with a [4Fe–4S] cluster bound. In the presence of HydE, HydF, HydG, and appropriate small molecules, the H-cluster is formed on HydA. Color scheme: Red, iron; yellow, sulfur; grey, carbon; blue, nitrogen; black, unidentified atom (X). Structure prepared using Pymol Molecular Graphics System (http://www.pymol.org) from PDB 1HFE. Genetic and biochemical studies have shown that three accessory proteins are required to synthesize and insert the H-cluster into the hydrogenase protein (HydA) (59):

20 HydE, HydG, and HydF (Figure 1-9). HydE and HydG are RS enzymes, while HydF contains GTPase activity (63-66). The X-ray crystal structure of HydE was recently solved to 1.35 Å, the highest resolution structure of any RS enzyme. The structure revealed a [2Fe–2S] cluster separated from the RS [4Fe–4S] cluster by ~20 Å in a spatial arrangement similar to that of the two [4Fe–4S] clusters in MoaA, which is involved in molybdopterin biosynthesis (36). It is not clear whether this [2Fe–2S] cluster—which may be a degradation product of a second [4Fe–4S] cluster observed spectroscopically in another study (64)—is actually required for maturation, because substitution of its coordinating Cys residues with those containing noncoordinating R-groups did not eliminate hydrogenase activity in an in vivo assay. Moreover, the ligands to the second cluster are not conserved among all HydE proteins. Although the substrate for HydE is unknown, Nicolet et al provided evidence that the protein can bind thiocyanate, which led them to speculate that it might be involved in generating the cyanide ligands to the H- cluster (67).

The protein HydG bears 27% sequence identity to the E. coli enzyme

ThiH, a RS protein that catalyzes a key step in the formation of the thiazole ring of the cofactor thiamine diphosphate (TDP). The 5’-dA• produced by ThiH is proposed to abstract the phenolic hydrogen atom from L-tyrosine, initiating a fragmentation reaction that liberates p-cresol and dehydroglycine. Dehydroglycine is then condensed with ThiFS thiocarboxylate and 1-deoxyxylulose 5-phosphate to give thiazole-phosphate in a reaction catalyzed by ThiG (68). The sequence similarity between HydG and ThiH inspired investigation by Pilet et al to ascertain whether the substrate for HydG was also L- tyrosine (69). HydG did in fact catalyze liberation of p-cresol from L-tyrosine, leading

21 the authors to postulate that HydG is the site for the synthesis of a dithiomethylamine ligand (–SCH2–NH–CH2S–)—derived from the presumed dehydroglycine product—onto a [2Fe–2S] cluster scaffold(69).

A different group investigating the role of HydG in the maturation of the H- cluster of hydrogenase also found that HydG catalyzes the cleavage of L-tryrosine. Not only was p-cresol found as a product, there was clear evidence for the formation of cyanide in almost equivalent amounts. The authors proposed that the cyanide produced could derive from a facile oxidative decarboxylation of dehydroglycine, but more interestingly, suggested that both cyanide and carbon monoxide could be produced in a single reaction via a decarbonylation of dehydroglycine, which they stated has chemical precedent (70). Therefore, it appears that the role of HydG is to use RS chemistry to catalyze formation of the cyanide ligands of the 2Fe subcluster, and as well as the carbon monoxide ligands.

The study by Sofia et al suggested that the nifB gene product, involved in an unknown step in the biosynthesis of FeMo-co, was a RS protein (1). A subsequent report by Curatti, Ludden, and Rubio showed that purified and reconstituted NifB was able to support in vitro reconstitution of FeMo-co in the presence of SAM (71). Recently, Wiig et al have generated a NifB fusion protein – wherein NifB was fused, through peptide sequence, to NifN to form the full NifEN-B – that allowed the stable purification of a

NifEN-B complex (72). This stable construct was shown to be fully active in FeMo-co maturation. The complex has been characterized in detail, including both structurally and spectroscopic ally (72-74). In addition to the detail structural work on, Wiig et al have finally shown the in vivo activity of this RS enzyme. Through the use of radio and

22 isotopic labeling, they were able to unequivocally show the long sought after source of the interstitial atom with FeMo-co is derived from the methyl group of SAM (75). They showed, that much like MiaB, RimO, RlmN, and Cfr, NifEN-B was able to activiate

SAM for two distinct chemistries; radical SAM chemistry and polar, methyl transfer chemistry.

23 1.4 Radical SAM enzymes lacking the canonical CxxxCxxC motif.

ThiC, an enzyme involved in TDP biosynthesis in prokaryotes, was not identified as an RS member by Sofia et al (76). The penultimate step in the de novo TDP biosynthetic pathway involves a condensation of the thiazole and pyrimidine moieties of the cofactor, each synthesized in two independent branches of the pathway, to furnish thiamine monophosphate, which is subsequently phosphorylated to the active cofactor

(76). In contrast to ThiH, which participates along with other proteins in the formation of the thiazole moiety, ThiC alone catalyzes formation of the pyrimidine moiety (76). The reaction is among the most complex in all of mechanistic enzymology, which is the conversion of 5-aminoimidazole ribonucleotide (AIR) to 4-amino-5-hydroxymethyl-2- methylpyrimidine phosphate (HMP). Figure 1-10 highlights the results of labelling studies, illustrating the complex nature of the reaction (76).

Figure 1-10. The reaction catalyzed by ThiC. Color-coding depicts the change in positioning of certain atoms during the rearrangement as determined by labelling experiments.

Recently, the enzymes from Arabidopsis thaliana (77), Salmonella enterica (78), and Caulobacter crescentus (79), have been characterized to be iron–sulfur (Fe/S)

24 proteins, the latter two of which were shown to catalyze in vitro formation of HMP in the presence of substrate, SAM, and dithionite (78, 79).

The recent X-ray crystal structure of apo-ThiC from C. crescentus with bound

HMP-P identified three structural domains: an N-terminal domain, a central domain, and a disordered C-terminal domain. The latter bears a conserved CxxCxxxxC motif, the Cys residues of which could ligate an Fe/S cluster. However, note that the sequence differs from the canonical RS CxxxCxxC motif, and its position in the protein at the C-terminus instead of near the N-terminus is also distinct (79). Typical RS enzymes contain the

CxxxCxxC motif in the N-terminal half of their primary structures (1). The structure revealed the protein to be dimeric, and the [4Fe–4S] cluster, shown to be present on the reconstituted enzyme by Mössbauer and electron paramagnetic resonance (EPR) spectroscopy, was modelled into the protein using the structure of biotin synthase as a template. SAM was modelled into the active site pocket to coordinate the unique iron of the [4Fe–4S] cluster in a bidentate fashion, in common with other RS enzymes, which places the 5’-carbon in a suitable position to abstract a hydrogen atom from the ribose ring of the substrate by a generated 5’-dA•.

Evidence for a mechanism involving organic radicals was provided in an EPR study. When SAM was added to the dithionite-reduced enzyme, a new signal centered at g = 2.002 emerged, which had linewidth and temperature-dependent properties that were consistent with an organic radical. A sample prepared in D2O allowed determination that the radical was centered on the α-carbon of an amino acid residue other than glycine or alanine. Exposure of the protein bearing the organic radical to oxygen led to rapid destruction of the EPR signal and cleavage of the polypeptide chain between Gly436 and

25 His437 (80). Whether this organic radical is an intermediate in this reaction remains to be resolved.

Elp3 from Methanocaldococcus jannaschii and HmdB from Methanococcus maripaludis S2 are two partially characterized enzymes similarly found to lack the canonical CxxxCxxC motif. Elp3 is a component of the Elongator complex, required for transcription elongation. Elongator is composed of six subunits, Elp1–Elp6. Elp3 is thought to be the catalytic subunit, given that it is one of the subunits that form the core of the complex, and that it displays histone acetyl transferase (HAT) activity (81). In addition to its C-terminal HAT domain, Elp3 has a domain potentially related to RS enzymes despite a CX4CX9CX2C motif deviating from the canonical CxxxCxxC RS motif. It was speculated that the RS domain might catalyze demethylation of methylated lysyl residues on histones (82). The RS domain of Elp3 from M. jannaschii (residues 63–

371) was subsequently purified and shown to bind SAM and small amounts of iron (83).

The Cys motif in this archaeal Elp3 (CxxxxCxxC) is different from both the canonical motif and that found in eukaryotic proteins.

Small amounts of Elp3 from Saccharomyces cerevisiae were recently isolated, allowing the involvement of possible Fe/S clusters in catalysis to be investigated.

Substitution of individual Cys residues by Ala residues in the proposed RS domain of

Elp3 resulted in phenotypes that were indistinguishable from those observed upon deletion of the entire ELP3 gene, suggesting that the proposed cluster is important for normal Elongator function. Further studies showed that the CysAla substitutions affected assembly of the Elongator complex, but had little effect on HAT activity or the ability of the complex to bind to RNA polymerase II in chromatin. In addition, no histone

26 demethylase activity was detected, and no evidence for the ability to bind SAM was found. The authors concluded that the Fe/S cluster, if present, serves a structural rather than catalytic role (84).

A more recent study has demonstrated at least partial involvement of the RS domain of mammalian Elp3 in active demethylation of 5-methyl cytosines of the paternal

DNA strand at the zygotic stage of fertilization and development (85). This event is believed to be vital in the reprogramming of germ cells to allow their transition to somatic cells. To show this, Okada et al developed molecular probes to allow determination of the methylation state of DNA in zygotes via time-lapse imaging, which they used in conjunction with RNA interference to allow cellular levels of candidate demethylases to be knocked down. Single interfering RNA (siRNA) molecules targeting

Elp1, Elp3, and Elp4 all affected the zygotic paternal methylation status. Interestingly, introduction of mRNA encoding substitutions of the Cys residues within the proposed RS domain of Elp3 affected the paternal methylation status, whereas substitutions in the

HAT domain of Elp3 did not (85). They suggested that demethylation might be mediated through a reaction that requires an intact RS domain.

The hmdB gene from Methanococcus maripaludis S2 was recently found to be adjacent on the chromosome to the hmdA gene. HmdA, found in hydrogenotrophic methanogens, catalyzes the reversible reduction of methenyl-tetrahydromethanopterin

(H4MPT+) to methylene-H4MPT and H+, and contains an octahedrally coordinated nonheme iron atom bearing two CO ligands, a protein cysteinyl ligand, an unknown ligand, and a guanylyl pyridinol cofactor ligand (86). The primary structure of HmdB contains a CxxxxxCxxC motif and is phylogentically related to ThiH, HydE, and HydG.

27 The purified protein was shown by UV–vis and EPR spectroscopy to contain a [4Fe–4S] cluster. In addition, it was capable of catalyzing cleavage of SAM to 5’-dA in the presence of dithionite, suggesting its inclusion in the RS superfamily (86). It was suggested that HmdB might participate in the synthesis of the iron-carbonyl linkage in the Hmd cofactor.

28 1.5 Advances in understanding the reductive cleavage of S-adenosylmethionine.

The fundamental chemical transformation common to all radical SAM enzymes is the reductive cleavage of SAM to generate the 5’-dA•, which in solution is thermodynamically highly unfavorable (Figure 1-11). Midpoint potentials for the irreversible one-electron reduction of a trialkylsulfonium ion have been measured to be ~

–1.8 V, while those for radical SAM proteins tend to be much higher (87). A study by

Wang and Frey investigated the energetics of SAM cleavage by lysine 2,3-aminomutase

(LAM), which uses RS chemistry to catalyze an interconversion of α and β-lysine when bound in an aldimine linkage to a required pyridoxal 5’-phosphate (PLP) cofactor (87).

They found that in the resting state of the enzyme—with SAM and PLP bound—the

[4Fe–4S] clusters exhibited a midpoint potential of –430 mV, and that the binding of lysine lowered the midpoint potential by ~150 mV. Similarly, the midpoint potential for the reductive cleavage of SAM in the enzyme/SAM/lysine complex was estimated to be –

990 mV from values obtained using the analog S-3’,4’-anhydroadenosyl-L-methionine.

Therefore, the enzyme active site environment raises the redox potential of SAM by ~810 mV while lowering the redox potential of the Fe/S cluster upon substrate binding, which corresponds to a decrease in the overall barrier for the reductive cleavage of SAM from

32 kcal/mol in solution to 9 kcal/mol. Additional energy for the process is believed to derive from ligation of the sulfur atom of the generated methionine to the unique iron of the cluster, which generates a hexacoordinate species and facilitates inner-sphere electron transfer (Figure 1-11) (87, 88).

Figure 1-11. Model for the reductive cleavage of SAM to generate a 5’-deoxyadenosyl radical.

A recent study by Nicolet et al provided additional support for the mechanism of reductive cleavage of SAM proposed by Frey and coworkers, and argued that the mechanism should be common to all RS enzymes containing the canonical CxxxCxxC motif (89). This conclusion stems from the X-ray structures of HydE with SAM bound and with both 5’-dA and methionine bound at 1.62 Å and 1.25 Å, respectively, and then using these structures in concert with computational methods to calculate the most likely reaction trajectory. Interestingly, their calculated barrier for SAM cleavage of 54.0 kJ/mol (12.9 kcal/mol) agrees well with the experimental estimate made by Wang and

Frey (9 kcal/mol). Moreover, they remarked that in all RS structures solved in complex with SAM, SAM was bound in essentially the same fashion in each case (89).

30 1.6 Future directions.

In the near future much of the focus on RS enzymes will undoubtedly involve characterizing new and novel enzymatic reactions. As metagenomics continues to add new and uncharacterized sequences to the databases, the number of RS enzymes will continue to grow beyond 40,000.

Many new discoveries will emanate from studies to identify gene clusters for the biosynthesis of a variety of natural products such as clorobiocin (90), moenomycin A

(91), pactamycin (92), gentamicin (93), nosiheptide (94), unusual lipids (95), and deazapurine-containing secondary metabolites (96), which are just a few of the more recent ones to be discovered. Indeed RS enzymes involved in the biosynthesis of the antibiotic butirosin B and the antibiotic precursor D-desosamine have already been well characterized and have added chemically challenging questions that remain to be addressed (42, 51, 54).

Of particular interest is the rapidly expanding class of RS methyltransferases, which are distinct from Cfr and RlmN. This class was highlighted in the study by Sofia et al, and its participant enzymes are annotated as being in the P-methylase family. This name derives from one of the founding members of this subclass of RS enzymes, which catalyzes the methylation of a phosphinate phosphorus atom in the biosynthesis of the herbicide bialaphos (97). Interestingly, these enzymes are annotated as cobalamin binding proteins, and a number of genetic and in vivo biochemical studies support that assignment

(27). A hypothetical mechanism for these RS methyltransferase reactions was advanced by van der Donk, in which he proposed that the added methyl group is transferred from

31 methylcobalamin to the substrate radical generated via hydrogen atom abstraction by the

5’-dA• in a radical process (98). In a recent study, TsrM—a member of the P-methylase family involved in the biosynthetic pathway of the antibiotic thiostrepton A—was purified to homogeneity and characterized with UV-visible spectroscopy (99). The authors also showed that the enzyme does indeed use methylcobalmin as a substrate to methylate tryptophan. They also found the TsrM did not make 5’dA during turnover as well as production of cob(II)alamin. The authors speculate that TsrM uses the RS cluster, not for reduction of SAM to 5’-dA•, but instead coordinate tryptophan for methyl radical addition methyl-cob(III)alamin. This would result in a methyl-tryptophan radical that would then be oxidized by the RS cluster. This mechanism is provocative, but does not account for other members of the methyl- cob(III)alamin dependent RS enzymes that modify sp3 carbons. The methyl radical addition to an sp2 center has been demonstrated several times, but this reaction is not possible with sp3 carbon centers. It is unlikely that this highly homologues subfamily of enzymes developed two distinct mechanism for the same reaction.

Two additional areas of future interest are the elucidation of the mechanisms for re-installing the sacrificed Fe/S clusters in RS enzymes that catalyze sulfur insertion, and the development of more robust bioinformatics methods for identifying possible substrats for RS proteins that do not share significant primary structure with known RS enzymes.

Many of the ~ 48,000 protein sequences identified as RS enzymes have no know function and little similarity to other RS proteins. To discover new chemistry, the substrates for an enzyme need to be identified. The emergence of the RS superfamily of enzymes has brought renewed vigor to mechanistic enzymology. Many of the known transformations

32 are simply astounding, and the future bodes well for discovering new ones that will remind us of the wonders of nature.

33 1.7 References

1. Sofia, H. J., Chen, G., Hetzler, B. G., Reyes-Spindola, J. F., and Miller, N. E.

(2001) Radical SAM, a novel protein superfamily linking unresolved steps in

familiar biosynthetic pathways with radical mechanisms: functional

characterization using new analysis and information visualization methods,

Nucleic Acids Res. 29, 1097-1106.

2. Frey, P. A., Hegeman, A. D., and Ruzicka, F. J. (2008) The radical SAM

superfamily, Crit. Rev. Biochem. Mol. Biol. 43, 63–88.

3. Walsby, C. J., Ortillo, D., Yang, J., Nnyepi, M. R., Broderick, W. E., Hoffman, B.

M., and Broderick, J. B. (2005) Spectroscopic approaches to elucidating novel

iron-sulfur chemistry in the "radical-SAM" protein superfamily, Inorg. Chem. 44,

727-741.

4. Chin, K.-C., and Cresswell, P. (2001) Viperin (cig5), an IFN-inducible antiviral

protein directly induced by human cytomegalovirus, Proc. Natl. Acad. Sci. USA

98, 15125-15130.

5. Dehwah, M. A., Wang, M., and Huang, Q. Y. (2010) CDKAL1 and type 2

diabetes: a global meta-analysis, Genet. Mol. Res. 9, 1109-1120.

6. Saxena, R., Voight, B. F., Lyssenko, V., Burtt, N. P., de Bakker, P. I., Chen, H.,

Roix, J. J., Kathiresan, S., Hirschhorn, J. N., Daly, M. J., Hughes, T. E., Groop,

L., Altshuler, D., Almgren, P., Florez, J. C., Meyer, J., Ardlie, K., Bengtsson

Boström, K., Isomaa, B., Lettre, G., Lindblad, U., Lyon, H. N., Melander, O.,

Newton-Cheh, C., Nilsson, P., Orho-Melander, M., Råstam, L., Speliotes, E. K.,

34 Taskinen, M. R., Tuomi, T., Guiducci, C., Berglund, A., Carlson, J., Gianniny, L.,

Hackett, R., Hall, L., Holmkvist, J., Laurila, E., Sjögren, M., Sterner, M., Surti,

A., Svensson, M., Tewhey, R., Blumenstiel, B., Parkin, M., Defelice, M., Barry,

R., Brodeur, W., Camarata, J., Chia, N., Fava, M., Gibbons, J., Handsaker, B.,

Healy, C., Nguyen, K., Gates, C., Sougnez, C., Gage, D., Nizzari, M., Gabriel, S.

B., Chirn, G. W., Ma, Q., Parikh, H., Richardson, D., Ricke, D., and Purcell, S.

(2007) Genome-wide association analysis identifies loci for type 2 diabetes and

triglyceride levels, Science 316, 1331-1336.

7. Scott, L. J., Mohlke, K. L., Bonnycastle, L. L., Willer, C. J., Li, Y., Duren, W. L.,

Erdos, M. R., Stringham, H. M., Chines, P. S., Jackson, A. U., Prokunina-Olsson,

L., Ding, C. J., Swift, A. J., Narisu, N., Hu, T., Pruim, R., Xiao, R., Li, X. Y.,

Conneely, K. N., Riebow, N. L., Sprau, A. G., Tong, M., White, P. P., Hetrick, K.

N., Barnhart, M. W., Bark, C. W., Goldstein, J. L., Watkins, L., Xiang, F.,

Saramies, J., Buchanan, T. A., Watanabe, R. M., Valle, T. T., Kinnunen, L.,

Abecasis, G. R., Pugh, E. W., Doheny, K. F., Bergman, R. N., Tuomilehto, J.,

Collins, F. S., and Boehnke, M. (2007) A genome-wide association study of type

2 diabetes in Finns detects multiple susceptibility variants, Science 316, 1341-

1345.

8. Steinthorsdottir, V., Thorleifsson, G., Reynisdottir, I., Benediktsson, R.,

Jonsdottir, T., Walters, G. B., Styrkarsdottir, U., Gretarsdottir, S., Emilsson, V.,

Ghosh, S., Baker, A., Snorradottir, S., Bjarnason, H., Ng, M. C., Hansen, T.,

Bagger, Y., Wilensky, R. L., Reilly, M. P., Adeyemo, A., Chen, Y., Zhou, J.,

Gudnason, V., Chen, G., Huang, H., Lashley, K., Doumatey, A., So, W. Y., Ma,

35 R. C., Andersen, G., Borch-Johnsen, K., Jorgensen, T., van Vliet-Ostaptchouk, J.

V., Hofker, M. H., Wijmenga, C., Christiansen, C., Rader, D. J., Rotimi, C.,

Gurney, M., Chan, J. C., Pedersen, O., Sigurdsson, G., Gulcher, J. R.,

Thorsteinsdottir, U., Kong, A., and Stefansson, K. (2007) A variant in CDKAL1

influences insulin response and risk of type 2 diabetes, Nat. Genet. 39, 770-775.

9. Wang, X., Hinson, E. R., and Cresswell, P. (2007) The interferon-inducible

protein viperin inhibits influenza virus release by perturbing lipid rafts, Cell Host

Microbe 2, 96–105.

10. Hinson, E. R., and Cresswell, P. (2009) The N-terminal amphipathic α-helix of

viperin mediates localization to the cytosolic face of the endoplasmic reticulum

and inhibits protein secretion, Proc. Natl. Acad. Sci. USA 284, 4705–4712.

11. Jiang, D., Guo, H., Xu, C., Chang, J., Gu, B., Wang, L., Block, T. M., and Guo,

J.-T. (2008) Identification of three interferon-inducible cellular enzymes that

inhibit the replication of hepatitis C virus, J. Virol. 82, 1665–1678.

12. Hinson, E. R., and Cresswell, P. (2009) The antiviral protein, viperin, localizes to

lipid droplets via its N-terminal amphipathic α-helix, Proc. Natl. Acad. Sci. U S A

106, 20452–20457.

13. Duschene, K. S., and Broderick, J. B. (2010) The antiviral protein viperin is a

radical SAM enzyme, FEBS Lett.

14. Shaveta, G., Shi, J., Chow, V. T. K., and Song, J. (2009) Structural

characterization reveals that viperin is a radical S-adenosyl-L-methionine (SAM)

enzyme, Biochem. Biophys. Res. Commun. 391, 1390–1395.

36 15. Toh, S.-M., Xiong, L., Arias, C. A., Villegas, M. V., Lolans, K., Quinn, J., and

Mankin, A. S. (2007) Acquisition of a natural resistance gene renders a clinical

strain of methicillin-resistant Staphylococcus aureus resistant to the synthetic

antibiotic linezolid, Mol. Microbiol. 64, 1506–1514.

16. Giessing, A. M. B., Jensen, S. S., Rasmussen, A., Hansen, L. H., Gondela, A.,

Long, K. S., Vester, B., and Kirpekar, F. (20090) Identification of 8-

methyladenosine as the modification catalyzed by the radical SAM

methyltransferase Cfr that confers antibiotic resistance in bacteria, RNA 15, 327–

336.

17. Giessing, A. M. B., Jensen, S. S., Rasmussen, A., Hansen, L. H., Gondela, A.,

Long, K. S., Vester, B., and Kirpekar, F. (2009) Identification of 8-

methyladenosine as the modification catalyzed by the radical SAM

methyltransferase Cfr that confers antibiotic resistance in bacteria, RNA 15, 327–

336.

18. LaMarre, J. M., Locke, J. B., Shaw, K. J., and Mankin, A. S. (2011) Low fitness

cost of the multidrug resistance gene cfr, Antimicrob. Agents Chemother. 55,

3714–3719.

19. Toh, S.-M., Xiong, L., Bae, T., and Mankin, A. S. (2008) The methyltransferase

YfgB/RlmN is responsible for modification of adenosine 2503 in 23S rRNA, RNA

14, 98–106.

20. Vazquez-Laslop, N., Ramu, H., Klepacki, D., and Mankin, A. S. (2010) The key

role of a conserved and modified rRNA residue in the ribosomal response to the

nascent peptide, EMBO J. 29, 3108–3117.

37 21. Kaminska, K. H., Purta, E., Hansen, L. H., Bujnicki, J. M., Vester, B., and Long,

K. S. (2010) Insights into the structure, function and evolution of the radical-SAM

23S rRNA methyltransferase Cfr that confers antibiotic resistance in bacteria,

Nuc. Acids. Res.

22. Yan, F., LaMarre, J. M., Röhrich, R., Wiesner, J., Jomaa, H., Mankin, A. S., and

Galoníc Fujimori, D. (2010) RlmN and Cfr are radical SAM enzymes involved in

methylation of ribosomal RNA, J. Am. Chem. Soc. 132, 3953-3964.

23. Boal, A. K., Grove, T. L., McLaughlin, M. I., Yennawar, N. H., Booker, S. J., and

Rosenzweig, A. C. (2011) Structural basis for methyl transfer by a radical SAM

enzyme, Science 332, 1089–1092.

24. Grove, T. L., Benner, J. S., Radle, M. I., Ahlum, J. H., Landgraf, B. J., Krebs, C.,

and Booker, S. J. (2011) A radically different mechanism for S-

adenosylmethionine-dependent methyltansferases, Science 332, 604–607.

25. Ugulava, N. B., Gibney, B. R., and Jarrett, J. T. (2001) Biotin synthase contains

two distinct iron–sulfur binding sites: chemical and spectroelectrochemical

analysis of iron–sulfur cluster interconversions, Biochemistry 40, 8343-8351.

26. Ugulava, N. B., Surerus, K. K., and Jarrett, J. T. (2002) Evidence from Mössbauer

sectroscopy for distinct [2Fe-2S]2+ and [4Fe-4S]2+ cluster binding sites in biotin

synthase from Escherichia coli, J. Am. Chem. Soc. 124, 9050-9051.

27. Booker, S. J. (2009) Anaerobic functionalization of unactivated C–H bonds, Curr.

Opin. Chem. Biol. 13, 58–73.

28. Hernández, H. L., Pierrel, F., Elleingand, E., García-Serres, R., Huynh, B. H.,

Johnson, M. K., Fontecave, M., and Atta, M. (2007) MiaB, a bifunctional radical-

38 S-adenosylmethionine enzyme involved in the thiolation and methylation of

tRNA, contains two essential [4Fe–4S] clusters, Biochemistry 46, 5140-5147.

29. Jarrett, J. T. (2005) The novel structure and chemistry of iron-sulfur clusters in the

adenosylmethionine-dependent radical enzyme biotin synthase, Arch. Biochem.

Biophys. 433, 312-321.

30. Cicchillo, R. M., Lee, K.-H., Baleanu-Gogonea, C., Nesbitt, N. M., Krebs, C., and

Booker, S. J. (2004) Escherichia coli lipoyl synthase binds two distinct [4Fe–4S]

clusters per polypeptide, Biochemistry 43, 11770-11781.

31. Pierrel, F., Douki, T., Fontecave, M., and Atta, M. (2004) MiaB protein is a

bifunctional radical-S-adenosylmethionine enzyme involved in thiolation and

methylation of tRNA, J. Biol. Chem. 279, 47555-47653.

32. Anton, B. P., Saleh, L., Benner, J. S., Raleigh, E. A., Kasif, S., and Roberts, R. J.

(2008) RimO, a MiaB-like enzyme, methylthiolates the universally conserved

Asp88 residue of ribosomal protein S12 in Escherichia coli, Proc. Natl. Acad. Sci.

USA 105, 1826–1831.

33. Lee, K.-H., Saleh, L., Anton, B. P., Madinger, C. L., Benner, J. S., Iwig, D. F.,

Roberts, R. J., Krebs, C., and Booker, S. J. (2009) Characterization of RimO, a

new member of the methylthiotransferase subclass of the radical SAM

superfamily, Biochemistry 48, 10162–10174.

34. Arragain, S., García-Serres, R., Blondin, G., Douki, T., Clemancey, M., Latour,

J.-M., Forouhar, F., Neely, H., Montelione, G. T., Hunt, J. F., Mulliez, E.,

Fontecave, M., and Atta, M. (2010) Post-translational modification of ribosomal

proteins: Structural and functional characterization of RimO from Thermotoga

39 maritima, a radical S-adenosylmethionine methylthiotransferase, J. Biol. Chem.

285, 5792–5801.

35. Hänzelmann, P., Hernandez, H. L., Menzel, C., Garcia-Serres, R., Huynh, B. H.,

Johnson, M. K., Mendel, R. R., and Schindelin, H. (2004) Characterization of

MOCS1A, an oxygen-sensitive iron-sulfur protein involved in human

molybdenum cofactor biosynthesis, J. Biol. Chem. 279, 34721-34732.

36. Hänzelmann, P., and Schindelin, H. (2004) Crystal structure of the S-

adenosylmethionine-dependent enzyme MoaA and its implications for

molybdenum cofactor deficiency in humans, Proc. Natl. Acad. Sci. USA 101,

12870-12875.

37. Hänzelmann, P., and Schindelin, H. (2006) Binding of 5'-GTP to the C-terminal

FeS cluster of the radical S-adenosylmethionine enzyme MoaA provides insights

into its mechanism, Proc. Natl. Acad. Sci. USA 103, 6829–6834.

38. Lees, N. S., Hänzelmann, P., Hernandez, H. L., Subramanian, S., Schindelin, H.,

Johnson, M. K., and Hoffman, B. M. (2009) ENDOR spectroscopy shows that

guanine N1 binds to [4Fe–4S] cluster II of the S-adenosylmethionine-dependent

enzyme MoaA: Mechanistic implications, J. Am. Chem. Soc. 131, 9184–9185.

39. Goodwin, P. M., and Anthony, C. (1998) The Biochemistry, physiology and

genetics of PQQ and PQQ-containing enzymes, In Advances in Microbial

Physiology (Poole, R. K., Ed.), pp 1–80, Elsevier, Amsterdam.

40. Houck, D. R., Hanners, J. L., and Unkefer, C. J. (1991) Biosynthesis of

pyrroloquinoline quinone. 2. Biosynthetic assembly from glutamate and tyrosine,

J. Am. Chem. Soc. 113, 3162–3166.

40 41. Wecksler, S. R., Stoll, S., Tran, H., Magnusson, O. T., Wu, S.-P., King, D., Britt,

R. D., and Klinman, J. P. (2009) Pyrroloquinoline quinone biogenesis:

Demonstration that PqqE from Klebsiella pneumoniae is a radical S-adenosyl-L-

methionine enzyme, Biochemistry 48, 10151–10161.

42. Yokoyama, K., Numakura, M., Kudo, F., Ohmori, D., and Eguchi, T. (2007)

Characterization and mechanistic study of a radical SAM dehydrogenase in the

biosynthesis of butirosin, J. Am. Chem. Soc. 129, 15147-15155.

43. Benjdia, A., Subramanian, S., Leprince, J., Vaudry, H., Johnson, M. K., and

Berteau, O. (2008) Anaerobic sulfatase-maturating enzymes – first dual substrate

radical S-adenosylmethionine enzymes, J. Biol. Chem. April 11, 2008, 17815-

17826.

44. Grove, T. L., Lee, K. H., St Clair, J., Krebs, C., and Booker, S. J. (2008) In vitro

characterization of AtsB, a radical SAM formylglycine-generating enzyme that

contains three [4Fe-4S] clusters, Biochemistry 47, 7523-7538.

45. Benjdia, A., Leprince, J., Sandström, C., Vaudry, H., and Berteau, O. (2009)

Mechanistic investigations of anaerobic sulfatase-maturating enzyme: Direct Cβ

H-atom abstraction catalyzed by a radical AdoMet enzyme, J. Am. Chem. Soc.

131, 8348–8349.

46. Haft, D. H. (2011) Bioinformatic evidence for a widely distributed, ribosomally

produced electron carrier precursor, its maturation proteins, and its

nicotinoprotein redox partners, BMC Genomics 12, 21.

41 47. Haft, D. H., and Basu, M. K. (2011) Biological systems discovery in silico:

radical S-adenosylmethionine protein families and their target peptides for

posttranslational modification, J Bacteriol 193, 2745-2755.

48. Flühe, L., Knappe, T. A., Gattner, M. J., Schäfer, A., Burghaus, O., Linne, U., and

Marahiel, M. A. (2012) Thioether bond formation during subtilosin A maturation

is catalyzed by the two [4Fe-4S]-cluster containing radical SAM enzyme AlbA,

Nat. Chem. Biol.

49. Flühe L, Burghaus O, Wieckowski BM, Giessen TW, Linne U, and MA., M. Two

[4Fe-4S] clusters containing radical SAM enzyme SkfB catalyze thioether bond

formation during the maturation of the sporulation killing factor., J Am Chem Soc.

. 135, 959-962.

50. Hayon, E., and Simic, M. (1974) Acid-Base Properties of Free-Radicals in

Solution, Accounts of Chemical Research 7, 114-121.

51. Yokoyama, K., Ohmori, D., Kudo, F., and Eguchi, T. (2008) Mechanistic study

on the reaction of a radical SAM dehydrogenase BtrN by electron paramagnetic

resonance spectroscopy, Biochemistry 47, 8950-8960.

52. Grove, T. L., Ahlum, J. H., Sharma, P., Krebs, C., and Booker, S. J. (2010) A

consensus mechanism for radical SAM-dependent dehydrogenation? BtrN

contains two [4Fe–4S] clusters, Biochemistry 49, 3783–3785.

53. Schlippe, Y. V. G., and Hedstrom, L. (2005) A twisted base? The role of arginine

in enzyme-catalyzed proton abstractions, Archives of Biochemistry and

Biophysics 433, 266-278.

42 54. Szu, P.-H., Ruszczycky, M. W. C. S.-H., Yan, F., and Liu, H.-W. (2009)

Characterization and mechanistic studies of DesII: A radical S-adenosyl-L-

methionine enzyme involved in the biosynthesis of TDP-D-desosamine, J. Am.

Chem. Soc. 131, 14030–14042.

55. Kren, V., and Martinkova, L. (2001) Glycosides in medicine: "The role of

glycosidic residue in biological activity", Current Medicinal Chemistry 8, 1303-

1328.

56. Thorson, J. S., Hosted, T. J., Jiang, J. Q., Biggins, J. B., and Ahlert, J. (2001)

Nature's carbohydrate chemists: The enzymatic glycosylation of bioactive

bacterial metabolites, Current Organic Chemistry 5, 139-167.

57. Ruszczycky, M. W., Choi, S. H., and Liu, H. W. (2010) Stoichiometry of the

Redox Neutral Deamination and Oxidative Dehydrogenation Reactions Catalyzed

by the Radical SAM Enzyme Desll, Journal of the American Chemical Society

132, 2359-2369.

58. Ruszczycky, M. W., Choi, S. H., Mansoorabadi, S. O., and Liu, H. W. (2011)

Mechanistic Studies of the Radical S-Adenosyl-L-methionine Enzyme DesII:

EPR Characterization of a Radical Intermediate Generated During Its Catalyzed

Dehydrogenation of TDP-D-Quinovose, Journal of the American Chemical

Society 133, 7292-7295.

59. Böck, A., King, P. W., Blokesch, M., and Posewitz, M. C. (2006) Maturation of

hydrogenases, Adv. Microb. Physiol. 51, 1-71.

43 60. Peters, J. W., Lanzilotta, W. N., Lemon, B. J., and Seefeldt, L. C. (1998) X-ray

crystal structure of the Fe-only hydrogenase (CpI) from Clostridum pasteurianum

to 1.8 angstrom resolution, Science 282, 1853-1858.

61. Nicolet, Y., Piras, C., Legrand, P., Hatchikian, C. E., and Fontecilla-Camps, J. C.

(1999) Desulfovibrio desulfuricans iron hydrogenase: the structure shows unusual

coordination to an active site Fe binuclear center, Structure 7, 13-23.

62. Silakov, A., Wenk, B., Reijerse, E., and Lubitz, W. (2009) (14)N HYSCORE

investigation of the H-cluster of [FeFe] hydrogenase: evidence for a nitrogen in

the dithiol bridge, Phys Chem Chem Phys 11, 6592-6599.

63. King, P. W., Posewitz, M. C., Ghirardi, M. L., and Seibert, M. (2006) Functional

studies of [FeFe] hydrogenase maturation in an Escherichia coli biosynthetic

system, J. Bacteriol. 188, 2163–2172.

64. Rubach, J. K., Brazzolotto, X., Gaillard, J., and Fontecave, M. (2005)

Biochemical characterization of the HydE and HydG iron-only hydrogenase

maturation enzymes from Thermatoga maritima, FEBS Lett. 579, 5055-5060.

65. Posewitz, M. C., King, P. W., Smolinski, S. L., Zhang, L., Seibert, M., and

Ghirardi, M. L. (2004) Discovery of two novel radical S-adenosylmethionine

proteins required for the assembly of an active [Fe] hydrogenase, J. Biol. Chem.

279, 25711–25720.

66. Brazzolotto, X., Rubach, J. K., Gaillard, J., Bambarellis, S., Atta, M., and

Fontecave, M. (2006) The [Fe-Fe]-hydrogenase maturation protein HydF from

Thermotoga maritima is a GTPase with an iron–sulfur cluster, J. Biol. Chem. 281,

769–774.

44 67. Nicolet, Y., Rubach, J. K., Posewitz, M. C., Amara, P., Mathevon, C., Atta, M.,

Fontecave, M., and Fontecilla-Camps, J. C. (2008) X-ray structure of the [FeFe]-

hydrogenase maturase HydE from Thermotoga maritima, J. Biol. Chem. 283,

18661–18872.

68. Kriek, M., Martins, F., Challand, M. R., Croft, A., and Roach, P. L. (2007)

Thiamine biosynthesis in Escherichia coli: Identification of the intermediate and

by-product dervied from tyrosine, Angew. Chem. Int. Ed. 46, 9223–9226.

69. Pilet, E., Nicolet, Y., Mathevon, C., Douki, T., Fontecilla-Camps, J. C., and

Fontecave, M. (2009) The role of the maturase HydG in [FeFe]-hydrogenase

active site synthesis and assembly, FEBS Lett 583, 506–511.

70. Driesener, R. C., Challand, M. R., McGlynn, S. E., Shepard, E. M., Boyd, E. S.,

Broderick, J. B., Peters, J. W., and Roach, P. L. (2010) [FeFe]-Hydrogenase

cyanide ligands derived from S-adenosylmethionine-dependent cleavage of

tyrosine, Angew. Chem. Int. Ed. 49, 1687–1690.

71. Curatti, L., Ludden, P. W., and Rubio, L. M. (2006) NifB-dependent in vitro

synthesis of the iron-molybdenum cofactor of nitrogenase, Proc. Natl. Acad. Sci.

USA 103, 5297–5301.

72. Wiig, J. A., Hu, Y. L., and Ribbe, M. W. (2011) NifEN-B complex of

Azotobacter vinelandii is fully functional in nitrogenase FeMo cofactor assembly,

Proceedings of the National Academy of Sciences of the United States of America

108, 8623-8627.

45 73. Kaiser, J. T., Hu, Y. L., Wiig, J. A., Rees, D. C., and Ribbe, M. W. (2011)

Structure of Precursor-Bound NifEN: A Nitrogenase FeMo Cofactor

Maturase/Insertase, Science 331, 91-94.

74. Rupnik, K., Hu, Y. L., Fay, A. W., Ribbe, M. W., and Hales, B. J. (2011)

Variable-temperature, variable-field magnetic circular dichroism spectroscopic

study of NifEN-bound precursor and "FeMoco", Journal of Biological Inorganic

Chemistry 16, 325-332.

75. Wiig, J. A., Hu, Y. L., Lee, C. C., and Ribbe, M. W. (2012) Radical SAM-

Dependent Carbon Insertion into the Nitrogenase M-Cluster, Science 337, 1672-

1675.

76. Jurgenson, C. T., Begley, T. P., and Ealick, S. E. (2009) The structural and

biochemical foundations of thiamin biosynthesis, Ann. Rev. Biochem. 78, 569–

603.

77. Raschke, M., Bürkle, L., Müller, N., Nunes-Nesi, A., Fernie, A. R., Arigoni, D.,

Amrhein, N., and Fitzpatrick, T. B. (2007) Vitamin B1 biosynthesis in plants

requires the essential iron–sulfur cluster protein, THIC, Proc. Natl. Acad. Sci.

USA 104, 19637–19642.

78. Martinez–Gomez, N. C., and Downs, D. M. (2008) ThiC is an [Fe–S] cluster

protein that requires AdoMet to generate the 4-amino-5-hydroxymethyl-2-

methylpyrimidine moiety in thiamin synthesis, Biochemistry 47, 9054–9056.

79. Chatterjee, A., Li, Y., Zhang, Y., Grove, T. L., Lee, M., Krebs, C., Booker, S. J.,

Begley, T. P., and Ealick, S. E. (2008) Reconstitution of ThiC in thiamine

46 pyrimidine biosynthesis expands the radical SAM superfamily, Nat. Chem. Biol.

4, 758–765.

80. Martinez–Gomez, N. C., Poyner, R. R., Mansoorabadi, S. O., Reed, G. H., and

Downs, D. M. (2009) Reaction of AdoMet with ThiC generates a backbone free

radical, Biochemistry 48, 217–219.

81. Svejstrup, J. Q. (2007) Elongator complex: how many roles does it play?, Curr.

Opin. Cell. Biol. 19, 331–336.

82. Chinenov, Y. (2002) A second catalytic domain the Elp3 histone

acetyltransferases: a candidate for histone demethylase activity?, Trends Biochem.

Sci. 27, 115–117.

83. Paraskevopoulou, C., Fairhurst, S. A., Lowe, D. J., Brick, P., and Onesti, S.

(2006) The Elongator subunit Elp3 contains a Fe4S4 cluster and binds S-

adenosylmethionine, Mol. Microbiol. 59, 795–806.

84. Greenwood, G., Selth, L. A., Dirac-Svejstrup, A. B., and Svejstrup, J. Q. (2009)

An iron–sulfur cluster domain in Elp3 important for the structural integrity of

elongator, J. Biol. Chem. 284, 141–149.

85. Okada, Y., Yamagata, K., Hong, K., Wakayama, T., and Zhang, Y. (2010) A role

for the elongator complex in zygotic paternal genome demethylation, Nature 463,

554–558.

86. McGlynn, S. E., Boyd, E. S., Shepard, E. M., Lange, R. K., Gerlach, R.,

Broderick, J. B., and Peters, J. W. (2010) Identification and characterization of a

novel member of the radical AdoMet enzyme superfamily and implications for

47 the biosynthesis of the Hmd hydrogenase active site cofactor, J. Bacteriol. 192,

595–598.

87. Wang, S. C., and Frey, P. A. (2007) Binding energy in the one-electron reductive

cleavage of S-adenosylmethionine in lysine 2,3-aminomutase, a radical SAM

enzyme, Biochemistry 46, 12889–12895.

88. Cosper, N. J., Booker, S. J., Ruzicka, F., Frey, P. A., and Scott, R. A. (2000)

Direct FeS cluster involvement in generation of a radical in lysine 2,3-

aminomutase, Biochemistry 39, 15668-15673.

89. Nicolet, Y., Amara, P., Mouesca, J.-M., and Fontecilla-Camps, J. C. (2009)

Unexpected electron transfer mechanism upon AdoMet cleavage in radical SAM

proteins, Proc. Natl. Acad. Sci. USA 106.

90. Anderle, C., Alt, S., Gulder, T., Bringmann, G., Kammerer, B., Gust, B., and

Heide, L. (2007) Biosynthesis of clorobiocin: investigation of the transfer and

methylation of the pyrrolyl-2-carboxyl moiety, Arch. Microbiol. 187, 227–237.

91. Ostash, B., Saghatelian, A., and Walker, S. (2007) A streamlined metabolic

pathway for the biosynthesis of moenomycin A, Chem. Biol. 14, 257–267.

92. Kudo, F., Kasama, Y., Hirayama, T., and Eguchi, T. (2007) Cloning of the

pactamycin biosynthetic gene cluster and characterization of a crucial

glycosyltransferase prior to a unique cyclopentane ring formation, J. Antibiot. 60,

492–503.

93. Kim, J.-Y., Suh, J.-W., Kang, S.-H., Phan, T. H., Park, S.-H., and Kwon, H.-J.

(2008) Gene inactivation study of gntE reveals its role in the first step of

48 pseudotrisaccharide modifications in gentamicin biosynthesis, Biochem. Biophys.

Res. Commun. 372, 730–734.

94. Yu, Y., Duan, L., Zhang, Q., Liao, R., Ding, Y., Pan, H., Wendt-Pienkowski, E.,

Tang, G., Shen, B., and Liu, W. (2009) Nosiheptide biosynthesis featuring a

unique indole side ring formation on the characteristic thiopeptide framework,

ACS Chem. Biol. 4, 855–864.

95. Rattray, J. E., Strous, M., Op den Camp, H. J., Schouten, S., Jetten, M. S., and

Damsté, J. S. (2009) A comparative genomics study of genetic products

potentially encoding ladderane lipid biosynthesis, Biol. Direct.

96. McCarty, R. M., Somogyi, Á., Lin, G., Jacobsen, N. E., and Bandarian, V. (48)

The deazapurine biosynthetic pathway revealed: In vitro enzymatic synthesis of

PreQ0 from guanosine 5'-triphosphate in four steps, Biochemistry 48, 3847–3852.

97. Seto, H., and Kuzuyama, T. (1999) Bioactive natural products with carbon-

phosphorus bonds and their biosynthesis, Nat. Prod. Rep. 16, 589–596.

98. van der Donk, W. A. (2006) Rings, radicals, and regeneration: the early years of a

bioorganic laboratory, J. Org. Chem. 71, 9561–9571.

99. Pierre, S., Guillot, A., Benjdia, A., Sandstrom, C., Langella, P., and Berteau, O.

(2012) Thiostrepton tryptophan methyltransferase expands the chemistry of

radical SAM enzymes, Nature Chemical Biology 8, 957-959.

Part I

Characterization of Radical SAM-Dependent Methyl Synthases.

50 Chapter 2 A Radically Different Mechanism for S-adenosylmethionine-dependent Methyltransferases Exhibited by the Antibiotic Resistance Protein Cfr and its Homologue RlmN

This Chapter was reproduced from “Grove, T.L., Benne,r J.S., Radle, M.I., Ahlum, J.H., Landgraf, B.J., Krebs, C., Booker, S.J. A radically different mechanism for S- adenosylmethionine-dependent methyltransferases. Science. 2011 Apr 29;332(6029):604-607.”

52 2.1 Abstract

The most widely recognized biological role of S-adenosyl-L-methionine (SAM) is as methyl group donor. SAM-dependent methyltransferases operate on small molecules, proteins, lipids and nucleic acids and play crucial roles in metabolism, cell signaling, and epigenetic programming. Most commonly, they employ a simple SN2 mechanism to methylate nucleophilic sites on their substrates. Recently, a class of enzymes that use

SAM in a distinct role—as the precursor to the potently oxidizing 5'-deoxyadenosyl 5’- radical (5'-dA•)—has emerged. In these radical SAM enzymes, 5'-dA• abstracts hydrogen atoms from substrates to initiate over forty distinct reaction types, including, ironically, the methylation of carbon atoms that are not inherently nucleophilic. Herein, we dissect the mechanisms of two such unusual SAM-dependent methylation reactions targeting the sp2-hybridized C2 and C8 positions of adenosine 2503 (A2503) in 23S rRNA. C2 methylation, catalyzed by the enzyme RlmN, is widespread in eubacteria and is believed to serve a "housekeeping" function, presumably in maintenance of translation fidelity, whereas C8 methylation, catalyzed by Cfr, is the basis for acquired resistance to at least five classes of antibiotics targeting the 50S ribosomal subunit of bacteria. Both reactions proceed by an unprecedented ping-pong-like mechanism: (1) the methyl group from one SAM molecule is initially appended to a strictly conserved cysteinyl residue on the enzyme, presumably by a standard SN2 displacement; (2) the SAM-derived C1 unit is subsequently attached to the RNA substrate in a novel radical-addition step initiated by the 5'-dA• formed from a second molecule of SAM; and (3) the covalent intermediate

53 from the radical-addition step is resolved, leading to formation of a disulfide between the methyl-carrying Cys residue and a second conserved cysteine.

54 2.2 Introduction.

A number of different posttranscriptional and posttranslational modifications adorn both protein and RNA components of the ribosome. Although the raisons d’être for most of these modifications are currently unknown, many are believed to confer stability and translational fidelity, and direct the proper assembly of this complex machine (1-3).

Among the most common modifications is a methyl group derived from S-adenosyl-L- methionine (SAM) that is appended to specific atoms residing in amino acid side chains or nucleotide bases or sugars. The overwhelming majority of these groups are installed via SN2 displacement mechanisms, involving attack of a nucleophile on the methyl group of SAM with concomitant release of S-adenosyl-L-homocysteine (SAH) (4, 5).

Intriguingly, two modifications of 23S rRNA of bacterial ribosomes, appended by the enzymes RlmN and Cfr, involve the SAM-dependent methylation of C2 and C8 of adenosine 2503 (A2503), respectively, which are electrophilic rather than nucleophilic sp2-hybridized carbons (Figure 2-1) (6-8). Moreover, the poor acidity of the protons attached to C2 and C8 preclude their removal by typical polar (acid/base) processes that are mediated by the side chains of the 20 common amino acids (9, 10).

The bacterial ribosome is the target of about half of all antibiotics currently in use, of which the majority bind to sites on the large (50S) subunit, disrupting functions associated with GTP hydrolysis, peptide-bond formation, and exit of the nascent polypeptide (1, 11). These latter two functions are inhibited by several classes of antibiotics, which bind to overlapping regions within the 50S subunit. As such, bacterial

55 interventions to resist a given antibiotic that targets this region, such as modification of the surrounding rRNA, often result in resistance to multiple drugs (1, 12).

Figure 2-1. Reactions catalyzed by RlmN and Cfr. RlmN catalyzes uniquely methylation at C2, whereas Cfr catalyzes methylation at C8 and C2, although C8 is the preferred target.

A2503 of 23S rRNA resides within the peptidyltransferase center (PTC) of the 50S subunit near the entrance to the exit tunnel for the nascent peptide (13-16), and has been shown in E. coli to be methylated at C2 (6, 17) (Figure 2-1). This natural modification — catalyzed by RlmN (YfgB) in E. coli — is believed to play a housekeeping function.

Although the yfgB gene is not essential, E. coli yfgB null strains lose to wild-type (WT) strains in co-growth competition experiments (17). Modification of C8 of A2503 has also been observed, and confers resistance to several classes of antibiotics, including phenicols, lincosamides, oxazolidinones, pleuromutilins, and streptogramin A (8, 18-20).

This activity is encoded by the product of the cfr gene, which was first identified on a plasmid isolated from Staphylococcus sciuri, an animal pathogen (19). More recently, it

56 has also been found on the chromosome of a methicillin resistant strain of S. aureus

(MRSA) obtained from a hospital isolate along with the ermB gene, which encodes a 23S

A2508 dimethylase (21). Co-expression of these genes renders the bacterium resistant to all currently used antibiotics that target the 50S subunit (21). Bioinformatics analysis indicates that RlmN is widespread throughout eubacteria, and that Cfr evolved directly from it as an antibiotic resistance mechanism (22). In fact, Cfr, which shares 33% sequence identity with E. coli RlmN, also catalyzes methylation at C2 of A2503, although it is not the preferred target (7).

RlmN and Cfr belong to a special class of SAM-dependent enzymes designated radical

SAM (RS), which reductively cleave the molecule to render a 5’-deoxyadenosyl 5’- radical (5’-dA•) rather than mediate alkylation reactions by polar mechanisms (23). The

5’-dA• intermediate is common to reactions of all RS enzymes studied to date, and is invariably used to abstract a hydrogen atom from the substrate. The radical peregrinations that follow formation of the various substrate radicals give rise to the bewildering array of reactions and products associated with the family. Bioinformatics analyses indicate that there are currently more than 40 distinct reaction types catalyzed by over 2400 of these enzymes, including sulfur insertion, methylthiolation, complex carbon-skeleton rearrangement, isomerization, functional group oxidation, generation of glycyl-radical cofactors, and methylation of unactivated carbon atoms (24-26). All of these enzymes contain at least one [4Fe–4S] cluster — typically coordinated by the Cys residues of a

CxxxCxxC motif — which supplies the electron for the reductive cleavage of SAM. In E. coli the iron–sulfur (Fe/S) cluster is reduced by an electron derived from the flavodoxin/flavodoxin reductase (Flv/Flx) reducing system, whereas in vitro it can also

57 be supplied by low-potential artificial reductants such as dithionite or illuminated deazaflavin (27).

Important in vitro studies by Yan et al provided valuable insight into the reaction mechanisms of RlmN and Cfr, showing that SAM acts as both a radical generator and as the source of the appended methyl group (28). Moreover, they showed that both enzymes act preferentially on naked 23S rRNA, of which helices 89 and 90-92 in domain V are critical for substrate recognition and turnover (28). Herein, we provide evidence that these enzymes employ an unprecedented chemical mechanism for the addition of a methyl group to an unactivated carbon atom, involving covalent catalysis and unique roles for two strictly conserved Cys residues.

2.3 Materials and Methods

Cloning of the E. coli rlmN gene. The gene encoding RlmN was amplified from

Escherichia coli K12 genomic DNA using the following primers: RlmN_For (5’- gcccggcatatgtctgaacaattagtcacacctgaaaacg-3’) and RlmN_Rev (5’- ccccgaattcccgaccgctttaatgtcgatggcttcaccc-3’). RlmN_For contained a NdeI restriction site

(underlined), while RlmN_Rev contained a EcoRI restriction site (underlined). Primer

RlmN_Rev removes the native stop codon in the yfgB gene, allowing the encoded protein to be produced with a C-terminal hexahistidine tag. The resulting PCR product was digested with NdeI and EcoRI and cloned into similarly digested pET-26b by standard methods. The final construct was verified by DNA sequencing at the Molecular Core

Facility (Penn State University) and designated pRlmN-Wt.

58 Construction of CysAla Variants of RlmN. Single CysAla substitutions at C118 and C355 were generated using the primers in the following table in conjunction with the

Stratagene QuikChange II kit:

Table 2-1. Primers for RlmN variants C118A and C355A Primer name Sequence

Forward 5’-cgaccgtgccacgctcGCCgtctcttcgcaggtgggg-3’ RlmN C118A Reverse 5’- ccccacctgcgaagagacGGCgagcgtggcacggtcg -3’

Forward 5’-ggtgatgatatcgatgctgccGCTggtcagttggcgggcg - 3’ RlmN C355A Reverse 5’- cgcccgccaactgaccAGCggcagcatcgatatcatcacc -3’

Cloning of Staphylococcus aureus cfr gene. The gene encoding cfr from S. aureus

(A5HBL2) was codon-optimized for expression in E. coli by GeneArt (Burlingame, CA), and supplied in the plasmid pMA. The gene included NdeI (5’ end) and EcoRI (3’ end) restriction sites to allow subcloing into pET28a. The resulting construct encodes a protein containing an N-terminal hexahistidine tag. The codon optimized gene sequence is below:

5’catatgaactttaacaacaaaaccaaatatggcaaaatcaggaatttctgcgtagca ataatgaaccggattatcgcatcaaacaaatcaccaacgccatttttaaacagcgtatc agccgctttgaagatatgaaagttctgccgaaactgctgcgtgaagatctgattaataa ctttggcgaaaccgtgctgaatattaaactgctggccgaacagaatagcgaacaggtta ccaaagttctgtttgaggtgagcaaaaatgaacgtgtggaaaccgtgaacatgaaatat aaagccggttgggaaagcttttgtattagcagccagtgtggttgtaattttggttgcaa attttgcgcaaccggtgatattggtctgaaaaaaaacctgaccgtggatgaaattaccg atcaggtgctgtattttcatctgctgggtcatcagattgatagcatcagctttatgggt atgggtgaagcactggcaaatcgtcaggtttttgatgcactggatagctttaccgatcc gaacctgtttgcactgtctccgcgtcgtctgagcattagcaccattggtattattccga gcattaaaaaaatcacccaggaatatccgcaggttaatctgacctttagcctgcatagc ccgtatagcgaagaacgtagcaaactgatgccgattaatgatcgctatccgatcgatga agtgatgaacatcctggatgaacatattcgtctgaccagccgcaaagtgtatattgcat atattatgctgcctggtgttaatgatagcctggaacatgcaaatgaagttgttagcctg

59 ctgaaaagccgttataaaagcggcaaactgtatcatgtgaacctgattcgctataatcc gaccattagcgcaccggaaatgtatggtgaagcaaatgaaggtcaggtggaagcctttt ataaagttctgaaaagcgcaggtattcatgttaccattcgcagccagtttggtattgat attgatgcagcatgtggtcagctgtatggcaattatcagaacagccagtaagaattcga gctc-3’

The final construct was designated pSaCfr after confirmation of the sequence by the

Molecular Core Facility (Penn State University).

Expression and Purification procedure. Plasmids encoding WT or mutant ygfB and cfr genes were transformed into BL-21(DE3), along with pDB1282 to allow for expression in the presence of Fe/S cluster assembly proteins (29). A single colony was used to inoculate a starter culture, which was subsequently grown overnight. M9 minimal media was inoculated with the overnight starter culture and grown to an O.D600 of 0.3 before induction of the genes encoded on pDB1282 with L-arabinose (0.2 %), while also adding

Fe and cysteine to final concentrations of 25 µM and 150µM, respectively. The bacteria were cultured to an O.D.600 of approximately 0.6 and then cooled to 18 °C. Expression was induced by addition of IPTG to a final concentration of 400 µM, and the cells were cultured for an additional 18 h at 18 °C.

All purification steps were carried out in a Coy anaerobic chamber. In a typical purification of RlmN and its variants, 20 g of cell paste is resuspended in 200 mL of lysis buffer [50 mM HEPES, pH 7.5, 300 mM KCl, 2 mM imidazole, 10 mM 2- mercaptoethanol (BME)] containing lysozyme at a final concentration of 1 mg mL-1.

After stirring at room temperature for 50 min, the solution is placed in an ice bath, cooled to ~4 °C (10 min), and subjected to ten 40 s bursts of sonic disruption (30% output) with intermittent pausing for 8 min to maintain a temperature less than 8 °C. The lysate is centrifuged for 1 h at 50,000 x g and 4 °C. The resulting supernatant is loaded onto a

60 Talon (Clontech) resin column equilibrated in lysis buffer. The column is washed twice with 100 mL of lysis buffer (50 mM HEPES, pH 7.5, 300 mM KCl, 2 mM imidazole, 10 mM BME), before elution with 50 mL of elution buffer (50 mM HEPES, pH 7.5, 300 mM KCl, 10 mM BME, 300 mM imidazole, 10% glycerol). Fractions displaying significant brown color are pooled and concentrated by ultracentrifugation using an

Amicon Centricon (Millipore; Billerica, MA) with a YM-10 membrane. The protein is exchanged into final buffer (10 mM HEPES pH 7.5, 500 mM KCl, 5 mM DTT, and 10%

Glycerol) using a PD-10 column and subjected to reconstitution of its iron–sulfur (Fe/S) cluster as previously described (30). The protein is concentrated, passed over a PD-10 column equilibrated in final buffer and loaded onto a HiPrep 16/60 S-200 column equilibrated in final buffer. A typical yield of protein is between 6-10 mg/g of cell paste for WT and single variant proteins. The purification of Cfr is conducted essentially as described above, with the exception that HEPES buffer, pH 7.5, is replaced with Tris–

HCl, pH 8.4.

Overproduction and purification of RlmN (d3-Met) and Cfr (d3-Met) was conducted as described above, with the exception that the corresponding plasmids were transformed into the E. coli methionine auxotroph B834(DE3)pLysS. In addition, the growth media was supplemented with 80 mg/L [methyl-d3]-L-methionine (99.9% D-atom).

Protein concentration was determined by the method of Bradford (31) using a correction factors of 0.76 (RlmN) and 0.78 (Cfr), which were established via amino acid analysis of the purified proteins.

General Methods. High-performance liquid chromatography (HPLC) with detection by mass spectrometry (LC-MS) was conducted on an Agilent Technologies (Santa Clara,

61 CA) 1200 system coupled to an Agilent Technologies 6410 QQQ mass spectrometer with simultaneous UV-vis analysis using an Agilent diode-array detector. The system was operated with the associated MassHunter software package, which was also used for data collection and analysis. Assay mixtures were separated on an Agilent Technologies

Zorbax Rapid Resolution XBD-C18 column (4.6 mm  50 mm, 1.8 µm particle size) equilibrated in 98% solvent A (40 mM aqueous ammonium acetate, pH 6.2) and 2% solvent B (acetonitrile). A gradient of 2–12% B was applied from 0.5 to 5 min, and then a gradient from 12–24% B was applied from 5 to 6.5 min. A subsequent increase to 50% B from 6.5 to 7.5 min was applied and then maintained at 50% for 0.5 min before returning to 2% B from 8 to 9 min. The column was allowed to re-equilibrate for 3 min under initial conditions before subsequent sample injections. Detection of products was performed using electrospray ionization in positive mode (ESI+) with MS2-scan or neutral loss MS as noted. Neutral loss MS takes advantage of the readily labile C-N bond between ribose and its associated base, with a mass of 132 that is specific to the ribose ring of RNA. This method allows small amounts of product to be observed as well as the exact mass of the corresponding charged base to be generated. It is also a convenient way to differentiate between modifications of the ribose from the modifications on the base, because the uncharged modified riboses are not detected in this mode, given that they have masses greater than 132. An elution profile of all relevant standards analyzed by HPLC with UV-vis detection, with the exception of m2A, is shown in Figure 2-2.

Synthesis and Purification of RNA 2500-2506. Dharmacon Inc. (Lafayette, CO) synthesized the 7-mer RNA substrate with proprietary methods containing the following bases: 2500-UCGAΨGU-2506. The 7-mer was deprotected with their standard protocol,

62 lyophilized, and resuspended in anaerobic water. The supplied extinction coefficient of

70,300 M-1cm-1 was used to determine the final concentration.

Synthesis and Purification of RNA 2018-2788. To generate the 771 nucleotide (nt) region of 23S rRNA corresponding to nt 2018-2788, a PCR product was amplified using the following primers: Forward 5′-GAA-ATT-AAT-ACG-ACT-CAC-TAT-AGG-GAA-

CTC-GCT-GTG-AAG-ATG-CAG-TGT-ACC-3′; Reverse 5′-

GGAGAACTCATCTCGGGGCAAGTTTCG-3, where the underlined region indicates the T7 promoter (28). The product was isolated and used in a reaction with T7 RNA polymerase under the following conditions in a final volume of 1 mL: 40 mM Tris-HCl, pH 8.0, 16 mM MgCl2, 10 mM DTT, 2 mM spermidine, 4 mM each NTP (dry powder was dissolved in water and neutralized with Tris base), 1 µg of PCR product, 0.2 U inorganic pyrophosphatase and hexahistidine-tagged T7 RNA polymerase (generous gift from Phil Bevilacqua). Reactions were incubated at 37 °C for 3 h. 20 U of RQ1 DNase

(Promega; Madison, WI) was added, and incubation at 37 °C was continued for 30 min.

After cooling on ice, LiCl and EDTA (pH 8.0) were added to final concentrations of 5 M and 50 mM, respectively. The reactions were placed at -20 °C overnight, and then centrifuged for 30 min at 13,000 rpm. The pellets were resuspended in 50 mM ammonium acetate (NH4OAc) (pH 5.2), and precipitated with 3 volumes of ethanol

(EtOH). The reactions were reprecipitated as described above, and the pellets were dissolved in 50 mM Tris-HCl (pH 8), 50 mM EDTA. The RNA was renatured by heating the solution for 8 min at 75 °C and then rapidly cooling it on ice for 10 min. The full transcripts were separated from abortive transcripts by size exclusion chromatography over an S200 column in water. Fractions containing the full transcript were pooled and

63 concentrated using a centricon 30 (MWCO 30 kDa) (Millipore; Billerica, MA). The final concentration was determined from an extinction coefficient (7.71 µM-1•cm-1) calculated from the sequence using the program DNAssist (32).

RNA digestion protocol. RNA was digested as previously described with slight modifications (Crain 1991). To determine the extent and type of modification of A2503 in activity assays, the RNA from each time point was digested to its constituent nucleotides. Subsequent to acid quenching of each time point (see below), two volumes of 200 mM (NH4OAc) (pH 7) were added to titrate the solution to a pH value of 5-6. P1 nuclease (US biological; Swampscott, MA) (0.4 U) was added and the reaction was incubated for 4 h at 45 °C. One volume of 1M NH4HCO3 (pH 8) was added along with 5

U of alkaline phosphatase (Promega) and the solution was incubated at 37 °C for 2 h.

The reactions were subjected to centrifugation to remove any precipitate and the supernatants were lyophilized to dryness. The resulting pellets were dissolved in 20 µL of H2O and analyzed by the above LC-MS method.

Activity Determinations of RlmN or Cfr. Reactions contained in a total volume of 100

µL: 100 mM Tris-HCl pH 8.4, 10 mM MgCl2, 1.5 mM SAM or d3-methyl-SAM, 25 µM

771-mer RNA substrate, 100 µM flavodoxin (Flv), 10 µM flavodoxin reductase (Flx), 2 mM NADPH, and either 2 µM RlmN or Cfr. Reaction mixtures lacking NADPH were incubated for 5-10 min at 37 °C, and a 10 µL aliquot was removed (t=0) and added to 10

µL of a solution containing 100 mM H2SO4 and 100 µM L-tryptophan (internal standard) to yield a final internal standard (IS) concentration of 50 µM. Reactions were initiated by the addition of NADPH and incubated for appropriate times before being quenched in acid as described above. The samples were subjected to the above RNA digestion

64 protocol and analyzed by LC-MS as described above. Standard curves were generated with 5’-dA and SAH purchased from Sigma-Aldrich. 8-methyl-adenosine was synthesized as described previously (28). An extinction coefficient of 15,400 M-1cm-1 was assumed for quantifying all adenosine-related nucleotides by UV-vis spectroscopy.

Activity Determination of RlmN or Cfr in 90% D2O. In order to determine the exchange of protons during a step in the mechanism we conducted assays in the presence of 90% D2O. Deuterium oxide (99.99% D) was first deoxygenated with 10 cycles of vacuum followed by purging with argon gas. The deoxygenated D2O was introduced into a Coy anaerobic chamber and used to prepare buffer containing 20 mM Tris–DCl, pD

~8.0, 200 mM KCl, 20 mM MgCl2, and 5 mM DTT by dilution (1 to 10) of a concentrated stock of the above buffer prepared in H2O. This buffer was then used to equilibrate Nick pre-poured gel filtration columns (GE Life Sciences; Piscataway, NJ).

Reaction mixtures (100 µL) containing 100 mM Tris-HCl (pH 8.4), 200 mM KCl, 10 mM MgCl2, 30 µM 771-mer RNA substrate, and either 10 µM RlmN or Cfr—all in

H2O—were passed through the individual columns, and 400 µL of the eluate was collected. The reaction was initiated by the addition of SAM (1 mM) followed by dithionite (2 mM), and quenched and analyzed as described above at appropriate times.

Single turnover experiments with labeled or unlabelled RlmN and Cfr. Reactions contained in a total volume of 50 µL: 100 mM Tris-HCl, pH 8.4, 10 mM MgCl2, 1.5 mM

SAM or d3-SAM, 10 µM 771-mer RNA substrate, and either 400 µM RlmN (or RlmN

(d3-met)) or 300 µM Cfr (or Cfr (d3-met)). Reaction mixtures lacking dithionite were incubated for 5-10 min at 37 °C, and a 20 µL aliquot was removed (t=0) and added to 20

µL of a solution containing 100 mM H2SO4. Reactions were initiated by the addition of 2

65 mM dithionite and incubated for 30 minutes before being quenched as described above.

The samples were analyzed as described above.

RlmN and Cfr catalyze Isotope exchange into 5’-deoxyadenosine. Reactions contained in a total volume of 100 µL: 100 mM Tris-HCl, pH 8.4, 10 mM MgCl2, 2 mM SAM, 20

µM 771-mer RNA substrate, 100 µM Flv, 10 µM Flx, 2 mM NADPH, and either 150 µM

RlmN (d3-met) or 150 µM Cfr (d3-met). Reaction mixtures lacking NADPH were incubated for 5-10 min at 37 °C. Reactions were initiated by the addition of 2 mM

NADPH and incubated for 240 minutes before being quenched as described above. The samples were analyzed as described above.

Determination of RNA bound to RlmN C118A. Digestion of RNA was carried out as described with slight modifications. As-isolated RlmNC118A (50 µL of a 411 µM solution) was with mixed with 50 µL of 200 mM NH4OAc and incubated for 5 min at 45

°C. P1 nuclease (1 U) was added and the solution was incubated further for 4 h at 45 °C, and then cooled to 37 °C. Ammonium bicarbonate (100 µL of 1 M solution) was added along with 5 U of alkaline phosphatase. The mixture was incubated for 1 h and then centrifuged to pellet the precipitate. The supernatant was removed and evaporated to dryness before resuspending in 20 µL water. The mixture of nucleosides was analyzed by

LC-MS as described above with simultaneous monitoring of the absorbance at 260 nm and the neutral loss product of 132 (loss of the ribose). The identification of cyclic GMP and cyclic AMP were determined by fragmentation of the 346 and 330 m/z ions and comparison of the fragment ions to that of known standards. Inosine was also confirmed by fragmentation of the ion having m/z 269 to the product ions 137 and 119, as detailed in the LC-MS massbank (http://www.massbank.jp/)

66 Trpsin digestion of RlmN and data acquisition by LC-MS. After the final purification steps, RlmN was diluted out to approximately 50 µM with 2X SDS-PAGE buffer. The solution was boiled for 5 min and centrifuged for 30 min, then loaded onto a 12% SDS-

PAGE gel. The band was excised from the destained gel with a new razor, and then cut into ~ 1 mm x 1mm cubes. The cubes were loaded into a 1.7 mL eppendorf tube and coomassie blue dye was removed from the protein by 25 mM NH4HCO3 in 50 % ACN.

Gel pieces were then dried with a speed-vac, then rehydrated in the 10 mM DTT and 25 mM NH4HCO3 for 1 hr at ~ 60 °C. The DTT solution was removed and the gel pieces washed 3 times with fresh 25 mM NH4HCO3. A solution of 100 mM iodoacetamide in

25 mM NH4HCO3 was added to cover the pieces, then the samples was wrapped in Al foil and placed in the dark at room temperature for 1 hr. The liquid was then removed and washed an additional 3 times with 25 mM NH4HCO3. The gel pieces were dehydrated with 25 mM NH4HCO3 in 50 % ACN twice, then covered with a solution of

25 mM NH4HCO3 with 12.5 ng/µL of sequencing grade modified Trypsin (Promega,

MA). The sample was placed on ice to 10 min, then overlaid with 25 mM NH4HCO3.

The digestion was then carried out for an additional 18 hr at 37 °C. To extract the peptides, the solution was removed from the gel pieces and place in a new eppendorf tube, and 5% formic acid in 50% aqueous ACN was added, vortex, spun down, and sonicated for 1 hr at 50 °C. The solution was removed and added to the original fraction.

The was repeated twice more, with all extractions being combined and lyophilized to dryness. The samples were resuspended in 1% TFA in 50% aqueous ACN. A blank containing approximately the same size gel piece with no protein was worked up simultaneously to account for auto-proteolysis of trypsin. Eight µL of each digested

67 fraction was injected using a Proxeon (Thermo-Fisher) EASY nLC II, trapped on 2 cm x

100 mm C18 column (Thermo-Fisher) and separated on a C18 reversed-phase column

(ProteoPep 2, 150 µm x 50 mm, 300; New Objectives), using a 55 min 5-45% FB linear gradient (FA = 0.1% formic acid, FB = CH3CN, 0.1% formic acid) at a flow rate of 500 nL min-1. Multiply charged peptide-ions were automatically chosen during a 30,000 resolution scan in the Orbitrap and fragmented by CID (isolation width of 2.0 m/z,

Normalized collision energy of 35, Activation Q of 0.25, and Activation time of 30 msec.) in the ion trap of a LTQ-Orbitrap XL-ETD mass spectrometer (Thermo-Fisher) with a Nano-Electrospray ionization source (New Objectives) using the following instrument settings: Spray Voltage of 4 kV, Capillary Temp. of 250 C, Capillary Voltage of 110 V, Tube Lens of 101 V, AGC targets of 1.0e+06 for the FT scan and 1.0e+04 for the MS/MS scan. The manufacturer’s recommendations were followed for all other settings. The data was collected and stored for later analysis.

The MS and MS/MS fragmentation data were analyzed using Proteome Discoverer 1.1

(Thermo-Fisher) using both Sequest and the Mascot version 2.27 (33). For both analyses, data were searched using an E. coli (BL21DE3) genome with the addition of the SA Cfr sequence into this FASTA database. For these analysis, theoretical peptides generated by a tryptic digest with a maximum of two missed cleavages were considered, and the precursor and product mass tolerances were set to ± 10 ppm and ± 1.2 Da, respectively.

Variable modifications of Cysteine of both methylation and carboxyamidomethyltion were allowed for. Data was validated using a reverse database decoy search to a false discovery rate of 0.5%.

1000 A 8 SAH m A 140

800 120 Absorbance(mAU)

G 5'-dA 100 600 SAM 80

400 U 60

Absorbance (mAU) Absorbance 40 200 pU C Adenine Trp 20

0 0 0 2 4 6 8 Time (min)

69 2.4 Results

Demonstration of in vitro methylation activity with 7-mer oligonucleotide substrate

Initial activity determinations were conducted with a seven nucleotide oligomer (7- mer) spanning positions 2500–2506 on rRNA, which included the naturally occurring pseudouridine modification at position 2504. Shown in Figure 2-3 are mass spectra of the resulting methylated adenosine (CH3-Ad) of a reaction containing 500 µM RlmN, 1.5 mM SAM, and 250 µM RNA substrate. A chromatographic peak at retention time 6.85 min (Figure 2-3, red trace) exhibits an MS peak at m/z = 282.1 (Figure 2-3 inset, red trace) and a fragmentation pattern consistent with that of m2A (7), indicating that the 7- mer can indeed be methylated by RlmN. MS peaks at 283.1 and 284.1 m/z (inset) derive from natural abundance 13C in the product. Use of the in vivo reducing system

(Flv/Flx/NADPH) in place of dithionite gives turnover, even after 18 h of incubation.

Testing for direct methyl transfer from SAM by isotope-tracing experiments

To confirm that the methyl group derives from SAM, as had been shown previously by

Yan et al (28), both RlmN and Cfr were incubated as described above, with the exception

that SAM was replaced with S-adenosyl-L-[methyl-d3]methionine (d3-SAM).

20 Tryptophan (IS, 50 µM)

No SAM visible Absorbance (mAU) Absorbance 5

0 1 2 3 4 5 6

Time (min)

Unexpectedly, the product of the RlmN reaction gives the same m/z and isotopic distribution (Figure 2.3 inset, solid black trace) as the product of the reaction with unlabeled SAM, suggesting that the methyl group transferred under single-turnover conditions (protein in excess of substrate) does not reflect the isotopic composition of

SAM added to the assay mixture. To test whether the methyl group could be derived from a SAM molecule that bound to the enzyme during its production in E. coli, a 200 µM sample of RlmN was quenched in acid and analyzed as described; SAM is not detected

(Figure 2.4). Similar results are obtained when Cfr (500 µM) is incubated with d3-SAM

(1.5 mM) and the 7-mer RNA substrate (250 µM) under turnover conditions (Figure 2.3,

72 dotted black trace). In this case, the product’s retention time is slightly less (6.8 min) than that of the product from the RlmN reaction, but identical to that of a synthetic m8A standard (Figures 2.3 and 2.5).

In addition, its m/z value and fragmentation pattern are also identical to those of the synthetic m8A standard and consistent with published results (7). Significantly, the methyl group transferred to the adenosine corresponding to A2503 is again found to lack deuterium from d3-SAM; this surprising isotope-tracing result is common to the two enzymes.

tryptophan (IS)

35 12 30 10

8 Intensity(x10 )

3 25 m A 8 20 6

3 ) Intensity (x10 Intensity 4 10

5 2

0 5.6 5.8 6.0 6.2 6.4 6.6 6.8

Time (min)

Figure 2-5. Overlay of m8A standard (black trace) and product of Cfr (red trace).

73 Testing for methyl transfer from the protein

To test the possibility that the methyl group transferred in the first turnover derives from the protein, both RlmN and Cfr were produced in, and isolated from, an E. coli methionine auxotroph cultured in the presence of [methyl-d3]-methionine (d3-Met), and then used in subsequent activity determinations. Shown in Figure 2.6 are mass spectra of the associated products. As described above, when unlabeled RlmN (Figure 2.6A) or Cfr

(Figure 2.6B) is incubated with d3-SAM (black traces) the methyladenosine (CH3-Ad) products are found to lack deuteria (m/z 282.1).

Figure 2-6. Single turnover experiments. 250 µM 7-mer RNA substrate with (A) 500 µM RlmN and 1 mM d3-SAM (black trace), 500 µM RlmN (d3-met) with 1 mM SAM (red trace), or (B) 500 µM Cfr and 1 mM d3-SAM (black trace), 500 µM Cfr (d3-met) with 1 mM SAM (red trace). Peaks at m/z 282.1 correspond to m2A (RlmN) or m8A (Cfr) with no deuterium enrichment, while peaks at m/z 284.1 correspond to m2A (RlmN) or m8A (Cfr) with two deuterium atoms. No significant peaks at m/z 285.1 are observed, which would correspond to m2A (RlmN) or m8A (Cfr) with three deuterium atoms.

By contrast, when RlmN (Figure 2-6A) or Cfr (Figure 2-6B) isolated from E. coli supplemented with d3-Met during growth [RlmN (d3-Met) and Cfr (d3-Met), respectively]

74 is incubated in the presence of unlabeled SAM (red traces), the products exhibit peaks corresponding to m/z = 284.1, two mass units higher. This result implies that the transferred methyl group contains two deuteria, and that it does indeed derive from the protein. The finding that a peak for m/z = 285.1 (corresponding to three deuteria) is not observed implies that methyl transfer takes place with loss of one deuterium atom. This result stands in stark contrast to the accepted mechanism of SAM-dependent methyl transfer, in which the methyl group is transferred without exchange of any of its three hydrogens (4, 5). The results with Cfr (d3-Met) are even more striking (Figure 2-6B, red trace). In this instance, the ladder of m/z values (284.1, 283.1, and 282.1) suggests the presence of a step that effects hydrogen exchange with solvent either before or after methyl transfer.

Identification of the source of hydrogen transferred to 5’-dA and analysis of abortive

SAM cleavage

In all RS enzymes characterized to date, SAM is reductively cleaved to a 5’-dA• intermediate. Many RS enzymes catalyze abortive cleavage of SAM in the presence of dithionite, which can complicate efforts to assess the role of the 5’-dA• (29, 34). As has been shown previously, RlmN and Cfr both cleave SAM abortively, even in the absence of substrate (Figure 2-7) (28, 35).

500

400

300

200 Concentration (µM) Concentration

100

0 10 20 30 40 50 60

Time (min)

Figure 2-8. Domain V of the 23S rRNA. Position A2503 is highlighted in red.

Studies on other RS enzymes have shown that abortive cleavage is reduced significantly when dithionite is replaced by the in vivo reducing system

(Flv/Flx/NADPH) (29, 36). However, as described above, the in vivo reducing system does not support turnover of the 7-mer RNA substrate. Studies by Yan et al have shown that a segment of 23S rRNA spanning nucleotides 2018–2788 (which encompasses

Domain V of 23 S rRNA) supports turnover as well as, if not better than, the entire 23S rRNA. (28). Therefore, a 771-nucleotide RNA (Figure 2-8) substrate was synthesized by in vitro transcription and used in single turnover activity determinations (500 µM Cfr or

RlmN and 10 µM RNA substrate).

Figure 2-9. Single turnover experiment with 771mer. (A) 300 µM RlmN (d3-met) and (B) 300 µM Cfr (d3-met). Both reactions contained 1.5 mM SAM (black trace), 20 µM 771-mer RNA substrate, and the Flv/Flx reducing system. The red dotted line in both traces corresponds to the mass spectrum of commercially available 5’-dA. Shown in Figure 2-9 are mass spectra of 5’-dA analyzed from reactions conducted with RlmN (d3-Met) (Figure 2-9A) and Cfr (d3-Met) (Figure 2-9B) in the presence of

78 unlabeled SAM. The dotted red line in both Figures corresponds to a 5’-dA standard at natural abundance, which displays the expected m/z = 252.1. The solid black line in both

Figures corresponds to 5’-dA generated during turnover. Astonishingly, the results demonstrate significant deuterium enrichment in 5’-dA, implying that the role of the 5’- dA• is not to abstract a hydrogen atom from C2 or C8 of the substrate as suggested by

Yan et al (28), but to abstract a hydrogen atom from a protein-bound methyl group.

Identifying the source of the methyl group on RlmN by mass spectrometry

To identify the amino acid that bears the methyl group to be transferred, MS was carried out on tryptic digests of RlmNWT that had been reduced and alkylated with iodoacetamide. One peptide (GDDIDAAC355GQLAGDVIDR) is found to contain a methylated Cys residue (m/z 909.42017, 45 spectra, 76%) or to be alkylated by iodoacetamide (m/z 930.92358, 14, 24%). The mass of the intact precursors are within 3 ppm of the theoretical values, and both the b and y ion series produced by fragmentation unambiguously define the site of the methyl modification as Cys355 (Figure 2-10).

Figure 2-10. MSn of C355 containng peptide. A) Roepstorff and Fohlman fragementation pattern of the peptide contain C355 from the trypsin digest of E. coli RlmN and B) the corresponding spectrum from the doubly charged peptide-ion separated by C18 reverse phase chromatography and fragmented by CID in the ion trap of a LTQ-Orbitrap XL-ETD mass spectrometer (Thermo- Fisher) with a Nano-Electrospray ionization. The precursor peptide had an m/z of 909.42017 (doubly charged) that eluted at 37.05 min. and was identified with an Xcorr score of 5.02 by Sequest. Y ions are indicated in blue and b ions in red as well as the charge state of each. The y ions that indentify the methylation is at C355 are y10 and y11. Y10 has a m/z = 1043.5496 corresponding to the sequence GQLAGDVIDR and y11 has a mz = 1160.5206 corresponding to the sequence C355GQLAGDVIDR, an increase of 116.971 mass units indentifying the C355 as the site of methylation. A mechanism for catalysis by RlmN that is consistent with observations is proposed in

Figure 2-11. In the initial, priming, step a methyl group from the first molecule of SAM is transferred to C355 via a typical SN2 displacement mechanism. Reductive cleavage of a second SAM molecule affords the 5’-dA•, which, in the subsequent step of the mechanism (1), abstracts a hydrogen atom from the protein-bound methyl group to yield a neutral carbon-centered radical. Attack of the carbon-centered radical on the sp2- hybridized C2 of A2503 in 23S rRNA results in formation of a carbon–carbon bond and

80 generation of a resonance-delocalized radical on the nucleotide base (2). Loss of an electron (3)—perhaps back to the Fe/S cluster—and abstraction of the proton from C2 affords the methylated product crosslinked to the protein via C355 (4). This crosslink is resolved by attack of the C118 thiolate onto C355 to yield a disulfide bond and an enamine (5), which collapses to the methylated product upon tautomerization and acquisition of a proton derived ultimately from solvent (6). An analogous mechanism for

Cfr-catalyzed methylation of C8 is depicted in Figure 2-11.

81 The mechanism allows rationalization of previously published results that were incompletely understood. Both RlmN and Cfr contain five conserved Cys residues, three of which are in the canonical C112xxxC116xxC119 (S. aureus Cfr numbering) RS motif. In vitro mutagenesis studies by Yan et al showed, as expected, that substitution of any of these Cys residues with Ala residues abrogated turnover (28), consistent with in vivo studies by Giessing et al and Kaminska et al (7, 22). Further studies by Kaminska et al showed that the remaining two conserved Cys residues (C105 and C338) are also absolutely required for generation of m8A by Cfr. Interestingly, introduction of the

C105A substitution resulted in a stop in a reverse transcription reaction used to assess modification of the RNA substrate, suggesting that some type of modification took place.

Failure, however, of the C105A variant to elicit resistance to florfenicol and tiamulin in vivo suggested that the 23S rRNA did not contain the m8A modification (22).

A B

1.2 0.8 1.0 0.6 0.8 0.6

0.4 Absorbance Absorbance 0.4 0.2 0.2

300 400 500 600 700 300 400 500 600 700

Wavelength (nm) Wavelength (nm) C D

0.8 1.2

0.6 0.8 0.4

0.4

Absorbance Absorbance 0.2

0.0

300 400 500 600 700 300 400 500 600 700

Wavelength (nm) Wavelength (nm)

Figure 2-12. UV-visible spectra of A) RlmN Wt (16.8 μM), B) RlmN C118A (12.6 μM), C) RlmN C355A (12 μM), and D) Cfr Wt (27 μM).

Interestingly, RlmNC118A used in our studies displays a max at 265 nm (Figure 2-12) rather than 278 nm as is observed for RlmNWT (Figure 2-12A), suggesting that it contains bound nucleic acid. Figure 2-13 depicts an elution profile of RlmNC118A subjected to digestion with P1 nuclease and alkaline phosphatase, which shows not only the canonical RNA bases but also pseudouridine and other modified bases, confirming

83 the presence of RNA. These characteristics are not observed for CfrWT (Figure 2-12D),

RlmNWT (Figure 2-12A), or RlmNC355A (Figure 2-12C), consistent with a model in which

C118 is required to resolve a covalently bound intermediate.

A G 800

600

cGMP cAMP 400 C U

pU I Absorbance (mAU) Absorbance 200 7 m G

6 m A

0 2 4 6 8

Time (min)

Figure 2-13.Elution profile of nucleosides generated by RNA digestion of AI RlmNC118A. Nucleoside identity was determined by neutral loss MS, comparison to known standards, and collision induced dissociation (CID) MS. pU, pseudouridine; C, cytidine; U, uridine; cGMP, (5’- 3’) cyclic guanosine monophosphate; m7G, 7-methyl guanosine; I, inosine; G, guanosine; cAMP, (5’-3’) cyclic adenosine monophosphate; A, adenosine; m6A, 6-methyl adenosine. The mechanism in Figure 2-11 makes a further prediction that can be tested experimentally. Under single turnover conditions using d3-SAM, RlmN, which methylates only C2 of A2503, should generate product containing essentially no deuterium.

25 20

) ) 3 20 3 15 15 10 700 10

5 Intensity Intensity (x10 5 Intensity (x10 600 0 0 280 282 284 286 294 296 298 300

) 500 3 m/z m/z 400

300 Intensity (x10 Intensity 200

100

6.6 6.8 7.0 7.2 7.4 7.6 7.8

Time (min)

Figure 2-14. Single turnover experiment with 300 µM Cfr in the presence of 1.5 mM d3-SAM, and 10 µM 771-mer RNA substrate. The peak at 6.7 min corresponds to m8A (1), with its associated mass spectrum in the above left inset. The peak at 7.7 min corresponds to 2,8- dimethyladenosine with its associated mass spectrum in the above right inset. The enrichment at m/z 298.1 in 2,8-dimethyladenosine indicates that a small amount of this molecule receives a methyl group from d3-SAM. By contrast, Cfr, which can methylate both C2 and C8 of A2503, should give rise to a dimethylated product containing both unlabeled and labeled methyl groups under similar conditions. This arises from two distinct modifications on the same base, of which the first (at C8) results in incorporation of the unlabeled methyl group associated with protein, and the second (at C2) results in incorporation of a labeled methyl group from d3-

85 SAM added in the reaction. Figure 2-14 depicts mass spectra of monomethylated and dimethylated CH3-Ad from Cfr reactions.

2.0 40

Intensity (x10 Intensity )

2 8 1.8 30 6

1.6 20 )

5 4 1.4

10 3 )

Intensity (x10 2

1.2 0 0 1.0 280 282 284 286 288

Intensity (x10 Intensity m/z 0.8

0.6

0.4

6.6 6.8 7.0 7.2 7.4 7.6 7.8 8.0

Time (min)

Figure 2-15. Single turnover experiment with 400 µM RlmN in the presence of 1.5 mM d3-SAM (red trace); or 400 µM (d3-met) RlmN in the presence of 1.5 mM SAM (black dotted trace), both using 10 µM 771-mer RNA substrate. The peak at 6.8 min corresponds to the m2A with its associated mass spectrum in the above inset. The mass spectra of the corresponding products (inset) indicate that RlmN catalyzes its reaction with minimal exchange with solvent. As can be observed, the monomethylated species displays m/z = 282.1, consistent with it having minimal, if any, deuterium enrichment. By contrast, the dimethylated species exhibits predominantly m/z = 296.1, consistent with no deuterium incorporation, as well as m/z 298.1, consistent with an unlabeled methyl group and a methyl group containing two deuteria. Interestingly the dimethylated species does not contain significant quantities of m/z = 297.1, which would be expected if exchange takes place readily during both methylations. Consistent with this observation, RlmN, which catalyzes only

86 C2 methylation, does not catalyze exchange of methyl-associated hydrogens with solvent hydrons (Figure 2-15).

2.5

) 1.2 ) 180 3 3 2.0

160 1.5 0.8 1.0

140 0.4

)

Intensity Intensity (x10 Intensity Intensity (x10 3 0.5 120

280 282 284 286 294 296 298 300 302 304 100 m/z m/z

80 Intensity (x10 Intensity

6.6 6.8 7.0 7.2 7.4 7.6 7.8 8.0

Time (min)

Figure 2-16. Single turnover experiment with 300 µM d3-met Cfr in the presence of 1.5 mM SAM, and 10 µM 771-mer RNA substrate. The peak at 6.7 min corresponds to a mixture of m8A 8 (1) and d2-m A with the associated mass spectrum in the above left inset. The peak at 7.7 min corresponds to a mixture of 2,8-dimethyladenosine and 2,8-d2-dimethyladenosine with the associated mass spectrum in the above right inset. The lack of complete transfer of deuterium from the enzyme to the product indicates a step in which substrate deuteria are exchanged with solvent hydrons.

Figure 2-16 further supports the contention that exchange takes place predominantly at

C8 in Cfr. When Cfr (d3-Met) is incubated with a 771-mer RNA substrate and unlabeled

SAM, the monomethylated species loses a significant amount of expected deuterium incorporation, while the majority of the dimethylated species contains d4- dimethyladenosine. If both positions exchange readily with solvent hydrons, then the

87 distribution of m/z values in the mass spectrum should be similar in shape to that observed for the monomethylated adenosine.

Figure 2-17. RlmN (black trace) or Cfr (red trace) reaction conducted in the presence of 90% D2O, 1 mM SAM and 771-mer RNA substrate. The mass spectra are extracted from the peaks eluting at 6.8 min (RlmN, black trace) or 6.7 min (Cfr, red trace). As shown previously, the product from the Cfr reaction exchanges more readily with solvent than that from the RlmN reaction. To support these conclusions further, experiments assessing RlmN- and Cfr-catalyzed exchange of solvent hydrons into products were conducted. Figure 2-17 displays mass spectra of CH3-Ad products from RlmN (black trace) or Cfr (red trace) reactions conducted in the presence of 90% D2O, 1 mM SAM and the 771-mer RNA substrate.

The mechanism in Figure 2-11 predicts that RlmN should catalyze incorporation of one deuterium atom at C2 of the CH3-Ad product. Surprisingly, very little of the expected

88 m/z = 283.1 is observed, suggesting that a monoprotic base removes the proton at C2 of the substrate (5) and uses it to protonate the enamine. As expected, methylation of C8 by

Cfr is accompanied by extensive exchange of solvent deuterium into the product, even to the extent of generating d3-methyladenosine (m/z = 285.1).

2.5 Discussion

In summary we have detailed an unprecedented mechanism for SAM-dependent methyltransfer to carbon atoms that are unactivated toward traditional polar electrophilic addition. This ping-pong mechanism necessitates the involvement of two SAM molecules: one as the ultimate source of the appended methyl group, and the other as the source of the potently oxidizing 5’-dA•, which activates the protein-bound methyl group for transfer via radical addition by abstracting one of its hydrogen atoms. Interestingly, this unusual strategy involves an initial deactivation of the methyl group of SAM, given that it is transformed from a sulfonium constituent to a thioether constituent. At present, it is unknown whether there are two distinct SAM binding sites on RlmN and Cfr as suggested by Kaminska et al based on modeling of the Cfr structure (22). This strategy of generating a covalent methylcysteinyl intermediate might allow mediation of chemistry that takes place at one location on the protein with chemistry that takes place on another.

Moreover, the choice of a Cys residue as the methyl carrier is energetically expedient with regard to the following step, because adjacent sulfur atoms are known to stabilize carbon-centered radicals and therefore decrease the homolytic bond-dissociation energy associated with their generation (37). The RlmN and Cfr reactions are redox neutral. One

89 electron, however, is removed from the system by the 5’-dA•, while removal of a second electron is required to facilitate abstraction of a C2 (RlmN) or C8 (Cfr) proton from the substrate, a net two-electron oxidation. The unique ability of Cys residues to form disulfide bonds allows for cleavage of the covalent intermediate with concomitant return of the required electrons to the product.

Although the proposed mechanisms of RlmN and Cfr are essentially identical, there are some differences between the two reactions. Cfr appears to have evolved directly from

RlmN, using the same mechanistic strategy to methylate C8 of A2503, but still retaining some ability to methylate C2. Cfr’s bifunctional activity suggests that its substrate- binding pocket is less rigid, allowing for at least two different conformations of A2503 that would put either C2 or C8 in resonance with the enzyme’s catalytic machinery. This is consistent with the observation that Cfr catalyzes exchange of solvent hydrons into product —most likely only when it catalyzes methylation at C8 —while RlmN does not.

A number of other RS enzymes have been predicted to catalyze methyltransfer to unactivated carbon atoms in pathways that involve the biosynthesis of a number of secondary metabolites including, among others, fosfomycin, bialaphos, fortimicin, thienamycin, gentamicin, and pactamycin (38), as well as catalyze key steps in the biosynthesis of specific bacteriochlorophylls (39). These enzymes are all annotated as cobalamin binding proteins, and cobalamin is believed to be the source of the appended methyl group. No detailed in vitro mechanistic studies on these enzymes have been forthcoming because of the difficulty associated with their stable isolation (40, 41). One class of RS enzymes that may be directly related, however, are the

90 methylthioltransferases, which also use SAM as the source of a 5’-dA• and an appended methyl group.

2.6 Acknowledgements

We thank J. Martin Bollinger Jr. for a critical reading of the manuscript and helpful discussions. We thanks Drs. Harry Noller and Laura Landcaster (UC Santa Cruz) for plasmid pKK3535 and Dr. Song Tan for E. coli strain B834(DE3)pLysS. We thank Dr.

Philip Bevilacqua for a plasmid encoding hexahistidine-tagged T7 RNA polymerase, and his lab for sharing their expertise in RNA manipulation.

91 2.7 References.

1. Poehlsgaard, J., and Douthwaite, S. (2005) The bacterial ribosome as a target for

antibiotics, Nat. Rev. Microbiol. 3, 870–881.

2. Kaczanowska, M., and Rydén–Aulin, M. (2007) Ribosome biogenesis and the

translation process in Escherichia coli, Microbiol. Mol. Biol. Rev. 71, 477–494.

3. Chow, C. S., Lamichhane, T. N., and Mahto, S. K. (2007) Expanding the

nucleotide repertoire of the ribosome with post-transcriptional modifications, ACS

Chem. Biol. 2, 610–619.

4. Hegazi, M. F., Borchardt, R. T., and Schowen, R. L. (1979) a-Deuterium and

carbon-13 isotope effects for methyl transfer catalyzed by catechol-O-methyl-

transferase. SN2-like transition state., J. Am. Chem. Soc. 101, 4359–4365.

5. Woodard, R. W., Tsai, M.-D., Floss, H. G., Crooks, P. A., and Coward, J. K.

(1980) Sterochemical course of the transmethylation catalyzed by catechol O-

methyltransferase, J. Biol. Chem. 255, 9124–9127.

6. Kowalak, J. A., Bruenger, E., and McCloskey, J. A. (1995) Posttranscriptional

modification of the central loop of domain V in Escherichia coli 23S ribosomal

RNA, J. Biol. Chem. 270, 17758–17764.

7. Giessing, A. M. B., Jensen, S. S., Rasmussen, A., Hansen, L. H., Gondela, A.,

Long, K. S., Vester, B., and Kirpekar, F. (2009) Identification of 8-

methyladenosine as the modification catalyzed by the radical SAM

methyltransferase Cfr that confers antibiotic resistance in bacteria, RNA 15, 327–

336.

92 8. Kehrenberg, C., Schwarz, S., Jacobsen, N. E., Hansen, L. H., and Vester, B.

(2005) A new mechanism for chloramphenicol, florfenicol and clindamycin

resistance: methylation of 23S ribosomal RNA at A2503, Mol. Microbiol. 57,

1064–1073.

9. Frey, P. A., and Hegeman, A. D. (2007) Enzymatic Reaction Mechanisms, Oxford

University Press, New York.

10. Silverman, R. B. (2002) The Organic Chemistry of Enzyme-Catalyzed Reactions,

Academic Press, New York.

11. Steitz, T. A. (2010) From the structure and function of the ribosome to new

antibiotics (Nobel Lecture), Angew. Chem. Int. Ed. 49, 4381–4398.

12. Weisblum, B. (1995) Erythromycin resistance by ribosome modification,

Antimicrob. Agents Chemother. 39, 577–585.

13. Ban, N., Freborn, B., Nissen, P., Penczek, P., Grassucci, R. A., Sweet, R., Frank,

J., Moore, P. B., and Steitz, T. A. (1998) A 9 Å resolution x-ray crystallographic

map of the large ribosomal subunit, Cell 93, 1105–1115.

14. Harms, J., Schluenzen, F., Zarivach, R., Bashan, A., Gat, S., Agmon, I., Bartels,

H., Francheschi, F., and Yonath, A. (2001) High-resolution structure of the large

ribosomal subunit from a mesophilic eubacterium, Cell 107, 679–688.

15. Schuwirth, B. S., Borovinskaya, M. A., Hau, C. W., Zhang, W., Vila-Sanjurjo, A.,

Hoton, J. M., and Cate, J. H. D. (2005) Structures of the bacterial ribosome at 3.5

Å resolution, Science 310, 827–834.

93 16. Selmer, M., Dunham, C. M., Murphy, F. V., IV, Weislbaumer, A., Petry, S.,

Kelley, A. C., Weir, J. R., and Ramakrishnan, V. (2006) Structure of the 70S

ribosome complexed with mRNA and tRNA, Science 313, 1935–1942.

17. Toh, S.-M., Xiong, L., Bae, T., and Mankin, A. S. (2008) The methyltransferase

YfgB/RlmN is responsible for modification of adenosine 2503 in 23S rRNA, RNA

14, 98–106.

18. Long, K. S., Poehlsgaard, J., Kehrenberg, C., Schwarz, S., and Vester, B. (2006)

The Cfr rRNA methyltransferase confers resistance to phenicols, lincosamides,

oxazolidinones, pleuromutilins, and streptogramin A antibiotics, Antimicrob.

Agents Chemother. 50, 2500–2505.

19. Schwarz, S., Werckenthin, C., and Kehrenberg, C. (2000) Identification of a

plasmid-borne chloramphenicol–florfenicol resistance gene in Staphylococcus

sciuri, Antimicrob. Agents Chemother. 44, 2530–2533.

20. Smith, L. K., and Mankin, A. S. (2008) Transcriptional and translational control

of the mlr operon, which confers resistance to seven classes of protein synthesis

inhibitors, Antimicrob. Agents Chemother. 52, 1703–1712.

21. Toh, S.-M., Xiong, L., Arias, C. A., Villegas, M. V., Lolans, K., Quinn, J., and

Mankin, A. S. (2007) Acquisition of a natural resistance gene renders a clinical

strain of methicillin-resistant Staphylococcus aureus resistant to the synthetic

antibiotic linezolid, Mol. Microbiol. 64, 1506–1514.

22. Kaminska, K. H., Purta, E., Hansen, L. H., Bujnicki, J. M., Vester, B., and Long,

K. S. (2010) Insights into the structure, function and evolution of the radical-SAM

94 23S rRNA methyltransferase Cfr that confers antibiotic resistance in bacteria,

Nuc. Acids. Res.

23. Sofia, H. J., Chen, G., Hetzler, B. G., Reyes-Spindola, J. F., and Miller, N. E.

(2001) Radical SAM, a novel protein superfamily linking unresolved steps in

familiar biosynthetic pathways with radical mechanisms: functional

characterization using new analysis and information visualization methods,

Nucleic Acids Res. 29, 1097-1106.

24. Frey, P. A., Hegeman, A. D., and Ruzicka, F. J. (2008) The radical SAM

superfamily, Crit. Rev. Biochem. Mol. Biol. 43, 63–88.

25. Booker, S. J. (2009) Anaerobic functionalization of unactivated C–H bonds, Curr.

Opin. Chem. Biol. 13, 58–73.

26. Atta, M., Mulliez, E., Arragain, S., Forouhar, F., Hunt, J. F., and Fontecave, M.

(2010) S-adenosylmethionine-dependent radical-based modification of biological

macromolecules, Curr. Opin. Struct. Biol. 20, 1–9.

27. Frey, P. A., and Booker, S. J. (2001) Radical mechanisms of S-

adenosylmethionine-dependent enzymes, Adv. Protein Chem. 58, 1-45.

28. Yan, F., LaMarre, J. M., Röhrich, R., Wiesner, J., Jomaa, H., Mankin, A. S., and

Galoníc Fujimori, D. (2010) RlmN and Cfr are radical SAM enzymes involved in

methylation of ribosomal RNA, J. Am. Chem. Soc. 132, 3953-3964.

29. Cicchillo, R. M., Iwig, D. F., Jones, A. D., Nesbitt, N. M., Baleanu-Gogonea, C.,

Souder, M. G., Tu, L., and Booker, S. J. (2004) Lipoyl synthase requires two

equivalents of S-adenosyl-L-methionine to synthesize one equivalent of lipoic

acid, Biochemistry 43, 6378-6386.

95 30. Grove, T. L., Lee, K. H., St Clair, J., Krebs, C., and Booker, S. J. (2008) In vitro

characterization of AtsB, a radical SAM formylglycine-generating enzyme that

contains three [4Fe-4S] clusters, Biochemistry 47, 7523-7538.

31. Bradford, 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.

32. Patterton, H.-G., and Graves, S. (2000) DNAssist: the integrated editing and

analysis of molecular biology sequences in Windows, Bioinformatics 16, 652–

653.

33. Perkins, D. N., Pappin, D. J., Creasy, D. M., and Cottrell, J. S. (1999) Probability-

based protein identification by searching sequence databases using mass

spectrometry data, Electrophoresis 20, 3551–3567.

34. Grove, T. L., Ahlum, J. H., Sharma, P., Krebs, C., and Booker, S. J. (2010) A

consensus mechanism for radical SAM-dependent dehydrogenation? BtrN

contains two [4Fe–4S] clusters, Biochemistry 49, 3783–3785.

35. Booth, M. P. S., Challand, M. R., Emery, D. C., Roach, P. L., and Spencer, J.

(2010) High-level expression and reconstitution of active Cfr, a radical-SAM

rRNA methyltransferase that confers resistance to ribosome-acting antibiotics,

Protein Expr. Purif. 74, 204–210.

36. Guianvarc'h, D., Florentin, D., Bui, B. T. S., Nunzi, F., and Marquet, A. (1997)

Biotin synthase, a new member of the family of enzymes which uses S-

adenosylmethionine as a source of deoxyadenosyl radical, Biochem. Biophys. Res.

Commun. 236, 402-406.

96 37. Wu, W., Lieder, K. W., Reed, G. H., and Frey, P. A. (1995) Observation of a

second substrate radical intermediate in the reaction of lysine 2,3-aminomutase:

A radical centered on the -carbon of the alternative substrate, 4-Thia-L-lysine.,

Biochem. 34, 10532-10537.

38. Kudo, F., Kasama, Y., Hirayama, T., and Eguchi, T. (2007) Cloning of the

pactamycin biosynthetic gene cluster and characterization of a crucial

glycosyltransferase prior to a unique cyclopentane ring formation, J. Antibiot. 60,

492–503.

39. Gomez Macqueo Chew, A., and Bryant, D. A. (2007) Chlorophyll biosynthesis in

bacteria: the origins of structural and functional diversity, Ann. Rev. Microbiol.

61, 113-129.

40. Seto, H., and Kuzuyama, T. (1999) Bioactive natural products with carbon-

phosphorus bonds and their biosynthesis, Nat. Prod. Rep. 16, 589–596.

41. van der Donk, W. A. (2006) Rings, radicals, and regeneration: the early years of a

bioorganic laboratory, J. Org. Chem. 71, 9561–9571.

Chapter 3

Cfr and RlmN Contain Only One Binding Site for S-Adenosylmethionine, which Supports Two Distinct Reactivities — Methyl Transfer by SN2 Displacement and Radical Generation

This chapter was reproduced from “Grove, T.L., Radle ,M.I., Krebs, C., Booker, S.J. Cfr and RlmN contain a single [4Fe-4S] cluster, which directs two distinct reactivities for S- adenosylmethionine: methyl transfer by SN2 displacement and radical generation. J. Am. Chem. Soc. 2011 Dec 14;133(49):19586-19589.

98 3.1 Abstract

The radical SAM (RS) proteins RlmN and Cfr catalyze methylation of carbons 2 and 8, respectively, of adenosine 2503 in 23S rRNA. Both reactions are similar in scope, entailing the synthesis of a methyl group partially derived from S-adenosylmethionine

(SAM) onto electrophilic sp2-hybridized carbon atoms via the intermediacy of a protein

S-methylcysteinyl (mCys) residue. Both proteins contain five conserved Cys residues, each of which is required for turnover. Three cysteines lie in a canonical RS CxxxCxxC motif and coordinate a [4Fe–4S]-cluster cofactor. The remaining two cysteines are at opposite ends of the polypeptide. Herein we show that each protein contains only the one

“radical SAM” [4Fe–4S] cluster, and that the two remaining conserved cysteines do not coordinate additional iron-containing species. In addition, we show that while wild-type

RlmN bears the C355 mCys residue in its as-isolated state, RlmN that is either engineered to lack the [4Fe–4S] cluster by substitution of the coordinating cysteines, or isolated from Escherichia coli cultured under iron-limiting conditions, does not bear a

C355 mCys residue. Reconstitution of the [4Fe–4S] cluster on wild-type apo RlmN followed by addition of SAM results in rapid production of S-adenosylhomocysteine

(SAH) and the mCys residue, while treatment of apo RlmN with SAM affords no observable reaction. These results indicate that in Cfr and RlmN, SAM bound to the unique iron of the [4Fe–4S] cluster displays two reactivities. It serves to methylate C355 of RlmN (C338 of Cfr), or it serves to generate the 5’-deoxyadenosyl 5’-radical, required for substrate-dependent methyl synthase activity.

3.2 Introduction

The simple addition of a methyl group to an acceptor molecule is one of the most basic and routine reactions in the cell. Although methyl groups can emanate from several complex cofactors (e.g. methylcobalamin and methylene- and methyltetrahydrofolate), the vast majority derive from the more understated molecule, S-adenosyl-L-methionine

(SAM).(1-3) Seminal model and enzymatic studies on SAM-dependent methyl-transfer reactions argue for a polar process, in which an appropriate nucleophile attacks the methyl group of SAM with concomitant elimination of S-adenosyl-L-homocysteine

(SAH). (4-6) Until recently, a direct SN2 displacement was the only means by which methyl groups derived from SAM were thought to be appended onto acceptor atoms.(7,

8) However, RlmN and Cfr catalyze methyl transfer as well as methyl synthesis via radical-dependent methylene transfer to electrophilic carbon atoms.(7, 8)

Escherichia coli (Ec) RlmN and Staphylococcus aureus (Sa) Cfr share 33% sequence identity and modify the same nucleotide in 23S rRNA, adenosine 2503 (A2503).

Although RlmN and Cfr act preferentially on naked rRNA,(9) A2503 resides ultimately in the peptidyltransferase center of the 50S subunit of the bacterial ribosome.(10-13)

Methylation of C2 by RlmN is found throughout eubacteria (Figure 3-1).(14) Although this activity is nonessential, Ec mutants that lack it lose to wild-type (wt) Ec in co-growth competition experiments.(15) Cfr, which is evolutionarily related to RlmN, also can catalyze methylation of C2 of A2503; however, C8 is its preferred target (Figure 3-

1).(16)

100 Methylation of C8 confers bacterial resistance to multiple classes of antibiotics, including phenicols, lincosamides, oxazolidinones, pleuromutilins, and streptogramin A, as well as the macrolides josamycin and spiramycin.(17) Recently, it has been shown that point mutations in the rlmN gene in Sa also decrease susceptibility to linezolid,(18, 19) a synthetic oxazolidinone antibiotic used to treat methicillin resistant Sa (MRSA) and infections from other human pathogens.(20, 21) Cfr and RlmN are members of the radical SAM (RS) superfamily, enzymes that cleave SAM reductively to a 5’- deoxyadenosyl 5’-radical (5’-dA•).(22-25) The common feature of RS enzymes is the use of the 5’-dA• to abstract key substrate hydrogen atoms. Cleavage of SAM requires an electron, which is supplied by a reduced [4Fe–4S]+ cluster. The Cys residues that coordinate this essential [4Fe–4S] cluster typically reside in a CxxxCxxC motif,(25) although exceptions have been reported.(26-28)

NH3 NH3 NH3 N N N N N N H3C N N N N CH N N CH O O 3 O 3 O RlmN O Cfr O 2 SAM 2 SAM O OH O OH O OH

Figure 3-1. Reactions catalyzed by RlmN and Cfr in vivo.

Our recent mechanistic studies of RlmN and Cfr have unraveled an unprecedented strategy for SAM-dependent methylation of an sp2-hybridized carbon atom.(7) When

RlmN or Cfr was incubated with S-adenosyl-L-[methyl-d3]methionine (d3-SAM) or unlabeled SAM under single turnover conditions, the isotopic composition of the methyl

101 group incorporated at C2 (RlmN) or C8 (Cfr) of the target adenosine nucleotide did not always reflect that of the methyl donor in the reaction, but rather the isotopic composition of L-methionine in the growth media of the Ec used for overproducing the protein.

SAM 5'-dA 5'-dA-H R R NH2 NH2 S+ B- N N B- N N CH3 H N N H N N BH BH S355 S355 CH2 H S355 CH2

23S rRNA 23S rRNA

SH118 S118 S118

NH2 NH2 NH2 BH N B- N N N B- N N H H N N S355 C N N N N H S355 C S355 C 2 H2 H BH BH 2 BH e- S118 23S rRNA 23S rRNA 23S rRNA S118 S118

NH2 NH2 - BH N N B N N N N S355 H2C N S355 H2C N HB- H BH

23S rRNA S118 23S rRNA S118 Figure 3-2. Mechanism proposed by Grove et al for the RlmN reaction.(7)

Consistent with this observation, analysis of RlmN by high-resolution mass spectrometry(7) and X-ray crystallography(29) showed the as-isolated (AI) protein to bear a methylcysteinyl (mCys) residue at C355. Moreover, methylation always involved incorporation of one hydrogen atom from the mCys residue into 5’-dA, indicating that the

5’-dA• activates the methyl group for radical addition rather than abstracts a hydrogen atom from the nucleotide to be modified. Cleavage of an ensuing methylene-bridged

102 protein–nucleic acid crosslink was proposed to be catalyzed by disulfide-bond formation—with participation of a second absolutely conserved Cys residue (C118 in

RlmN; C105 in Cfr), as shown in Figure 3.2.(7)

A related study by Yan et al. also found that the 5’-dA• radical does not abstract a hydrogen atom from the substrate nucleotide. In that study, however, the authors failed to recognize the initial transfer of a methyl group from SAM to C355 of RlmN (C338 of

Cfr), and therefore proposed that the 5’-dA• abstracts a hydrogen atom from the methyl group of a second, simultaneously bound, SAM molecule, activating it for radical addition to the target nucleotide.(8)

Figure 3-3. Active site structure of RlmN with SAM present.

The recent X-ray crystal structure of RlmN with SAM bound (2.05 Å)(29) is consistent with the mechanism proposed by Grove et al.(7) The structure shows C355 to reside in a

103 flexible loop containing residues 350–358, which are strictly conserved (Figure 3-3). This loop is disordered in the absence of SAM, but is visible in the RlmN+SAM structure.

Only one SAM-binding site was identified in the structure, wherein SAM is coordinated to the unique iron of the [4Fe–4S] cluster via its -amino and -carboxy groups, as is observed for other structurally characterized RS proteins.(30) Although it was speculated that the related protein, Cfr, contains two Fe/S clusters,(31) this stoichiometry is not supported by the structure of RlmN. The structure shows C355 to be S-methylated, and to be located ~6 Å from the sidechain of C118 and C5’ of SAM. This arrangement is consistent with disulfide-bond formation between C355 and C118 to resolve the protein– nucleic acid covalent adduct as well as abstraction of a hydrogen atom from the C355 mCys by the 5’-dA• as proposed by Grove et al.(7) The absence of any other obvious

SAM binding site led to the proposal that SAM coordinated to the [4Fe–4S] must exhibit dual functionality; it acts as both a methylating agent and as the source of the 5’-dA• intermediate.(29) In this work we use Mössbauer spectroscopy in concert with quantitative analyses of iron and sulfide to show that both RlmN and Cfr contain only one [4Fe–4S] cluster, which is coordinated by cysteines in a canonical CxxxCxxC motif.

This finding implies that the two remaining conserved Cys residues do not coordinate

Fe/S species and are free to participate in other modes of catalysis. In addition, we overproduce RlmN in Ec under iron-limiting conditions and show that the AI protein is not S-methylated at C355 and does not catalyze methylation of C355 when incubated with SAM. Reconstitution of the protein with iron and sulfide followed by addition of

SAM results in rapid methylation of C355, consistent with the proposal that SAM coordinated to the [4Fe–4S] cluster has a dual function.

104 3.3 Materials and Methods.

Construction of CysAla variants of RlmN. Single CysAla substitutions at C118,

C355, and the C125/129/132 triple variant were generated using the primers listed in

Table S1 in conjunction with the Stratagene QuikChange II kit as described previously.(32) Bold uppercase letters indicate the altered codon. Underlined bases for primers used to construct the triple variant indicate complementary overhangs.

Table 3-1. Primers for constructing RlmN variants Primer Sequence Forward 5’- cgaccgtgccacgctcGCCgtctcttcgcaggtgggg -3’ RlmN C118A Reverse 5’- ccccacctgcgaagagacGGCgagcgtggcacggtcg -3’ Forward 5’-ggtgatgatatcgatgctgccGCTggtcagttggcgggcg - 3’ RlmN C355A Reverse 5’- cgcccgccaactgaccAGCggcagcatcgatatcatcacc -3’ 5’-ccgtgccacgctctgcgtctcttcgcaggtggggGCTgcgctggagGCT Forward RlmN aaattcGCTtc-3’ C125/129/132A 5’-gcaggttgcggttaaagccctgctgggcggtggaAGCgaatttAGCctc Reverse cagcgcAGCcc-3’

Gene expression and protein overproduction. Plasmids encoding wild-type (wt) or mutant yfgB (encodes RlmN) or cfr genes were transformed into BL-21(DE3) along with plasmid pDB1282 as previously described.(7) Gene expression and protein overproduction were conducted as previously described.(7)

Production of apo RlmNwt. BL-21(DE3) transformed with pRlmNWt was plated onto a kanamycin (50 µg/mL) -containing LB plate. A single colony was used to inoculate a

Luria-Bertani (LB) starter culture, which was grown to stationary phase. M9 minimal media flasks (4 L), supplemented with 400 µL 10,000 micronutrient mix, were inoculated with 40 mL of the starter culture. Flasks were shaken at 210 rpm and 37 °C.

105

At an OD600 nm of 0.3, 200 mL 10 amino acid mix was added and shaking was continued. At an OD600nm of 0.5, 1,10-phenanthroline was added to the media to a final concentration of 75 µM. Flasks were shaken for an additional 30 min before being cooled to ~18 °C in an ice bath. Expression was induced with 100 µM (final concentration) IPTG, and the cultures were shaken for 18 h at 18 °C. Cell paste was harvested by centrifugation at 10,000 g and flash frozen with liquid N2. The cell paste was stored in a liquid N2 dewar until purification.

Purification of RlmNwt and RlmN variants. All purification steps were carried out in a

Coy anaerobic chamber. In a typical purification, 20 g cell paste is resuspended in 150 mL lysis buffer [50 mM HEPES, pH 7.5, 300 mM KCl, 2 mM imidazole, 10 mM 2- mercaptoethanol (BME)] containing lysozyme at a final concentration of 1 mg/mL. After stirring at room temperature for 50 min, the solution is placed in an ice bath, cooled to ~4

°C, and subjected to ten 40 s bursts of sonic disruption (30% output) with intermittent pausing for ~8 min to maintain a temperature less than 8 °C. The lysate is centrifuged for

1 h at 50,000 g and 4 °C. The resulting supernatant is loaded onto a column of Talon

(Clontech) Co2+ resin equilibrated in lysis buffer for purification by immobilized metal affinity chromatography (IMAC). The column is washed twice with 100 mL lysis buffer

(50 mM HEPES, pH 7.5, 300 mM KCl, 2 mM imidazole, 10 mM BME) before eluting with 50 mL elution buffer (50 mM HEPES, pH 7.5, 300 mM KCl, 10 mM BME, 300 mM imidazole, 10% glycerol). Fractions displaying significant brown color are pooled and concentrated by ultrafiltration using an Amicon Centricon (Millipore; Billerica, MA) with a YM-10 membrane. The protein is exchanged into final buffer (10 mM HEPES, pH 7.5, 500 mM KCl, 5 mM dithiothreitol (DTT), and 10% glycerol) using a PD-10

106 column (GE Biosciences). RlmN purified through this sequence is termed “as-isolated”

(AI). Reconstitution of AI RlmNwt or variants is carried out in 100 mM HEPES, pH 7.5,

500 mM KCl, 100 mM DTT, and 10% glycerol, with a protein concentration of 50 µM.

A 50 mL solution of the protein is incubated on ice for 30 min before 500 µM FeCl3 (or

57 FeSO4) is slowly added over 1-2 min with gentle stirring. The protein is incubated for an additional 60 min before addition of 600 µM (final concentration) Na2S in 6 equal portions over a period of 3 h. The protein is incubated further for ~15 h before being subjected to centrifugation to remove particulate matter. The supernatant is concentrated by ultrafiltration and exchanged into final buffer using a PD-10 column. The protein is then subjected to molecular-sieve chromatography on a HiPrep 16/60 S-200 column equilibrated in final buffer to afford purified holo RlmN. RlmN purified through this sequence is termed “reconstituted” (RCN).

Table 3-2. UV–vis characteristics and Fe and S2- content of wt and variant RlmN or Cfr proteins

A281/A400 Iron per sulfide per ratioa polypeptide polypeptide AI RC AI RC AI RC RlmNwt 4.11 4.06 4.0±0.1 4.1±0.3 5.0±0.2 5.2±0.7 b apo RlmNwt ND 4.7 0.04±0.001 2.2±0.1 0.07±0.01 1.2±0.3 b b b RlmNC125-129-132A 2.25 ND 0.8±0.1 ND 1.1±0.3 ND c b b b b RlmNC118A ND ND 4.0±0.2 4.6±0.1 ND ND b b RlmNC355A 3.93 4.07 4.1±0.1 4.1±0.1 ND ND Cfrwt 2.99 3.6 4.5±0.2 3.8±0.1 6.0±0.3 4.7±0.4 a For Cfr, the ratio is taken from A278/A400 bNot determined cIt was previously shown that the C118A variant binds RNA, therefore the ratio was not determined.

107 Purification of Cfr. The purification of Cfr is conducted essentially as described above, with the exception that 10% glycerol is used in the lysis buffer and 25% glycerol is used in the elution and final buffer.

Trypsin digestion of RlmN and data acquisition by LC-Q-Tof: In-gel trypsin digestions of RlmN samples were carried out as previously described.(7) Mass spectra were acquired on a quadrupole time-of-flight instrument (Q-Tof Premier; Waters, Milford,

MA) with a nano-LC front end (nanoAcquity UPLC; Waters). Samples (2-5 µL) were loaded on an Acquity Symmetry C18 trapping column (180 µm  20 mm, 5 µm) (Waters) at 5 µL/min for 3 min and separated on an Acquity BEH 130 C18 analytical column (100

µm  100 mm, 1.7 µm) (Waters) held at 35 °C. A 0.4 µL/min gradient was delivered by the binary pump over 60 min and consisted of 3% - 40% acetonitrile in water containing

0.1% formic acid, and was followed by a 5 min wash with 95% acetonitrile, 0.1% formic acid. A 100 fmol/µL solution of Glu-fibrinopeptide (exact mass 1569.6694) in 50% aqueous methanol containing 0.1% formic acid was used as a lock mass reference. The

Glu-fibrinopeptide solution was infused at 0.4 µL/min. The instrument was operated in the positive-ion, V-optics mode with a capillary voltage of 3.25 kV and a source temperature of 90 °C. Tandem mass spectra were acquired using data-dependent acquisition. Following a 1 sec MS scan over a 300 – 1990 m/z range, the peaks were selected for MS2 from the include list, and then the 4 most intense peaks with charge states 2, 3, 4, and 5 were selected in the absence of specified peaks. The ions were selected within a ± 0.3 m/z window from the include list and a ± 3 m/z window for other ions; their tandem mass spectra were acquired, and these ions were then excluded from acquisition for 180 s. Tandem mass spectra were acquired over a 90 – 1990 m/z range

108 with a collision energy ramp from 15 to 60 V. All data were processed using

ProteinLynx Global Server 2.1 software (Micromass).

Trypsin digestion of RlmN and data acquisition by LC-MS: Digestion of RlmN samples were carried out as above. Samples were separated on a Agilent Technologies

Zorbax Rapid Resolution XBD-C18 column (4.6 mm  50 mm, 1.8 µm particle size) equilibrated in 95% solvent A (0.1 % formic acid) and 5% solvent B (acetonitrile) with a flow rate of 0.5 mL/min. After injection, solvent conditions remained constant for 8 min before a gradient of 5–38% B was applied over 32 min, and then a sharp increase from

38–80% B was applied from 40 to 44 min. The column was then returned to 5 % B from

44 to 49 min with a re-equilibration time of 6 min before subsequent sample injections.

An Agilent Technologies 6410 QQQ mass spectrometer was operated in positive mode using only the MS2 scan function. The MS2 scan range was set from m/z of 400 – 2000 with data collection at every 0.1 amu. After collection of data, EIC were generated using the corresponding Masshunter Qualitative Analysis software to extract m/z values corresponding to the carbamidomethyl containing peptide (m/z: 931.0) or the mCys containing peptide (m/z: 909.6) with mass tolerances set to ± 0.1 amu. The [M+2H] peaks were the most abundant species in the mass spectra as the [M+1H] ion was almost nonexistent. The [M+3H] ion was also apparent, but to lower extent than the [M+2H] ion.

General Methods: High-performance liquid chromatography (HPLC) with detection by mass spectrometry (LC-MS) was conducted on an Agilent Technologies (Santa Clara,

CA) 1200 system, which was fitted with an autosampler for sample injection and coupled to an Agilent Technologies 6410 QQQ mass spectrometer. The system was operated with

109 the associated MassHunter software package, which was also used for data collection and analysis. Assay mixtures were separated on an Agilent Technologies Zorbax Rapid

Resolution XBD-C18 column (4.6 mm  50 mm, 1.8 µm particle size) equilibrated in

98% solvent A (40 mM aqueous ammonium acetate — titrated to pH 6.2 with glacial acetic acid — and 5% methanol) and 2% solvent B (acetonitrile). A gradient of 2–12% B was applied from 0.5 to 5 min, and then a gradient from 12–24% B was applied from 5 to

6.5 min. A subsequent increase to 50% B from 6.5 to 7.5 min was applied and then maintained at 50% for 0.5 min before returning to 2% B from 8 to 9 min. The column was allowed to re-equilibrate for 3 min under initial conditions before subsequent sample injections. Detection of substrates and products (Table S3) was performed using electrospray ionization in positive mode (ESI+) with multiple reaction monitoring

(MRM).

Table 3-3. Retention time and fragmentation products monitored by LC-MS

Retention Time Parent Ion* Product Ion 1† Product Ion 2† SAH 3.6 min 385.4 (100) 136 (24) 134 (24) Tryptophan (IS) 5.1 min 188 (130) 146.1 (10) 118 (21) 5’-dA 5.9 min 252.1 (90) 136 (13) 119 (50) m8A 6.3 min 282.1 (90) 150 (15) 106 (60) 8 d2-m A 6.3 min 284.1 (90) 152 (15) 108 (60)

*Respective fragmentor voltages are in parenthesis. †Respective collision energies are in parenthesis.

SAH production by apo RlmNwt. Reactions contained in a total volume of 100 µL: 100 mM Tris-HCl, pH 8.4, 10 mM MgCl2, 1.5 mM SAM or d3-SAM and 150 µM apo

RlmNwt, or apo RlmNwt reconstituted (apo RlmNwtRCN). In assays containing the

F/FR/NADPH reducing system, their final concentrations were 200 µM, 20 µM, and 2 mM respectively. Reaction mixtures were incubated in the absence of SAM (or d3-SAM)

110 for 5-10 min at ambient temperature, and a 10 µL aliquot was removed (t=0) and added to 10 µL of a solution containing 100 mM H2SO4 and 100 µM L-tryptophan (internal standard) to yield a final internal standard (IS) concentration of 50 µM. Upon addition of

SAM or d3-SAM, additional aliquots were removed at designated times and treated as described above. All final concentrations were multiplied by a dilution factor of 2 to calculate the original concentrations in the assay mixtures. Standard curves were generated with SAH purchased from Sigma-Aldrich.

Turnover of apo RlmNwtRCN. Reactions contained in a total volume of 60 µL: 100 mM Tris-HCl, pH 8.4, 10 mM MgCl2, 1.5 mM SAM or d3-SAM, 30 µM 771-mer RNA substrate, the F/FR/NADPH reducing system, and 150 µM apo RlmNwtRCN. Reaction mixtures lacking NADPH were incubated for 5-10 min at 37 °C, and 10 µL aliquots were removed (t=0) and added to 10 µL of a solution containing 100 mM H2SO4 and 100 µM

L-tryptophan (IS) to yield a final IS concentration of 50 µM. Reactions were initiated by addition of 2 mM NADPH and incubated for appropriate times before being quenched as described above. The samples were subjected to the RNA digestion protocol and analyzed by LC-MS as described above. All traces are normalized to the intensity of the

IS peak.

RNA digestion protocol. RNA was digested as previously described with slight modifications.(33) One volume of 200 mM ammonium acetate (pH 5) was added to acid- quenched time points to titrate the solution to a pH value of 4 – 5. P1 nuclease (1 U) was added and the reactions were incubated for ~4 h at 45 °C. One volume of 1 M NH4HCO3

(pH 8) was added to each reaction along with 5 U of alkaline phosphatase, and the reactions were incubated at 37 °C for an additional 2 h. The resulting precipitate was

111 removed by centrifugation and the supernatant was dried to completeness in vacuo. The pellet was dissolved in ddH2O to the original time-point volume and analyzed by the above LC-MS method.

EPR spin quantification of Mössbauer samples. All X-band EPR spectra where collected at 13 K with the following parameters: power, 101 µW; modulation amplitude,

10 G; modulation frequency, 100 kHz; center field, 3400 G; sweep width, 2000 G; time constant, 163.84 s; and conversion time, 163.84 s. A 1 mM Cu-EDTA standard was used to determine spin concentrations of the samples. Its spectrum was recorded under conditions identical to those used for recording samples. Protein concentration in each sample was as follows: AI RlmNWt, 945 µM; AI CfrWt, 300 µM; AI RlmNC125A-C129A-

C132A, 540 µM; AI RlmNC118A, 472 µM; AI RlmNC355A, 569 µM.

3.4 Results and Discussion

To determine the stoichiometry and configuration of iron-containing species associated with Cfr and RlmN, both proteins were purified from Ec cultured in minimal media supplemented with 57Fe, and then analyzed by Mössbauer spectroscopy. The UV–visible spectra of both AI proteins are consistent with the presence of [4Fe–4S] clusters (Figure

3-4), in agreement with previous studies.(9, 31) The 4.2-K/53-mT spectrum of as-isolated

(AI) wild-type (wt) RlmN (RlmNwt) (Figure 3-5A) is dominated by a broad quadrupole doublet with parameters typical of [4Fe-4S]2+ clusters [isomer shift () of 0.44 mm/s and quadrupole splitting parameter (EQ) of 1.14 mm/s], and accounts for 93% of the total intensity. The weak absorption at 0.6 mm/s is most likely associated with a small

112 amount (3%) of [2Fe-2S]2+ clusters. Together with the ratio of 4.0 Fe per polypeptide, determined by quantitative analyses of acid-labile iron and sulfide (Table 3-2), the

Mössbauer spectrum reveals a stoichiometry of 0.9 [4Fe-4S] clusters per AI RlmNwt.

Reconstitution of AI RlmNwt with additional Fe and sulfide (RCN RlmNwt) does not result in an increase of Fe and sulfide associated with the protein (4.1 Fe per polypeptide). Consistent with this observation, the 4.2-K/53-mT Mössbauer spectrum of

RCN RlmNwt (Figure 3-5B) is virtually identical to that of AI RlmNwt and reveals that

RCN RlmNwt harbors 1.0 [4Fe-4S] cluster. This cluster is that coordinated by C125,

C129, and C132, the indicated residues in the canonical CxxxCxxC RS motif.

The above cluster stoichiometry is corroborated by studies on an RlmN variant, in which C125, C129, and C132 have been changed to Ala (RlmNC125A-C129A-C132A), non-

113 coordinating residues. AI RlmNC125A-C129A-C132A has only 0.8 Fe per polypeptide.

Spectroscopic studies reveal an upper limit of 0.02 [4Fe–4S] clusters per protein, with the majority of Fe being unspecifically bound. Therefore, we conclude that RlmNC125A-C129A-

C132A does not harbor an additional Fe/S cluster.

In addition to C125, C129, and C132, RlmN has two additional, strictly conserved, cysteines that are absolutely required for complete turnover: C118 and C355.(34) To show that they do not ligate Fe/S species, CysAla variants of those residues

(RlmNC118A and RlmNC355A) were engineered and studied with a combination of biochemical and spectroscopic methods. RlmNC118A and RlmNC355A contain 4.0–4.6 Fe per polypeptide in both their AI and RCN forms (Table 3-2), and display spectroscopic features and cluster stoichiometries that are virtually identical to those of RlmNwt,

114 supporting the presence of one [4Fe–4S] cluster per polypeptide (Figures 3-8) ligated by

C125, C129, and C132.

The 4.2-K/53-mT spectrum of AI Cfrwt (Figures 3-5C) is dominated by a broad

2+ quadrupole doublet indicative of [4Fe-4S] clusters ( = 0.44 mm/s and EQ = 1.10 mm/s), which accounts for 86% of the total intensity. In addition, the spectrum of the sample exhibits a small amount (15% of total Fe) of a broad, poorly resolved, feature that extends from –2 to 2 mm/s. Because an identical EPR sample does not show any significant signals associated with features of Fe/S clusters with S = 1/2 ground state

(Figure 3-6), we assign the broad feature to unspecifically bound iron. Together with the ratio of 4.5 Fe per polypeptide (Table 3-2), the Mössbauer spectrum reveals a stoichiometry of 1.0 [4Fe-4S] cluster per AI Cfrwt.

Figure 3-6. EPR spectra of RlmN and Cfr Mössbauer samples.

Reconstitution of AI Cfrwt with additional Fe and sulfide followed by purification

(RCN Cfrwt) results in a slight decrease of Fe. The 4.2-K/53-mT Mössbauer spectrum of

115

RCN Cfrwt (Figure 3-5D) can be simulated with a quadrupole doublet with identical parameters, and accounts for 98% of the total intensity. Thus, the results suggest that

RCN Cfrwt harbors 1.0 [4Fe-4S] cluster, which is coordinated by C112, C116, and C119.

We therefore conclude, that in contrast to radical SAM proteins that catalyze methylthiolation, which contain two distinct Fe/S clusters,(22, 35, 36) RlmN and Cfr only contain the cluster housed in the canonical CxxxCxxC motif.

A 0.30 B 1.4 1.2 0.30 RlmN C118A Ai RlmN C355A Ai 1.2 RlmN C118A Rc 0.25 RlmN C355A Rc 1.0 0.25 1.0 0.20 0.8 0.20 0.8 0.15 0.6 0.15

0.6

Absrobance Absrobance 0.10 0.10 0.4 0.4

0.2 0.05 0.2 0.05

0.0 0.00 0.0 0.00 300 400 500 600 700 300 400 500 600 700 Wavelength (nm) Wavelength (nm)

Figure 3-7. UV-visible spectra of A) AI RlmNC118A (10 μM, black trace, left Y-axis) and AI RlmNC355A (6 μM, red trace, right Y-axis); and B) RCN RlmNC118A (12 μM, black trace, left Y- axis) and RCN RlmNC355A (7 μM, red trace, right Y-axis). The A280/A400 ratios for RlmNC355A are given in Table S2.

While previous studies have firmly established one role for the [4Fe–4S] cluster in the unique methylation reactions catalyzed by RlmN and Cfr,(7-9, 29) — reductive cleavage of SAM to generate the 5’-dA• intermediate — the role of the [4Fe–4S] cluster, if any, in the first step of the reaction — methylation of a conserved cysteine (C355 in RlmN and

C338 in Cfr) — has not been addressed. Interestingly, when the RlmNC125A-C129A-C132A

116 triple variant was analyzed by ESI+-MS after alkylation with iodoacetamide and digestion with trypsin, the peptide containing C355 was found exclusively acetylated, indicating that it did not bear a mCys residue (Figure 3-9B).

Figure 3-8. 4.2-K/53-mT Mössbauer spectra of AI RlmNC118A (A), AI RlmNC355A (B), RCN RlmNC118A (C), and RCN RlmNC355A (D). Experimental spectra are shown as vertical bars. The solid lines are quadrupole doublet simulations ( = 0.44 mm/s and EQ = 1.14 mm/s) accounting for 89% (A), 94% (B), 95% (C), and 97% (D) of the total intensity, indicating 0.9 (A), 1.0 (B), 1.1 (C), and 1.0 (D) [4Fe–4S] clusters per polypeptide (see Table S2).

This observation contrasts with results obtained for RlmNwt, wherein almost 90% of the AI protein contained a mCys residue, and very little of it was alkylated by iodoacetamide during preparation for analysis by mass spectrometry (Figure 2A).(7) To assess the effect of the iron–sulfur (Fe/S) cluster on generation of the mCys residue in

RlmN, apo RlmNwt was produced by expression of the Ec yfgB gene — which encodes

RlmN — in a modified M9 minimal medium containing 75 µM o-phenanthroline to

117 chelate available iron (see supporting information for experimental details).(37) A large fraction of RlmN was produced in inclusion bodies when expression was carried out in this manner. The soluble fraction was isolated anaerobically under conditions identical to those for isolation of holo RlmN, and a portion was set aside for reconstitution of its

[4Fe–4S] cluster with Fe and sulfide by standard methods.(38)

Figure 3-10A shows UV–vis spectra of apo RlmNwt isolated from Ec cultured in o- phenanthroline-containing medium (solid black line), and RlmNwt reconstituted with iron and sulfide (apo RlmNwtRCN) (solid red line). As can be observed, the telltale features of [4Fe–4S] clusters are significantly diminished in apo RlmNwt, consistent with the finding of 0.04 Fe and

0.07 sulfides per polypeptide (Table 3-2), but reappear in the apo RlmNwtRCN sample. Analysis of trypsin-digested apo RlmNwt by LC-MS showed that it did not bear a C355 mCys modification, consistent with the premise that SAM bound to the [4Fe–4S] cluster acts as a methylating agent (Figure 3-9E). To ensure that in vivo methylation of apo RlmNwt was not the factor that distinguished soluble protein from insoluble protein, apo RlmNwt inclusion bodies were also analyzed by LC-MS; fragments obtained from trypsin digests of the inclusion body fraction did not bear a C355 mCys modification (Figure 3-9D).

118

119

Figure 3-10. Characterization of apo RlmNwtRCN. A) UV–vis traces of apo RlmNwt (5 µM, solid black line) and apo RlmNwtRCN (12 µM, solid red line). B) Methyl transfer catalyzed by apo RlmNwt (150 µM) or apo RlmNwtRCN (150 µM). Production of SAH by apo RlmNwt (purple diamonds) or apo RlmNwtRCN (red circles) in the presence of 1.5 mM SAM; or apo RlmNwtRCN (150 µM) in the presence of 1.5 mM d3-SAM (black triangles). Error bars indicate one standard deviation from the average of three assays. C) Q-Tof MS analysis of tryptic peptides from apo

RlmNwtRCN after incubation in the presence of SAM (black trace) or d3-SAM (red trace). Indicated m/z values correspond to the +2 charge state.

Reconstitution of apo RlmNwt followed by further purification by size-exclusion chromatography afforded a protein containing 2.2 irons and 1.2 sulfides, respectively, per polypeptide (Table 3-2). This stoichiometry implies that 50% of the RCN protein contained a [4Fe–4S] cluster, given that its UV-vis spectrum does not contain telltale signatures of [2Fe–2S] clusters, and most, if not all, of the adventitiously bound iron was removed by molecular-sieve chromatography. When SAM was added to apo RlmNwt, methyl transfer to C355 did not occur, as evidenced by no significant production of SAH

(Figure 3-10B, purple diamonds). By contrast, when SAM or d3-SAM was added to apo

RlmNwtRCN, ~0.5 equiv of SAH was produced within the first time point (15 s).

Subsequent time points did not show increased amounts of SAH, suggesting that 50% of apo RlmNwt was reconstituted to its holo form (Figure 3-10B, red circles). Importantly, the quantity of SAH produced was directly proportional to the RlmN cluster stoichiometry, providing further evidence that the [4Fe–4S] cluster is required for methyl

120 transfer. In addition, apo RlmNwtRCN was capable of catalyzing time-dependent methylation of the target nucleotide in a 771-base synthetic RNA substrate (Figure 3-11)

40 20 A B 2 2 d -m Adenosine m Adenosine 3

30 15

20 10

10 5

Normalized Intensity Intensity Normalized Normalized Intensity

IS IS 0 0 5.0 6.0 7.0 8.0 5.0 6.0 7.0 8.0

Time (min) Time (min)

Figure 3-11. Time-dependent formation of m2A from an assay containing apo RlmNwtRCN. The enzyme was incubated with the 771-mer RNA substrate and 1.5 mM A) SAM or B) d3-SAM. Time points are 0, 0.5, 1, 10, and 30 min.

Analysis of apo RlmNwtRCN by LC–MS after incubating it with SAM reveals a peak exhibiting a mass-to-charge ratio (m/z) of 909.6 (+2 charge state) for the C355-containing peptide fragment obtained after trypsin digestion, indicating that it bears a methyl group

(Figures 3-9F and 3-10C, black trace). Moreover, the peak at m/z = 909.6 shifts to 911.2

(+2 charge state; thus the shift is ~1.5 units instead of the expected 3 for the +1 charge

121 state) when apo RlmNwtRCN is incubated with d3-SAM and similarly analyzed (Figure

3-10C, red trace). This modification is specific to C355 of RlmN, given that MS/MS analysis of the peptides shows that the y+ ion series is shifted by +15 for fragments that

+1 bear C355 (y11 and above) (Figure 3-12).(7)

4 1+ y11

1+ (10 Intensity

) y11 1163.6 3 6 3 1160.59 1+ 1+ y12 y12 4 2 1234.65 1+ 1231.62 1+ 1+ y14 1+ y y y 13 14 1420.71 13 1305.68

1 1302.66 1417.68 2

) Intensity (10 Intensity 0 0 1100 1200 1300 1400 1500 m/z

Figure 3-12. MS/MS of apo RlmNwt tryptic peptides after reconstitution and subsequent + incubation of the protein in the presence of SAM or d3-SAM. Selected y ions from collision induced dissociation (CID) of either the m/z = 909.6 (mCys peptide, M+2H) or the m/z = 911.2 (d3-mCys peptide, M+2H) species, indicating a difference of m/z = +3 for 1+ the sample treated with d3-SAM. Ion y11 corresponds to the sequence 355 1+ C GQLAGDVIDR, while ion y10 (not observed in this spectrum) corresponds to the 1+ sequence GQLAGDVIDR. The y11 ion displays m/z = 1160.59 (1163.6 for deuterated 1+ methyl group), corresponding to a m/z increase of 116.97 over that of the y10 ion, the mass of a methylCys residue.

The effect of C118 — another Cys residue that plays a key role in RlmN catalysis — on methylation of C355 was also assessed. As observed in Figure 2C, a large fraction of the AI C118A variant contains a mCys modification, consistent with this residue playing an insignificant role in SAM binding and methylation of C355. Last, the inclusion of the flavodoxin/flavodoxin reductase/NADPH reducing system had no effect on C355

122 methylation, indicating that methyl transfer to C355 does not require a reduced [4Fe–4S] cluster (Figure 3-13), and therefore most likely proceeds by a polar nucleophilic 100apoRlmN Rc w/ Fdr/Fdx/NADPH SAM displacement, as do all characterized SAM-dependent methyltransferases.(1) apoRlmN Rc w/ Fdr/Fdx/NADPH d3-SAM 80

40 SAH [µM] SAH

0 40 80 120 Time (s)

Figure 3-13. Methyl transfer catalyzed by apo RlmNwtRCN (150 µM) in the presence of the in vivo reducing system. Production of SAH by apo RlmNwtRCN in the presence of 1.5 mM SAM and the F/FR/NADPH reducing system (black triangles); or apo RlmNwtRCN in the presence of 1.5 mM d3-SAM and the F/FR/NADPH reducing system (green diamonds).

Herein we have shown that RlmN and Cfr catalyze two separate and distinct reactions, exploiting both polar and radical reactivities of SAM within a single polypeptide. Indeed, these proteins both possess methyltransferase and methylsynthase activities. They contain only one [4Fe–4S] cluster — ligated by cysteines in CxxxCxxC motifs — to which SAM binds via its -amino and carboxy groups.(29) In this conformation, methyl transfer to unmodified C355 of RlmN (C338 of Cfr) takes place rapidly. Release of the product,

SAH, permits binding of a second molecule of SAM to the exact same site; however,

123 methyl transfer to the reactive cysteine is now blocked. Upon binding of the RNA substrate and the addition of an electron to the [4Fe–4S]2+ cluster, SAM is now induced to fragment into methionine and the 5’-dA•, the latter initiating catalysis by abstracting a hydrogen atom from the mCys residue. Our studies show that methyl transfer to the target

Cys residue does not require the presence of substrate. Whether substrate binding accelerates this step is yet to be determined. However, our previous studies showed that substrate binding does indeed effect radical formation when the physiological reducing system is used to supply the requisite electron.(7) Excitingly, this mode of catalysis is a clever twist on the “principle of economy in the evolution of binding sites,” wherein

Nature evolves only a single substrate-binding site for reactions that involve two or more substrates with similar structural elements, and then mediates transfer of a group from one to the other by means of an intermediate enzyme–functional group covalent adduct.(39) The Cfr and RlmN reactions are the first recognized instances in which a single substrate-binding site activates one molecule both for polar and radical-based chemistry

124 3.5 Acknowledgements

We thank Tatiana N. Laremore of the PSU MS facility for collecting high-resolution MS data.

125 3.6 References

1. Frey, P. A., and Hegeman, A. D. (2007) Enzymatic Reaction Mechanisms, Oxford

University Press, New York.

2. Markham, G. D. (2010) S-adenosylmethionine, In Encyclopedia of Life Sciences,

John Wiley & Sons, Inc.

3. Silverman, R. B. (2002) The Organic Chemistry of Enzyme-Catalyzed Reactions,

Academic Press, New York.

4. Hegazi, M. F., Borchardt, R. T., and Schowen, R. L. (1979) a-Deuterium and

carbon-13 isotope effects for methyl transfer catalyzed by catechol-O-methyl-

transferase. SN2-like transition state., J. Am. Chem. Soc. 101, 4359–4365.

5. Iwig, D. F., Grippe, A. T., McIntyre, T. A., and Booker, S. J. (2004) Isotope and

elemental effects indicate a rate-limiting methyl transfer as the initial step in the

reaction catalyzed by Escherichia coli cyclopropane fatty acid synthase,

Biochemistry 43, 13510-13524.

6. Woodard, R. W., Tsai, M.-D., Floss, H. G., Crooks, P. A., and Coward, J. K.

(1980) Sterochemical course of the transmethylation catalyzed by catechol O-

methyltransferase, J. Biol. Chem. 255, 9124–9127.

7. Grove, T. L., Benner, J. S., Radle, M. I., Ahlum, J. H., Landgraf, B. J., Krebs, C.,

and Booker, S. J. (2011) A radically different mechanism for S-

adenosylmethionine-dependent methyltansferases, Science 332, 604–607.

126 8. Yan, F., and Fujimori, D. G. (2011) RNA methylation by radical SAM enzyme

RlmN and Cfr proceeds via methylene transfer and hydride shift, Proc. Natl.

Acad. Sci. U S A 108, 3930–3934.

9. Yan, F., LaMarre, J. M., Röhrich, R., Wiesner, J., Jomaa, H., Mankin, A. S., and

Galoníc Fujimori, D. (2010) RlmN and Cfr are radical SAM enzymes involved in

methylation of ribosomal RNA, J. Am. Chem. Soc. 132, 3953-3964.

10. Ban, N., Freborn, B., Nissen, P., Penczek, P., Grassucci, R. A., Sweet, R., Frank,

J., Moore, P. B., and Steitz, T. A. (1998) A 9 Å resolution x-ray crystallographic

map of the large ribosomal subunit, Cell 93, 1105–1115.

11. Harms, J., Schluenzen, F., Zarivach, R., Bashan, A., Gat, S., Agmon, I., Bartels,

H., Francheschi, F., and Yonath, A. (2001) High-resolution structure of the large

ribosomal subunit from a mesophilic eubacterium, Cell 107, 679–688.

12. Schuwirth, B. S., Borovinskaya, M. A., Hau, C. W., Zhang, W., Vila-Sanjurjo, A.,

Hoton, J. M., and Cate, J. H. D. (2005) Structures of the bacterial ribosome at 3.5

Å resolution, Science 310, 827–834.

13. Selmer, M., Dunham, C. M., Murphy, F. V., IV, Weislbaumer, A., Petry, S.,

Kelley, A. C., Weir, J. R., and Ramakrishnan, V. (2006) Structure of the 70S

ribosome complexed with mRNA and tRNA, Science 313, 1935–1942.

14. The_UniProt_Consortium. (2007) The Universal Protein Resource (UniProt),

Nucleic Acids Res. 35, D193-D197.

15. Toh, S.-M., Xiong, L., Bae, T., and Mankin, A. S. (2008) The methyltransferase

YfgB/RlmN is responsible for modification of adenosine 2503 in 23S rRNA, RNA

14, 98–106.

127 16. Giessing, A. M. B., Jensen, S. S., Rasmussen, A., Hansen, L. H., Gondela, A.,

Long, K. S., Vester, B., and Kirpekar, F. (2009) Identification of 8-

methyladenosine as the modification catalyzed by the radical SAM

methyltransferase Cfr that confers antibiotic resistance in bacteria, RNA 15, 327–

336.

17. Smith, L. K., and Mankin, A. S. (2008) Transcriptional and translational control

of the mlr operon, which confers resistance to seven classes of protein synthesis

inhibitors, Antimicrob. Agents Chemother. 52, 1703–1712.

18. Gao, W., Chua, K., Davies, J. K., Newton, H. J., Seemann, T., Harrison, P. F.,

Holmes, N. E., Rhee, H.-W., Hong, J.-I., Hartland, E. L., Stinear, T. P., and

Howden, B. P. (2010) Two novel point mutations in clinical Staphylococcus

aureus reduce linezolid susceptibility and switch on the stringent response to

promtoe persistent infection, PLoS Pathog. 6, e1000944.

19. LaMarre, J. M., Howden, B. P., and Mankin, A. S. (2011) Inactivation of the

indigenous methyltransferase RlmN in Staphylococcus aureus increases linezolid

resistance, Antimicrob. Agents Chemother. doi:10.1128/AAC.00183-11.

20. Diekema, D. J., and Jones, R. N. (2000) Oxazolidinones: A review, Drugs 59, 7–

16.

21. Diekema, D. J., and Jones, R. N. (2001) Oxazolidinone antibiotics, The Lancet

358, 1975–1982.

22. Booker, S. J. (2009) Anaerobic functionalization of unactivated C–H bonds, Curr.

Opin. Chem. Biol. 13, 58–73.

128 23. Frey, P. A., and Booker, S. J. (2001) Radical mechanisms of S-

adenosylmethionine-dependent enzymes, Adv. Protein Chem. 58, 1-45.

24. Frey, P. A., Hegeman, A. D., and Ruzicka, F. J. (2008) The radical SAM

superfamily, Crit. Rev. Biochem. Mol. Biol. 43, 63–88.

25. Sofia, H. J., Chen, G., Hetzler, B. G., Reyes-Spindola, J. F., and Miller, N. E.

(2001) Radical SAM, a novel protein superfamily linking unresolved steps in

familiar biosynthetic pathways with radical mechanisms: functional

characterization using new analysis and information visualization methods,

Nucleic Acids Res. 29, 1097-1106.

26. Chatterjee, A., Li, Y., Zhang, Y., Grove, T. L., Lee, M., Krebs, C., Booker, S. J.,

Begley, T. P., and Ealick, S. E. (2008) Reconstitution of ThiC in thiamine

pyrimidine biosynthesis expands the radical SAM superfamily, Nat. Chem. Biol.

4, 758–765.

27. Martinez–Gomez, N. C., and Downs, D. M. (2008) ThiC is an [Fe–S] cluster

protein that requires AdoMet to generate the 4-amino-5-hydroxymethyl-2-

methylpyrimidine moiety in thiamin synthesis, Biochemistry 47, 9054–9056.

28. McGlynn, S. E., Boyd, E. S., Shepard, E. M., Lange, R. K., Gerlach, R.,

Broderick, J. B., and Peters, J. W. (2010) Identification and characterization of a

novel member of the radical AdoMet enzyme superfamily and implications for

the biosynthesis of the Hmd hydrogenase active site cofactor, J. Bacteriol. 192,

595–598.

129 29. Boal, A. K., Grove, T. L., McLaughlin, M. I., Yennawar, N. H., Booker, S. J., and

Rosenzweig, A. C. (2011) Structural basis for methyl transfer by a radical SAM

enzyme, Science 332, 1089–1092.

30. Vey, J. L., and Drennan, C. L. (2011) Structural insights into radical generation

by the radical SAM superfamily, Chem. Rev. 111, 2487–2506.

31. Booth, M. P. S., Challand, M. R., Emery, D. C., Roach, P. L., and Spencer, J.

(2010) High-level expression and reconstitution of active Cfr, a radical-SAM

rRNA methyltransferase that confers resistance to ribosome-acting antibiotics,

Protein Expr. Purif. 74, 204–210.

32. Grove, T. L., Lee, K. H., St Clair, J., Krebs, C., and Booker, S. J. (2008) In vitro

characterization of AtsB, a radical SAM formylglycine-generating enzyme that

contains three [4Fe-4S] clusters, Biochemistry 47, 7523-7538.

33. Crain, P. F. (1990) Preparation and enzymatic hydrolysis of DNA and RNA for

Mass spectrometry, Methods Enzymol. 193, 782–790.

34. Kaminska, K. H., Purta, E., Hansen, L. H., Bujnicki, J. M., Vester, B., and Long,

K. S. (2010) Insights into the structure, function and evolution of the radical-SAM

23S rRNA methyltransferase Cfr that confers antibiotic resistance in bacteria,

Nuc. Acids. Res.

35. Atta, M., Mulliez, E., Arragain, S., Forouhar, F., Hunt, J. F., and Fontecave, M.

(2010) S-adenosylmethionine-dependent radical-based modification of biological

macromolecules, Curr. Opin. Struct. Biol. 20, 1–9.

36. Booker, S. J., Cicchillo, R. M., and Grove, T. L. (2007) Self-sacrifice in radical S-

adenosylmethionine proteins, Curr. Opin. Chem. Biol. 11, 543-552.

130 37. Parkin, S. E., Chen, S., Ley, B. A., Mangravite, L., Edmondson, D. E., Huynh, B.

H., and Bollinger Jr., J. M. (1998) Electron injection through a specific pathway

determines the outcome of oxygen activation at the diiron cluster in the F208Y

mutant of Escherichia coli ribonucleotide reductase Protein R2, Biochemistry 37,

1124-1130.

38. Cicchillo, R. M., Lee, K.-H., Baleanu-Gogonea, C., Nesbitt, N. M., Krebs, C., and

Booker, S. J. (2004) Escherichia coli lipoyl synthase binds two distinct [4Fe–4S]

clusters per polypeptide, Biochemistry 43, 11770-11781.

39. Frey, P. A. (1992) Nucleotidyltransferases and phosphotransferases:

Stereochemistry and covalent intermediates, In The Enzymes, 3rd Ed. (Sigman, D.

S., Ed.), Academic Press, Inc., San Diego.

Chapter 4

A Kinetically Competent Substrate Radical Intermediate in Catalysis by the Antibiotic Resistance Protein Cfr

This chapter was reproduced from “Tyler L. Grove, Jovan Livada, Erica L. Schwalm, Michael T. Green, Squire J. Booker, and Alexey Silakov. Accepted to Nature Chemical Biology.

132 4.1 Abstract

Until recently, biological alkylation reactions such as methylation were all presumed to proceed by polar mechanisms, involving the attack of a nucleophile onto an electrophilic alkylating agent. The methylation reactions catalyzed by the radical S- adenosylmethionine (SAM) enzymes, RlmN and Cfr, modify carbons 2 and 8, respectively, of adenosine 2503 (A2503) of bacterial 23S rRNA, carbon centers that are not well-suited to form carbanion-like intermediates capable of acting as nucleophiles in

SN2 displacements. These reactions proceed by a mechanism involving organic radicals generated by the reductive cleavage of SAM to a 5’-deoxyadenosyl 5’-radical (5’-dA•), and the synthesis of a methyl group onto the target carbon center. C8 methylation by Cfr confers bacterial resistance to an array of clinically important antibiotics that target the large subunit of the ribosome, including the synthetic oxazolidinone linezolid. The key element of the proposed mechanism for Cfr is the addition of a methylene radical — generated by a hydrogen-atom abstraction from the methyl group of an S-methylated conserved cysteine residue (mCys) — onto C8 of A2503 to form a protein–nucleic acid crosslinked species containing an unpaired electron. Herein we provide direct spectroscopic evidence for this intermediate, showing a spin-delocalized radical with maximum spin density at N7 of the adenine ring. Electron paramagnetic resonance and electron nuclear double resonance spectroscopies in concert with strategic isotopic labeling of the substrate reveal a change in hybridization at C8 from sp2 to sp3 as well as the attachment of the methyl carbon of the mCys residue to C8 of the nucleotide base. In

133 addition, we use rapid-freeze quench EPR to show that the radical forms and decays with rate constants that are consistent with the rate of formation of the methylated product.

134 4.2 Introduction

The rise of antibiotic-resistance among bacterial pathogens is becoming a global health crisis, and the inability to produce new antibiotics at a pace keeping up with the growing spread of resistance mechanisms, especially across bacterial species, threatens to undermine the world’s current arsenal of antibiotics (1, 2). Approximately one-half of all clinically relevant antibiotics target some aspect of bacterial ribosome function (3, 4). As with other classes of antibiotics, bacteria evolve and/or share mechanisms to evade those that bind to the ribosome, often involving nucleotide changes in the surrounding rRNA or amino acid changes in ribosomal proteins, as well as posttranscriptional and/or posttranslational modifications of these ribosomal components (3, 5). One such posttranscriptional modification is methylation of C8 of adenosine 2503 (A2503) of 23S rRNA (6-10) (Figure 4-1). This nucleotide resides in the peptidyltransferase center of active ribosomes and is essential for proper ribosome function (11-14). In bacteria,

A2503 carries a natural C2 modification, which is installed by the chromosomally encoded product of the rlmN gene (15, 16). This C2 modification is believed to enhance translational fidelity (17). By contrast, C8 methylation, catalyzed by the product of the cfr gene in Staphylococcus aureus (SA), is generally a plasmid-borne acquired activity that confers resistance to multiple classes of antibiotics, including phenicols, lincosamides, oxazolidinones, pleuromutilins, streptogramin A, and the macrolides josamycin and spiramycin (6-9).

Cfr activity was first identified in 2000 on a plasmid isolated from the animal pathogen

Staphylococcus sciuri (8). It has also been found in staphylococci that infect humans,

135 including methicillin-resistant Staphylococcus aureus (MRSA). In 2007, a chromosomally located cfr ortholog was identified in a strain of MRSA obtained from a patient in Colombia with fatal ventilator-associated pneumonia. This strain was also resistant to one of the newest and most promising antibiotics currently in use, the synthetic oxazolidinone linezolid (18, 19). Since the 2007 report, new cases of cfr- positive staphylococcal isolates from hospitalized patients in the United States (US) (20-

23), Spain (24, 25), Italy (26), Mexico (27), and Ireland (28) have been reported, suggesting this mechanism of antibiotic resistance is readily spreading, even across different staphylococcal species. More recently, cfr genes in human isolates of

Enterococcus faecalis(29) have been found, and is widespread in the order

Bacillales(30), suggesting the existence of a large environmental reservoir of this mode of antibiotic resistance. The global spread of this mode of antibiotic resistance warrants detailed analysis of all aspects of the Cfr reaction to prepare to fend off a looming and potentially disastrous world health crisis.

Figure 4-1. Mechanistic proposal for catalysis by Cfr. The numbering of the carbons in the adenosine ring is displayed in 1.

136 Cfr and RlmN belong to a rapidly growing class of metalloenzymes dubbed the radical

S-adenosylmethionine (SAM) superfamily. These enzymes employ [4Fe–4S] clusters as cofactors and SAM as either a cofactor or cosubstrate to generate a potent 5’- deoxyadenosyl 5’-radical (5’-dA•) by reductive cleavage of SAM, affording L- methionine as a coproduct. In almost all radical SAM (RS) enzymes studied, the role of the 5’-dA• is to abstract a target hydrogen atom (H•) from a carbon center of a noncovalently bound substrate to generate a substrate radical that undergoes further transformation (31-33). However, the associated homolytic bond-dissociation energies

(BDEs) for the sp2-hybridized carbon centers functionalized by RlmN (C2 of adenosine:

105 kcal/mol) and Cfr (C8 of adenosine: 113 kcal/mol) are even greater than that for methane (104 kcal/mol) (34, 35). Recent studies by Grove et al (36) and Yan and

Fujimori (37), however, showed that Cfr and RlmN employ a unique mechanism of catalysis, involving H• abstraction from the methyl moiety that is ultimately appended to the substrate with addition of the ensuing methylene radical to C8 or C2 of the adenine base. In the mechanism proposed by Grove et al for Cfr, which is supported by an X-ray crystal structure of the mechanistically analogous enzyme, RlmN with bound SAM, SAM binds in contact with the unique iron ion of the [4Fe–4S] cluster. Cysteine 338 attacks the activated methyl group of SAM, resulting in formation of a methylcysteinyl (mCys) residue and S-adenosylhomocysteine (SAH), which subsequently dissociates from the enzyme’s active site. Following binding of another molecule of SAM and reduction of the [4Fe–4S] cluster from the +2 to the +1 oxidation state, this second SAM molecule undergoes reductive cleavage to a 5’-dA•, which abstracts a H• from the mCys residue.

The resulting methylene radical adds to C8 of A2503 to generate a covalent protein–RNA

137 crosslink containing an unpaired electron (Figure 4- 1, Species 3). Upon loss of the electron and abstraction of the C8 proton by a general base, the covalent crosslink is resolved by disulfide-bond formation, involving a second strictly conserved cysteine

(C105), with partial return of the proton removed from C8 to the nascent methyl group

(36, 38). The key, but unusual, step in this proposed mechanism is the radical addition step that affords the covalent protein–RNA intermediate. The evidence for this crosslinked species is, however, indirect and consists of two observations. When the Cys residue purported to resolve the crosslink is substituted by an Ala residue (C118A) in

RlmN, the protein is isolated with covalently bound rRNA (36). More recent studies by

McCusker et al. using both the C118A and C118S RlmN variants show unambiguously that a covalent crosslink between C355 of RlmN and C2 of A2503 can be formed in vitro

(39). In in vivo studies of Cfr, the analogous variant (C105A) gave rise to a strong pause at A2503 in reverse transcription assays of the RNA substrate — as is seen when C8 of

A2503 undergoes methylation — but antibiotic resistance was not conferred, indicating a modification to this nucleotide that was not a methyl group (40). Herein we use continuous wave (CW) electron paramagnetic resonance spectroscopy (EPR) coupled with pulsed electron nuclear double resonance (ENDOR) spectroscopy to provide direct evidence for this novel nucleic acid radical intermediate in the Cfr mechanism, clearly showing a change in hybridization at C8 and attachment of the methyl carbon of the mCys residue to the nucleotide base. Moreover, rapid-freeze-quench EPR coupled with product analysis in similar single-turnover reactions provide evidence both for the chemical and kinetic competence of this unusual species.

138 Materials and Methods

Materials. [methyl-13C]-methionine, DNase I (from bovine pancreas), and S-

2 adenosylhomocysteine were purchased from Sigma–Aldrich and used as received. H8-

2 adenosine 5’-triphosphate (ammonium salt; 97% H8), was purchased from Cambridge

Isotopes and used as received. X-band quartz EPR tubes (3.8 mm) were purchased from

Wilmad-LabGlass. [methyl-13C]-SAM was synthesized as previously described (41).

Cloning of Staphylococcus aureus cfr Gene. The gene encoding cfr from S. aureus

(A5HBL2) was codon-optimized for expression in E. coli by GeneArt (Burlingame, CA), and supplied in the plasmid pMA. The gene was then amplified from this plasmid using the following primers: Cfr_For (5’- gcccggcatatgaactttaacaacaaaaccaaatatggc -3’) and

Cfr_Rev (5’- gcgcgaattccctggctgttctgataattgcc -3’). Cfr_For contained an NdeI restriction site (underlined), while Cfr_Rev contained an EcoRI restriction site (underlined). Primer

Cfr_Rev removes the native stop codon in the cfr gene, allowing the encoded protein to be produced with a C-terminal hexahistidine tag. The resulting PCR product was digested with NdeI and EcoRI and cloned into similarly digested pET-26b by standard methods. The final construct was verified by DNA sequencing at the Molecular Core

Facility (Penn State University) and designated pET26b-Cfr-Wt.

Synthesis and Purification of 155mer RNA and deu155mer (nucleotides 2454-2608 of

E. coli 23S rRNA). To generate a 155 nucleotide (nt) strand corresponding to nt 2454-

2608 of 23S rRNA, a PCR product was amplified from plasmid pKK3535 (36) using the following primers: Forward 5′- CGG-AAA-TTA-ATA-CGA-CTC-ACT-ATA-GGC-

TGA-TAC-CGC-CCA-AGA-GTT-CAT-ATC-G -3′; Reverse 5′- mCmCG-AAC-TGT-

139 CTC-ACG-ACG-TTC-TAA-ACC-3, where the underlined region indicates the T7 promoter. The product was isolated by agarose gel-electrophoresis and used in a large scale PCR (20 rxn  200 µL volume each). The resulting product was precipitated with

2.5 M sodium acetate (pH 5) and used in a reaction with T7 RNA polymerase under the following conditions in a final volume of 10 mL: 30 mM Tris-HCl, pH 8.0, 26 mM

MgCl2, 0.01 % Triton X-100, 20 mM DTT, 2 mM spermidine, 5 mM each NTP (dry powder was dissolved in water, neutralized with Tris base, and quantified by UV-visible spectroscopy), 5 ng/µL of PCR product, and hexahistidine-tagged T7 RNA polymerase

(5% of the final volume of reaction). Reactions were incubated at 37 °C for 1 h. 20 U of

DNase (NEB; Boston, MA) was added, and incubation at 37 °C was continued for 60 min. The precipitate was removed by centrifugation and the supernatant was separated from digested DNA and remaining NTPS by size-exclusion chromatography on an S-200 column equilibrated in 10 mM HEPES, pH 7.5, 50 mM KCl, 10 mM MgCl2, and 10% glycerol. Fractions containing the full transcript were pooled and concentrated using an

Amicon Centricon with a YM-10 membrane (Millipore; Billerica, MA). The final concentration was determined from an extinction coefficient (1.87 µM-1•cm-1) calculated from the sequence using the program DNAssist. Production of the deu155mer was

2 carried out exactly as above except that ATP was replaced with H8-ATP.

Purification of wt Cfrapo. All purification steps were carried out in a Coy anaerobic chamber. In a typical purification, 20 g cell paste is resuspended in 150 mL lysis buffer

[50 mM HEPES, pH 7.5, 300 mM KCl, 4 mM imidazole, 10 mM MgCl2, 10% glycerol,

10 mM 2-mercaptoethanol (BME)] containing lysozyme and DNase I at final concentrations of 1 mg/mL and 0.1 mg/mL, respectively. After stirring at room

140 temperature for 30 min, the solution is placed in an ice bath, cooled to ~4 °C, and subjected to six 40 s bursts of sonic disruption (30% output) with intermittent pausing for

~8 min to maintain a temperature less than 8 °C. The lysate is centrifuged for 1 h at

50,000 g and 4 °C. The resulting supernatant is loaded onto a column of Talon

(Clontech) Co2+ resin equilibrated in lysis buffer for purification by immobilized metal affinity chromatography (IMAC). The column is washed twice with 100 mL lysis buffer before eluting with 50 mL of elution buffer (50 mM HEPES, pH 7.5, 300 mM KCl, 10 mM BME, 300 mM imidazole, 10 mM MgCl2, 30% glycerol). Fractions displaying significant brown color are pooled and concentrated by ultrafiltration using an Amicon

Centricon (Millipore; Billerica, MA) with a YM-10 membrane. The protein is exchanged into final buffer (10 mM HEPES, pH 7.5, 500 mM KCl, 5 mM DTT, 10 mM MgCl2, and

25% glycerol) using a PD-10 column (GE Biosciences). Reconstitution of wt Cfrapo is carried out as previously described (42). The protein is then subjected to molecular-sieve chromatography on a HiPrep 16/60 S-200 column equilibrated in final buffer to afford wt

CfrapoRCN. Protein concentrations were determined by the method of Bradford (43) using a correction factor of 0.78 (36).

Assays. Assays monitoring SAH production by wt CfrapoRCN were conducted and analyzed as described in Chapter 3 (42).

EPR Sample Preparation and Measurement. All experiments were conducted on Cfr containing a C-terminal hexahistidine tag. All manipulations of the enzyme, including preparation of EPR and ENDOR samples, were conducted inside an anaerobic chamber from Coy Laboratories. Samples to be analyzed by EPR contained the following in a total volume of 150 µL: 50 mM Tris-HCl, pH 8.4, 10 mM MgCl2, 2 mM SAM or [methyl-

141

13 C]-SAM, 150 µM 155-mer RNA or deu155mer substrate, and 200 µM wt CfrapoRCN.

Samples were incubated for 5 min at 22 °C before being rapidly mixed with 3 µL of dithionite (114 mM in 1 M Tris-HCl). Reactions were placed in 3.8 mm quartz EPR tubes and incubated for ~ 15 s before being freeze-quenched in cryogenic liquid isopentane. All

EPR measurements were acquired on a Bruker Elexsys E580 X-band spectrometer equipped with a SuperX-FT microwave bridge. CW EPR measurements were performed using an ER 4122 SHQE SuperX high-sensitivity cavity in combination with an ER

4112-HV Oxford Instruments variable temperature helium flow cryostat. For pulsed

EPR/ENDOR measurements a Bruker EN 4118X-MD4 dielectric ENDOR resonator was used in concert with an Oxford CF935 helium flow cryostat. MW pulses generated by the

MW bridge were amplified by an Applied Systems Engineering 1 kW traveling wave tube (TWT) amplifier (model 117x), which allows use of an 8 ns π/2 pulse at minimum attenuation.

X-band pulsed ENDOR spectra were recorded using the standard Bruker data acquisition system. RF pulses were generated by the Bruker ”DICE” system (first generation) and ampliﬁed by a 150 W Amplifier Research RF ampliﬁer. This set up

1 allows generation of RF π-pulses of 12 μs at 15 MHz (νL( H)) at 10 dB total attenuation.

All ENDOR measurements were recorded in a stochastic regime, as implemented in the

Bruker Dice system and Bruker XEpr software. Two pulse sequences were used: Davies and Mims ENDOR (44). Corresponding pulse sequences are shown in Figure 4-2. In addition, Figure 4-2 describes the manner in which HF coupling constants and nuclear

Zeeman frequency affects the peak position.

142 EPR and ENDOR measurements were performed on samples with organic radical concentrations ranging from 20 to 100 μM for the labeled material and 100-200 μM for the unlabeled material.

Figure 4-2. Mims and Davies pulse sequences used in this work (upper left), and illustration of distortions induced by the respective sequences relative to a hypothetical axial ENDOR spectrum

(S=1/2, I=1/2, Az=AII> Ax=Ay=A) . The lower part of the figure illustrates peak positions for two cases: weak (2νL> [A > AII]) and strong (2νL< [A > AII]) hyperfine coupling. This distinction arises from the fact that in standard ENDOR experiments no information about the sign of the corresponding nuclear transition frequency is obtained. Thus, in the strong coupling case, negative frequency signals are "folded" to the positive axis.

Rapid freeze-quench: Rapid freeze-quench (RFQ) was conducted as previously described (45). Briefly, the components of each syringe apparatus were taken into the anaerobic chamber 24 h before loading the syringes with the reaction mixtures. For each

143 time series, consisting of eight points, a total of 3.0 mL of reaction mixture was used, which was distributed between 2 mL and 1 mL syringes. In the 2 mL syringe, the following was added: 50 mM Tris-HCl, pH 8.4, 5 mM MgCl2, 2 mM SAM, 300 µM

155mer RNA substrate, and 275 µM Cfr. In the 1 mL syringe, the following was added:

50 mM Tris-HCl, pH 8.4, 5 mM MgCl2, and 20 mM dithionite. These volumes are sufficient to prepare all time points and result in dilution of the reaction only by 33% after mixing. Each syringe was then capped and taken out of the anaerobic chamber to be placed in the Update Instruments Syringe driver. The instrument was programed to mix the two syringes together at 2 cm/s and then expel 357 µL of reaction mixture into a collection tube filled with liquid cryogenic ethane (~ -150 ° C) at specific reaction times.

The ethane was then removed from each sample by vacuum while maintaining the collection tube at ~ - 140 °C using a liquid N2/isopentane bath. After cryogenic ethane removal, the frozen mixture becomes a fine powder, which is then packed into individual

3.8 mm quartz EPR tubes. CW-EPR spectra were recorded as described above and the concentration of each sample was determined by comparison of the sample’s double integral to that of the flavodoxin semiquinone radical (37.2 µM). All concentrations were multiplied by two to account for the packing factor (45).

Chemical Quench of Cfr Reaction: Reaction tube 1 contained in a total volume of 225

µL: 50 mM Tris-HCl, pH 8.4, 5 mM MgCl2, 2 mM SAM, 300 µM 155mer RNA substrate, and 275 µM Cfr. Reaction tube 2 contained in a total volume of 100 µL: 50 mM Tris-HCl, pH 8.4, 5 mM MgCl2, and 20 mM dithionite. The contents of Reaction tube 1 (200 µL) were then rapidly mixed with those of Reaction tube 2 to initiate the reaction. Aliquots (20 µL) were removed at designated times and then added to 20 µL of

144 a solution containing 50 mM H2SO4 and 100 µM L-tryptophan (IS) to quench the reaction, yielding a final IS concentration of 50 µM. After quenching the reaction, 40 µL of 2  P1 nuclease quench buffer was added (250 mM NaAcetate, pH 6.0, 45 mM NaCl, and 4 mM ZnCl2). P1 nuclease (0.5 U) and Antarctic phosphatase (5 U) were added, and the RNA was then digested at 37 °C for 12 h. The precipitate was removed by centrifugation and the supernatants directly analyzed by LC-MS, as previously described

(36, 42). This entire procedure was performed in duplicate.

Production of Flavodoxin Semiquinone Radical for EPR Quantification: Flavodoxin semiquinone (FSQ) was produced by incubating 50 µM flavodoxin, 100 mM Tris-HCl, pH 8.4, and 150 mM KCl with 25 µM dithionite at 37 °C for 30 min in an anaerobic chamber. The reaction was diluted 1:5, added to a septum-sealed anaerobic cuvette, and its UV-visible spectrum recorded. The concentration of FSQ was determined from its

-1 -1 UV/vis spectrum (580 nm = 4,570 M • cm ) (46) (Figure 4-3).

145

Spectral Interpretation. Spectral simulations were performed using a spin Hamiltonian formalism in a high field approximation, in which the nuclear Zeeman, hyperfine and quadrupole interactions are considered as perturbations to the electron Zeeman interaction. CW EPR simulations were calculated using the "pepper" routine from the

EasySpin package for Matlab (47) using the following g-matrix principal components: g1,2 = 2.0044 and g3= 2.0079. We note that because the g-matrix components cannot be resolved directly from the EPR spectra, the presented g-values have no physical meaning and were used in the fit merely to account for minor asymmetric distortions of the lineshape. Simulations included all nuclei with HF coupling constants above 15 MHz. All other nuclei were simulated by introducing inhomogeneous line broadening via g-strain.

146 ENDOR Simulations. ENDOR simulations were performed independently for each nucleus by home-written routines that are a part of our (Alexey Silakov) package "Kazan

Viewer" (48). Because the MW pulses used have a bandwidth that is larger than the overall spectral width of the system, the simulations were performed including all orientations in powder averaging. Because no quadrupole splitting could be resolved in the deuterium ENDOR signals, we excluded this interaction from the simulation of 2H signals.

Density Functional Theory. All calculations were performed using Gaussian 03 Rev. E. within the spin-unrestricted Density Functional Theory (DFT) level (49). The geometries of all models were optimized utilizing the BP86 functional (50, 51) without any restrictions. In the optimizations, Ahlrichs triple- valence basis set (TZV) (52) with one set of polarization functions was used for all atoms (TZVP). The single point calculations with subsequent extraction of EPR parameters were performed using the B3LYP functional (53, 54) with TZVP basis sets on all atoms. Both geometry optimization and single point calculations were performed using Gaussian's implementation of continuous solvation model COSMO (conductor-like screening model) (55) in the PCM (polarizable continuum models) framework termed (56) as C-PCM with ε=4.0. Two types of models were used, with a full cysteine residue and a truncated version with cysteine modeled as

S-CH3. No substantial difference was found between those models. Thus, we present only the smaller model to simplify the representation.

147 4.3 Results and Discussion

To generate a substrate that would trigger formation of the organic radical, a 155- nucleotide RNA strand comprising nucleotides 2464-2608 (155mer) was synthesized by

T7 run-off transcription (see Online Methods). This strand incorporates helices 89, 90-92, and 93, shown by previous studies to be essential for high activity (57). A sample containing 200 µM wild-type (wt) Cfr, 2 mM SAM and 150 µM of the 155mer was incubated for 5 min at 22 °C before being mixed by hand with 3 µL of a 114 mM solution of sodium dithionite to initiate the reaction. The reaction was quickly loaded into an EPR tube before being freeze-quenched in cryogenic liquid isopentane, yielding a total reaction time of ~ 15 s. Analysis of the sample by CW X-band EPR revealed a spectrum characteristic of a radical strongly coupled to a single proton (Figure 4-4A). The optimal conditions for observing the signal were similar to those of other characterized organic radicals, requiring relatively high temperatures (≥ 70 K) and low microwave powers

Table 4-1. HF coupling constants (all in MHz) extracted from the experimental data as compared with calculated values from spin-unrestricted DFT methods. exp DFT * * * HFC A1 A2 A3 Aiso assign. A1 A2 A3 Aiso H1 80 82±2 85 82.3 C8-H +86.4 +88.0 +95.6 +90.0 H2 15 14 13±1 12.7 C2-H -23.4 -15.7 -6.1 -15.1 H3 9 9 7±2 8.3 NH2 -15.3 -10.4 +0.9 -8.3 H4 6.5 6.5 4±2 5.7 NH2 -11.5 -11.2 -2.1 -8.2 H5 2 2 -2:0** 0.1:0.7** 1'-H -1.5 -1.3 + 2.8 -0.0 C1 60 60 64 61.3 C8-13C +59.6 +59.9 +70.6 +63.4 N1 60±3 -5±2 -5±2 16.7 N7 +61.5 -4.5 -5.0 +17.3 C8-CH2 -6.1 -4.4 +1.9 -2.9 C8-CH2 -6.0 -4.5 +1.6 -3.0

148

149 (20 µW) to prevent saturation. A simulation of the spectrum (red trace) indicated a highly

1 isotropic H hyperfine (HF) coupling tensor (A1,2,3 = [80, 82, 85] MHz) for its three

principal components. Importantly, it was necessary to include an additional contribution

14 from an N nucleus with highly anisotropic HF coupling (A1,2,3 = [60, -5, -5] MHz) to

reproduce the shape of the EPR signal.

To characterize the nature of this radical, a spectrum of a similar sample, in which the

155mer substrate was replaced with a substrate isotopolog containing perdeuterated

adenosine nucleotides (deu155mer), was recorded. The analysis of this sample by CW

EPR reveals a dramatic narrowing of the spectrum in comparison with the unlabeled

sample, indicating substitution of the strongly coupled proton with a deuteron (Figure 4-

4B). This behavior is consistent with the observation that the HF coupling constants of

the substituting 2H are scaled with respect to the original 1H HF coupling constants by the

1 2 quotient of the gyromagnetic ratios of these two nuclei, gn( H)/gn( H)= 6.51.

Table 4-2. Comparison of HF coupling constants (in MHz) obtained experimentally or calculated for two models that differ by the protonation state of the N7 atom.

nucleus Experiment DFT, non-protonated DFT, protonated

Ax Ay Az Aiso Ax Ay Az Aiso Ax Ay Az Aiso H-C8 80.0 82.0 85.0 82.3 +86.4 +88.0 +95.6 +90.0 +80.6 +82.3 +89.2 +84.0 H-C2 14.0 14.0 12.0 13.3 -23.4 -15.7 -6.1 -15.1 -13.7 -9.5 -2.6 -8.6

NH2 9.0 9.0 7.0 8.3 -15.3 -10.4 +0.9 -8.3 -20.5 -14.1 0.1 -11.5

NH2 6.5 6.5 4.0 5.7 -11.5 -11.2 -2.1 -8.2 -16.6 -14.3 -1.8 -10.9 1'-H -2.0 2.0 2.0 0.7 -1.5 -1.3 +2.8 -0.0 -1.2 -0.8 +3.8 +0.6 met(13C) 60.0 60.0 64.0 61.3 +59.6 +59.9 +70.6 +63.4 +42.7 +43.0 +57.0 +47.5 N7 -60.0 5.0 5.0 -16.7 -5.0 -4.5 +61.5 +17.3 -2.8 -2.7 +45.9 +13.5

met-H -6.1 -4.5 1.9 -2.9 -4.8 -2.8 +2.4 -1.7 met-H -6.1 -4.6 1.6 -3.0 -5.6 -3.9 1.2 -2.8 N7-H - - - - -45.0 -31.4 -2.0 -26.1

150 Our previous studies on RlmN suggest that both RlmN and Cfr contain only one SAM binding site, of which a major determinant is the unique iron ion of the [4Fe–4S] cluster to which the -amino and carboxylate groups of SAM coordinate (38, 42). We showed that this one site supports both transfer of a methyl group from SAM to the target Cys residue and generation of a 5’-dA• to abstract a H• from the resulting mCys residue.

When overproduced under our normal expression conditions, both RlmN and Cfr are isolated almost exclusively (≥ 95%) with the mCys modification. However, we showed that when RlmN is overproduced in its apo form (i.e. no Fe/S cluster present), by adding o-phenanthroline to the growth media of the expression host at induction, the protein is isolated almost exclusively without a mCys modification (≥ 98%). When apo RlmN was incubated with SAM, no reaction took place. However, when the [4Fe–4S] cofactor was first reconstituted into the apo protein, rapid formation of SAH took place upon addition of SAM (42). Figure 4-5 depict a similar experiment with Cfr. When Cfr is overproduced and isolated in its apo form and then incubated with SAM, no reaction takes place. By contrast, when the [4Fe–4S] cofactor is first reconstituted into apo Cfr (designated

Cfraporcn) before incubating it with SAM, rapid formation of SAH ensues, indicating transfer of a methyl group to C338.

151

Using the methodology described above, we reconstituted the [4Fe–4S] cluster on apo

13 Cfr and then incubated the protein with S-adenosyl-L-[methyl- C]methionine to generate

13 Cfraporcn containing a [methyl- C]Cys residue at C338. To show that the methylene carbon of the mCys residue is connected to C8 of the adenine ring of Species 3 (Figure

4- 1), the organic radical was generated with the 13C-labeled protein and analyzed by CW

EPR and ENDOR spectroscopies. Because 13C contains a nuclear spin (I = ½), the expectation is that the EPR signal should be split or broadened, depending on the magnitude of the HF coupling. To observe the anticipated 13C splitting in the absence of the strong 1H splitting, the sample was prepared using the deu155mer substrate. We term this EPR sample, containing both [methyl-13C]mCys 338 and the deu155mer RNA substrate, [2H, 13C]. Its spectrum is markedly split in comparison to the sample generated with unlabeled Cfr and the deu155mer substrate (Figure 4-4C). In the spectrum of the

152

13 sample prepared with [methyl- C]mCys 338 Cfraporcn and the unlabeled 155mer substrate, termed [13C], the splitting due to the strongly coupled proton (Figure 4-4A) is superimposed on that from 13C (Figure 4-4C).

Because the magnitude of the 13C HF coupling is similar to that of the proton, the superpositioning creates a pseudo 1:2:1 pattern in the EPR spectrum (Figure 4-4D).

From simulations of these EPR spectra, the corresponding 13C HF coupling constant is observed to be ~60 MHz with only 2 MHz axial anisotropy (see Table 1), indicating the presence of spin density directly on this nucleus. We therefore conclude that there is substantial spin density around C8 of the adenine ring, indicating that the strong 1H HF coupling in the EPR spectrum described above derives from the C8 proton, while the 14N contribution derives from N7.

To gain a better understanding of the spin distribution in the radical species, we performed additional pulsed ENDOR experiments targeted specifically to observe the 1H and 2H signals. The Mims ENDOR pulse sequence is most suitable for analysis of weak, fairly isotropic, HF interactions, while the Davies ENDOR pulse sequence is used for nuclei that are more strongly coupled, and which possess relatively large anisotropy

(Figure 4-2). We chose the [2H, 13C] sample for this study over the [2H] sample because of the similarity of the EPR spectra of the doubly labeled and unlabeled samples.

This strategy results in a similar orientation selection pattern in all ENDOR experiments allowing direct comparison of the signals obtained. Mims ENDOR analysis of the [2H, 13C] sample reveals two distinct 2H signals that are absent in control measurements of the unlabeled sample (Figure 4-6B). One of these signals (D1) is centered at 6.5 MHz and split by 4.54 MHz, which is 2 the Larmor frequency (L) of

153 deuterium. This behavior indicates that these peaks derive from a strongly coupled deuteron. The observed HF coupling constant of ~12.5 MHz precisely matches the splitting observed in the EPR signal of the unlabeled sample, accounting for the scaling factor of 6.51.

154

The hyperfine coupling constants derived from simulation of the CW EPR spectra can also be used directly to simulate these ENDOR signals (Figure 4-6B). Therefore, this signal can be assigned unambiguously to the C8 deuteron of A2503 in the deu155mer substrate. This observation, together with those described above, indicates that C8 has undergone a change in hybridization from sp2 to sp3 as depicted in Figure 4-1 (Species

3).

2 2 The second H signal (D2) is centered at L( H) = 2.27 MHz and split by 2.0 MHz, indicating a case of weak coupling. The corresponding 1H doublet with 13 MHz splitting is also found in the Davies ENDOR spectra of the proton region (see H2 in Figure 4-6A).

Given that the CW-EPR experiments presented above establish that the spin density is located on the adenine ring, this signal clearly derives from the C2 deuteron of A2503.

The isotropic character of this 1H HF coupling is also an indication of substantial spin density at C2, given that, for π-type radicals, there is a direct relationship between the isotropic 1H HF coupling constant of a ring proton and the spin density population on the carbon to which it is bonded (58). Therefore, these results indicate that the spin density is delocalized throughout the adenine ring.

A subtraction of the Mims ENDOR spectrum of the sample prepared with the deu155mer from that of the sample containing the unlabeled 155mer reveals a doublet centered at the Larmor frequency of a 1H that vanishes upon selective 2H labeling

(Figure 4-6C). The shape of the signal suggests that the corresponding HF coupling is dipolar with a magnitude of ~2 MHz. Because this splitting is small, no definite assignment of all three principal values of the HF coupling tensor could be achieved.

155 Nevertheless, this coupling most likely derives from the closest proton on the ribose ring

(C1’-H), because this nucleus is the only remaining unassigned proton/deuteron in the vicinity of the spin density. The corresponding deuteron signal in the sample with the deu155mer substrate has a splitting that is too small to be observed in these experiments

(2 MHz/6.51 = 0.3 MHz). Additionally, Davies ENDOR measurements reveal two distinct signals in the proton region that remain unchanged upon 2H labeling. The two signals are similar, and have extracted HF coupling constants that differ by ~1.5 MHz.

Based on the proposed structure of the organic radical, these signals could correspond either to the pair of protons on the N6 amine or the bridging methylene protons of the mCys residue.

To obtain a more precise depiction of the electronic structure of Species 3 and verify assignments made by EPR and ENDOR spectroscopies, DFT calculations were performed using the Gaussian 03 package (49) on the unrestricted B3LYP level (50, 51), and using Aldrich’s triple zeta basis set with a single set of polarization functions for all atoms (TZVP) (52). Table 1 includes comparisons of the calculated HF coupling constants with those derived experimentally. The agreement is good, given that the sets of isotropic HF values deviate within 15% of each other. However, aromatic ring protons, such as C2-H, typically experience strong anisotropic HF interactions with a rhombic character (59), as indeed predicted by the DFT calculations. Surprisingly, the line width of the D2 and H2 ENDOR signals in the experimental sample are too narrow to account for any significant anisotropy. Therefore it is likely that the structure of the adenosine ring around C2 is perturbed upon docking to Cfr, which may be related to the ability of the protein to catalyze C2 methylation after the preferred C8 methylation. Based on the

156 performed DFT calculations, we were able to assign the HF signals of H3 and H4 in the

6 ENDOR spectrum to the amine protons (N H2), because the protons on the added methylene carbon were calculated to experience a relatively weak isotropic HF interaction. Overall, the experimentally deduced picture of the spin distribution is confirmed by the DFT calculations. Considering that N7 has the largest spin population of all atoms (Figure 4-7), we can formally assign the radical as residing on this atom, as shown in Figure 4-1. The reaction catalyzed by Cfr and RlmN is potentially analogous to the Minisci reaction, wherein protonated heteroaromatic bases are substituted by nucleophilic carbon-centered radicals (60). A Minisci-like reaction would suggest protonation of N7 before or during radical addition to afford a radical cation species. We have explored this possibility via DFT analysis. Table S1 displays predicted HF coupling constants for both protonated and non-protonated models. The non-protonated model reproduces the experimental data much better, even when accounting for relatively poor

13 couplings for C2-H and NH2. Most notably, in the protonated model the calculated C

HF coupling for the added methylene group and the 14N HF coupling of N7 are considerably smaller than the experimental values. It is important to note that the proton at N7 in the protonated model is expected to exhibit a large HF coupling constant, given that it is in direct contact with the spin density located on N7.reproduction of the anisotropy of the 1H hyperfine

157

Figure 4-7. Spin density distribution in the A2503 radical species, as modeled by spin- unrestricted B3LYP (DFT) methods. Blue cyan areas correspond to negative spin density, and orange to positive spin density. Numbers correspond to most significant spin populations based on Mulliken population analysis. Atom color coding: white-H, gray-C, blue-N, red-O and yellow- S. Indeed, calculations on the protonated model reveal a mostly axial HF coupling tensor with the principal values Ax=-45 MHz, Ay=-30 MHz, Az=-2 MHz. Therefore, the presence of the proton would significantly perturb the shape of the CW EPR spectra.

Figure 4-8 shows a comparison between the measured EPR spectra and a simulation that includes the HF coupling constants predicted for the proton at N7, keeping the remaining parameters identical to those in the simulations shown in Figure 4-4. The simulated spectra for the protonated model do not reproduce the measured EPR spectra.

158 To determine whether the characterized radical is consistent with a species that is on the reaction pathway to the m8A product, as described in Figure 4-1, its chemical and kinetic competence was assessed using rapid-freeze-quench (RFQ) EPR in combination with chemical-quench LC/MS. Cfr (275 µM), SAM (2 mM), 155mer RNA (300 µM),

MgCl2 (5 mM) and 50 mM Tris–HCl, pH 8.4, were loaded into one syringe of an Update

Instruments RFQ apparatus, while dithionite (20 mM), MgCl2 (5 mM) and 50 mM Tris–

HCl, pH 8.4, were loaded into a second syringe. Designated volumes of the two syringes were rapidly mixed in a 2:1 (syringe 1:syringe 2) ratio — such that the concentration of

Cfr after mixing was reduced by 33% — and then expelled at various times into cryogenic liquid ethane (~-150 °C). After removing the ethane by evaporation, the remaining powder was packed into EPR tubes at cryogenic temperature, and the resulting samples were stored in liquid N2 until ready for analysis. A parallel experiment was conducted using the exact same components at the indicated concentrations, but quenching was performed by hand rather than using an RFQ apparatus.

159

160

Figure 4-9. Kinetic Competence of Radical Intermediate: A. Representative CW-EPR spectra from the rapid freeze-quench time course (green trace, 180 s; red trace, 45 s; blue trace, 15 s). B. Time-dependent formation and decay of the A2503 radical species (solid triangles) and time- dependent formation of m8A (open triangles). The EPR data were fitted to an ABC kinetic model, while the m8A data were fitted to a pseudo-first order exponential equation. The extracted rate constants are reported in the text. Error bars, where indicated, denote the standard deviation from two independent reactions. CW-EPR spectra were collected as described in the Materials and Methods.

In Figure 4-9B (closed triangles), the time-dependent concentration of the organic radical is plotted, while in Figure 4-9A, representative EPR spectra at indicated times

(matched by colored arrows in Figure 4-9B) are displayed. As can be observed, the

-1 -1 radical forms and decays, affording the rate constants k1 = 0.024 s and k2 = 0.016 s upon fitting the data to an irreversible ABC kinetic model, wherein A is the substrate, B is the organic radical, and C is the product. The net rate constant for these two steps, calculated as the product of k1 and k2 divided by the sum of k1 and k2, is 0.0096 s-1. The open triangles in Figure 4B correspond to the time-dependent formation of m8A.

161 A fit of this data to a first-order single-exponential equation affords a rate constant of

0.0093 s-1, which is in agreement with the net rate constant calculated from RFQ EPR.

Therefore, this radical species is chemically and kinetically competent, and consistent with an intermediate species on the reaction pathway.

4.4 Conclusions

RlmN and Cfr exhibit both methyltransferase and methylsynthase activity. The methyltransferase activity has been well characterized and involves a direct nucleophilic attack of a strictly conserved Cys residue on the activated methyl group of SAM, resulting in a mCys intermediate that has been observed by high-resolution mass spectrometry and X-ray crystallography (36, 38). By contrast, intermediates in the unconventional methylsynthase step of the reactions have been more elusive.

Spectroscopic and kinetic studies presented herein provide clear evidence for the second step in methylsynthase activity, involving abstraction of a H• from the mCys residue and addition of the ensuing methylene radical onto C8 of A2503 to afford a spin delocalized radical formally residing at N7.

162 4.5 Coordinates from Density Functional Theory

Non-protonated model (presented in the main text):

BP86 optimization, final energy -1441.36456 A.U.

C -4.8477670 -1.3796160 0.6812370

C -3.6775200 -1.3609030 -0.1443480

C -3.2580830 -0.0675260 -0.5985030

C -5.0572880 0.8484960 0.5286960

H -5.6078810 1.7566200 0.7902710

N -2.1694800 0.0803810 -1.3829770

H -1.8983030 1.0046260 -1.7051940

H -1.6232710 -0.7279460 -1.6633970

N -3.9738090 1.0266690 -0.2416960

N -5.5613540 -0.3078680 1.0295600

N -3.1652580 -2.5688330 -0.3875660

N -5.0690890 -2.6925520 1.0264110

C -4.0073520 -3.5086370 0.3503560

H -4.4927660 -4.2164820 -0.3483000

C -3.0925300 -4.2996610 1.3162130

H -2.3014420 -4.7469060 0.6990770

H -2.6212350 -3.6118380 2.0325950

C -4.1562780 -4.9279220 3.8969580

H -4.9592710 -4.1811760 3.8703410

H -4.4591470 -5.7506500 4.5580730

163 S -3.8658840 -5.6907950 2.2493660

C -8.5623600 -2.9412420 0.6693060

O -7.2312760 -3.2173290 0.2031450

C -6.3519710 -3.2579770 1.3748390

C -7.1191130 -2.5705990 2.5394540

C -8.3508590 -1.9684490 1.8394770

H -9.0443010 -3.8640800 1.0499360

H -6.1702500 -4.3091320 1.6487660

H -8.0898550 -0.9707060 1.4486690

O -9.4509420 -1.9006750 2.7538650

H -10.2544820 -1.8598790 2.1915570

C -9.4023010 -2.3648390 -0.4592710

H -9.4850780 -3.0802470 -1.2956500

H -8.9410560 -1.4322370 -0.8269250

O -10.7006180 -2.1059160 0.1289690

H -11.1772090 -1.4845030 -0.4489260

O -7.5196560 -3.5847330 3.4722130

H -8.4217790 -3.3216100 3.7597210

H -6.5036620 -1.8052170 3.0360980

H -3.2269960 -4.4848960 4.2779200

Protonated model:

BP86 optimization, final energy -1441.79258 A.U.

C -4.8263630 -1.3672660 0.5729120

164 C -3.6224870 -1.3104660 -0.1970420

C -3.1818930 -0.0173780 -0.6171870

C -5.0615580 0.8533020 0.4220770

H -5.6421480 1.7508820 0.6540160

N -2.0624800 0.1955220 -1.3297390

H -1.8293480 1.1519000 -1.5909860

H -1.4282740 -0.5452180 -1.6108140

N -3.9447820 1.0516460 -0.2838640

N -5.5688680 -0.3129890 0.8889410

N -3.1779540 -2.5715370 -0.3819900

N -5.0646520 -2.6784400 0.8960910

C -4.0057420 -3.5394630 0.3431450

H -4.4361860 -4.2782940 -0.3541940

C -3.1153310 -4.2756780 1.3980890

H -2.2548560 -4.6786520 0.8456740

H -2.7506310 -3.5535280 2.1411900

C -4.2418560 -5.0293080 3.9175930

H -5.0342060 -4.2706120 3.9028120

H -4.5897210 -5.8898790 4.5041360

S -3.8988250 -5.7009890 2.2423810

C -8.5707470 -2.9464800 0.6858810

O -7.2437230 -3.1978570 0.1740710

C -6.3485460 -3.2339190 1.3096990

C -7.0579360 -2.5175030 2.4952630

C -8.3378640 -1.9581270 1.8404090

165 H -9.0101750 -3.8793610 1.0881890

H -6.1508760 -4.2785120 1.6009040

H -8.1285840 -0.9553330 1.4312930

O -9.3890350 -1.9200350 2.8048780

H -10.2236350 -1.8723240 2.2910890

C -9.4601080 -2.3988730 -0.4189290

H -9.5814060 -3.1361640 -1.2307480

H -9.0153760 -1.4765080 -0.8321640

O -10.7216000 -2.1263840 0.2298910

H -11.2832470 -1.6229880 -0.3856470

O -7.3797620 -3.5085550 3.4744570

H -8.2772690 -3.2715060 3.7987380

H -6.4317690 -1.7248760 2.9323980

H -3.3204230 -4.6356370 4.3638430

H -2.3371160 -2.8581630 -0.8800060

166 4.6 References

1. Choffnes, E. R., Relman, D. A., and Mack, A. (2010) Antibiotic Resistance: Implications

for Global Health and Novel Intervention Strategies, National Academies Press,

Washington, D. C.

2. Fischbach, M. A., and Walsh, C. T. (2009) Antibiotics for emerging pathogens, Science

325, 1089–1093.

3. Poehlsgaard, J., and Douthwaite, S. (2005) The bacterial ribosome as a target for

antibiotics, Nat. Rev. Microbiol. 3, 870–881.

4. Steitz, T. A. (2010) From the structure and function of the ribosome to new antibiotics

(Nobel Lecture), Angew. Chem. Int. Ed. 49, 4381–4398.

5. Weisblum, B. (1995) Erythromycin resistance by ribosome modification, Antimicrob.

Agents Chemother. 39, 577–585.

6. Kehrenberg, C., Schwarz, S., Jacobsen, N. E., Hansen, L. H., and Vester, B. (2005) A

new mechanism for chloramphenicol, florfenicol and clindamycin resistance: methylation

of 23S ribosomal RNA at A2503, Mol. Microbiol. 57, 1064–1073.

7. Long, K. S., Poehlsgaard, J., Kehrenberg, C., Schwarz, S., and Vester, B. (2006) The Cfr

rRNA methyltransferase confers resistance to phenicols, lincosamides, oxazolidinones,

pleuromutilins, and streptogramin A antibiotics, Antimicrob. Agents Chemother. 50,

2500–2505.

8. Schwarz, S., Werckenthin, C., and Kehrenberg, C. (2000) Identification of a plasmid-

borne chloramphenicol–florfenicol resistance gene in Staphylococcus sciuri, Antimicrob.

Agents Chemother. 44, 2530–2533.

167 9. Smith, L. K., and Mankin, A. S. (2008) Transcriptional and translational control of the

mlr operon, which confers resistance to seven classes of protein synthesis inhibitors,

Antimicrob. Agents Chemother. 52, 1703–1712.

10. Giessing, A. M. B., Jensen, S. S., Rasmussen, A., Hansen, L. H., Gondela, A., Long, K.

S., Vester, B., and Kirpekar, F. (2009) Identification of 8-methyladenosine as the

modification catalyzed by the radical SAM methyltransferase Cfr that confers antibiotic

resistance in bacteria, RNA 15, 327–336.

11. Ban, N., Freborn, B., Nissen, P., Penczek, P., Grassucci, R. A., Sweet, R., Frank, J.,

Moore, P. B., and Steitz, T. A. (1998) A 9 Å resolution x-ray crystallographic map of the

large ribosomal subunit, Cell 93, 1105–1115.

12. Harms, J., Schluenzen, F., Zarivach, R., Bashan, A., Gat, S., Agmon, I., Bartels, H.,

Francheschi, F., and Yonath, A. (2001) High-resolution structure of the large ribosomal

subunit from a mesophilic eubacterium, Cell 107, 679–688.

13. Schuwirth, B. S., Borovinskaya, M. A., Hau, C. W., Zhang, W., Vila-Sanjurjo, A., Hoton,

J. M., and Cate, J. H. D. (2005) Structures of the bacterial ribosome at 3.5 Å resolution,

Science 310, 827–834.

14. Selmer, M., Dunham, C. M., Murphy, F. V., IV, Weislbaumer, A., Petry, S., Kelley, A.

C., Weir, J. R., and Ramakrishnan, V. (2006) Structure of the 70S ribosome complexed

with mRNA and tRNA, Science 313, 1935–1942.

15. Kowalak, J. A., Bruenger, E., and McCloskey, J. A. (1995) Posttranscriptional

modification of the central loop of domain V in Escherichia coli 23S ribosomal RNA, J.

Biol. Chem. 270, 17758–17764.

16. Toh, S.-M., Xiong, L., Bae, T., and Mankin, A. S. (2008) The methyltransferase

YfgB/RlmN is responsible for modification of adenosine 2503 in 23S rRNA, RNA 14,

98–106.

168 17. Vazquez-Laslop, N., Ramu, H., Klepacki, D., and Mankin, A. S. (2010) The key role of a

conserved and modified rRNA residue in the ribosomal response to the nascent peptide,

EMBO J. 29, 3108–3117.

18. Diekema, D. J., and Jones, R. N. (2001) Oxazolidinone antibiotics, The Lancet 358,

1975–1982.

19. Toh, S.-M., Xiong, L., Arias, C. A., Villegas, M. V., Lolans, K., Quinn, J., and Mankin,

A. S. (2007) Acquisition of a natural resistance gene renders a clinical strain of

methicillin-resistant Staphylococcus aureus resistant to the synthetic antibiotic linezolid,

Mol. Microbiol. 64, 1506–1514.

20. Bonilla, H., Huband, M. D., Seidel, J., Schmidt, H., Lescoe, M., McCurdy, S. P.,

Lemmon, M. M., Brennan, L. A., Tait-Kamradt, A., Puzniak, L., and Quinn, J. P. (2010)

Multicity outbreak of linezolid-resistant Staphylococcus epidermidis associated with

clonal spread of a cfr-containing strain, Clin. Infect. Dis. 51, 796–800.

21. Mendes, R. E., Deshpande, L. M., Castanheira, M., Dipersio, J., Saubolle, M. A., and

Jones, R. N. (2008) First report of cfr-mediated resistance to linezolid in human

staphylococcal clinical isolates recovered in the United States, Antimicrob. Agents

Chemother. 52, 2244–2246.

22. Farrell, D. J., Mendes, R. E., Ross, J. E., and Jones, R. N. (2009) Linezolid surveillance

program results for 2008 (LEADER Program for 2008), Diagn. Microbiol. Infect. Dis.

65, 392–403.

23. Farrell, D. J., Mendes, R. E., Ross, J. E., Sader, H. S., and Jones, R. N. (2011) LEADER

program results for 2009: an activity and spectrum analysis of linezolid using 6,414

clinical isolates from 56 medical centers in the United States, Antimicrob. Agents

Chemother. 55, 3684–3690.

169 24. Morales, G., Picazo, J. J., Baos, E., Candel, F. J., Arribi, A., Peláez, B., Andrade, R., de

la Torre, M.-A., Ferres, J., and Sánchez-Garcia, M. (2010) Resistance to linezolid is

mediated by the cfr gene in the first report of an outbreak of linezolid-resistant

Staphylococcus aureus, Clin. Infect. Dis. 50, 821–825.

25. Sánchez-Garcia, M., de la Torre, M.-A., Morales, G., Peláez, B., Tolón, M. J., Domingo,

S., Candel, F. J., Andrade, R., Arribi, A., García, N., Martínez Sagasti, F., Fereres, J., and

Picazo, J. (2010) Clinical outbreak of linezolid-resistant Staphylococcus aureus in an

intensive care unit, JAMA 303, 2260–2264.

26. Mendes, R. E., Deshpande, L. M., Farrell, D. J., Spanu, T., Fadda, G., and Jones, R. N.

(2010) Assessment of linezolid resistance mechanisms among Staphylococcus

epidermidis causing bacteraemia in Rome, Italy, J. Antimicrob. Chemother. 65, 2329–

2335.

27. Mendes, R. E., Deshpande, L., Rodriguez-Noriega, E., Ross, J. E., Jones, R. N., and

Morfin-Otero, R. (2010) First report of Staphylococcal clinical isolates in Mexico with

linezolid resistance caused by cfr: evidence of in vivo cfr mobilization, J. Clin. Microbiol.

48, 3041–3043.

28. Shore, A. C., Brennan, O. M., Ehricht, R., Monecke, S., Schwarz, S., Slickers, P., and

Coleman, D. C. (2010) Identification and characterization of the multidrug resistance

gene cfr in a Panton-Valentine leukocidin-positive sequence type 8 methicillin-resistant

Staphylococcus aureus IVa (USA300) isolate., Antimicrob. Agents Chemother. 54, 4978–

4984.

29. Diaz, L., Kiratisin, P., Mendes, R. E., Panesso, D., Singh, K. V., and Arias, C. A. (2012)

Transferable plasmid-mediated resistance to linezolid due to cfr in a human clinical

isolate of Enterococcus faecalis, Antimicrob. Agents Chemother. 56, 3917–3922.

170 30. Hansen, L. H., Planellas, M. H., Long, K. S., and Vester, B. (2012) The order Bacillales

hosts functional homologs of the worrisome cfr antibiotic resistance gene, Antimicrob.

Agents Chemother. 56, 3563–3567.

31. Booker, S. J. (2009) Anaerobic functionalization of unactivated C–H bonds, Curr. Opin.

Chem. Biol. 13, 58–73.

32. Challand, M. R., Driesener, R. C., and Roach, P. L. (2011) Radical S-adenosylmethionine

enzymes: mechanism, control and function, Nat. Prod. Rep. 28, 1696–1721.

33. Frey, P. A., Hegeman, A. D., and Ruzicka, F. J. (2008) The radical SAM superfamily,

Crit. Rev. Biochem. Mol. Biol. 43, 63–88.

34. Kim, S., Meehan, T., and Schaefer, H. F., III. (2007) Hydrogen-atom abstraction from the

adenine–uracil base pair, J. Phys. Chem. A. 111, 6806–6812.

35. Zierhut, M., Roth, W., and Fischer, I. (2004) Dynamics of H-atom loss in adenine, Phys.

Chem. Chem. Phys. 6, 5178–5183.

36. Grove, T. L., Benner, J. S., Radle, M. I., Ahlum, J. H., Landgraf, B. J., Krebs, C., and

Booker, S. J. (2011) A radically different mechanism for S-adenosylmethionine-

dependent methyltansferases, Science 332, 604–607.

37. Yan, F., and Fujimori, D. G. (2011) RNA methylation by radical SAM enzyme RlmN

and Cfr proceeds via methylene transfer and hydride shift, Proc. Natl. Acad. Sci. U S A

108, 3930–3934.

38. Boal, A. K., Grove, T. L., McLaughlin, M. I., Yennawar, N. H., Booker, S. J., and

Rosenzweig, A. C. (2011) Structural basis for methyl transfer by a radical SAM enzyme,

Science 332, 1089–1092.

39. McCusker, K. P., Medzihradszky, K. F., Shiver, A. L., Nichols, R. J., Yan, F., Maltby, D.

A., Gross, C. A., and Galonic Fujimori, D. (2012) Covalent intermediate in the catalytic

171 mechanism of the radical S-adenosyl-L-methionine methyl synthase RlmN trapped by

mutagenesis, J. Am. Chem. Soc. 134, 18074-18081.

40. Kaminska, K. H., Purta, E., Hansen, L. H., Bujnicki, J. M., Vester, B., and Long, K. S.

(2010) Insights into the structure, function and evolution of the radical-SAM 23S rRNA

methyltransferase Cfr that confers antibiotic resistance in bacteria, Nuc. Acids. Res. 38,

1652–1663.

41. Iwig, D. F., and Booker, S. J. (2004) Insight into the polar reactivity of the onium

chalcogen analogues of S-adenosyl-L-methionine, Biochemistry 43, 13496-13509.

42. Grove, T. L., Radle, M. I., Krebs, C., and Booker, S. J. (2011) Cfr and RlmN contain a

single [4Fe–4S] cluster, which directs two distinct reactivities for S-adenosylmethionine:

methyl transfer by SN2 displacement and radical generation, J. Am. Chem. Soc. 133,

19586–19589.

43. Bradford, 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.

44. Schweiger, A., and Jeschke, G. (2001) Principles of pulse electron paramagnetic

resonance, Oxford University Press, New York.

45. Bollinger, J. M., Jr., Tong, W. H., Ravi, N., Huynh, B. H., Edmondson, D. E., and

Stubbe, J. A. (1995) Use of rapid kinetics methods to study the assembly of the diferric-

tyrosyl radical cofactor of E. coli ribonucleotide reductase, Methods Enzymol. 258, 278-

303.

46. Fujii, K., Galivan, J. H., and Huennekens, F. M. (1977) Activation of methionine

synthase: further characterization of flavoprotein system, Arch. Biochem. Biophys. 178,

662-670.

172 47. Stoll, S., and Schweiger, A. (2006) EasySpin, a comprehensive software package for

spectral simulation and analysis in EPR, Journal of Magnetic Resonance 178, 42-55.

48. Silakov, A. (2011) Kazan Viewer, https://sites.google.com/site/silakovalexey/kazan-

viewer.

49. Frisch, M. J., Trucks, G. W., Schlegel, H. B., Scuseria, G. E., Robb, M. A., Cheeseman,

J. R., Montgomery, J., J. A., Vreven, T., Kudin, K. N., Burant, J. C., Millam, J. M.,

Iyengar, S. S., Tomasi, J., Barone, V., Mennucci, B., Cossi, M., Scalmani, G., Rega, N.,

Petersson, G. A., Nakatsuji, H., Hada, M., Ehara, M., Toyota, K., Fukuda, R., Hasegawa,

J., Ishida, M., Nakajima, T., Honda, Y., Kitao, O., Nakai, H., Klene, M., Li, X., Knox, J.

E., Hratchian, H. P., Cross, J. B., Bakken, V., Adamo, C., Jaramillo, J., Gomperts, R.,

Stratmann, R. E., Yazyev, O., Austin, A. J., Cammi, R., Pomelli, C., Ochterski, J. W.,

Ayala, P. Y., Morokuma, K., Voth, G. A., Salvador, P., Dannenberg, J. J., Zakrzewski, V.

G., Dapprich, S., Daniels, A. D., Strain, M. C., Farkas, O., Malick, D. K., Rabuck, A. D.,

Raghavachari, K., Foresman, J. B., Ortiz, J. V., Cui, Q., Baboul, A. G., Clifford, S.,

Cioslowski, J., Stefanov, B. B., Liu, G., Liashenko, A., Piskorz, P., Komaromi, I., Martin,

R. L., Fox, D. J., Keith, T., Al-Laham, M. A., Peng, C. Y., Nanayakkara, A.,

Challacombe, M., Gill, P. M. W., Johnson, B., Chen, W., Wong, M. W., Gonzalez, C.,

and Pople, J. A. (2004) Gaussian, Revision E.01 ed. ed., Gaussian, Inc, Wallingford, CT.

50. Becke, A. D. (1988) Density-functional exchange-energy approximation with correct

asymptotic-behavior, Phys. Rev. A 38, 3098–3100.

51. Perdew, J. P., and Wang, Y. (1988) Jellium work function for all electron densities, Phys.

Rev. B 38, 12228–12232.

52. Schafer, A., Huber, C., and Ahlrichs, R. (1994) Fully optimized contracted Gaussian-

basis sets of triple zeta valence quality for atoms Li to Kr, J. Chem. Phys. 100, 5829–

5835.

173 53. Becke, A. D. (1993) Density-functional thermochemistry 3. The role of exact exchange,

J. Chem. Phys. 98, 5648–5652.

54. Lee, C. T., Yang, W. T., and Parr, R. G. (1988) Development of the Colle-Salvetti

correlation-energy formula into a functional of the electron-density, Phys. Rev. B 37, 785-

789.

55. Klamt, A., and Schuurmann, G. (1993) Cosmo - a new approach to dielectric screening in

solvents with explicit expressions for the screening energy and its gradient, J. Chem. Soc.

Perkin Trans. 2, 799–805.

56. Barone, V., and Cossi, M. (1998) Quantum calculation of molecular energies and energy

gradients in solution by a conductor solvent model, J. Phys. Chem. A. 102, 1995–2001.

57. Yan, F., LaMarre, J. M., Röhrich, R., Wiesner, J., Jomaa, H., Mankin, A. S., and Galoníc

Fujimori, D. (2010) RlmN and Cfr are radical SAM enzymes involved in methylation of

ribosomal RNA, J. Am. Chem. Soc. 132, 3953-3964.

58. Weil, J. A., Bolton, J. R., and Wertz, J. E. (1994) Electron Paramagnetic Resonance.

Elementary Theory and Practical Applications, John Wiley & Sons, Inc., New York.

59. Close, D. M., and Nelson, W. H. (1989) ESR and ENDOR study of adenosine single

crystals X-irradiated at 10 K, Radiat. Res. 117, 367-378.

60. Minisci, F. (1973) Novel applications of free-radical reactions in preparative organic

chemistry, Synthesis, 1–24.

Chapter 5

Generation and Characterization of the Proposed Disulfide Intermediate in the Reactions Catalyzed by the Radical SAM Methylsynthases, RlmN and Cfr

This chapter was reproduced from “Tyler L. Grove, Tatiana Laremore and Squire J. Booker. In preperation

175 5.1 Abstract

RlmN and Cfr are members of the Radical S-adenosyl-L-methionine (RS) superfamily of proteins, which utilize radical based chemistry to install methyl groups at carbons 2 and 8, respectively, of adenosine 2503 (A2503) in bacterial 23S rRNA. These methylations require the involvement of five invariant cysteine residues. Three cysteines reside in the CX3CX2C motif that contains the cysteines that ligate the requisite [4Fe-4S] cluster found in all RS enzymes, while a fourth (Cys355 in RlmN) is transiently methylated to a methyl-Cys residue (mCys). The resulting methylene carbon of the mCys residue is then appended to C2 (RlmN) or C8 (Cfr) of A2503 through a protein–nucleic acid thioether crosslink by a radical mediated process, which is proposed to be resolved by disulfide-bond formation via the participation of a fifth cysteine (Cys118 in RlmN).

The involvement of Cys118, to date, has only been inferred by the inability of C118A or

C118S variants of RlmN to resolve the thioether-crosslink, as determined by their failure to form m2A or m8A products and the observation that they copurify with covalently bound RNA. Herein, we show by high-resolution mass spectrometry (HRMS) that an intramolecular disulfide bond both on RlmN and Cfr indeed can be formed between the indicated cysteinyl residues in their proposed mechanisms. We also determine the midpoint potentials of these disulfide bonds in RlmN and Cfr to be -268 and -289 mV, respectively, which are similar to the redox potential of the E. coli cytoplasm during aerobic respiration. Interestingly, only small-molecule two-electron reductants such as glutathione and dithiothreitol are capable of reducing the disulfide bonds back to free thiols, while protein disulfide oxidoreductases such as thioredoxin and DsbA have no

176 effect. Moreover, one electron reductants, such as the Flavodoxin/Flavodoxin

Reductase/NADPH in vivo reducing system, are incapable of reducing the disulfide bond proposed to be formed concomitant with each turnover. Most importantly, when all two- electron reductants are removed from in vitro assays and only the in vivo reducing system is supplied in the mixture, multiple turnover takes place, suggesting perhaps that a true disulfide bond is not formed during turnover. Lastly, a mechanism is proposed in which a radical disulfide anion–reminiscent of that proposed in the catalytic mechanism of ribonucleotide reductase–rather than a full disulfide bond is formed, which is then reduced by one electron to yield the dithiol-containing active form of the enzyme.

177 5.2 Introduction

The bacterial ribosome is a complex machine, which deftly orchestrates the vital process of stitching amino acids together to form proteins. Its two major subunits are composed of 52 proteins and 3 strands of rRNA, which are frequently adorned with both posttranslational and posttranscriptional modifications that most often play a role in facilitating the assembly of the ribosome and enhancing the efficiency and fidelity of translation.. The majority of these modifications are methyl groups that are attached to various amino acid side-chains of ribosomal proteins or the ribose rings or bases of the nucleotides that compose rRNA. One such methyl modification, found almost universally in bacteria, occurs at C2 of adenosine 2503 (A2503) of the 23S rRNA, and is installed by the radical S-adenosylmethionine (SAM) enzyme RlmN (1). A2503 is located within the peptidyl transferase center (PTC) of the intact ribosome, and methylation of C2 of this nucleotide is believed to enhance translational fidelity (2). Recent studies indicate that

RlmN also is responsible for installing a methyl modification at C2 of A37 in several

Escherichia coli (Ec) tRNAs, and perhaps tRNAs in other organisms (3)

178

Figure 0-1. Proposed mechanism of methylation by RlmN (A) and Cfr (B). A second methyl modification at C8 of the same nucleotide has been found in strains of species of staphylococci that both infect humans and animals, as well as Enterococcus faecalis, Ec, Proteus vulgaris and the entire order Bacillales. This methyl modification is installed by the radical SAM (RS) protein Cfr, which is typically encoded extrachromosomally on plasmids or chromosomally together with insertion sequences. Importantly, C8 methylation of A2503 confers resistance to several classes of antibiotics that target the PTC of the ribosome, including phenicols, lincosamides, oxazolidinones, pleuromutilins, and streptogramin A, as well as the macrolides josamycin and spiramycin (4-6). Staphyoloccocus aureus (Sa) Cfr shares 33% amino acid sequence

179 identity with Ec RlmN, and it is believed that Cfr evolved directly from RlmN. In fact, after methylating C8 of A2503, Cfr will also methylate C2 if the site is available, while

RlmN will only methylate C2.

Both Cfr and RlmN share a unique, radical based catalytic mechanism involving the participation of five strictly conserved cysteines (7, 8). Three of these cysteines reside in the canonical CxxxCxxC RS signature sequence, and ligate the requisite [4Fe–4S] cluster found in all RS enzymes (9-12), while the remaining two cysteines (C118 and

C355 in RlmN; C105 and C338 in Cfr) are found on opposing ends of the polypeptide chain. Mechanisms for methylation of C2 and C8 by RlmN and Cfr which account for in vivo and in vitro observations, and which include key roles for the two conserved cysteines that are not cluster ligands, were recently proposed (Figure 1A and Scheme

Figure 1B, respectively) (13, 14). In the RlmN mechanism, one molecule of SAM, bound to the available coordination site of the [4Fe–4S] cluster, is used to methylate C355, which is 4 Å away in the crystal structure of RlmN with SAM bound. Subsequent to release of S-adenosylhomocysteine (SAH), a second molecule of SAM binds to the same site to which the first molecule binds. Upon binding of RNA to the protein and reduction of its [4Fe–4S] cluster to the +1 oxidation state, SAM undergoes reductive fragmentation to a 5’-dA• (15-17), which abstracts a hydrogen atom from the methylcysteine (mCys) residue. The resulting methylene radical then adds to C2 of the target adenosine nucleotide, affording a protein/nucleic acid crosslink containing an unpaired electron extensively delocalized around the aromatic ring (18). Loss of an electron—to an unknown accepter, proposed to be the oxidized [4Fe–4S] cluster (Step 3)—and a proton, results in a thioether linkage proposed to be cleaved via disulfide bond formation via the

180 participation of Cys118. Tautomerization of the resulting enamine affords the final product. The intramolecular disulfide bond formed during turnover would then need to be reduced by 2 electrons to allow subsequent rounds of turnover to take place, beginning with reformation of the mCys residue.

Biochemical and recent spectroscopic observations are consistent with the proposed protein/nucleic acid cross-linked species. When the Cys residue purported to resolve the crosslink is substituted by an Ala residue (C118A) in RlmN, the protein is isolated with covalently bound rRNA (13). More recent studies by McCusker et al. using both the C118A and C118S RlmN variants show unambiguously that a covalent crosslink between C355 of RlmN and C2 of A2503 can be formed in vitro (19). In in vivo studies of Cfr, the analogous variant (C105A) gave rise to a strong pause at A2503 in reverse transcription assays of the RNA substrate—as is seen when C8 of A2503 undergoes methylation—but antibiotic resistance was not conferred, indicating a modification to this nucleotide that was not a methyl group (20). Most recently, Grove et al. provided evidence via electron paramagnetic resonance and electron nuclear double resonance spectroscopies for the paramagnetic intermediate (Scheme 1B) formed upon addition of the methylene radical of the mCys residue onto C8 of the target adenosine in the Cfr reaction. Further isotopic labeling showed unambiguously that the unpaired electron was delocalized throughout the adenine base and that the greatest concentration of spin was found at N7. Moreover, C8 was observed to be sp3-hybridized and bonded to the methylene carbon of the mCys residue as determined from strong hyperfine interactions between a 13C-labeled carbon of the mCys residue and the C8 hydrogen of the base.

181 Another key intermediate in the RlmN and Cfr reactions is the intramolecular disulfide- bond generated concomitant with turnover. The crystal structure of RlmN + SAM shows the two cysteines to be 6 Å apart, a distance that is slightly long for facile disulfide-bond formation, but sufficiently close that subtle conformational changes may allow it to take place. Herein we show that indeed the disulfide species can be generated chemically both in RlmN and Cfr. We also show that this form of the enzyme is capable of being reduced by glutathione (GSH), which allowed determination of midpoint potentials of -268 and -

289 mV for RlmN and Cfr, respectively Interesting, in the absence of two-electron reductants such as GSH and dithiothreitol (DTT), both RlmN and Cfr still catalyze greater than 30 turnovers with no measureable difference from the rate in the presence of two-electron reductants. We therefore amend the mechanism originally proposed by

Grove et al amended to include a disulfide radical anion species, which is more consistent with all observations.

5.3 Materials and Methods

General Methods: All protein manipulations were carried out in an anaerobic chamber under a 95:5 N2/H2 atmosphere, with O2 levels being maintained at < 1 ppm using a constant purge palladium catalyst system (Coy Laboratories, MI). All high- performance liquid chromatography coupled mass spectrometry (LC-MS) analysis of

SAH, 5’-dA, and methyl-adenosine products were carried out as previously described

(16, 19). The 155mer RNA substrate, that was used in activity assays, was prepared and

182 digested as previously described (21). Preparation of non-reducing SDS-PAGE gel loading buffer performed as previously described (22).

DTT free RlmN or Cfr: Preparation of RlmNapoRCN and CfrapoRCN were performed as previously described (19, 21). Two-electron reductants – such as DTT and

BME – were removed by passing 1 mL of the purified, concentrated proteins through a

PD-10 gel filtration column (GE Healthcare Life Sciences) equilibrated in 50 mM

HEPES, pH 7.5, 500 mM KCl, 4 mM MgCl2, and 25% glycerol. The final protein concentrations were determined by Bradford’s assay using the appropriate correction factors (16, 19). The AI wt RlmN and AI wt Cfr were overproduced and purified as described previously before being passed over a PD-10 column as described above.

Preperation of oxTrx and oxDsbA: Thioredoxin (Trx) and DsbA, from E. coli, were cloned and purified as previously described (22-24). The oxidized form of Trx and

DsbA were prepared by overnight incubation in the presence of oxygen. The proteins were then taken in a glovebox and 1 mL of concentrated protein was passed over a PD-10 column equilibrated in an anaerobic buffer composed of 50 mM HEPES, pH 7.5, 100 mM KCl, and 10 % glycerol. The protein concentration of DsbA and Trx was determined using the published extinction coefficient of 21,740 M-1cm-1 and 13,700 M-

1cm-1, respectively. Free thiol concentrations of each protein stock were determined by

Ellman’s reagent and were below the detection limit of this method (25). The reduced form of TrxC35A variant – used in disulfide trapping assays – of Trx was constructed, purified and reduced as previously described (22).

Effect of oxTrx and oxDsbA on the redox state of Cfr and RlmN: To assess the ability of the protein disulfide oxidoreductases Trx and DsbA on the redox state of wt

183

RlmNapoRCN or wt CfrapoRCN, 300 µM oxidized Thioredoxin (oxTrx) or 300 µM oxDsbA was added to a mixture of 70 µM wt RlmNapoRCN or wt CfrapoRCN, 50 mM

Tris-HCl, pH 8.4, 100 mM KCl and incubated for 1, 10, 30, or 60 min at room temperature. At the designated time, a 15 µL sample was removed and mixed with 1 µL of SAM (final concentration of 2 mM). The mixture was incubated for an additional 5 min at room temperature before being quenched with 16 µL of a 100 mM H2SO4 solution containing 100 µM tryptophan. SAH production was monitored for each sample as described above. After the 60 min incubation of RlmN or Cfr in the presence of DsbA,

15 µL aliquots of the remaining solution was mixed with 15 µL of 1, 2, 3, 6, 12, or 25 equivalents of redTrxC35A to trap the disulfide form of the enzymes. The samples were incubated for an additional 30 min before being quenched with 30 µL of non-reducing

SDS-PAGE loading buffer. The protein samples were separated on a 12% non-reducing

SDS-PAGE gel. Proteins were visualized by coomassie brilliant blue staining.

Time-dependent formation of CfrapoRCN Disulfide: In a final reaction volume of

100 µL contained the following: 30 µM CfrapoRCN, 50 mM Tris-HCl, pH 8.4, 100 mM

KCl. Before initiation of the reaction, a zero time point was removed and quenched by adding it to an equal amount of 2 Χ non-reducing SDS PAGE buffer. The reaction was initiated by adding diamide (TCI America) to final concentration of 1 mM. Designated time points were removed and quenched by mixing with 2X non-reducing SDS-PAGE gel loading buffer.

Preparation and MS detection of disulfide bond: To directly detect the intramolecular disulfide bond within wt RlmNapoRCN and wt CfrapoRCN, 30 µM protein

184 was mixed with 50 mM Tris-HCl, pH 8.4, 100 mM KCl and 1 mM diamide for 30 min at room temperature. Subsequently, 100 mM iodoacetamide is added with an additional 30 min incubation. The solutions are then mixed 1:1 with 2x non-reducing SDS-PAGE loading buffer (SDS-PAGE loading buffer lacking reductants) and separated on a 12 % non-reducing SDS-PAGE gel. The gel was stained using coomassie brilliant blue to visualize the protein bands. Bands corresponding to RlmN or Cfr were excised from the gel, diced into ~ 1 mm x 1 mM pieces, and distained with a solution of 50 mM NH4HCO3 in 50 % acetonitrile. The destaining process was completed by alternating addition of 50 mM NH4HCO3 in 50 % acetonitrile and 50 mM NH4HCO3. Once the dye was removed, and additional labeling with 100 mM CAM in of 50 mM NH4HCO3 was performed. The excess CAM was then removed by washing the gel pieces with 50 mM NH4HCO3 in 50

% acetonitrile. The gel pieces were further dehydrated by lyophilization before a 12 µg mL-1 solution of trypsin/gluC in 50 mM NH4HCO3 was added to the gel pieces. The digestion was carried out at 37 ° C for 16 hr. The solution was removed from the gel pieces by careful pipetting and placed in a new eppendorff tube. The gel pieces were further extracted 5 times using a solution of 5 % formic acid in 50 % acetonitrile. The combined extracts were dried to approximately 50 µL.

Mass Spectrometry: The concentrated peptide mixtures were first diluted 1:100, then 2 µL of the peptide solution was loaded onto an Acclaim PepMap100 trapping column (100 µm × 2 cm, C18, 5 µm, 100 Å, Thermo) at a flow rate of 20 µL•min-1 using water as a mobile phase. The peptides were separated with an Acclaim PepMap RSLC column (75 µm × 15 cm, C18, 2 µm, 100 Å, Thermo) with a 30-min (Cfr) or 90-min

(RlmN) 4% - 50% linear gradient of acetonitrile in water containing 0.1% formic acid.

185 The gradient was delivered by a Dionex Ultimate 3000 nano-LC system (Thermofisher

Scientific) at 300 nL•min-1.

An LTQ Orbitrap Velos mass spectrometer was set to acquire data using the following data-dependent parameters: For Cfr digests, full FT MS scans at R 60,000 was followed by 10 ion-trap MS/MS scans on the most intense precursors with CID activation.

Precursors with charge states of +2 and higher were selected for MS/MS. The monoisotopic precursor selection was enabled with an isolation window of 2 m/z. For the

RlmN digests, full FT MS scans at R 60,000 was followed by Orbitrap MS/MS scans (R

7,500). The precursors were selected from the parent mass list and then the 5 most intense precursors were fragmented. The isolation window was set to 1 m/z with a low mass setting of 0.4 m/z and high mass setting of 0.6 m/z from the precursor m/z. This was done to minimize co-isolation of the doubly charged CAM- and d2-CAM-labeled precursors.

Data Processing: The mass spectra were processed using Proteome Discoverer

1.3 (P.D. 1.3, Thermofisher Scientific). The protein identities were confirmed by searching the data with SEQUEST against UniProtKB/Swiss-Prot database (release

2012_10) containing 538,259 entries using standard search parameters. The disulfide- linked peptides were identified using P.D.1.3 by searching a custom database containing

8 sequences: the peptide sequences of RlmN and Cfr with possible cleavages by trypsin and gluC. The search parameters were as follows: precursor tolerance 10 ppm, fragment tolerance 0.8 Da (ion trap), dynamic modifications included carbamidomethyl (+ 57.021

Da, C), d2-carbamidomethyl (+59.034 Da, C), oxidation (+15.995 Da, M), and sample- specific custom modifications of cysteine residues listed in Table 5-1.

186

Table 5-1.Custom peptide sequence used in Proteome Discoverer 1.3. peptide sequences searched in RlmN samples mi m/z composition

* VETVYIPEDDRATLCVSSQVGC ALE 2751.2731 C116 H186 N30 O43 S2

* TVYIPEDDRATLCVSSQVGC ALE 2523.1622 C106 H170 N28 O39 S2

* DDRATLCVSSQVGC ALE 1820.8033 C72 H120 N22 O29 S2

* RATLCVSSQVGC ALE 1590.7494 C64H110 N20 O23 S2

* ATLCVSSQVGC ALE 1434.6483 C58 H98 N16 O22 S2

AACGQLAGDVIDR 1285.6085 C52 H87 N17 O19 S1

AACGQLAGDVID 1129.5074 C46 H75 N13 O18 S1

AACGQLAGD 802.3280 C31 H50 N10 O13 S1

GDDIDAACGQLAGDVIDR 1800.7950 C72 H116 N22 O30 S1

DIDAACGQLAGDVIDR 1628.7464 C66 H108 N20 O26 S1

GDDIDAACGQLAGD 1317.5143 C51 H79 N15 O24 S1 peptide sequences searched in Cfr samples mi m/z composition

SQFGIDIDAACGQLYGNYQNSQGNSSSVDK 3163.3789 C132 H198 N38 O51 S1

IDAACGQLYGNYQNSQGNSSSVDK 2516.0874 C103 H157 N31 O41 S1

AACGQLYGNYQNSQGNSSSVDK 2287.9764 C93 H141 N29 O37 S1

SQFGIDIDAACGQLYGNYQNSQGNSSSVD 3035.2839 C126 H186 N36 O50 S1

IDAACGQLYGNYQNSQGNSSSVD 2387.9924 C97 H145 N29 O40 S1

AACGQLYGNYQNSQGNSSSVD 2159.8814 C87 H129 N27 O36 S1

* * * AGWESFCISSQC GC NFGC K 2194.8329 C91 H130 N26 O30 S4

* * * SFCISSQC GC NFGC K 1751.6524 C70 H105 N21 O24 S4

* denotes Cys was also search with a CAM modification Quality of the MS2 data was checked manually for each cross-linked peptide. The assignment of peaks resulting from the fragmentation of the “Cys modification” peptide was done manually with an aid of MS Product tool (Protein Prospector, http://prospector.ucsf.edu).

187

Determination of the Disulfide Redox Potentials of RlmN and Cfr: A 100 µM solution

(225 µL final volume) of wt RlmNapoRCN, C118S RlmNapoRCN or wt CfrapoRCN containing 100 mM Tris–HCl, pH 8.4, and 300 mM KCl was prepared in the presence or absence of 100 µM diamide. Aliquots (15 µL) of the solution were placed in 14 different tubes containing 5 µL of a 25 mM solution of varying ratios of reduced to oxidized glutathionine (GSH:GSSG) (Table 2). The solutions were incubated for 30 min at room temperature before addition of SAM to a final concentration of 2 mM to allow for methyl transfer to the nucleophilic cysteine. The reactions were quenched by addition of 20 µL of a 100 mM H2SO4 solution containing 100 µM tryptophan (IS). Reactions were analyzed by LC-MS for the production of SAH and 5’-dA as previously described above.

Data were fitted to a modified version of the Nernst equation derived previously

(Equation 1)(22).

188

Table 5-2. Ratios of GSH and GSSG used for redox titrations. [GSH]1 [GSSH]1 Potential* 95 5 -0.398 90 10 -0.377 80 20 -0.366 70 30 -0.346 60 40 -0.339 50 50 -0.331 40 60 -0.323 30 70 -0.314 20 80 -0.302 15 85 -0.294 10 90 -0.283 4 96 -0.259 2 98 -0.241 1 99 -0.223 1Values are in millimolar *Redox potentials were calculated using the midpoint potential of GSH of -240 mV with adjustment for pH (ref)

[ ] [ ( )] [ ( )] { }

Equation 5-1.

Activity Assays in the Presence or Absence of Two-Electron Reductants: The reactions contained the following in a final volume of 100 µL: 10 µM AI RlmN or wt AI

Cfr, 50 mM Tris-HCl, pH 8.4, 2 mM MgCl2, 2 mM SAM, 300 µM 155mer RNA substrate, 200 µM flavodoxin, and 20 µM flavodoxin reductase; in some instances, 10 mM DTT, 10 mM GSH, or 10 mM GSSG was also included. A 10 µL aliquot was removed (T = 0) and quenched with 10 µL of a solution containing 50 mM H2SO4 and

100 µM L-tryptophan, before initiating the reaction by addition of NADPH to a final

189 concentration of 2 mM. Reactions were initiated by addition of 2 mM NADPH. At designated times, 10 µL aliquots were removed and treated in the same fashion. After quenching the reactions, a 40 µL aliquot of 2 P1 nuclease quench buffer (250 mM

NaAcetate, pH 6.0, 45 mM NaCl, and 4 mM ZnCl2), 0.5 U of P1 nuclease, and 5 U of antarctic phosphatase were added, and the RNA was digested at 37 °C for 12 h.

Precipitate formed during the reaction was removed by centrifugation, and the supernatants were analyzed by LC-MS for m2A m8A as previously described (21). A correction factor of 0.6 was necessary to apply to the concentrations determined for m2A, because its response is 1.66 times greater than that for m8A, for which we have an authentic synthetic standard.

Reduction of the Disulfide Bonds in RlmN and Cfr by Various Reductants:

Reactions contained the following components in a final volume of 100 µL: 70 µM of wt

RlmNapoRCN or 70 µM of wt CfrapoRCN , 50 mM Tris-HCl, pH 8.4, 100 mM KCl, 70

µM diamide, and in some instances 200 µM Fdx and 20 µM Fdr. The reactions were incubated for 30 min at 22 °C before adding one of the following reductants: 10 mM

DTT, 10 mM GSH, 1 mM reduced Trx, or 2 mM NADPH when Fdx and Fdr are included. After incubating for an additional 30 min, SAM was added to a final concentration of 1 mM to allow for methyl transfer to take place. After a 5 min incubation, 20 µL aliquots were removed and added to an equal volume of 100 mM

H2SO4 containing 100 µM trp to quench the reactions, which were subsequently analyzed for SAH as described previously (14). A control reaction was carried out as described above with the exception that water was added in the place of reductant.

190 5.4 Results

In vitro formation of disulfide bond.

Our previous studies showed that overproduction of RlmN or Cfr under conditions that result in efficient incorporation of their Fe/S clusters results in isolation of the proteins containing a mCys residue (13). This modification, key to the enzymes’ mechanism of action, blocks the artificial introduction of a disulfide-bond between the modified Cys and the resolving Cys, thereby precluding the possibility of assessing whether this form of the enzyme is chemically competent to be returned to the active form, which contains either two free cysteine thiols or one free cysteine thiol and a mCys residue. More recent studies showed that both RlmN and Cfr could be isolated in the absence of their mCys modifications if they were overproduced in Ec under iron-limiting conditions. This characteristic arises because the exact same SAM binding site, in which its -amino and carboxylate groups coordinate to the unique iron ion of the [4Fe–4S] cluster, supports both methyl transfer to the cysteine nucleophile and generation of the 5’-dA• (14, 21).

Moreover, upon reconstitution of the Fe/S cluster in apo RlmNwt→RCN or apo Cfrwt→RCN, both enzymes catalyzed rapid methylation of the appropriate cysteine nucleophile, and were also fully capable of catalyzing formation of m2A or m8A (14, 21). We therefore used this strategy to produce RlmN and Cfr in a form in which their catalytically active cysteines exist as free thiols to allow for the possibility of artificially introducing an intramolecular disulfide bond by incubating them in the presence of diamide and other oxidants.

191

Both wt CfrapoRCN and wt RlmNapoRCN migrate as monomers on non-reducing SDS-

PAGE in the absence or presence of 2-electron reductants (Figure 5-2).

M 1 M 2 97.0 66.0

45.0 30.0

20.0

Figure 5-2. non-reducing SDS-PAGE gel of wt RlmNapoRCN and wt CfrapoRCN in the absence of reductants: lane M, molecular weight marker; lane 1, CfrapoRCN; lane 2, RlmNapoRCN. Standards on the left side indicate protein size in kDa.

As the crystal structure of RlmN shows, a disulfide bond formed between Cys355 and

Cys118 would be buried deep in the active site, with poor access to the surface of the protein where it could be acted upon by protein disulfide oxidoreductases such as Trx or

DsbA (Figure 5-3). Indeed, when 70 µM wt RlmNapoRCN or 70 µM wt CfrapoRCN was incubated with 300 µM oxTrx no loss in the ability of either enzyme to transfer a methyl group to Cys355 or Cys338, respectively, was observed, indicating that the catalytically active cysteines remained in their free thiol forms. However, a 20 % decrease in the amount of SAH produced by RlmN was observed after a 60 min incubation of the protein with oxDsbA (Figure 5-4A), while only a 2 % decrease in SAH production was observed

192 when Cfr was incubated under identical conditions (Figure 5-4B). To differentiate between formation of an intramolecular disulfide-bond as opposed to an intermolecular disulfide-bond between two Cys338 (Cfr) or two Cys355 (RlmN) residues, which reside on flexible and mobile loops,, a variant of Trx, TrxC35A, was used in an effort to trap the intermolecular disulfide bond. The TrxC35A variant contains the nucleophilic cysteine that initiates disulfide-bond reduction via formation of a heterodisulfide bond, but lacks the resolving cysteine that effects cleavage of the heterodisulfide bond, resulting in crosslinking of the two proteins (27, 28). Indeed, when TrxC35A was incubated with

RlmNapoRCN that was first oxidized by DsbA, a new band of about 60 kDa appears on a non-reducing SDS-PAGE gel (Figure 5-4C). In addition, when this oxidized form of

RlmNapoRCN was reacted with increasing equivalents of TrxC35A, the 60 kDa band increased in intensity (Figure 5-4C, marker labeled R-T35A). Interestingly, the increase in intensity of the 60 kDa band is directly proportional to the decrease in the intensity of another band present on the gel. This band displays an apparent molecular mass of 95 kDa, the size of two RlmN monomers cross-linked by Cys355 (Figure 5-4C, marker labeled R-R). This same behavior, although less pronounced, was also observed with CfrapoRCN (Figure 5-4D), indicating that large protein disulfide oxidoreductases are ineffective at generating the proposed intramolecular disulfide bond intermediate due to poor access, and are therefore unlikely to participate in the catalytic mechanisms of these enzymes by reducing the intramolecular disulfide bond proposed to form during turnover.

The disulfide-bonded dimeric forms of RlmN and Cfr were confirmed by MS, as were their disulfide-bonded heterdimeric forms in complex with Trx.

193

Figure 5-3. Active site of RlmN in the presence of SAM demonstrating the proximity of Cys118 and mCys355.

We therefore focused attention on small-molecule thiol-oxidizing agents, such as diamide, which might have better access to the catalytically active cysteines (29). The mechanism by which diamide oxidizes free thiols is shown in Figure 5-5 (30, 31). Its reaction is irreversible and, unlike protein and small-molecule oxidoreductases that contain sulfhydryl groups, it does not exist in a dynamic equilibrium with thiol- containing compounds.

194

O CH3 H3C CH3 O CH3 N CH3 H R2 H C N N H3C N N SH 3 N N SH N N CH S S R + N N CH3 + R 3 + 1 O N CH3 2 H R1 CH3 O CH3 O S OH R1

Figure 5-4. Reaction catalyzed by diamide in the presence of free thiols.

When diamide was incubated with CfrapoRCN, an apparent decrease in the size of the protein was observed by non-reducing SDS-PAGE analysis (Figure 5-6A). Importantly, this apparent increase in mobility is consistent with formation of an intramolecular disulfide bond, as has been observed previously for other proteins acted upon by diamide, and was not seen when CfrapoRCN was incubated with DsbA (Figure 5-4D)(32).

Additionally, disulfide-bond formation is rapid, with > 90 % of the shift occurring in the first minute after addition of diamide (Figure 5-6A, lane 2), and is accompanied by the loss in ability to transfer a methyl group from SAM to C338 in CfrapoRCN. No apparent shift in mobility was observed when RlmN was incubated with 1 equiv of diamide

(Figure 5-6B); however, the protein similarly lost its ability to catalyze methyl transfer to its nucleophilic cysteine (C355). It should be noted that bands corresponding to molecular masses of 95 kDa (RlmN, Figure 4B) and 80 kDa (Cfr, Figure 4A) are also observed when RlmN and Cfr are treated with diamide, which represent forms of the proteins containing intermolecular disulfide bonds between two Cys355 residues or two

Cys338 residues. The addition of more than 6 equiv of diamide does not induce additional formation of these intermolecular disulfide bonds (Figure 5-6B, lane 4).

195

Figure 5-5. Effect of oxidized DsbA on intramolecular disulfide bond formation. Time dependent production of SAH by A) 70 µM wt RlmNapoRCN and B) 70 µM wt CfrapoRCN in the presence of 300 µM oxDsbA. C) non-reducing SDS-PAGE gel of wt RlmNapoRCN and D) wt CfrapoRCN after a 60 min incubation with 300µM oxDsbA followed by reaction with 1, 2, 3, 6, 12, 25 equivalents of TrxC35A. lane M, molecular weight marker; lane 1, mixture in the presence of 1 mM DTT; lane 2, 1 eq. TrxC35A; lane 3, 2 eq. TrxC35A; lane 4, 3 eq. TrxC35A; lane 5, 6 eq. TrxC35A; lane 6, 12 eq. TrxC35A; lane 7, 25 eq. TrxC35A. Arrows on the left side indicate bands of importance. R-R, RlmN disulfide dimer; R-TC35A, RlmN-TrxC35A mixed disulfide; R, RlmN monomer; DsbA, oxidized DsbA; C-C, Cfr disulfide dimer; C-TC35A, Cfr-TrxC35A mixed disulfide; C, Cfr monomer. Standards on the left side indicate protein size in kDa.

Given that a disulfide bond between Cys355 (Cys338) and Cys118 (Cys105) blocks methyl transfer, reduction of oxRlmNapoRCN or oxCfrapoRCN by 10 mM DTT should restore this activity, as measured by increased SAH production. Figure 5-7 shows that disulfide- bond formation is reversible. When RlmN was mixed with 1 eq of diamide, no peak

196 corresponding to SAH appears (Figure 5-7, green trace retention time, 3.6 min). When the protein is then incubated with 10 mM DTT and subsequently reacted with SAM, a peak at 3.6 min appears (Figure 5-7, blue trace) which is equivalent to the amount of

SAH produced before diamide treatment. In addition, when the Fe/S cluster is degraded to a [3Fe-4S] cluster, both RlmN and Cfr are incapable of methyl transfer from SAM to the requisite Cys, even in the presence of DTT.

Figure 5-6. non-reducing SDS-PAGE gels of reaction between diamide and wt Cfr (A) or wt

RlmN (B). A) Time dependent formation of a disulfide bond on wt CfrapoRCN via 1 mM of diamide. The reactions were quenched with non-reducing SDS-PAGE buffer: lane M, molecular weight marker; lane 1, before diamide addition; lane 2, 1 min; lane 3, 5 min; lane 4, 10 min; lane

30, 1 min; lane 6, 60 min. B) Reaction of wt RlmN apoRCN with 1, 2, 6, 10 equivalents of diamide. The reactions were quenched with non-reducing SDS-PAGE buffer: lane M, molecular weight marker; lane 1, 1 eq.; lane 2, 2 eq.; lane 3, 6 eq.; lane 4, 10 eq. Standards on the left side indicate protein size in kDa.

197

HRMS Characterization of the Disulfide Bond

oxidizing the protein with 6 eq. of diamide. The disulfide-bonded forms of the proteins were then treated with 100 mM IAM to block all free cysteinyl thiols, before being denatured with non-reducing SDS-PAGE buffer and separated on a 12% SDS-PAGE gel (Figure 5-8). As was shown above, the predominant

-bonded; only a small

198 portion of the protein formed the intermolecular disulfide bond. Formation of an intramolecular disulfide bond on Cfr induces a mobility shift; therefore the oxidized and reduced forms of the enzyme can be readily separated by a non-reducing SDS-PAGE gel allowing characterization of both forms of the enzyme individually. Since the disulfide- bonded form of RlmN does not induce a shift in apparent mobility, separation of the two species is not possible. The upper band and lower band from the diamide-oxidized

-8) were excised and reacted with additional IAM to label any remaining cysteines that were protected by the fold of the protein or the Fe/S cluster.

RlmN Cfr M apoRCN M apoRCN 97.0

66.0

45.0

30.0

20.0

14.0

Figure 5-8. Generation of a disulfide in wt RlmNapoRCN and wt CfrapoRCN via diamide oxidation for in-gel digestion and HRMS/MS analysis. The bands corresponding to RlmNapoRCN (~47 kDa), oxidized Cfr apoRCN (~43 kDa, bottom band), and Cfr apoRCN (~45 kDa, top band) were excised, labeled, and digested as described in the Materials and Methods. Note the a small amount of dimer formation in the RlmN and Cfr samples (~ 95 kDa and ~ 80 kDa, respectively).

199

Figure 5-9. Selective labeling results of peptides analyzed by HRMS from the top band (A) and bottom band (B) from the CfrapoRCN reaction with diamide. Bands were excised from the gel, reduced with DTT, labeled with d2-iodoacetamide, and digested with trypsin/gluC proteases. Letters above amino acids denote the type of modification. If two labels are present, the label closest to the sequence is in a higher concentration. C, CAM; O, oxidation; d, d2-CAM. Green highlighted areas indicated sequence coverage. . The cysteines that initially formed a disulfide bond would be protected from this additional treatment with IAM. The gel pieces were then reduced with DTT, which reduces the disulfides back to free thiols allowing them to react with added d2- iodoacetamide. The labeled proteins were then in-gel digested with trypsin/gluC cocktail of proteases. Utilizing this labeling strategy, cysteines that were protected by disulfide bonds generated in the reaction with diamide would carry a d2-CAM modification.

Figure 5-9 shows the results generated by HRMS for the selective labeling of top band

(A) and bottom band (B) of the CfrapoRCN oxidized samples. The top band was fully labeled with only iodoacetamide (C above each cysteine) and therefore Cys105 and

200 Cys338 were not in a disulfide bond. The results generated by HRMS for the bottom band, however, show that both Cys338 and Cys105 are predominately labeled with d2-CAM

(d above each cysteine) and were therefore protected from carbidomethylation after oxidation with diamide. Interestingly, all cysteines in the top band (Figure 5-9A) of the diamide-oxidized Cfr sample were labeled with CAM, including the cluster ligands. This may have resulted from the top band containing a form of Cfr that was soluble, yet not folded correctly, and therefore did not reconstitute an Fe/S cluster. It may have also resulted from a correctly folded protein that lacked an Fe/S, which rendered the active site of Cfr incapable of forming the disulfide bond. If the latter is true, then it would imply that the [4Fe–4S] on Cfr, and by analogy RlmN, is involved in disulfide bond formation. This issue is not currently resolved.

MS/MS analysis of the intact tryptic peptides cross-linked through the disulfide bond would give direct confirmation of the disulfide-bonded form of the proteins. In addition, RlmNapoRCN does not shift in apparent molecular weight when oxidized by diamide. Analysis of the two forms, as was done with Cfr, would be inconclusive. Cys118 and Cys355 would contain both d2-CAM and CAM modifications. As above, the bands corresponding to the disulfide-bonded forms of RlmNapoRCN and CfrapoRCN were excised from an SDS-PAGE gel, digested with the trypsin/gluC cocktail of proteases.

The extracted peptides were then subjected to LTQ-Orbitrap hybrid tandem MS/MS analysis. The mixtures of digested peptides were first separated on a UPLC C-18 column, then each parent disulfide-containing peptide was isolated with the MS by matching the

201 molecular mass of the predicted sequence with that of the detected sequence.

202

Figure 5-10. Results of LTQ-Orbitrap MS/MS analysis of peptides from wt RlmNapoRCN and wt CfrapoRCN trypsin/gluC in-gel digestions. A) HRMS of RlmN disulfide containing peptide with a monoisotopic m/z of 1080.1646. B) HRMS of Cfr disulfide containing peptide with a monoisotopic m/z of 1230.2710. C) MS/MS analysis of the RlmN disulfide linked peptide with m/z of 1080.1646. Fragmentation patterns show the link is between Cys118 and Cys355. The table to the right is a list of m/z values for b and y ions labeled in this spectrum. D) MS/MS analysis of the Cfr disulfide linked peptide with m/z of 1230.2710. Fragmentation patterns show the link is between Cys105 and Cys338. The table to the right is a list of m/z values for b and y ions labeled in this spectrum.

This HRMS method allowed characterization of the intact disulfide containing peptides from both RlmNapoRCN and CfrapoRCN (5-10). In Figure 5-10A, the disulfide containing peptide from RlmN, with a monoisotopic m/z of 1080.1646 (z = 3), could be isolated and fragmented. The fragmentation pattern (Figure 5-10C) definitively shows the crosslink between Cys118 of one peptide and Cys355 of the second peptide. In addition, the MS/MS analysis also shows the Cys125 contains a carbidomethylation

(underlined). In Figure 5-10B, the disulfide containing peptide from CfrapoRCN was also isolated with a monoisotopic m/z of 1230.2710 (z = 3). The fragmentation pattern definitively shows the disulfide crosslink is between Cys105 and Cys338 of Cfr (Figure 5-

10D). Taken together, these results establish that a disulfide bond can be formed chemically on both RlmNapoRCN and CfrapoRCN.

Redox potential of disulfide bond

Generation of a disulfide bond by an unnatural oxidant such as diamide may mean the resulting form of the enzyme is not physiologically relevant. One way to demonstrate physiologically relevance is to show catalytic/kinetic competence; the enzyme can go from the oxidized form to an active form with natural reductants. An additional method, and more applicable to a disulfide bond, is to determine if the redox potential is in the

203 range of in vivo reductant. Since both RlmN and Cfr have a direct indicator of the disulfide redox state in SAH production, forming the disulfide bond with diamide, then mixing the protein with different ratios of oxidized (GSSG) and reduced (GSH) glutathione should allow measurement of the midpoint potential of the disulfide bond on both RlmNapoRCN and CfrapoRCN (33).

RlmN and Cfr were first mixed with one equivalent of diamide, incubated for 30 min, and then mixed with 5 µL of 1-14 of the redox buffers indicated in Table 5-2. The mixtures were incubated for an additional 30 min, before 1 mM SAM was added. The reactions were quenched with acid and the production of SAH was monitored versus the redox potential of each solution (Figure 5-11). The data were then fit to a derivative of the Nerst Equation (Eq. 1), which correlates the active form (free thiol form of RlmN and

Cfr) of the enzyme with the amount of SAH produced (22). From these fits, a redox potential of –268 for oxidized RlmNapoRCN and –289 for oxidized CfrapoRCN. These redox potentials fall within the physiological range of the reducing environment of Ec.

Additionally, if the disulfide bond were to be reduced in vivo then it would be expected that GSH (> 4 mM in vivo concentration) would be the redox partner (34-36).

As can be seen in Figure 5-12, when the protein are treated only with the redox buffers in

Table 5-2, without first being oxidized by diamide, the same redox dependent response is observed. More importantly, a fit to the data results in the same redox potentials of Cfr and RlmN (-319 and –273, respectively). The data generated for the Cfr titration in the absence of diamide arises from Cfr’s instability. Interestingly, Cfr is much more stable in the oxidized form.

204

RlmNapoRCN and wt CffapoRCN is low because of the poor stability of these proteins. Dashed lines represent the fit of the data using Equation 1.

205 Mediation of electrons to the disulfide bond

As shown above, formation of a physiological, intramolecular disulfide bond within

RlmNapoRCN and CfrapoRCN is possible. If this disulfide form of these enzymes is the penultimate state in the reaction mechanism, then electrons to reduce the disulfide must derive from one of two paths; 1) passed from the surface of the protein to the active site by mediation through the RS [4Fe-4S] cluster as is the case for Ferredoxin:Thioredoxin

Reductase (FTR) in plants (37, 38) or 2) the in vivo, small molecule two electron reductant glutathione. As shown above, glutathione and DTT are both capable of interacting and reducing the disulfide bond. The ability of the in vivo reducing,

F/FR/NADPH, was also tested for its capacity to reduce the disulfide in one electron steps. When 70 µM RlmNapoRCN and CfrapoRCN was oxidized by 1 eq. of diamide, then reacted with 2 mM NADPH less than 1% of the enzyme (~ 1 µM) is returned to the active form (Figure 5-13). This small amount of active enzyme generated by the

F/FR/NADPH reducing may seem promising, but this reaction is exceedingly slow as only about 1 µM of active enzyme is generated after 30 min of reaction. By comparison, if GSH is added back in place of NADPH (Figure 5-13, green trace), greater than 90% of RlmNapoRCN SAH production is restored. This result was surprising, but shows that under these conditions, the only electron carriers capable of reducing the disulfide bond are DTT and GSH. In addition, reduced Trx was not capable of reducing the disulfide, again mostly likely due to its lack of accessibility.

206

Figure 5-12. Reduction of disulfide bond by F/FR/NADPH reducing system. wt RlmNapoRCN was mixed with diamide and then 2 mM SAM (black dotted line), or diamide then F/FR/NADPH, then 2 mM SAM (blue line). wt CfrapoRCN was mixed with diamide (red trace), or diamide then F/FR/NADPH, then 2 mM SAM (black trace). For comparsin wt RlmNapoRCN was mixed with diamide then 10 mM GSH, then 2 mM SAM (green trace). The peak at 3.6 min corresponds to SAH, while the peak at 5.2 min corresponds to tryptophan (IS).

Multiple Turnover of RlmN and Cfr absence or presence of reductants

The catalytic role of the disulfide bond can be addressed by removing all two electron reductants from the system and performing multiple turnovers with the F/FR/NADPH system. Above it was shown that the F/FR/NADPH system was inadequate to reduce the disulfide state of Cfr and RlmN. Therefore it is predicated that after the first turnover, both RlmN and Cfr would stall in an inactive state disulfide-bonded state. Only if GSH or DTT were supplied the enzymes would they be capable of multiple turnovers. Since

207 RlmN and Cfr do not catalyze significant abortive cleavage of SAM in the presence the

F/FR/NADPH system, one of these two outcomes should be obvious (13).

As shown in Figure 5-14A, RlmN is capable of catalyzing greater than 30 turnovers in the absence (Figure 5-14A, blue squares) or presence of GSH (Figure 5-14A, black triangles), as well as the presence of DTT (Figure 5-14A, red circles). In addition,

-1 2 V•[ET] rates for formation m A were determined. In the presence of 10 mM GSSG,

2 -1 -1 - RlmN catalyzes production of m A with V•[ET] of 0.93 min . This rate drops 0.63 min

1 in the presence of 10 mM GSH. The lower activity most likely arise from the interaction of GSH, in high concentrations, with the open coordination site of the RS [4Fe-4S] cluster therefore inhibiting the binding of SAM for subsequent turnovers (39). When

-1 -1 RlmN catalyzes the reaction in the presence of 10 mM DTT, V•[ET] is 0.84 min .

Surprisingly, these results indicate that the presence of GSH or DTT do not have a direct role in catalysis. Cfr also follows this trend (Figure 5-14 B). In the of presence of 10 mM GSSG (Figure 5-14B, blue squares), Cfr catalyzes the production m8A with a rate of 1.07 min-1, while in the presence of 10 mM GSH (Figure 5-14B, black triangles) it drops to 0.82 min-1

208

209 5.5 Discussion

The radical SAM dependent enzymes RlmN and Cfr catalyze a unique methyl synthase reaction, which requires radical based chemistry. The methyl synthesis is directed to a specific adenosine (2503 in E. coli nomenclature) located within domain V of the 23S bacterial ribosomal RNA, and is presumably installed during transcription of the RNA or early in ribosome assembly, given that methylation is significantly more efficient with naked 23 S rRNA (8). The methylation reaction was shown actually to be a methyl synthesis rather than a methyl transfer, given that only two protons of the transferred methyl carbon of SAM are retained in the product. The remaining proton derives either from solvent or from the hydrogen originally located at C2 or C8 of adenosine (13, 40). A mechanism for catalysis by RlmN and Cfr was proposed by Grove et al (Figure 5-1) that supports all of the data in the literature (13). The mechanism begins with the free thiol of Cys355 (Cys338 in Cfr) acting as a nucleophile and attacking the methyl group of SAM bound to the RS [4Fe-4S] cluster. This key step has been supported by a 2.05 Å RlmN+SAM crystal structure (Figure 5-3). The structure revealed that the methylCys355 residue was already in place, but more importantly showed that Cys355 is in close proximity (~ 4 Å) to the methyl group of SAM and appropriately positioned for an inline nucleophilic attack (41). Additional support for this unique methyl transfer was demonstrated by Grove et al in which RlmN was purified under iron limiting conditions. Under these conditions, RlmN was purified without a

[4Fe-4S] cluster, which yielded protein lacking a methyl-Cys modification at Cys355 (14).

Importantly, methyl transfer from SAM to Cys355 was only restored after the [4Fe–4S]

210 cluster was reconstituted in the RS domain of the protein. It was proposed that both

RlmN and Cfr followed the “principle of economy in the evolution of binding sites,” which allowed them activate SAM – within the same binding site—for both polar chemistry (i.e. methyl transfer) and radical based cleavage of SAM to 5’-deoxy-5’-yl- adenosine radical (14). Recently, Cfr was also shown to adopt this strategy (21).

The marquee and most provocative intermediate the mechanistic proposal by Grove et al was the radical base-covalent adduct between A2503 and the methylCys355 of RlmN

(Figure 5-1, step 2-3). The radical addition to A2503 has been compared to the well know Minisci reaction, as it involves both an electrophilic amidine carbon (C2 of A2503) as well as a nucleophillic carbon radical (18). This cross-linked intermediate was originally proposed because of isotope labeling and also a Cys118Ala variant of RlmN isolated with bound ribonucleotides that could only be release by nuclease treatment (13).

Recently, McCusker et al have provided additional support for this cross-linked species.

When the treated the variants of Cys118 in RlmN in vitro, they were able to isolated this cross-linked species by gel electrophoresis as well as identify it by MS/MS analysis (42).

A major issue with both of these observations is the lack of chemical competence of the thioether crosslink. The variant forms of RlmN were completely inactive and therefore the crosslink could have arrived through aberrant. Since these variants are inactive, the kinetic competence of the thioether crosslinked species could not be tested. Importantly, a recent study by Grove et al with Cfr has shown that not only is the radical thioether crosslinked intermediate formed, but is kinetically competent. The radical intermediate was characterized by both EPR and ENDOR spectroscopies in concert with isotopically labeled adenosine-containing rRNA substrates as well as the methyl group appended to

211

Cys338. Grove et al also showed that the trapped radical intermediate forms and decays with rate constants that are consistent with rate of formation of m8A (21).

This thioether crosslinked intermediate was originally proposed to be cleaved by formation of an intramolecular disulfide bond. Characterization of this key disulfide intermediate in both RlmN and Cfr has proven elusive. A clear route to trapping this state would be to essentially run the reaction in reverse. By isolating Cfr and RlmN under iron limiting conditions, as has been done before, allows preparation of the conical cysteines as free thiols (14). By oxidizing this form of the enzyme with thiol oxidoreductases, the disulfide form could be generated.

RlmN and Cfr were both produced in the apo form, which upon reconstitution in presence of iron, sulfur and DTT allowed generation of the RlmNapoRCN and CfrapoRCN forms of the enzymes. After removal of the excess DTT, by buffer exchange, the proteins were mixed and reacted with both oxidized Trx and oxidized DsbA.

Thioredoxin is a common thiol oxidioreductase utilized in studies characterizing of disulfide containing proteins (27, 43-46). DsbA is an E. coli periplasmic disulfide oxidoreductases with a highly elevated midpoint potential (-187 mW), which makes it a very potent oxidant (23, 24, 47). If Cfr and RlmN indeed utilize and intramolecular disulfide bond in their mechanism, then the disulfide would need to be reduced after each turnover.

Monitoring the oxidation state of the disulfide on RlmNapoRCN and CfrapoRCN can easily be accomplished by determining the amount of SAH that is produced by the

212 enzymes. Previously, it was shown that both wt RlmNapoRCN and CfrapoRCN could be reacted with SAM and an immediate burst of SAH, that was equivalent the concentration the RS [4Fe-4S], was produced (14). If Cys355 in RlmN or Cys338 in Cfr is participating in a disulfide bond with Cys118 or Cys105, respectively, then they will be blocked for methyl transfer. When wt RlmNapoRCN and CfrapoRCN were mixed with oxTrx, no appreciable loss in methyl transfer activity is observed. By contrast, when oxDsbA is mixed with wt RlmNapoRCN after 60 min about 20 % of methyl transfer is lost (Figure 5-

4A). The effect on CfrapoRCN is not as high, but an effect is observed (Figure 5-4B). To determine is this effect is a physiological intramolecular disulfide bond, both wt

RlmNapoRCN and CfrapoRCN were oxidized with oxDsbA for 60 min, then mixed with various equivalents of TrxC35A. The C35A variant of Trx is incapable of resolving a disulfide bond and has been shown previously that it is capable of forming mixed disulfide crosslinks by reacting with the oxidized disulfide bond on proteins. As can be seen in Figure 5-4C and Figure 5-4D, wt RlmNapoRCN and CfrapoRCN are forming a mixed disulfide with TrxC35A, with the amount increasing with the increasing equivalents of TrxC35A as well as correlating with the amplitude of loss in methyl transfer activity; i.e. RlmN forms more crosslink than Cfr. An issue with this result is readily apparent on the gels though, as a band that corresponds to the disulfide dimer decreases in intensity with the increase in intensity of the crosslink species. Forming the dimer in RlmNapoRCN and CfrapoRCN is not in itself a problem, as crosslinked species form with free cysteines in the absence of reductant. The issue is that the intermolecular

213 disulfide bond gives rise to the mixed disulfide with TrxC35A. Therefore, Trx and DsbA are not capable of directly forming the intramolecular disulfide bond on Cfr or RlmN.

Since it appears that protein oxidoreductases are not capable of accessing the active site of RlmN and Cfr, a small molecule thiol oxidant may work. When wt CfrapoRCN is reacted with diamide an immediate mobility shift is visible on a non-reducing SDS-

PAGE gel (Figure 5-6A). This shift is encouraging as it is common feature of disulfide bond containing proteins, including DsbA and DsbB (23, 24, 32, 48). Importantly, this shift occurs rapidly and accompanies loss in methyl transfer activity in both wt

RlmNapoRCN and CfrapoRCN. On the other hand, wt RlmNapoRCN has no visual phenotype to monitor the state of oxidation (Figure 5-6B). Oxidation of wt

RlmNapoRCN and CfrapoRCN with diamide abrogates methyl transfer, but can be restored by the addition of DTT meaning diamide is not causing a denaturation of the proteins and the disulfide bond formation is reversible. (Figure 5-6).

Since oxidation of a protein by diamide is not a direct way to visualize the disulfide bond on both RlmNapoRCN and CfrapoRCN, an LTQ-Orbitrap MS was employed to search for and identify the disulfide crosslinked peptides. To do this experiment both

RlmNapoRCN and CfrapoRCN were first oxidized with diamide, labeled with iodoacetamide, separated on a non-reducing SDS-PAGE gel, and finally digested with a mixture of trypsin and gluC. Due the large size of the predicted disulfide crosslinked peptides, a gluC was used in addition to trypsin to cut proteins on the c-terminal side of glutamate residues. When these peptide fragments are isolated and analyzed by HRMS coupled to MS/MS analysis, both the RlmN and Cfr disulfide crosslinked peptides can be

214 characterized (Figure 5-10). Also, the MS/MS analysis of the peptides shows that the peptides are linked through the correct cysteine residues; namely Cys118 and Cys355 in

RlmN and Cys105 and Cys338 in Cfr (Figure 5-10C, D). In addition, wt CfrapoRCN can be separated on the state of this disulfide bond by non-reducing SDS-PAGE gels.

Therefore, if the two forms of the protein are separated on a gel, treated with DTT, additional labeling of free cysteines with d2-iodoacetamide, and digestion as above then redox state of the cysteines can be directly addressed as cysteines in a disulfide bond would be protected from the first labeling with iodoacetamide and instead would contain d2-CAM. Indeed, when this experiment is performed the MS/MS analysis of the isolated peptide fragments show that the top band in gel of oxidized wt CfrapoRCN only contains

CAM modifications (Figure 5-9A) indicating that no cysteines in this band were in disulfide bonds. The bottom band on the other hand has some d2-CAM modifications at

Cys110, Cys112, and Cys116 as well as predominately CAM modifications at these residues.

More importantly, Cys118 and Cys338 are predominantly labeled with the d2-CAM indicating they were protected from CAM modification by a disulfide bond. It is interesting that the top band of Cfr did not contain d2-CAM modifications for the cluster ligands as the bottom band did (Cys112, and Cys116). This is most likely due to the top band containing Cfr that was either 1) not successfully reconstituted with [4Fe-4S] cluster, as has been shown for RlmN, or 2) protein that was isolated in a soluble form but is improperly folded (14).

An important aspect of forming a disulfide bond is proving that it is physiologically relevant. To show that the intramolecular disulfide bond formed on wt RlmNapoRCN and

215

CfrapoRCN, the proteins were first oxidized with diamide and then mixed with 15 different rations of glutathione (GSH) and oxidized glutathione (GSSG) to pose the redox potential of the solution from -223 mV to -398 mV which was calculated based on the change of - 54 mV for every increase in one pH unit (Table 5-2) (49). Under these conditions a clear sigmoidal trend is found in SAH production for both wt RlmNapoRCN and CfrapoRCN with fits - using a derivative of the Nernst equation – of the data yielding

– 268 mV and – 289 mV, respectively (Figure 5-11A, B). In addition, when the redox titration are conducted in the absence of diamide with both wt RlmNapoRCN and

CfrapoRCN, the same trend and midpoint potentials are observed (Figure 5-11 C, D).

Importanly, when the variant C118S RlmNapoRCN is titrated with this redox series no loss in production of SAH is seen (Figure 5-11C, squares). These results indicate that

GSH is capable of accessing the active site disulfide of RlmN and Cfr.

With the disulfide in hand, the next logical step is to determine how this disulfide gets reduced in vivo to form the active enzyme. GSH has been shown to be capable of performing this reaction, but is the in vivo reducing system also capable of reducing the disulfide form of the enzyme by mediating electrons to the disulfide bond through the

Fe/S cluster. The [4Fe-4S] cluster containing protein Ferrodoxin/Thioredoxin reductase does this exact reaction in plants by ferrying electrons from Ferrodoxin to Trx through the active site Fe/S cluster and an invariant pair of cysteine residues located at opposite ends of the polypeptide sequence (7). To test this proposal, wt RlmNapoRCN and

CfrapoRCN were oxidized with diamide and mixed with F/FR/NADPH. Under these

216 conditions, only about 1% of the methyl transfer is restored to RlmN and Cfr, indicating that this is not feasible reaction (Figure 5-12).

Since the in vivo reducing is not sufficient to reduce the penultimate intermediate in the reactions catalyzed by Cfr and RlmN, but is capable of starting the reaction and making the thioether crosslinked intermediate then it would be expected that reactions lacking GSH or DTT would only catalyze the first turnover and stop after product release. Surprisingly, when wt Ai RlmN and Cfr are purified away from DTT and adding back to reactions that either have no DTT, 10 mM GSSG—which completely inhibits methyl transfer—or 10 mM GSH the reactions proceed unabated for ~ 30 turnovers

(Figure 5-13). The only inhibition of turnover is seen when 10 mM GSH is supplied in the reaction. Under these conditions, an inhibitory effect could be imagined where GSH binds to the open coordination site of the RS cluster. This effect has been demonstrated with lysine 2,3 – aminonmutase, the conical member of the RS superfamily (39).

Figure 5-14. Methylation of C2 in A2503 by RlmN with radical disulfide anion formation driving release of product.

217 With these results in hand, a modified mechanism can be proposed for methylation of

A2503 by RlmN and Cfr. In Figure 5-14 a modified mechanism for the RlmN dependent methylation of C2 is outlined. The major difference between this mechanism and the previously proposed mechanism is the pathway with which the electron takes out of the thioether crosslinked radical (Figure 5-14, steps 3-5). Previously, it was proposed that the electron would either reduce the RS Fe/S cluster or be transferred to an unknown acceptor after which acid base chemistry could occur (Figure 5-1A). The mechanism proposed in Scheme 3 suggests that the radical is not transferred out of the system first, but instead deprotonation occurs as a result of the radicals proximity to C2. Under normal conditions an sp3 carbon-based proton is not particularly acidic, but it is well known radicals depress the pKa of adjacent protons by as much as 10 units (50). This would allow for facile deprotonation by basic residues within the protein active site. In addition, the deprotonation would drive the rearrangement of the electronic structure of the radical intermediate, producing a formal radical on C2 (Figure 5-14, steps 3, 4).

After rearrangement, radical fragmentation of the thioether bond – aided by the approaching Cys118 – would result in tautomerization of the product and release. The radical disulfide anion would require an electron from the [4Fe-4S] cluster to regenerate the active state of the enzyme. This mechanism predicts two things. 1) the decay rate of the radical thioether crosslinked intermediate that was recently trapped in Cfr would have a significant isotope effect as the rate limiting step in the reaction is before the radical intermediate and base removal of the proton at C8 is the likely next step in the reaction.

2) Variants of C118 of RlmN would generate the thioether crosslikned radical intermediate that was trapped in Cfr. This is because radical release from the system is

218 dependent on the formation of the radical anion disulfide bond and not loss of the electron to the RS cluster. Both of these proposals are readily testable.

219 5.6 Conclusions

Our results, herein, provide strong evidence that a disulfide bond is not formed during turnover of RlmN and Cfr. In addition, we show that the F/FR/NADPH system can provide the electrons required by RlmN and Cfr for turnover. We also demonstrate that under these conditions, RlmN and Cfr are capable of catalyzing ~ 30 turnovers of this complicated methyl syntheses reaction with V•[ET]-1rates of about 1 min-1. This is the first time steady state rates have been determined for RS dependent methylation.

From these results we are also able to propose a modified reaction mechanism for Cfr and

RlmN, which includes formation of a radical disulfide anion, reminiscent of the intermediate formed the α-subunit of RNR during ribonucleotide reduction.

220 5.7 References

1. Toh, S.-M., Xiong, L., Bae, T., and Mankin, A. S. (2008) The methyltransferase

YfgB/RlmN is responsible for modification of adenosine 2503 in 23S rRNA, Rna-a

Publication of the Rna Society 14, 98–106.

2. Vazquez-Laslop, N., Ramu, H., Klepacki, D., and Mankin, A. S. (2010) The key role of a

conserved and modified rRNA residue in the ribosomal response to the nascent peptide,

EMBO J. 29, 3108–3117.

3. Benitez-Paez, A., Villarroya, M., and Armengod, M. E. (2012) The Escherichia coli

RlmN methyltransferase is a dual-specificity enzyme that modifies both rRNA and tRNA

and controls translational accuracy, Rna-a Publication of the Rna Society 18, 1783-1795.

4. Kehrenberg, C., Schwarz, S., Jacobsen, N. E., Hansen, L. H., and Vester, B. (2005) A

new mechanism for chloramphenicol, florfenicol and clindamycin resistance: methylation

of 23S ribosomal RNA at A2503, Mol. Microbiol. 57, 1064–1073.

5. Long, K. S., Poehlsgaard, J., Kehrenberg, C., Schwarz, S., and Vester, B. (2006) The Cfr

rRNA methyltransferase confers resistance to phenicols, lincosamides, oxazolidinones,

pleuromutilins, and streptogramin A antibiotics, Antimicrob. Agents Chemother. 50,

2500–2505.

6. Toh, S.-M., Xiong, L., Arias, C. A., Villegas, M. V., Lolans, K., Quinn, J., and Mankin,

A. S. (2007) Acquisition of a natural resistance gene renders a clinical strain of

methicillin-resistant Staphylococcus aureus resistant to the synthetic antibiotic linezolid,

Mol. Microbiol. 64, 1506–1514.

7. Kaminska, K. H., Purta, E., Hansen, L. H., Bujnicki, J. M., Vester, B., and Long, K. S.

(2010) Insights into the structure, function and evolution of the radical-SAM 23S rRNA

methyltransferase Cfr that confers antibiotic resistance in bacteria, Nuc. Acids. Res.

221 8. Yan, F., LaMarre, J. M., Röhrich, R., Wiesner, J., Jomaa, H., Mankin, A. S., and Galoníc

Fujimori, D. (2010) RlmN and Cfr are radical SAM enzymes involved in methylation of

ribosomal RNA, J. Am. Chem. Soc. 132, 3953-3964.

9. Cheek, J., and Broderick, J. B. (2001) Adenosylmethionine-dependent iron-sulfur

enzymes: versatile clusters in a radical new role, J. Biol. Inorg. Chem. 6, 209-226.

10. Cosper, M. M., Cosper, N. J., Hong, W., Shokes, J. E., Broderick, W. E., Broderick, J. B.,

Johnson, M. K., and Scott, R. A. (2003) Structural studies of the interaction of S-

adenosylmethionine with the [4Fe-4S] clusters in biotin synthase and pyruvate formate-

lyase activating enzyme, Protein Science 12, 1573-1577.

11. Henshaw, T. F., Cheek, J., and Broderick, J. B. (2000) The [4Fe-4S]+1 Cluster of

Pyruvate Formate-Lyase Activating Enzyme Generates the Glycyl Radical on Pyruvate

Formate-Lyase: EPR-Detected Single Turnover, J Am Chem Soc 122, 8331-8332.

12. Krebs, C., Broderick, W. E., Henshaw, T. F., Broderick, J. B., and Huynh, B. H. (2002)

Coordination of adenosylmethionine to a unique iron site of the [4Fe–4S] of pyruvate

formate–lyase activating enzyme: a Mössbauer spectroscopic study, J. Am. Chem. Soc.

124, 912-913.

13. Grove, T. L., Benner, J. S., Radle, M. I., Ahlum, J. H., Landgraf, B. J., Krebs, C., and

Booker, S. J. (2011) A radically different mechanism for S-adenosylmethionine-

dependent methyltansferases, Science 332, 604–607.

14. Grove, T. L., Radle, M. I., Krebs, C., and Booker, S. J. (2011) Cfr and RlmN contain a

single [4Fe–4S] cluster, which directs two distinct reactivities for S-adenosylmethionine:

methyl transfer by SN2 displacement and radical generation, J Am Chem Soc 133,

19586–19589.

15. Booker, S. J. (2009) Anaerobic functionalization of unactivated C–H bonds, Curr. Opin.

Chem. Biol. 13, 58–73.

222 16. Challand, M. R., Driesener, R. C., and Roach, P. L. (2011) Radical S-adenosylmethionine

enzymes: mechanism, control and function, Nat. Prod. Rep. 28, 1696–1721.

17. Cheek, J., and Broderick, J. B. (2002) Direct H atom abstraction from spore photoproduct

C-6 initiates DNA repair in the reaction catalyzed by spore photoproduct lyase: evidence

for a reversibly generated adenosyl radical intermediate, J. Am. Chem. Soc. 124, 2860-

2861.

18. Minisci, F., Citterio, A., Perchinunno, M., and Bertini, F. (1975) Nucleophilic Character

of Alkyl Radicals .11. New Methods of Homolytic Aromatic Alkylation by

Intramolecular and Intermolecular Reactions of N-Oxides of Trialkylamines (+), Gazzetta

Chimica Italiana 105, 1083-1092.

19. McCusker, K. P., Medzihradszky, K. F., Shiver, A. L., Nichols, R. J., Yan, F., Maltby, D.

A., Gross, C. A., and Galonic Fujimori, D. (2012) Covalent intermediate in the catalytic

mechanism of the radical S-adenosyl-L-methionine methyl synthase RlmN trapped by

mutagenesis, J. Am. Chem. Soc. 134, 18074-18081.

20. Kaminska, K. H., Purta, E., Hansen, L. H., Bujnicki, J. M., Vester, B., and Long, K. S.

(2010) Insights into the structure, function and evolution of the radical-SAM 23S rRNA

methyltransferase Cfr that confers antibiotic resistance in bacteria, Nuc. Acids. Res. 38,

1652–1663.

21. Grove, T. L., Livida, J., Schwalm, E. S., Green, M. T., Booker, S. J., and Silakov, A.

(2013) A Intermediate Substrate Radical Intermediate in Catalysis by the Antibiotic

Resistance Protein Cfr, Nat. Chem. Biol. xx, xx-xx.

22. Saunders, A. H., and Booker, S. J. (2008) Regulation of the activity of Escherichia coli

quinolinate synthase by reversible disulfide-bond formation, Biochemistry 47, 8467-

8469.

223 23. Wunderlich, M., and Glockshuber, R. (1993) Redox Properties of Protein Disulfide

Isomerase (Dsba) from Escherichia-Coli, Protein Science 2, 717-726.

24. Wunderlich, M., Jaenicke, R., and Glockshuber, R. (1993) The Redox Properties of

Protein Disulfide-Isomerase (Dsba) of Escherichia-Coli Result from a Tense

Conformation of Its Oxidized Form, Journal of Molecular Biology 233, 559-566.

25. Krause, G., Lundstrom, J., Barea, J. L., Delacuesta, C. P., and Holmgren, A. (1991)

Mimicking the Active-Site of Protein Disulfide-Isomerase by Substitution of Proline 34

in Escherichia-Coli Thioredoxin, Journal of Biological Chemistry 266, 9494-9500.

26. Ellman, G. L. (1959) Tissue Sulfhydryl Groups, Archives of Biochemistry and

Biophysics 82, 70-77.

27. Kallis, G. B., and Holmgren, A. (1980) Differential Reactivity of the Functional

Sulfhydryl-Groups of Cysteine-32 and Cysteine-35 Present in the Reduced Form of

Thioredoxin from Escherichia-Coli, Journal of Biological Chemistry 255, 261-265.

28. Wynn, R., Cocco, M. J., and Richards, F. M. (1995) Mixed Disulfide Intermediates

during the Reduction of Disulfides by Escherichia-Coli Thioredoxin, Biochemistry 34,

11807-11813.

29. Kosower, N. S., and Kosower, E. M. (1995) Diamide - an Oxidant Probe for Thiols,

Biothiols, Pt A 251, 123-133.

30. Kosower, E. M., Kosower, N. S., Kinon, B. J., and Correa, W. (1972) Glutathione .7.

Differentiation among Substrates by Thiol-Oxidizing Agent, Diamide, Biochimica Et

Biophysica Acta 264, 39-&.

31. Kosower, E. M., and Kanetylondner, H. (1976) Glutathione .13. Mechanism of Thiol

Oxidation by Diazenedicarboxylic Acid-Derivatives, Journal of the American Chemical

Society 98, 3001-3007.

224 32. Hoppe, G., Talcott, K. E., Bhattacharya, S. K., Crabb, J. W., and Sears, J. E. (2006)

Molecular basis for the redox control of nuclear transport of the structural chromatin

protein Hmgb1, Experimental Cell Research 312, 3526-3538.

33. Aslund, F., Berndt, K. D., and Holmgren, A. (1997) Redox potentials of glutaredoxins

and other thiol-disulfide oxidoreductases of the thioredoxin superfamily determined by

direct protein-protein redox equilibria, Journal of Biological Chemistry 272, 30780-

30786.

34. Fahey, R. C., Brown, W. C., Adams, W. B., and Worsham, M. B. (1978) Occurrence of

Glutathione in Bacteria, Journal of Bacteriology 133, 1126-1129.

35. Fahey, R. C., and Sundquist, A. R. (1991) Evolution of Glutathione Metabolism,

Advances in Enzymology and Related Areas of Molecular Biology 64, 1-53.

36. Meister, A. (1988) Glutathione Metabolism and Its Selective Modification, Journal of

Biological Chemistry 263, 17205-17208.

37. Balsera, M., Uberegui, E., Susanti, D., Schmitz, R. A., Mukhopadhyay, B., Schurmann,

P., and Buchanan, B. B. (2013) Ferredoxin:thioredoxin reductase (FTR) links the

regulation of oxygenic photosynthesis to deeply rooted bacteria, Planta 237, 619-635.

38. Schurmann, P., and Buchanan, B. B. (2008) The ferredoxin/thioredoxin system of

oxygenic photosynthesis, Antioxidants & Redox Signaling 10, 1235-1273.

39. Hinckley, G. T., and Frey, P. A. (2006) Cofactor dependence of reduction potentials for

[4Fe–4S]2+/1+ in lysine 2,3-aminomutase, Biochemistry 45, 3219-3225.

40. Yan, F., and Fujimori, D. G. (2011) RNA methylation by radical SAM enzyme RlmN

and Cfr proceeds via methylene transfer and hydride shift, Proc Natl Acad Sci U S A 108,

3930–3934.

225 41. Boal, A. K., Grove, T. L., McLaughlin, M. I., Yennawar, N. H., Booker, S. J., and

Rosenzweig, A. C. (2011) Structural basis for methyl transfer by a radical SAM enzyme,

Science 332, 1089–1092.

42. McCusker, K. P., Medzihradszky, K. F., Shiver, A. L., Nichols, R. J., Yan, F., Maltby, D.

A., Gross, C. A., and Fujimori, D. G. (2012) Covalent Intermediate in the Catalytic

Mechanism of the Radical S-Adenosyl-L-methionine Methyl Synthase RlmN Trapped by

Mutagenesis, Journal of the American Chemical Society 134, 18074-18081.

43. Hutchison, R. S., and Ort, D. R. (1995) Measurement of equilibrium midpoint potentials

of thiol/disulfide regulatory groups on thioredoxin-activated chloroplast enzymes,

Methods Enzymol. 252, 220–228.

44. Kallis, G.-B., and Holmgren, A. (1980) Differential reactivity of the functional sulfhydryl

groups of cysteine-32 and cysteine-35 present in the reduced form of thioredoxin from

Escherichia coli, J. Biol. Chem. 255, 10261-10265.

45. Lundström, J., and Holmgren, A. (1993) Determination of the reduction–oxidation

potential of the thioredoxin-like domains of protein disulfide-isomerase from the

equilibrium with glutathione and thioredoxin, Biochemistry 32, 6649-6655.

46. Prinz, W. A., Åslund, F., Holmgren, A., and Beckwith, J. (1997) The role of thioredoxin

and glutaredoxin pathways in reducing protein disulfide bonds in the Escherichia coli

cytoplasm, J. Biol. Chem. 272, 15661-15667.

47. Grauschopf, U., Winther, J. R., Korber, P., Zander, T., Dallinger, P., and Bardwell, J. C.

A. (1995) Why is DsbA such an oxidizing disulfide catalyst?, Cell 83, 1038–1042.

48. Kobayashi, T., and Ito, K. (1999) Respiratory chain strongly oxidizes the CXXC motif of

DsbB in the Escherichia coli disulfide bond formation pathway, Embo Journal 18, 1192-

1198.

49. Rost, J., and Rapoport, S. (1964) Reduction-Potential of Glutathione, Nature 201, 185-&.

226 50. Simic, M., and Hayon, E. (1971) Intermediates Produced from One-Electron Oxidation

and Reduction of Hydroxylamines - Acid-Base Properties of Amino, Hydroxyamino, and

Methoxyamino Radicals, Journal of the American Chemical Society 93, 5982-&.

Part II

Characterization of Radical SAM-Dependent Dehydrogenases

228

Chapter 6

In vitro Characterization of AtsB, a Radical SAM Formylglycine Generating Enzyme that Contains Three [4Fe–4S] Clusters†

This chapter was reproduced from “Grove, T.L., Lee, K.H., St. Clair, J., Krebs, C., Booker, S.J. In vitro characterization of AtsB, a radical SAM formylglycine-generating enzyme that contains three [4Fe-4S] clusters. Biochemistry. 2008 Jul 15;47(28):7523-3758.

229 6.1 Abstract

Sulfatases catalyze the cleavage of a variety of cellular sulfate esters via a novel mechanism that requires the action of a protein-derived formylglycine cofactor.

Formation of the cofactor is catalyzed by an accessory protein, and involves the two- electron oxidation of a specific cysteinyl or seryl residue on the relevant sulfatase.

Although some sulfatases undergo maturation via mechanisms in which oxygen serves as an electron acceptor, AtsB, the maturase from Klebsiella pneumoniae, catalyzes the oxidation of Ser72 on AtsA, its cognate sulfatase, via an oxygen-independent mechanism.

Moreover, it does not make use of pyridine– or flavin–nucleotide cofactors as direct electron acceptors. In fact, AtsB has been shown to be a member of the radical S- adenosylmethionine superfamily of proteins, suggesting that it catalyzes this oxidation via an intermediate 5’-deoxyadenosyl 5’-radical that is generated by a reductive cleavage of S-adenosyl-L-methionine. In contrast to AtsA, very little in vitro characterization of

AtsB has been conducted. Herein we show that co-expression of the Klebsiella pneumoniae atsB gene with a plasmid that encodes genes that are known to be involved in iron–sulfur cluster biosynthesis yields soluble protein that can be characterized in vitro. The as-isolated protein contained 8.7 ± 0.4 irons and 12.2 ± 2.6 sulfides per polypeptide, which existed almost entirely in the configuration [4Fe–4S]2+, as determined by Mössbauer spectroscopy, suggesting that it contained at least two of these clusters per polypeptide. Reconstitution of the as-isolated protein with additional iron and sulfide indicated the presence of 12.3 ± 0.2 of the former and 9.9 ± 0.4 of the latter per polypeptide. Subsequent characterization of the reconstituted protein by Mössbauer

230 spectroscopy indicated the presence of only [4Fe–4S] clusters, suggesting that reconstituted AtsB contains three per polypeptide. Consistent with this stoichiometry, an as-isolated AtsB triple variant containing CysAla substitutions at each of the cysteines in its CX3CX2C radical SAM motif, contained 7.3 ± 0.1 irons and 7.2 ± 0.2 sulfides per polypeptide while the reconstituted triple variant contained 7.7 ± 0.1 irons and 8.4 ± 0.4 sulfides per polypeptide, indicating that it was unable to incorporate an additional cluster.

UV-visible and Mössbauer spectra of both samples indicated the presence of only [4Fe–

4S] clusters. AtsB was capable of catalyzing multiple turnovers, and displayed a

-1 Vmax/[ET] of ~0.36 min for an 18 amino acid peptide substrate using dithionite to supply the requisite electron, and ~0.039 min-1 for the same substrate using the physiologically relevant flavodoxin reducing system. Simultaneous quantification of formylglycine and

5’-deoxyadenosine as a function of time indicates an approximate 1:1 stoichiometry. Use of a peptide substrate in which the target serine is changed to cysteine also gives rise to turnover, supporting approximately four-fold the activity of that observed with the natural substrate.

231

6.2 Introduction

Sulfatases catalyze the cleavage of a variety of organosulfate esters, resulting in the release of inorganic sulfate and the corresponding alcohol (1-4). Three classes of these enzymes are currently recognized and are delineated by the cofactors they employ in catalysis (5). Group I enzymes are often referred to as arylsulfatases, but their substrate profile extends beyond that designation since they catalyze sulfate hydrolysis from myriad sulfated compounds. In addition to a calcium or magnesium ion, they require a

C -formylglycine (2-amino-3-oxopropionic acid) cofactor, which is generated via a co- translational or post-translational modification of a specific cysteinyl or seryl residue on the enzyme (Figure 6-1A) (6). Group II enzymes catalyze sulfate release via an oxidative

-ketoglutarate and molecular oxygen as co-substrates. Concomitant with sulfate release, one equiv each of carbon dioxide, succinate, and an aldehyde or ketone is produced (5, 7). Group III enzymes constitute the least characterized sulfatase designation. They employ a dinuclear zinc center to activate water for nucleophilic attack on the sulfated substrate in a fashion analogous to that of members of the metallo- -lactamase family (8).

Group I sulfatases have been the most extensively studied and characterized (1-4, 9-11).

Although mostly eukaryotic, these enzymes are also found in archaea and bacteria. In bacteria, sulfatase genes are often expressed under conditions of sulfate starvation, and the gene products are believed to play a role in sulfate scavenging or the metabolism of

232 sulfated compounds (12, 13). In eukaryotes sulfatases catalyze key steps in the breakdown of sulfate-containing compounds, such as the hydrolysis of sulfate ester linkages in glycosaminoglycans, steroids, lipids, and various other molecules. Catalytic deficiencies in eight identified sulfatases lead to severe metabolic disorders, deriving from their inability, or reduced ability, to hydrolyze specific sulfate-containing metabolites (3). A rare autosomal recessive disease, multiple sulfatase deficiency (MSD), is a pleiotropic condition in which the essential formylglycine (FGly) cofactor is not formed on most, if not all, sulfatases (6, 14, 15).

233 The FGly modification is found in a signature sequence (C/S-X-P/A-S/X-R-X-X-X-

L/X-T/X-G/X-R/X) that is highly conserved from prokaryotes to eukaryotes, wherein the initial cysteine or serine amino acid (shown in bold type) is that which is modified (6, 16).

All eukaryotic Group I sulfatases contain a FGly cofactor derived from a cysteinyl residue, as do many bacterial sulfatases; however, some bacterial sulfatases also contain

FGly cofactors that are derived from a seryl residue, giving rise to the classification Ser- type or Cys-type sulfatase (2, 17, 18). The manner in which the cofactor is used in organosulfate hydrolysis is novel. Crystallographic studies of human arylsulfatase A

(HARSA) (19), as well as an arylsulfatase from Pseudomonas aeruginosa (9), show that the

FGly cofactor exists in its hydrated form as a geminal diol (Figure 6-1A). It is proposed that one of the hydroxyls of the hydrated cofactor performs a nucleophilic attack on the sulfur atom of the organosulfate substrate with concomitant transfer of the sulfate group to the cofactor and release of the corresponding alcohol. Deprotonation of the remaining hydroxyl group of the cofactor allows for elimination of inorganic sulfate and regeneration of the cofactor. Subsequent hydration of the cofactor re-establishes its active form (Figure 6-1B) (4, 9, 19). Consistent with this mechanism, variants of HARSA and

HARSB containing a Ser substitution at the Cys residue that is converted to FGly become irreversibly sulfated at the Ser residue when treated with [35S]-p-nitrocatechol sulfate, suggesting that these two variant human sulfatases can perform the nucleophilic attack on the organosulfate substrate but cannot eliminate the adduct (20). Note that, as opposed to some bacterial sulfatases, eukaryotic sulfatase maturation proteins cannot convert Ser residues into FGly.

234 Two mechanisms are currently recognized for generation of the FGly cofactor. In higher eukaryotes FGly generation takes place in the endoplasmic reticulum and is catalyzed by sulfatase modifying factor 1 (SUMF1) (14, 15). This protein is the product of the SUMF1 gene, which is the locus for mutations that give rise to MSD. The detailed mechanism by which this enzyme catalyzes its reaction has not been elucidated; however,

SUMF1 contains a redox-active disulfide bond–but no redox-active metals or other cofactors–and is reported to require both dioxygen and reducing equivalents for catalysis.

The intermediacy of a cysteine sulfenic acid on the substrate has been proposed (10).

A second mechanism for FGly generation is found almost exclusively in prokaryotes, and does not involve dioxygen. The prototype of this FGly-generating system is the enzyme AtsB from Klebsiella pneumoniae, which catalyzes FGly formation on its cognate protein AtsA, a Ser-type sulfatase (21-23). This sulfatase, like most, if not all, Ser- type sulfatases, contains a periplasmic localization sequence, the removal of which drastically reduces the efficiency of FGly modification in vivo (2, 22). AtsB, however, does not harbor a periplasmic localization signal sequence, and is believed to act on AtsA either co-translationally or during translocation into the periplasm. Cys-type sulfatases, by contrast, are believed to be located predominantly in the cytoplasm (2).

AtsB is composed of 395 amino acids, and its primary structure predicts a molecular mass of 44,237 Da (24). Although it was once believed to be a transcriptional regulator of

AtsA, the absence of clear DNA binding domains, as determined by bioinformatics analysis, cast doubt on this role in sulfatase function (2). In fact, AtsB contains 13 Cys residues, about ten of which are conserved among known and predicted FGly maturation

235

(2, 24) proteins within this class . Three of the cysteines are organized in a CX3CX2C motif, which is the predominant signature sequence that denotes a superfamily of enzymes–the radical SAM (RS) superfamily–that use S-adenosyl-L-methionine (SAM) as a precursor to a 5’-deoxyadenosyl 5’-radical (5’-dA•) (25). The 5’-dA• initiates catalysis on an appropriate substrate by abstracting a key hydrogen atom, affording 5’-deoxyadenosine

(5’-dA) and a substrate-derived organic radical intermediate. The generation of the 5’- dA• from SAM requires the input of one electron, which emerges from a reduced [4Fe–

4S] cluster ([4Fe–4S]+) that is ligated to the protein by the three cysteines within the

(26-29) hallmark CX3CX2C motif . The one iron atom of the cluster not ligated by a protein-

- -carboxylate groups of SAM, allowing for close spatial arrangement and optimal geometry for electron transfer into the sulfonium group (30-37).

Like many iron–sulfur (Fe/S) proteins, RS proteins tend to be exquisitely sensitive to oxygen, and upon heterologous overproduction often assemble into inclusion bodies, especially when incorporation of the cluster lags well behind production of the polypeptide scaffold. AtsB is no exception; these tendencies have hampered its in vitro characterization. However, characterization of AtsB in vivo or in cell lysates has confirmed its identity as a RS protein (23). AtsB was shown to associate with a GST-fused

AtsA fragment, and this association was stimulated in the presence of SAM. Moreover,

AtsB catalyzed FGly formation on the GST-fused AtsA fragment only in the presence of

SAM, although the efficiency of conversion was calculated to be ~10% (23). Herein we describe the first purification and in vitro characterization of AtsB from K. pneumoniae.

236 The enzyme is shown by analytical methods and Mössbauer spectroscopy to contain three

[4Fe–4S] clusters per polypeptide as predicted by an early bioinformatics study (2). In addition, AtsB is shown to catalyze formation of 5’-dA and FGly in a 1:1 ratio using an

18-amino acid (aa) peptide substrate. Substitution of the target Ser residue in the peptide substrate with Ala abolished formation of both 5’-dA and FGly; however, substitution of the Ser residue with Cys gave rise to a peptide substrate that was converted to product with a faster Vmax/[ET]. Finally, a working model for this interesting transformation that is consistent with the data described is presented.

6.3 Materials and Methods

Materials. All DNA modifying enzymes and reagents were purchased from New

England Biolabs (Beverly, MA), as was DeepVent polymerase and its associated 10X buffer. Oligonucleotide primers were obtained from Integrated DNA Technologies

(Coralville, IA). E. coli BL21 (DE3) and expression vector pET-28a were purchased from Novagen (Madison, WI). Phenylhydrazine, 5’-deoxyadenosine (5’-dA), sodium sulfide (nonahydrate), dithiothreitol (DTT), reduced nicotinamide adenine dinucleotide

-mercaptoethanol, L-tryptophan, and L-(+)-arabinose were purchased from Sigma–Aldrich Chemicals (St. Louis, MO) and used as is. Bovine serum albumin (BSA) standard (2 mg/mL) and the Bradford reagent for protein concentration determination were purchased from Pierce (Rockford, IL). Klebsiella pneumoniae genomic DNA (ATCC 700721D) was obtained from the American Type Culture

Collection (Manassas, VA). Nickel nitrilotriacetic acid (Ni-NTA) resin was purchased

237 from Qiagen (Valencia, CA), while Talon metal affinity resin was purchased from

Clontech (Moutain View, CA). Sephadex G-25 resin, and NICK and NAP pre-poured gel-filtration columns were purchased from GE Biosciences (Piscataway, NJ). All other buffers and chemicals were of the highest grade available.

(S,S)-Adenosyl-L-methionine (SAM) was synthesized enzymatically and purified as described previously (38). E. coli flavodoxin (Flv) and flavodoxin reductase (Flx) were purified from E. coli BL21(DE3) containing plasmids pTYB1-Flv and pTYB1-Flx, which harbor the genes for E. coli Flv and Flx, respectively, cloned into the NdeI and SapI sites of plasmid pTYB1 (New England Biolabs). This construct affords production of Flv and

Flx as intein fusions, which undergo splicing in the presence of DTT such that both target proteins are produced in their native forms.

General Methods. High performance liquid chromatography (HPLC) was conducted on an Agilent (Foster City, CA) 1100 system that was fitted with an autosampler for sample injection and a variable wavelength detector. The system was operated with the associated ChemStation software package, which was also used for data collection and analysis. Iron and sulfide analyses were performed by the methods of Beinert as previously described (39-41). Sonic disruption of E. coli cells was carried out with a 550 sonic dismembrator from Fisher Scientific (Pittsburgh, PA) using a horn containing a

1/2” tip. The cable connecting the horn to the power supply was threaded through a port in a Coy anaerobic chamber (Grass Lake, MI) to allow sonic disruption to be performed under an oxygen-free atmosphere. DNA sequencing was conducted at the Pennsylvania

State University Nucleic Acids Facility (University Park, PA), and peptide synthesis was

238 performed at The Pennsylvania State University College of Medicine Core Facility

(Hershey, PA).

Spectroscopic Methods. UV-visible spectra were recorded on a Cary 300 spectrometer (Varian; Walnut Creek, CA), employing the associated WinUV software package for operating the instrument and manipulating the data. Mössbauer spectra were recorded on a spectrometer from WEB Research (Edina, MN), which was equipped with an SVT-400 cryostat from Janis Research Company, Wilmington, MA). Spectra were collected in constant acceleration mode in transmission geometry. Isomer shifts are

-Fe at room temperature. Spectra were analyzed with the program WMOSS from WEB Research. Matrix-assisted laser desorption/ionization

(MALDI) mass spectrometry was performed on a MALDI micro MX instrument from

Waters Corporation (Milford, MA). Spectra were combined and refined with the assistance of the associated MassLynx software package.

Construction of the atsB Expression Vector. The atsB gene was amplified from

Klebsiella pneumoniae genomic DNA using polymerase chain reaction (PCR) technology. The forward amplification primer (5’-CGC CGC CCC CAT ATG CTG AAT

GCC CTG CGC CAG CAG C-3’) included an NdeI restriction site (underlined) flanked by a nine base pair (bp) GC clamp and the first 31 bases of the atsB gene. The reverse primer (5’-CGC CGC CCC GAA TTC CTA CGC AGT ATG CGC AGT CCC AAC

AAA CGC-3’) contained an EcoRI restriction site (underlined) flanked by a nine base

GC clamp and the last 30 bases of the atsB gene, including the stop codon. The PCR was performed with a Stratagene Robocycler thermocycler (La Jolla, CA) as described previously (42). The product was isolated and digested with NdeI and EcoRI and ligated

239 into similarly digested expression vector pET-28a by standard procedures (43). The correct construct was verified by DNA sequencing and designated pAtsBWt.

Construction of the AtsB C35A–C39A–C42A Triple Variant. The gene for the AtsB

C35A–C39A–C42A triple variant was constructed using the Stratagene QuikChange II site-directed mutagenesis kit according to the manufacturer’s specifications, and as described previously (42, 44). The following modifications were implemented to allow three mutations to be constructed in one PCR. The forward primer (5’-CCG TTT CAT

ATT CTG ATG AAG CCG ATT GGC CCC GCC GCC AAT CTC GCC GCC CGC

TAT GCC-3’) contained 60 bases, the last 24 of which were complementary to the last

24 bases of the reverse primer (5’-CAT CTT GTT GAC CGG CGT TTC GTC CTG

CGG GTA ATA GGC ATA GCG GGC GGC GAG ATT GGC-3’) (shown in bold type; base changes are underlined). These primers were added to a typical QuikChange II reaction to a final concentration of 20 µM, and ten cycles of the following program were initiated: 95°C for 1 min; 50 °C for 1.5 min; 72°C for 1 min. This initial thermocycling reaction allows creation of 96-base primers from the initial PCR, deriving from overlap extension of the forward and reverse primers after they anneal. The template plasmid, pAtsBWt (100 ng), was then added to the reaction and 15 cycles of the following program were initiated: 95°C for 1 min; 55 °C for 1.5 min; and 72°C for 10 min. Upon completion, the reaction was incubated for 10 min at 72°C before cooling to 4°C.

Subsequent to this step, the procedure followed the manufacturer’s specifications.

Mutations were verified by DNA sequencing, and the resulting plasmid was designated pAtsBMut1.

240 Expression of the atsB Gene, and Production and Purification of AtsB. For expression of the atsB gene, plasmid pAtsBWt was transformed into E. coli BL21 (DE3)/pDB1282 by standard methods (43). A single colony was used to inoculate 200 mL of M9 minimal media containing 100 µg/mL ampicillin and 50 µg/mL kanamycin. The culture was allowed to grow to late-log phase, upon which 20 mL portions were used to inoculate four 6 L flasks, each of which contained 4 L of the same media. The cultures were allowed to grow with moderate shaking (180 rpm) to an optical density (OD) at 600 nm of 0.3, after which solid L-(+)-arabinose was added to each flask to a final concentration of 0.2 % (w/v), while cysteine and ferric chloride were added to final concentrations of

200 µM and 50 µM, respectively. At an OD600 of 0.6 the flasks were cooled to 18 °C before a sterile solution of IPTG was added to each flask to a final concentration of 200

µM. Expression was allowed to take place during a further incubation at 18 °C for 20-23 h, upon which the cultures were harvested by centrifugation at 10,000  g for 10 min at 4

°C. Typical yields were 60-65 g of frozen cell paste from 16 L of M9 minimal media.

The frozen cell paste was stored at –80°C until needed for further use.

AtsB was purified by immobilized metal affinity chromatography (IMAC) using

Talon metal affinity chromatography media. All steps were performed inside a Coy

Laboratory Products, Inc (Grass Lake, MI) anaerobic chamber under an atmosphere of nitrogen (90%) and hydrogen (10%) gas. The oxygen concentration was monitored with a

H2/O2 meter (Coy Laboratory Products) that was housed inside of the chamber, and was maintained below 1 ppm via the use of palladium catalysts (Coy Laboratory Products).

All buffers were prepared using distilled and deionized water that was boiled for several hours before introduction into the anaerobic chamber, and then allowed to cool while

241 stirring uncapped for 48 h. They were titrated to their final pH values using an Accumet

(AP61) pH meter from Fisher Scientific (Fairlawn, NJ), which was housed inside of the anaerobic chamber. All plastic ware was autoclaved and introduced inside of the chamber while still hot, and allowed to equilibrate with the atmosphere for at least 24 h.

Centrifugation was performed outside of the anaerobic chamber; however, samples were loaded into air-tight centrifuge bottles before they were removed. Concentration by ultrafiltration was conducted inside of the anaerobic chamber using an Amicon

(Millipore; Billerica, MA) stirred cell that was connected to an argon tank via a port in the chamber. Although the purification was carried out at room temperature, all solutions were kept in ice–water baths, and during protein concentration the Amicon stirred cell was loosely jacketed with ice packs.

In a typical purification, 30 g of cell paste was resuspended in 150 mL of lysis buffer

(50 mM potassium HEPES, pH 7.5, 300 mM KCl, 10 mM BME, and 20 mM imidazole) containing lysozyme and peptide 3 (see below) at final concentrations of 1 mg/mL and 10

µM, respectively. After stirring at room temperature for 30 min, the solution was placed in an ice–water bath and allowed to cool to ~4 °C before subjecting it to four 1 min bursts of sonic disruption (40% output). Between bursts, the solution was allowed to stir in the ice–water bath for 10 min to re-equilibrate the temperature to ≤8 °C. The lysate was centrifuged for 1 h at 50,000  g and 4 °C, and the resulting supernatant was loaded onto a Talon cobalt affinity column (2.5 x 10 cm). The column was washed with 200 mL of lysis buffer and then eluted with 100 mL of elution buffer (50 mM potassium HEPES, pH

7.5, 300 mM KCl, 20% glycerol, 10 mM BME, and 300 mM imidazole). Fractions that displayed significant brown color were collected, and peptide 3 (see below) was added to

242 the pooled sample to a final concentration of 50 µM to stabilize the protein. The protein solution was concentrated to ~3 mL by stirred ultrafiltration using a Millipore YM-3 membrane, and exchanged into gel-filtration buffer (50 mM potassium HEPES, pH 7.5,

500 mM KCl, 10 mM DTT, 30% glycerol) using a Sephadex G-25 column. The brown fractions were pooled and concentrated to ~3 mL. The protein was snap-frozen in liquid nitrogen in 200 µL aliquots and stored at –80°C. When necessary, reconstitution of the

Fe/S clusters of AtsB was carried out as previously described, but in the presence of equimolar concentrations of peptide 3 (42).

Amino acid analysis of AtsB. Amino acid analysis of AtsB was carried out at the

Molecular Structure Facility at the University of California at Davis as well as the

Molecular Biology Core Facility of the Dana Farber Institute (Boston, MA). The protein was exchanged by gel-filtration (NICK pre-poured column) into 50 mM HEPES buffer, pH 7.5, containing 100 mM NaCl. The eluate was divided into 150 µL fractions, which were lyophilized to dryness using a Savant SpeedVac concentrator (Thermo Scientific;

Waltham, MA). One fraction was used to determine the protein concentration by the procedure of Bradford (45) before lyophilization. The remaining fractions were shipped for amino acid analysis, which was performed on three different samples.

Quantification of Turnover. Turnover was measured using four synthetic peptides as substrates: NH2-Tyr-Tyr-Thr-Ser-Pro-Met-Ser-Ala-Pro-Ala-Arg-Ser-Met-Leu-Leu-Thr-

Gly-Asn-COOH (peptide 1); Acetyl-Tyr-Tyr-Thr-Ser-Pro-Met-Ser-Ala-Pro-Ala-Arg-Ser-

Met-Leu-Leu-Thr-Gly-Asn-COOH (peptide 2); NH2-Tyr-Tyr-Thr-Ser-Pro-Met-Cys-Ala-

Pro-Ala-Arg-Ser-Met-Leu-Leu-Thr-Gly-Asn-COOH (peptide 3); and NH2-Tyr-Tyr-Thr-

Ser-Pro-Met-Ala-Ala-Pro-Ala-Arg-Ser-Met-Leu-Leu-Thr-Gly-Asn-COOH (peptide 4).

243 Note: The underlined amino acids in each peptide are at the position that undergoes modification to FGly. Assays conducted using sodium dithionite as the reductant contained the following in a final volume of 350 µL: 50 mM HEPES, pH 7.5, 2 mM dithionite, 1 mM peptide substrate, 1 mM SAM, 50 or 100 µM AtsB, and 1 mM L- tryptophan (internal standard). Reactions were initiated by addition of dithionite after incubation of the other components of the assay mixture at 37 °C for 5 min. At designated times, 25 µL aliquots were removed and added to an equal volume of 1M

HCl. Precipitated protein was pelleted by centrifugation, and the resulting supernatant

-cyano-4- hydroxycinnamic acid (5 mg/mL in 70% acetonitrile/30% 0.1% TFA), and 2 µL of this mixture was spotted onto the MALDI-TOF target plate. Spectra (70-100, 10 scans per spectrum) were acquired in reflector mode with 21 kV accelerating voltage, 65% grid voltage, and a 500 ns delay. The resulting data were transformed into X,Y format and then plotted using the Igor Pro software package (Wavemetrics; Lake Oswego, OR).

Assays in which Flv and Flx replaced sodium dithionite contained the same components as described above, except that dithionite was omitted and Flv, Flx, and NADPH were added to final concentrations of 25 µM, 5 µM, and 2 mM respectively. In addition, the

AtsB concentration was increased to 100 µM, and reactions were initiated by addition of

SAM.

A second, HPLC-based, method was also developed to quantify the FGly product, the peptide substrate, and 5’-dA. Activity assays were conducted as described above, except they were quenched by addition of H2SO4 and hydroxylamine to final concentrations of

50 mM and 500 mM, respectively. The quenched solutions were subjected to

244 centrifugation for 15 min at 14,000  g to pellet precipitated material, and 20 µL aliquots were injected onto a Zorbax SB-CN (4.6 x 250 mm, 5 µm) column (Agilent) that had been equilibrated in 95% solvent A (0.4% trifluoroacetic acid, 100 mM hydroxylamine, titrated to pH 1.78 with triethylamine). Simultaneous linear gradients of 5-13% methanol and 0-21% acetonitrile were applied from 0 to 7 min, and then held constant for 24 min before returning to the original conditions. Detection was monitored at 260 nm for the first 10.4 min of the run, and then at 220 nm until the end of the run. Under these conditions, peptides 2 and 3 eluted at retention times 20.4 and 17.6 min.

6.4 Results

Cloning and Expression of the Klebsiella pneumoniae atsB gene, and Purification of

AtsB. The atsB gene of K. pneumoniae was cloned into the NdeI and EcoRI sites of expression vector pET-28a by standard methods. This construct allows production of the encoded protein with an N-terminal hexahistidine tag that is separated from the natural start codon by a spacer of ten amino acids, facilitating rapid purification of the protein under anaerobic conditions. Subsequent sequencing of the gene revealed several changes in the DNA sequence that were not present in the published sequence of the atsB gene, and that resulted in amino acid changes. These amino acid changes include Gly140Ser,

Val197DAsp, Thr297Ala, Pro50Ser, and Ser284Leu. The first three amino acid changes have been observed previously and noted on the expasy proteomics server

(http://ca.expasy.org/uniprot/Q9X758). The last two changes we confirmed by sequencing three independent clones from two independent cloning trials.

245

Because AtsB was predicted, and a homolog subsequently shown (46), to contain Fe/S clusters, the atsB gene was co-expressed with genes on plasmid pDB1282, which encodes an operon from Azotobacter vinelandii that is known to be involved in Fe/S cluster biosynthesis (47, 48). Plasmid pDB1282 confers resistance to ampicillin, while plasmid pAtsBWt, which is derived from commercially available pET-28a, confers resistance to kanamycin. The genes on plasmid pDB1282 were under the control of an arabinose- inducible promoter, and were induced prior to the atsB gene, which was under the control of an IPTG-inducible promoter, to allow production of the proteins involved in Fe/S cluster biosynthesis before production of AtsB. Bacterial growth and protein overproduction was carried out in M9 minimal media to obtain maximum incorporation of 57Fe into the protein for analysis by Mössbauer spectroscopy.

246 AtsB was purified by cobalt–IMAC inside an anaerobic chamber under an atmosphere of hydrogen (10%) and nitrogen (90%) gas. Initial attempts made use of Ni-

NTA resin; however, the resulting protein was noticeably inhomogeneous (less than 90% pure) and contained two significant contaminants. Talon metal affinity resin (cobalt form) proved to be a fruitful alternative to Ni-NTA resin, which was used to purify hexahistidine-tagged AtsA. In Figure6-2, an SDS-PAGE analysis of the purification of

AtsB (lane 2) and AtsA (lane 3) is displayed. The amino acid sequences of hexahistidine tagged AtsB and AtsA predict proteins of molecular masses 46,432 and 66,319, and are consistent with their migratory properties by SDS-PAGE. From 30 g of cell paste, ~ 30 mg of protein that is ≥95% pure is routinely isolated. The fairly low yield partly stems from loss of protein during concentration steps; AtsB has a tendency to aggregate and form a membranous film on ultrafiltration filters. Inclusion of a peptide substrate during concentration greatly increases solubility, and allows the protein to be concentrated to

≥500 µM. An AtsB triple variant, containing CA substitutions at those cysteines within the RS signature CX3CX2C motif (Cys35, Cys39, Cys42) was also constructed and purified as described for the WT protein.

Amino acid analysis was performed on AtsB to ascertain whether it is necessary to apply a correction factor to the Bradford protein determination assay. Two independent analyses were conducted at two different facilities. To minimize variability, it was necessary to ship defined quantities of lyophilized AtsB in appropriate vials, rather than in frozen solution, to which acid was added directly for hydrolysis. Results indicate that the Bradford assay, using BSA as a standard, underestimates the concentration of AtsB

247 by a factor of 1.07 ± 0.05 (University of California–Davis) or 1.05 ± 0.09 (Dana Farber

Cancer Institute), indicating that, within experimental error, no correction is needed.

Spectroscopic Characterization of Wild-type and C35A–C39A–C42A AtsB. The UV- vis spectrum of as-isolated (AI) AtsB is shown in Figure 6-2A (solid line), and is consistent with the presence of [4Fe–4S] clusters. The distinguishing features include the shoulder at 320 nm, the peak at 395 nm, and the broad tailing that extends beyond 700 nm. The spectrum of WT AtsB (Figure6-4A) is dominated by an intense quadrupole doublet.

248 Iron and sulfide analyses of the AI protein indicate that each polypeptide contains 8.7 ±

0.4 iron atoms and 12.2 ± 2.6 sulfide atoms. To analyze the type and stoichiometry of

Fe/S clusters more rigorously, Mössbauer spectroscopy was conducted on 57Fe-enriched wild-type (WT) and C35A–C39A–C42A triple variant forms of AtsB.

Spectra collected over a wider range of Doppler velocities reveal that the sample contains only a small amount of one or more Fe species exhibiting a magnetically split subspectrum. The EPR spectrum of an identical sample reveals a small amount of a [3Fe-

4S]+ cluster (10 M), which corresponds to 1% of the total Fe contained in the sample

(3  10 M / 3.95 mM Fe). The Mössbauer spectra of a [3Fe-4S]+ cluster are the superposition of three magnetically split subspectra, corresponding to the three distinct Fe sites (49). Since their features are relatively broad and similar for different [3Fe-4S]+ clusters (49), the 1% contribution of the [3Fe-4S]+ clusters is beyond the detection limit of

Mössbauer spectroscopy. The spectrum can be analyzed assuming one broad quadrupole doublet with parameters typical of [4Fe-4S]2+ clusters: isomer shift (δ) 0.44 mm/s and quadrupole splitting parameter (ΔEQ) 1.17 mm/s. This quadrupole doublet accounts for

94 ± 3% of total Fe (solid line in Figure 6-4A). The weak absorption at 0.7 mm/s (see arrow) is consistent with the position of the high-energy line of the spectrum of [2Fe-

2S]2+ clusters. The sample, thus, may contain a small amount (<3%) of this cluster type.

In combination with the ratio of 8.7 iron atoms per AtsB polypeptide, these experiments indicate that AI WT AtsB contains 2.0 [4Fe-4S] clusters per polypeptide. Upon reconstitution of the protein with additional 57Fe and sulfide, followed by gel-filtration on

Sephadex G-25 resin, AtsB was found to contain 12.3 ± 0.2 irons and 9.9 ± 0.4 sulfides

249 per polypeptide. The UV-vis spectrum of the reconstituted (RCN) protein (Figure 6-3A, dashed line) shows an increased absorption at 395 nm, and a lower A280/A395 ratio as compared to the AI protein (2.31 versus 2.69, respectively).

Figure 6-4. Mössbauer spectra of (A) AI WT AtsB; (B) RCN WT AtsB; (C) AI C35A–C39A– C42A AtsB triple variant; and (D) RCN C35A–C39A–C42A AtsB triple variant. All spectra were collected at 4.2 K in an external 53-mT magnetic field. The solid lines in (A) – (D) represent a quadrupole doublet with the following parameters: δ = 0.44 mm/s and ΔEQ = 1.17 mm/s.

250 The 4.2-K/53-mT Mössbauer spectrum (Figure 6-4B) of reconstituted WT AtsB is identical within experimental uncertainty to that of AI WT AtsB. The solid line is a simulation using the same parameters (δ 0.44 mm/s; ΔEQ = 1.17 mm/s; 94% of total Fe).

Therefore, virtually all of the iron present in the reconstituted sample is of the configuration [4Fe–4S]2+, indicating that RCN WT AtsB contains three [4Fe–4S] clusters per polypeptide.

The finding of three [4Fe–4S] clusters per polypeptide is corroborated by site- directed mutagenesis studies, wherein all Cys residues in the CX3CX2C motif are changed to Ala. The UV-vis spectrum of the AI triple variant (Figure 6-3B) is quite similar to that of WT AtsB; it maintains the shoulder at 320 nm, the peak at 395 nm, and the broad tailing that extends beyond 700 nm, suggesting that although all ligands to the

RS cluster have been removed, [4Fe–4S] clusters are still associated with the protein.

Analyses for iron and sulfide indicate the presence of 7.3 ± 0.1 of the former and 7.2 ±

0.2 of the latter, consistent with the presence of two [4Fe–4S] clusters associated with this protein. The Mössbauer spectrum of the triple variant is dominated by a quadrupole doublet with parameters identical to those of WT AtsB (δ 0.44 mm/s; ΔEQ = 1.17 mm/s;

94% of intensity, Figure6-4C). A stoichiometry of 1.7 [4Fe-4S] clusters per polypeptide is deduced from the relative intensity of the quadrupole doublet [(94 ± 3)% of total Fe] and the Fe/AtsB ratio of 7.3, indicating the presence of two remaining [4Fe-4S] clusters in the AtsB triple variant. When the triple variant was subjected to reconstitution conditions, 7.7 ± 0.1 and 8.4 ± 0.4 sulfides per polypeptide were found associated with the protein, indicating that, unlike the WT protein, the triple variant was unable to

251 incorporate an additional Fe/S cluster. The Mössbauer spectrum of the RCN triple variant

(Figure 6-4D) is almost identical to that of the AI triple variant. It shows a quadrupole doublet of the same parameters as the AI triple variant, which accounts for 93% of the spectral intensity.

Time-Dependent Formation of Formylglycine and 5’-Deoxyadenosine. In Figure 6-5

MALDI mass spectra of the AtsB-catalyzed reaction are displayed. Each reaction was performed using 2 mM dithionite as the requisite reductant, and contained 1 mM peptide

1 and 100 µM AtsB. Spectrum A in Figure 6-5 is that of an assay mixture containing all components except AtsB (t = 0), which was then mixed with 100 mM phenylhydrazine before it was spotted onto a MALDI plate. The dominant feature in the spectrum has an m/z value of 1961.5, which corresponds to the molecular mass of the unmodified substrate. After addition of AtsB and a further incubation for 60 min, the peak at 1961.5 decreases dramatically (almost completely), and a new peak having an m/z value of

2049.5 appears, which is consistent with the phenylhydrazine adduct of a FGly- containing product (Spectrum B). Displayed in Spectrum C is a reaction in which the

Flv/Flx/NADPH reducing system is substituted for dithionite. Although a phenylhydrazine adduct of an FGly product is observed, the quantity of the product is substantially reduced as compared to that observed in the presence of dithionite. In the absence of phenylhydrazine, two product peaks are observed (Figure 6-6, solid line).

One product peak corresponds to the expected value of 1958.6, indicating the presence of the aldehyde, while a second, much larger, peak corresponds to an m/z value of 1940.6.

252

253 The difference in m/z values between these two peaks is 18.0, suggesting that the equivalent of a molecule of water is eliminated from the FGly product. Consistent with this conclusion, the shift only occurs under turnover conditions using an appropriate substrate. In this particular experiment the ratio of the peak at m/z 1958.6 to that at m/z

1940.6 is ~0.3; however, it was found to be variable from experiment to experiment, and probably reflects subtle differences in preparing and spotting samples for analysis by

MALDI-MS, since the ratios were consistent for samples spotted on the same plate. In fact, in some experiments the overwhelming majority of product was found at m/z

1958.6. Since the N-termini of the peptide substrates used were not blocked, we suggest that the peak at m/z value 1940.6 corresponds to Schiff base formation between the N- terminal amino group and the FGly residue. When a peptide substrate containing an acetylated N-terminus was used (peptide 2), the corresponding FGly-containing product, but not the substrate, was barely detectable (data not shown). This lack of detection was not due to poor turnover, since the phenylhydrazine derivative of the product was observed with ease, and the substrate peak became less intense as a function of extent of turnover. Time dependent quantification of FGly and 5’-dA was performed to extract rate constants for the reaction, and is shown in Figure 6-7. Formation of 5’-dA was quantified by methods previously established in our laboratory (47), while the ratio of the sum of the peaks at m/z 1940.6 (dehydrated product) and 1958.6 (FGly product) to that at m/z 1960.6 (peptide substrate) was used to quantify FGly, with the assumption that both

254 substrate and product peptides are desorbed to similar extents.

Figure 6-6. Quantification of FGly formation by MALDI MS in the absence of phenylhydrazine. The reaction contained 100 µM AI WT AtsB, 1 mM SAM, 2 mM peptide 1 (m/z: 1960.6), and 2 mM dithionite. Samples were quenched in 1 M HCl, and then spotted on a MALDI target plate as described in Materials and Methods. t = 0 (dotted line); t = 20 min (solid line). m/z 1940.6 corresponds to the dehydration product from Schiff formation between the FGly product (m/z 1958.6) and presumably the N-terminal amino group. A careful study of the characterization of FGly adducts in peptides by MALDI time- of-flight (TOF) MS has shown that this assumption is not valid when evaluating formation of the phenylhydrazine adduct of the FGly-containing product, since it is desorbed to a much greater extent than the substrate (50). However, since the substrate and product peptides differ by only the equivalent of a molecule of water, these two species should behave similarly by MALDI-MS. The following equation, [P] = ((x + y)/(x + y +

(z – 0.25y))[S], describes calculation of the FGly product concentration, [P], for each time point, wherein x, y, and z are the peaks at m/z values 1940.6, 1958.6 and 1960.6, respectively, and [S] is the starting substrate concentration.

255

Figure 6-7. Stoichiometry of formation of 5’-dA and FGly. The reaction contained 100 µM AI WT AtsB, 1 mM SAM, 1 mM peptide 1, and 2 mM dithionite. FGly formation (open squares) and substrate peptide loss (closed squares) was detected by MALDI MS in the absence of phenylhydrazine and quantified from the intensities of the appropriate peaks according to the following equation: [P] = ((x + y)/(x + y + (z – 0.25y))[S], where P and S correspond to FGly and substrate peptide concentrations, respectively, and x, y, and z correspond to the peak intensities at m/z values 1940.6, 1958.6, and 1960.6, respectively. The appearance of 5’-dA (closed triangles) was monitored by HPLC as previously described in Materials and Methods. The time-dependent formation of FGly (open squares) is approximately linear over the 20 min span of the assay, as is the loss of the peptide substrate (closed squares). This observation of linearity, even when greater than 60% of the substrate is consumed, suggests that the reaction is most likely irreversible, the Km for the substrate is relatively low, and the products of the reaction do not inhibit it to a significant extent. Within the

256 time frame of the assay greater than 600 µM product is formed in the presence of 100 µM of enzyme, indicating that unlike the RS enzymes lipoyl synthase (47, 51) biotin synthase

(52-54), and MiaB protein (55), AtsB catalyzes multiple turnovers. The extent of turnovers may even be underestimated, since the fraction of enzyme in the active state was not

-1 determined. Values for Vmax/[ET] of 0.32 ± 0.01 and 0.36 ± 0.03 min for formation of

FGly and loss of peptide 1, respectively, were determined from linear fits of the corresponding data. The time-dependent quantification of 5’-dA from the same assay

-1 afforded a Vmax/[ET] of 0.32 ± 0.01 min , similar to that for production of FGly.

Stoichiometry of the AtsB Reaction. A second, HPLC-based, method was also developed to monitor time-dependent formation of FGly and 5’-dA. Reactions were quenched in the presence of hydroxylamine, which was also included in the mobile phase during product analysis by HPLC, as it facilitated separation of the substrate peptide from the FGly-containing product peptide. HPLC traces of the AtsB reaction, quenched at t = 0

(solid black line), t = 5 min (dotted line), t = 10 min (dashed line), and t = 20 min (solid gray line), are displayed in Figure6-8A. The peak corresponding to the substrate (peptide

2) decreases as a function of time, and three new peaks, labeled 1, 2, and 3, appear at retention times 10.6, 12.1, and 12.7 min, respectively. In addition, the peak corresponding to 5’-dA (retention time, 5.8 min) increases as a function of extent of reaction. A control reaction is shown in Figure6-8B, in which the AtsB C35A–C39A–C42A triple variant was substituted for the WT protein.

257

258

259 The spectra at t = 0 and t = 20 min overlay almost perfectly, and no formation of 5’-dA is observed. The peak at retention time 12.1 min is present at t = 0, but only increases in magnitude in the presence of an active catalyst. When peptide 2 was replaced with peptide 4, which contains an Ala substitution at the target Ser residue, no production of peptide-related product peaks are observed. Figure6-9A shows the time-dependent formation of FGly and 5’-dA (open squares and closed triangles, respectively; left Y- axis), and the concomitant time-dependent loss of substrate, peptide 2 (closed squares; right Y-axis), using dithionite as the requisite reductant. The concentration of the FGly product or remaining substrate was calculated from the ratio of the corrected areas

(tryptophan standard) of the product peaks (sum of peaks 1, 2, and 3) to the area of the substrate peak, multiplied by 1000 µM, the initial substrate concentration. The concentrations of 5’-dA and the peptide product at each time point are of similar magnitudes, as are the rate constants (Vmax/[ET]) for formation of each, 0.38 ± 0.01 and

0.36 ± 0.01 min-1, respectively. Moreover, the rate constants agree well with those obtained using MADLI-MS to quantify FGly formation. After 20 min of reaction, greater than six turnovers per AtsB are observed. The results of FGly quantification by HPLC are consistent with those from quantification by MALDI-MS, and indicate that the generation of one FGly modification requires the expenditure of one equiv of SAM.

Since dithionite is not a physiologically relevant reductant, experiments were performed to ascertain whether E. coli flavodoxin could support turnover. Flavodoxin is a flavin mononucleotide-containing protein that supplies reducing equivalents to a number of proteins (56), including RS proteins (47, 57-59) and methionine synthase (60). The proteins

260 from E. coli and K. pneumoniae are 96% identical, suggesting that the E. coli enzyme should be a suitable substitute. When the Flv/Flx/NADPH reducing system is used to supply the requisite electron, the rate of product formation diminishes dramatically, exhibiting values of Vmax/[ET] of 0.039 ± 0.002 for formation of both 5’-dA (Figure6-9B, closed triangles) and the FGly-containing peptide (Figure6-9B, open squares).

AtsB Can Act as a Cys-type FGE. As discussed in the introduction, AtsB from K. pneumoniae is a Ser-type anaerobic FGE, as its normal substrate contains a conserved seryl residue that is converted to FGly. However, anaerobic FGEs from Pseudomonas aeruginosa and Clostridium perfringens are Cys-type (2, 16). In vivo studies in E. coli indicate that the sulfatase from P. aeruginosa can be activated by endogenous FGEs (17), whereas activation of AtsA from K. pneumoniae, a Ser-type sulfatase, requires co- transformation of the atsA and atsB genes (21). Given that the FGE from C. perfringens is also a RS protein (46), the mechanism of FGly generation in anaerobic Ser-type and Cys- type FGEs is likely to be similar. Figure 6-10 shows turnover by AtsB in the presence of peptide 3, which contains a Cys substitution at the target Ser. Using dithionite as the reductant, AtsB catalyzed formation of the FGly modification with a Vmax/[ET] of 1.14 ±

0.12 min-1 (Figure6-10A, open squares), which is more than three-fold greater than that observed using peptide 2 (0.32 ± 0.01 min-1). There also appears to be some uncoupling of SAM cleavage with respect to FGly generation; the Vmax/[ET] for 5’-dA is 1.59 ± 0.21 min-1 (Figure6-10A, closed triangles), and the ratio of 5’-dA to FGly grows larger as a function of time. In the presence of the Flv/Flx/NADPH reducing system, the Vmax/[ET]

261 for formation of both 5’-dA (Figure6-10B, closed triangles) and the FGly modification

(Figure6-10B, open squares) is 0.018 ± 0.001 min-1.

262

263 6.5 Discussion

Purification and characterization of AtsB. The AtsB-catalyzed oxidation of an alcohol to an aldehyde is an intriguing reaction, because it avoids use of cofactors that are typically employed in these types of transformations, such as those that contain pyridine- or flavin-nucleotides. Given our interest in characterizing RS enzymes that contain multiple Fe/S clusters (42, 51), and the prediction by bioinformatics methods that AtsB contains three (2), we undertook to clone the atsB gene from K. pneumoniae and characterize its protein product and reaction. Although catalytically active AtsB has been difficult to purify, as evidenced by its intense study only in cell lysates (22, 23), we employed a previously developed strategy, in which the atsB gene was co-expressed with genes from an Azotobacter vinelandii operon that are known to be involved in Fe/S cluster biosynthesis (47, 48, 61). This strategy allowed the hexahistidine-tagged protein to be purified under anaerobic conditions, affording about 5 mg of product per liter of growth media. The inclusion of a peptide substrate during cell lysis as well as steps that involve concentration by ultrafiltration greatly enhanced the protein’s stability, solubility, and final yield. Analysis of iron and sulfide on a preparation of AI, 57Fe-labeled, protein indicated that it contained 8.7 ± 0.4 irons and 12.2 ± 2.6 sulfides per polypeptide; however, these stoichiometries varied from ~8 to 12 irons and sulfides among different preparations. Characterization of the 57Fe-labeled protein after reconstitution with additional 57Fe and sulfide provided strong evidence for the presence of three [4Fe–4S] clusters; the protein contained 12.3 ± 0.2 irons and 9.9 ± 0.4 sulfides per polypeptide, of which 94% of the iron was shown by Mössbauer spectroscopy to exist in the form of

264 cubane [4Fe–4S]2+ clusters (2.9 clusters per polypeptide). The finding that 7.3 ± 0.1 irons and 7.2 ± 0.2 sulfides remained with the C35A-C39A-C42A triple variant, even after reconstitution, and that 94% of all of the iron was in the form of [4Fe–4S]2+ clusters (1.7 clusters per polypeptide), clearly supports the argument for three [4Fe–4S] clusters per polypeptide.

Stoichiometry of the AtsB reaction. The oxidation of a seryl residue to FGly results in a mass change of -2, and also affords a functional group not typically observed in proteins. In previous studies MALDI-MS was often used to qualitatively assess turnover, exploiting the reactivity of the functional group with phenylhydrazine or 2,4- dinitrophenylhydrazine (16, 22, 23, 46, 50). The adduct formed allows clear separation of the substrate and product peaks; however, the ratio of product to remaining substrate cannot be used to quantify extent of turnover, because the adduct-containing product is desorbed significantly better than the unmodified substrate (50). We, therefore, quantified product formation in the absence of derivatization using an 18-aa peptide substrate taken from a region of AtsA from K. pneumoniae that contained the target seryl residue and other highly conserved aa residues. When a substrate containing a free amine at the N-terminus was employed, two product peaks were observed by MALDI-MS. One corresponded to the peptide containing an FGly modification generated from a seryl residue, while the other appeared to be a dehydration product of the FGly-modified peptide, wherein Schiff base formation occurred between the free amine and the aldehyde. Acetylation of the N- terminal amine suppressed formation of the putative cyclized product, but also negatively affected the ability of the peptide to be monitored by MALDI-MS, especially the product

265 peptide containing the FGly modification. Rate constants, however, could be determined for the reaction containing the unblocked peptide substrate by accounting for the two products formed. Quantification of FGly formation by MALDI-MS was only performed using dithionite as the reductant; however, it is worthwhile to note that formation of 5’- dA and the FGly peptide occurred with the same rate constants (0.32 ± 0.01 min and 0.32

± 0.01 min-1, respectively), indicating that the AtsB-catalyzed oxidation of a seryl residue to a FGly residue requires the expenditure of one molecule of SAM.

A second, HPLC-based, assay was also developed to quantify turnover and verify the conclusions reached using the MS-based assay. This method also had associated caveats, the most significant being that during analysis by HPLC three product peaks were formed. Although we have yet to characterize these peaks definitively, they only appear under turnover conditions and are associated with the peptide substrate and not SAM cleavage products. Our best guess at this time is that they represent different adducts of the FGly modification. Nevertheless, each of these peaks reproducibly behaved similarly as a function of time, and the ratio of the sum of their areas to that of the peptide substrate afforded rate constants (0.36 ± 0.01 min-1) that were similar to those obtained by

MALDI-MS (0.32 ± 0.01 min-1). Again, this rate constant agrees well with that obtained for formation of 5’-dA (0.38 ± 0.01 min-1). Moreover, the same 1:1 stoichiometry was also observed in the presence of the physiological reducing system (Flv/Flx/NADPH), although the rate constants were considerably smaller (0.039 ± 0.002 min-1 for formation of both 5’-dA and FGly).

266 The slower turnover observed in the presence of the Flv/Flx/NADPH reducing system is not unexpected, since the redox potentials of Fe/S clusters in RS enzymes tend to be on par with, or higher than, that of dithionite, but significantly less than that of flavodoxin

(57, 62, 63). The slower turnover rate is observed because the equilibrium concentration of enzyme poised to initiate cleavage of SAM in the presence of reduced flavodoxin is considerably less than that in the presence of dithionite. In fact, the presence of reduced flavodoxin typically does not allow EPR-detectable reduction of the Fe/S cluster(s) of RS enzymes. Nevertheless, the Vmax/[ET] for AtsB using flavodoxin as reductant is not dissimilar to single-turnover rate constants reported for biotin synthase (0.07 min-1) (54) and lipoyl synthase (0.175 min-1) (47) obtained under similar conditions.

One issue not addressed in this study pertains to the stoichiometry of reducing equivalents in the reaction. The AtsB reaction is a two-electron oxidation of an alcohol to an aldehyde. One oxidizing equivalent is provided by the 5’dA• generated from the cleavage of SAM. This cleavage reaction occurs in a reductive fashion, requiring the input of one electron. The nature of the ultimate acceptor of the second electron is not readily apparent, although the electron might reside transiently on one of the three Fe/S clusters. One possibility is that the second electron is used as the required reducing equivalent for subsequent rounds of SAM cleavage during catalytic turnover. If this were the case, however, it might be expected that in reactions in which the Flv/Flx/NADPH reducing system is employed, a lag, corresponding to slow initial reduction of the RS

[4Fe–4S] cluster by reduced Flv, would be followed by a faster phase, corresponding to internal reduction of the RS cluster as a result of turnover. This is not observed, however,

267 in these preliminary studies. In the presence of the Flv/Flx/NADPH reducing system there appears to be no lag; the rate of product formation is linear throughout the course of the activity determination. AtsB may operate similarly to the RS protein HemN, wherein an unidentified protein or small molecule is proposed to serve as an electron acceptor (64).

Alternative electron acceptors, such as oxidized DTT, might serve this function in our assays.

Possible roles for additional clusters in the AtsB reaction. K. pneumoniae AtsB contains thirteen cysteinyl residues, twelve of which have been characterized as residing

(2) 35 39 42 in three conserved motifs . Motif I consists of the cysteines within the C X3C X2C

RS sequence, and is found near the N-terminus of the protein. Another cysteine, originally considered to reside in Motif I, is found 83-87 aa C-terminal to Cys42. The remaining eight cysteines are found in the C-terminal region of the protein, residing in

270 276 291 331 334 340 344 357 Motifs II (C X5C X14C ) and III (C X2C X5C X3C X11-17C ). We envision that some of these cysteinyl residues contribute ligands to the two additional [4Fe–4S] clusters observed in this study; however, with the exception of the RS cysteine sequence, we do not know whether cysteines in any given motif all contribute ligands to the same cluster. For example, one cluster might contain two cysteinyl ligands from Motif II and two cysteinyl ligands from Motif III, while another cluster might contain one cysteinyl ligand from Motif II and two or three from Motif III. Several cysteinyl residues within the three motifs have been changed to alanyl residues to assess their importance in catalysis by AtsB. Variant proteins containing CysAla substitutions at Cys39, Cys42,

Cys270, Cys331, and Cys334 were unable to activate AtsA when their genes were co-

268 expressed in E. coli with the atsA gene. Biochemical characterization of AtsA isolated from co-transformed cells indicated no evidence of the FGly cofactor (23).

269 Our original interest in AtsB stemmed from the possibility that the protein might use a second Fe/S cluster in a sacrificial manner, as is proposed for LS, BS, and MiaB protein

(51). In this scenario, hydrogen-atom abstraction from C-3 of the target seryl residue would be followed by sulfur insertion from the second Fe/S cluster to yield a geminal hydroxy thiol intermediate, which would collapse to the aldehyde with concomitant elimination of sulfide. This mechanism predicts that AtsB should catalyze only one turnover, since it would act as both a catalyst and a substrate (51). Results presented in this study show, however, that in the presence of dithionite, approximately eight turnovers are observed during a 20-min incubation, and indeed, 50 µM AtsB is capable of converting 1 mM substrate to product within 60 min of incubation at 37°C. Therefore, a mechanism involving cryptic sulfur insertion is ruled out.

A working hypothesis for the mechanism of the AtsB reaction that is consistent with the results of this study is shown in Figure 6-11. In this mechanism, one of the remaining two Fe/S clusters is ligated only by three cysteines, providing an open coordination site to which the hydroxyl group of the target seryl residue can bind, as is found in aconitase (65) and proposed in serine dehydratase (66) and other Fe/S cluster-dependent enzymes within the hydro-lyase family (67). This coordination would facilitate two processes: it would decrease the pKa for deprotonation of the alcohol prior to, or subsequent to, abstraction of a hydrogen atom from C-3 of the target seryl residue by a 5’-dA• (68); and it would allow for facile inner-sphere electron transfer to the cluster (68). The mechanism in Figure 6-11

2 shows deprotonation of the coordinated hydroxyl group as the first step in the reaction, because it would facilitate subsequent hydrogen-atom abstraction from C-3 of the target

270 Ser residue as a result of resonance delocalization (Step 3). Gas-phase homolytic bond- dissociation energies (BDEs) have been measured and/or calculated for carbon-centered methanol and methoxide radicals. The observed BDE to generate the former is 91.8 ± 1.2 kcal mol-1 (69, 70), and is calculated to decrease by 16.5 kcal mol-1 in generating the latter

(70). In addition, ligation of the hydroxyl group to an iron atom of the Fe/S cluster is also likely to lower the BDE significantly. Abstraction of a C-3 hydrogen atom from the target seryl residue by a 5’-dA• is then followed by an inner-sphere electron transfer to generate the aldehyde product, with concomitant one-electron reduction of the coordinating cluster

(Step 4). The last step would be transfer of one electron to an appropriate acceptor (Step

5), which in vivo may be flavodoxin, itself. The results presented herein do not support a mechanism in which the electron is returned to the RS Fe/S cluster to allow for subsequent SAM cleavage in the absence of external reductant, although more detailed studies are needed to address this issue unambiguously.

The first example of in vitro turnover of an anaerobic FGE has already been reported, in which a “Cys-type” FGE from Clostridium perfringens was produced recombinantly in

E. coli and characterized (46). As expected from its sequence homology to K. pneumoniae

AtsB, which had been shown to be a member of the RS family of proteins, the protein contained 3.0 ± 0.2 mol of iron per mol of monomer in its AI form, and displayed a UV- vis spectrum that was consistent with the presence of Fe/S clusters.

Upon reconstitution of the AI protein with a ten-fold excess of iron and sulfide followed by size exclusion chromatography to remove excess small molecules, the protein was found to contain 6.0 ± 0.3 mol of iron per monomer. Whether this excess iron was in the

271 form of a [4Fe–4S] cluster was not rigorously determined; however, the absorption band at 420 nm in the UV-visible spectrum of the protein was observed to increase. We (42), and others (71), have found that proteins with adventitiously bound iron that arises from their reconstitution with iron and sulfide often display UV-vis spectra that are consistent with the presence of [4Fe–4S] clusters. Although this protein, when reconstituted, was capable of catalyzing formation of FGly at the target cysteinyl residue in a 23 aa peptide substrate, rates of turnover were not reported (46). The significant presence of starting material subsequent to long incubation times under turnover conditions (6 h with 40 µM protein) would suggest that this protein is intrinsically less active than AtsB, or that the preparation contains an abundance of inactive protein.

AtsB Can Act as a Cys-type Anaerobic FGE. It has been suggested that the pathways for generating FGly from seryl versus cysteinyl residues by anaerobic FGEs are different

(22). This conclusion derives from two significant observations from in vivo activation studies of the Cys-type FGE from P. aeruginosa, and AtsB from K. pneumoniae, a Ser- type FGE. Expression of the P. aeruginosa atsA gene in an arylsulfatase-deficient strain of P. aeruginosa restored the ability of the organism to grow on aromatic sulfate esters as its source of sulfur. When the strain was transformed with a plasmid encoding a

CysSer substitution at the target cysteinyl residue, growth on aromatic sulfate esters was abolished (17). Similar results were obtained when the P. aeruginosa atsA gene was transformed in E. coli. Isolation and characterization of the heterologously produced

AtsA showed it to contain the FGly cofactor and to be catalytically active, suggesting that

E. coli contains Cys-type FGEs (17). By contrast, AtsA from K. pneumoniae is not

272 activated upon simple expression of its atsA gene; activation requires simultaneous expression of the K. pneumoniae atsB gene, suggesting that E. coli does not harbor a Ser- type FGE (21).

It is now clear that both Cys-type and Ser-type anaerobic FGEs are RS enzymes, and that they presumably catalyze their respective reactions by very similar mechanisms (23,

46). In the present study we show that AtsB can indeed activate peptide substrates containing a CysSer substitution at the seryl residue that is oxidized to FGly. In fact, the Cys-containing peptide affords a Vmax/[ET] that is approximately four-fold greater, suggesting that Cys-type sulfatases may be easier to activate by a radical mechanism than

Ser-type sulfatases. The mechanism in Figure 6-11 is consistent with our finding that

AtsB can activate peptides containing SerCys substitutions at the position of relevance, as well as the observation that Cys-type anaerobic FGEs cannot activate Ser-type sulfatases, or activate them very poorly. In Step 1, the pKa of a metal-bound cysteinyl residue would be expected to be significantly lower than that of a metal-bound seryl residue. The general base that removes that proton in a Cys-type anaerobic FGE might not be sufficiently strong to remove the proton from a metal-coordinated alcohol. In addition, abstraction of a hydrogen atom from C-3 of the substrate residue would be easier for the cysteinyl-containing substrate than the seryl-containing substrate, since the homolytic BDE for generating a carbon-centered methanethiol radical is on the order of

92 kcal mol-1, while the value for generating a carbon-centered methanol radical is on the order of 95 kcal mol-1 (72). The greater polarizability of the thiolate anion as compared to the alkoxide anion, both of which would be generated upon deprotonation of the respective metal-coordinated heteroatom, should allow for increased stabilization of the

273 carbon-centered methanethiolate radical as compared to the carbon-centered methoxide radical (70).

Very recently–subsequent to submission of the work described herein for publication–another group produced complementary data that suggest that Cys-type and

Ser-type anFGEs function by similar mechanisms (73). In this study, the authors cloned a gene from Bacteroides thetaiotaomicron that encodes a Ser-type RS anFGE, and showed that the reconstituted protein was able to catalyze formation of 5’-dA in the presence of

23-mer peptides containing either Cys or Ser at the position of modification, with the cysteinyl-containing peptide supporting the greater amount of SAM cleavage.

Interestingly, 5’-dA formation was greatest in the absence of a peptide substrate; however, a peptide substrate containing an alanyl residue at the position of modification almost completely suppressed it. Similar behavior was observed with the Cys-type anFGE from C. perfringens, except that 5’-dA formation in the absence of peptide was considerably lower. Additional in vitro and in vivo assays to qualitatively assess FGly formation showed that both anFGEs were able to activate Ser-type and Cys-type sulfatases; however, the absence of kinetic resolution did not allow accurate quantification of the efficiency (73).

Relationship of AtsB to other RS Proteins. Enzymes within the RS superfamily catalyze a diverse array of reactions, including key steps in the biosynthesis of cofactors, coenzymes, antimicrobial agents and herbicides, the biosynthesis and repair of DNA, the regulation of cellular function, and general metabolism (25). All of these reactions involve the common intermediate, 5’-deoxyadenosyl 5’-yl (5’-dA•), generated from a reductive cleavage of SAM (28, 29, 74-76). In an early review article on this superfamily, these

274 enzymes were grouped into three classes, which were based on the net stoichiometry of

5’-dA and product generated (77). Class I enzymes were those that catalyzed the reversible cleavage of SAM, wherein a product radical abstracts a hydrogen atom from 5’-dA, initiating reformation of SAM presumably via attack of the 5’-dA• on L-methionine with concomitant transfer of an electron back to the cluster. Characterized members of Class I enzymes include lysine 2,3-aminomutase (27, 75), a SAM-dependent glutamate mutase (78), and spore photoproduct lyase (79-82), which catalyzes the repair of the unique thymine dimer photoproduct, 5-thyminyl-5,6-dihydrothymine, found in UV-irradiated DNA of spores from Bacillus subtilis. Class II enzymes were those that catalyzed an irreversible cleavage of SAM, but used the 5’-dA• to generate a glycyl radical cofactor via

-carbon of a specific glycine residue on a cognate protein (83). Glycyl radical cofactors initiate turnover by abstracting an appropriate hydrogen atom from a small-molecule substrate or a protein cysteinyl residue, and are regenerated after each catalytic event. Therefore, the expenditure of one

SAM molecule allows for multiple turnovers via creation of a radical-based cofactor.

Several Class II RS enzymes have been studied in vitro at varying levels of detail. The best characterized are the activating enzymes of the anaerobic ribonucleotide reductase and pyruvate formate–lyase, both from E. coli (28, 29, 76, 84).

At the time of the early review very little was known about Class III enzymes. These enzymes cleave SAM irreversibly for each hydrogen atom abstracted by a 5’-dA• in a given reaction sequence. At this writing, six members have been characterized in moderate detail: biotin synthase (BS) (85), lipoyl synthase (LS) (42, 86), MiaB protein (55, 87),

275 MoaA protein (32), coproporphyrinogen III oxidase (HemN) (31), and BtrN (88). The first three proteins catalyze sulfur insertion into unactivated C–H bonds, representing the final step in the biosynthesis of biotin, lipoic acid cofactor, and the hypermodified tRNA nucleoside 2-methylthio-N6-(isopentenyl)adenosine, respectively (51). In these reactions evidence suggests that the inserted sulfur atom derives from a second Fe/S cluster on the enzyme itself, and one equiv of SAM is cleaved irreversibly to 5’-dA for each sulfur– carbon bond formed (51). The last three enzymes catalyze key steps in the biosynthesis of molybdopterin, the anoxic biosynthesis of heme, and the biosynthesis of butirosin, a 2- deoxystreptamine-containing aminoglycoside antibiotic (88). BtrN catalyzes a reaction that is analogous to that of AtsB, which is the two-electron oxidation of a hydroxyl group to a ketone in the conversion of 3-amino-2,3-dideoxy-scyllo-inosamine to 3-amino-2,3- dideoxy-scyllo-inosose. Recent mechanistic analysis of this enzyme indicates a 1:1 stoichiometry of 5’-dA and product formation, as is observed in the AtsB reaction.

However, the authors suggest that the extra electron is returned to the RS [4Fe–4S] cluster for subsequent rounds of SAM cleavage. Currently there is no experimental evidence that indicates that BtrN contains more than one [4Fe–4S] cluster per polypeptide; however, the primary structure of the protein contains eight cysteinyl residues. The studies presented herein indicate that AtsB also is a Class III RS enzyme.

Although it catalyzes multiple turnovers, it cleaves S-adenosyl-L-methionine irreversibly and stoichiometrically with respect to product formed. As is the case with HemN, which uses the 5’-dA• to initiate a two-electron oxidative decarboxylation of coproporphyrinogen-III (31), further studies are required to identify the acceptor of the second electron.

276 6.6 Acknowledgements

We would like to thank Heike Betz at the Proteomics and Mass Spectrometry Core

Facility, PSU for her assistance with the MALDI-MS.

277 6.7 References

1. Parenti, G., Meroni, G., and Ballabio, A. (1997) The sulfatase gene family, Curr

Opin Genet Dev 7, 386-391.

2. Schirmer, A., and Kolter, R. (1998) Computational analysis of bacterial sulfatases

and their modifying enzymes, Chem. Biol. 5, R181-R186.

3. von Figura, K., Schmidt, B., Selmer, T., and Dierks, T. (1998) A novel protein

modification generating an aldehyde group in sulfatases: its role in catalysis and

disease, Bioessays 20, 505-510.

4. Hanson, S. R., Best, M. D., and Wong, C.-H. (2004) Sulfatases: structure,

mechanism biological activity, inhibition, and synthetic utility, Angew. Chem. Int.

Ed. 43, 5736-5763.

5. Gadler, P., and Faber, K. (2007) New enzymes for biotransformations: microbial

alkyl sulfatases displaying stereo- and enantioselectivity, Trends Biotechnol. 25,

83-88.

6. Schmidt, B., Selmer, T., Ingendoh, A., and von Figura, K. (1995) A novel amino

acid modification in sulfatases that is defective in multiple sulfatase deficiency,

Cell 82, 271–278.

7. Müller, I., Kahnert, A., Pape, T., Sheldrick, G. M., Meyer–Klaucke, W., Dierks,

T., M., K., and Usón, I. (2004) Crystal structure of the alkylsulfatase AtsK:

insights into the catalytic mechanism of the Fe(II) alpha-ketoglutarate-dependent

dioxygenase superfamily, Biochemistry 43, 3075-3088.

278 8. Hagelueken, G., Adams, T. M., Wiehlmann, L., Widow, U., Kolmar, H.,

Tümmler, B., Heinz, D. W., and Chubert, W.-D. (2006) The crystal structure of

SdsA1, an alkylsulfatase from Pseudomonas aeruginosa, defines a third class of

sultatases, Proc. Natl. Acad. Sci. USA 103, 7631-7636.

9. Boltes, I., Czapinska, H., Kahnert, A., von Bülow, R., Dierks, T., Schmidt, B.,

von Figura, K., Kertesz, M. A., and Usón, I. (2001) 1.3 Å structure of

arylsulfatase from Pseudomonas aeruginosa establishes the catalytic mechanism

of sulfate ester cleavage in the sulfatase family, Structure 9, 483-491.

10. Dierks, T., Dickmanns, A., Preusser-Kunze, A., Schmidt, B., Mariappan, M., von

Figura, K., Ficner, R., and Rudolph, M. G. (2005) Molecular basis for multiple

sulfatase deficiency and mechanism for formylglycine generation of the human

formylglycine-generating enzyme, Cell 121, 541-552.

11. Ghosh, D. (2007) Human sulfatases: a structural perspective to catalysis, Cell.

Mol. Life Sci. 64, 2013-2022.

12. Dodgson, K. S., White, G. F., and Fitzgerald, J. W. (1982) Sulfatases of Microbial

Origin, Vol. 2, CRC Press, Inc., Boca Raton,.

13. Murooka, Y., Ishibashi, K., Yasumoto, M., Sasaki, M., Sugino, H., Azakami, H.,

and Yamashita, M. (1990) A sulfur- and tyramine-regulated Klebsiella aerogenes

operon containing the arylsulfatase (atsA) gene and the atsB gene, J. Bacteriol.

172, 2131-2140.

14. Dierks, T., Schmidt, B., Borissenko, L. V., Peng, J., Preusser, A., Mariappan, M.,

and von Figura, K. (2003) Multiple sulfatase deficiency is caused by mutations in

279 the gene encoding the human Cα-formylglycine generating enzyme, Cell 113,

435-444.

15. Cosma, M. P., Pepe, S., Annunziata, I., Newbold, R. F., Grompe, M., Parenti, G.,

and Ballabio, A. (2003) The multiple sulfatase deficiency gene encodes an

essential and limiting factor for the activity of sulfatases, Cell 113, 445-456.

16. Berteau, O., Guillot, A., Benjdia, A., and Rabot, S. (2006) A new type of bacterial

sulfatase reveals a novel maturation pathway in prokaryotes, J. Biol. Chem. 281,

22464-22470.

17. Dierks, T., Miech, C., Hummerjohann, J., Schmidt, B., Kertesz, M. A., and von

Figura, K. (1998) Posttranslational formation of formylglycine in prokaryotic

sulfatases by modification of either cysteine or serine, J. Biol. Chem. 273, 25560-

25564.

18. Miech, C., Dierks, T., Selmer, T., von Figura, K., and Schmidt, B. (1998)

Arylsulfatase from Klebsiella pneumoniae carries a formylglycine generated from

a serine, J. Biol. Chem. 273, 4835-4837.

19. Lukatela, G., Krauss, N., Theis, K., Selmer, T., Gieselmann, V., von Figura, K.,

and Saenger, W. (1998) Crystal structure of human arylsulfatase A: the aldehyde

function and the metal ion at the active site suggest a novel mechanism for sulfate

ester hydrolysis, Biochemistry 37, 3654-3664.

20. Recksiek, M., Selmer, T., Dierks, T., Schmidt, B., and von Figura, K. (1998)

Sulfatases, trapping of the sulfated enzyme intermediate by substituting the active

site formylglycine, J. Biol. Chem. 273, 6096-6103.

280 21. Szameit, C., Miech, C., Balleininger, M., Schmidt, B., von Figura, K., and Dierks,

T. (1999) The iron sulfur protein AtsB is required for posttranslational formation

of formylglycine in the Klebsiella sulfatase, J. Biol. Chem. 274, 15375-15381.

22. Marquordt, C., Fang, Q. H., Will, E., Peng, J. H., von Figura, K., and Dierks, T.

(2003) Posttranslational modification of serine to formylglycine in bacterial

sulfatases: recognition of the modification motif by the iron-sulfur protein AtsB,

J. Biol. Chem. 278, 2212-2218.

23. Fang, Q., Peng, J., and Dierks, T. (2004) Post-translational formylglycine

modification of bacterial sulfatases by the radical S-adenosylmethionine protein

AtsB, J. Biol. Chem. 279, 14570-14578.

24. The_UniProt_Consortium. (2007) The Universal Protein Resource (UniProt),

Nucleic Acids Res. 35, D193-D197.

25. Sofia, H. J., Chen, G., Hetzler, B. G., Reyes-Spindola, J. F., and Miller, N. E.

(2001) Radical SAM, a novel protein superfamily linking unresolved steps in

familiar biosynthetic pathways with radical mechanisms: functional

characterization using new analysis and information visualization methods,

Nucleic Acids. Res. 29, 1097-1106.

26. Layer, G., Heinz, D. W., Jahn, D., and Schubert, W. D. (2004) Structure and

function of radical SAM enzymes, Curr. Opin. Chem. Biol. 8, 468-476.

27. Frey, P. A., Hegeman, A. D., and Reed, G. H. (2006) Free radical mechanisms in

enzymology, Chem. Rev. 106, 3302-3316.

281 28. Fontecave, M., Mulliez, E., and Ollagnier-de Choudens, S. (2001)

Adenosylmethionine as a source of 5'-deoxyadenosyl radicals, Curr. Opin. Chem.

Biol. 5, 506-511.

29. Cheek, J., and Broderick, J. B. (2001) Adenosylmethionine-dependent iron-sulfur

enzymes: versatile clusters in a radical new role, J. Biol. Inorg. Chem. 6, 209-226.

30. Lepore, B. W., Ruzicka, F. J., Frey, P. A., and Ringe, D. (2005) The x-ray crystal

structure of lysine-2,3-aminomutase from Clostridium subterminale, Proc. Natl.

Acad. Sci. USA 102, 13819-13824.

31. Layer, G., Moser, J., Heinz, D. W., Jahn, D., and Schubert, W. D. (2003) Crystal

structure of coproporphyrinogen III oxidase reveals cofactor geometry of Radical

SAM enzymes, EMBO J. 22, 6214-6224.

32. Hänzelmann, P., and Schindelin, H. (2004) Crystal structure of the S-

adenosylmethionine-dependent enzyme MoaA and its implications for

molybdenum cofactor deficiency in humans, Proc. Natl. Acad. Sci. USA 101,

12870-12875.

33. Berkovitch, F., Nicolet, Y., Wan, J. T., Jarrett, J. T., and Drennan, C. L. (2004)

Crystal structure of biotin synthase, an S-adenosylmethionine-dependent radical

enzyme, Science 303, 76-79.

34. Walsby, C. J., Ortillo, D., E., B. W., Broderick, J. B., and Hoffman, B. M. (2002)

An anchoring role for FeS clusters: chelation of the amino acid moiety of S-

adenosylmethionine to the unique iron site of the [4Fe–4S] cluster of pyruvate

formate–lyase activating enzyme, J. Am. Chem. Soc. 124, 11270-11271.

282 35. Walsby, C. J., Hong, W., Broderick, W. E., Cheek, J., Ortillo, D., Broderick, J. B.,

and Hoffman, B. M. (2002) Electron-nuclear double resonance spectroscopic

evidence that S-adenosylmethionine binds in contact with the catalytically active

[4Fe-4S]+ cluster of pyruvate formate-lyase activating enzyme, J. Am. Chem. Soc.

124, 3143-3151.

36. Krebs, C., Broderick, W. E., Henshaw, T. F., Broderick, J. B., and Huynh, B. H.

(2002) Coordination of adenosylmethionine to a unique iron site of the [4Fe–4S]

of pyruvate formate–lyase activating enzyme: a Mössbauer spectroscopic study, J.

Am. Chem. Soc. 124, 912-913.

37. Chen, D., Walsby, C., Hoffman, B. M., and Frey, P. A. (2003) Coordination and

mechanism of reversible cleavage of S-adenosylmethionine by the [4Fe-4S]

center in lysine 2,3-aminomutase, J. Am. Chem. Soc. 125, 11788-11789.

38. Iwig, D. F., and Booker, S. J. (2004) Insight into the polar reactivity of the onium

chalcogen analogues of S-adenosyl-L-methionine, Biochemistry 43, 13496-13509.

39. Kennedy, M. C., Kent, T. A., Emptage, M., Merkle, H., Beinert, H., and Münck,

E. (1984) Evidence for the formation of a linear [3Fe-4S] cluster in partially

unfolded aconitase, J. Biol. Chem. 259, 14463-14471.

40. Beinert, H. (1983) Semi-micro methods for analysis of labile sulfide and of labile

sulfide plus sulfane sulfur in unusually stable iron-sulfur proteins, Anal. Biochem.

131, 373-378.

41. Beinert, H. (1978) Micro methods for the quantitative determination of iron and

copper in biological material, Methods Enzymol. 54, 435-445.

283 42. Cicchillo, R. M., Lee, K. H., Baleanu-Gogonea, C., Nesbitt, N. M., Krebs, C., and

Booker, S. J. (2004) Escherichia coli lipoyl synthase binds two distinct [4Fe-4S]

clusters per polypeptide, Biochemistry 43, 11770-11781.

43. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A

Laboratory Manual, Vol. 3, 2nd ed., Cold Spring Harbor Laboratory Press,

Plainview, New York.

44. Ramamurthy, V., Swann, S. L., Paulson, J. L., Spedaliere, C. J., and Mueller, E.

G. (1999) Critical aspartic acid residues in pseudouridine synthases, J. Biol.

Chem. 274, 22225-22230.

45. Bradford, 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.

46. Benjdia, A., Leprince, J., Guillot, A., Vaudry, H., Rabot, S., and Berteau, O.

(2007) Anaerobic sulfatase-maturating enzymes: radical SAM enzymes able to

catalyze in vitro sulfatase post-translational modification, J. Am. Chem. Soc. 129,

3462.

47. Cicchillo, R. M., Iwig, D. F., Jones, A. D., Nesbitt, N. M., Baleanu-Gogonea, C.,

Souder, M. G., Tu, L., and Booker, S. J. (2004) Lipoyl synthase requires two

equivalents of S-adenosyl-L-methionine to synthesize one equivalent of lipoic

acid, Biochemistry 43, 6378-6386.

48. Johnson, D., Unciuleac, M., and Dean, D. (2006) Controlled expression and

functional analysis of iron-sulfur cluster biosynthetic components within

Azotobacter vinelandii, J. Bacteriol. 188, 7551-7561.

284 49. Emptage, M. H., Kent, T. A., Huynh, B. H., Rawlings, J., Ormejohnson, W. H.,

and Münck, E. (1980) Nature of the Iron-Sulfur Centers in a Ferredoxin from

Azotobacter vinelandii - Mössbauer Studies and Cluster Displacement

Experiments, J Biol Chem 255, 1793-1796.

50. Peng, J., Schmidt, B., von Figura, K., and Dierks, T. (2003) Identification of

formylglycine in sulfatases by matrix-assisted laser desorption/ionization time-of-

flight mass spectrometry, J. Mass Spectrom. 38, 80-86.

51. Booker, S. J., Cicchillo, R. M., and Grove, T. L. (2007) Self-sacrifice in radical S-

adenosylmethionine proteins, Curr. Opin. Chem. Biol. 11, 543-552.

52. Jameson, G. N. L., Cosper, M. M., Hernández, H. L., Johnson, M. K., and Huynh,

B. H. (2004) Role of the [2Fe–2S] cluster in recombinant Escherichia coli biotin

synthase, Biochemistry 43, 2022-2031.

53. Tse Sum Bui, B., Benda, R., Schünemann, V., Florentin, D., Trautwein, A. X.,

and Marquet, A. (2003) Fate of the [2Fe-2S]2+ cluster of Escherichia coli biotin

synthase during reaction: a Mössbauer characterization, Biochemistry 42, 8791-

8798.

54. Ugulava, N. B., Sacanell, C. J., and Jarrett, J. T. (2001) Spectroscopic changes

during a single turnover of biotin synthase: destruction of a [2Fe–2S] cluster

accompanies sulfur insertion, Biochemistry 40, 8352-8358.

55. Pierrel, F., Douki, T., Fontecave, M., and Atta, M. (2004) MiaB protein is a

bifunctional radical-S-adenosylmethionine enzyme involved in thiolation and

methylation of tRNA, J. Biol. Chem. 279, 47555-47653.

285 56. Sancho, J. (2006) Flavodoxins: sequence, folding, binding, function and beyond,

Cell. Mol. Life Sci. 63, 855-864.

57. Mulliez, E., Padovani, D., Atta, M., Alcouffe, C., and Fontecave, M. (2001)

Activation of Class II Ribonucleotide Reductase by flavodoxin: A Protein

Radical-Driven Electron Transfer to the Iron–sulfur Center, Biochemistry 40,

3730-3736.

58. Birch, O. M., Fuhrmann, M., and Shaw, N. M. (1995) Biotin synthase from

Escherichia coli, an investigation of the low molecular weight and protein

components required for activity in vitro, J. Biol. Chem. 270, 19158-19165.

59. Knappe, J., and Schmitt, T. (1976) A novel reaction of S-adenosyl-L-methionine

correlated with the activation of pyuvate formate–lyase, Biochem. Biophys. Res.

Commun. 71, 1110-1117.

60. Hoover, D. M., Jarrett, J. T., Sands, R. H., Dunham, W. R., Ludwig, M. L., and

Matthews, R. G. (1997) Interaction of Escherichia coli cobalamin-dependent

methionine synthase and its physiological partner: binding of flavodoxin leads to

axial ligand dissociation from the cobalamin, Biochemistry 36, 127-138.

61. Frazzon, J., and Dean, D. R. (2003) Formation of iron–sulfur clusters in bacteria:

an emerging field in bioinorganic chemistry, Curr. Opin. Chem. Biol. 7, 166-173.

62. Wang, S. C., and Frey, P. A. (2007) Binding energy in the one-electron reductive

cleavage of S-adenosylmethionine in lysine 2,3-aminomutase, a radical SAM

enzyme, Biochemistry, 12889-12895.

286 63. Ugulava, N. B., Gibney, B. R., and Jarrett, J. T. (2001) Biotin synthase contains

two distinct iron–sulfur binding sites: chemical and spectroelectrochemical

analysis of iron–sulfur cluster interconversions, Biochemistry 40, 8343-8351.

64. Layer, G., Verfurth, K., Mahlitz, E., and Jahn, D. (2002) Oxygen-independent

coproporphyrinogen-III oxidase HemN from Escherichia coli, J. Biol. Chem. 277,

34136-34142.

65. Beinert, H., Kennedy, M. C., and Stout, C. D. (1996) Aconitase as iron-sulfur

protein, enzyme, and iron-regulatory protein, Chem. Rev. 96, 2335-2373.

66. Cicchillo, R. M., Baker, M. A., Schnitzer, E. J., Newman, E. B., Krebs, C., and

Booker, S. J. (2004) Escerhichia coli L-serine deaminase requires a [4Fe–4S]

cluster in catalysis, J. Biol. Chem. 279, 32418-32425.

67. Fling, D. H., and Allen, R. M. (1996) Iron–sulfur proteins with nonredox

functions, Chem. Rev. 96, 2315-2334.

68. Lippard, S. J., and Berg, J. M. (1994) Principles of Bioinorganic Chemistry,

University Science Books, Mill Valley, CA.

69. Golden, D. M., and Benson, S. W. (1969) Free-radical and molecule

thermochemistry from studies of gas-phase iodine-atom reactions, Chem. Rev. 69,

125-134.

70. Steigerwald, M. L., Goddard III, W. A., and Evans, D. A. (1979) Theoretical

studies of the oxy anionic substitutent effect, J. Am. Chem. Soc. 101, 194-1997.

71. Cosper, M. M., Jameson, G. N. L., Hernández, H. L., Krebs, C., Huynh, B. H.,

and Johnson, M. K. (2004) Characterization of the cofactor composition of

Escherichia coli biotin synthase, Biochemistry 43, 2007-2021.

287 72. Henry, D. J., Parkinson, C. J., Mayer, P. M., and Radom, L. (2001) Bond

dissociation energies and radical stabilization energies associated with substituted

methyl radicals, J. Phys. Chem. A 105, 6750-6756.

73. Benjdia, A., Subramanian, S., Leprince, J., Vaudry, H., Johnson, M. K., and

Berteau, O. (2008) Anaerobic sulfatase-maturating enzymes – first dual substrate

radical S-adenosylmethionine enzymes, J. Biol. Chem. April 11, 2008,

doi:10.1074/jbc.M710074200.

74. Jarrett, J. T. (2003) The generation of 5'-deoxyadenosyl radicals by

adenosylmethionine-dependent radical enzymes, Curr. Opin. Chem. Biol. 7, 174-

182.

75. Frey, P. A., and Magnusson, O. T. (2003) S-Adenosylmethionine: a wolf in

sheep's clothing, or a rich man's adenosylcobalamin?, Chem. Rev. 103, 2129-

2148.

76. Frey, P. A., and Booker, S. J. (2001) Radical mechanisms of S-

adenosylmethionine-dependent enzymes, Adv. Protein Chem. 58, 1-45.

77. Frey, P. A., and Booker, S. (1999) Radical intermediates in the reaction of lysine

2,3-aminomutase, In Advances in Free Radical Chemistry (Zard, S. Z., Ed.), pp 1-

43, JAI Press Inc., Stamford, CT.

78. Ruzicka, F. J., and Frey, P. A. (2007) Glutamate 2,3-aminomutase: a new member

of the radical SAM superfamily of enzymes, Biochim. Biophys. Acta 1774, 286-

296.

79. Slieman, T. A., Rebeil, R., and Nicholson, W. L. (2000) Spore photoproduct (SP)

lyase from Bacillus subtilis specifically binds to and cleaves SP (5-thyminyl-5,6-

288 dihydrothymine) but not cyclobutane pyrimidine dimers in UV-irradiated DNA, J.

Bacteriol. 182, 6412-6417.

80. Rebeil, R., and Nicholson, W. L. (2001) The subunit structure and catalytic

mechanism of the Bacillus subtilis DNA repair enzyme spore photoproduct lyase,

Proc. Natl. Acad. Sci. USA 98, 9038-9043.

81. Cheek, J., and Broderick, J. B. (2002) Direct H atom abstraction from spore

photoproduct C-6 initiates DNA repair in the reaction catalyzed by spore

photoproduct lyase: evidence for a reversibly generated adenosyl radical

intermediate, J. Am. Chem. Soc. 124, 2860-2861.

82. Buis, J. M., Cheek, J., Kalliri, E., and Broderick, J. B. (2006) Characterization of

an active spore photoproduct lyase, a DNA repair enzyme in the radical S-

adenosylmethionine superfamily, J. Biol. Chem. 281, 25994-26003.

83. Wagner, A. F., Frey, M., Neugebauer, F. A., Schafer, W., and Knappe, J. (1992)

The free radical in pyruvate formate-lyase is located on glycine-734, Proc. Natl.

Acad. Sci. 89, 996-1000.

84. Knappe, J., and Wagner, A. F. (2001) Stable glycyl radical from pyruvate

formate-lyase and ribonucleotide reductase (III), Adv. Protein Chem. 58, 277-315.

85. Jarrett, J. T. (2005) The novel structure and chemistry of iron-sulfur clusters in the

adenosylmethionine-dependent radical enzyme biotin synthase, Arch. Biochem.

Biophys. 433, 312-321.

86. Cicchillo, R. M., and Booker, S. J. (2005) Mechanistic investigations of lipoic

acid biosynthesis in Escherichia coli: both sulfur atoms in lipoic acid are

289 contributed by the same lipoyl synthase polypeptide, J. Am. Chem. Soc. 127,

2860-2861.

87. Hernández, H. L., Pierrel, F., Elleingand, E., García-Serres, R., Huynh, B. H.,

Johnson, M. K., Fontecave, M., and Atta, M. (2007) MiaB, a bifunctional radical-

S-adenosylmethionine enzyme involved in the thiolation and methylation of

tRNA, contains two essential [4Fe–4S] clusters, Biochemistry 46, 5140-5147.

88. Yokoyama, K., Numakura, M., Kudo, F., Ohmori, D., and Eguchi, T. (2007)

Characterization and mechanistic study of a radical SAM dehydrogenase in the

biosynthesis of butirosin, J. Am. Chem. Soc. 129, 15147-15155.

290 Chapter 7

A General Cofactor Requirement for Radical S-adenoyslmethionine- dependent Dehydrogenases? BtrN contains two [4Fe–4S] Clusters

This chapter was reproduced from “Grove, T.L., Ahlum, J.H., Sharma, P., Krebs, C., Booker, S.J. A consensus mechanism for Radical SAM-dependent dehydrogenation? BtrN contains two [4Fe-4S] clusters. Biochemistry. 2010 May 11;49(18):3783-3785.

291 7.1 Abstract

BtrN catalyzes the third step in the biosynthetic pathway for the generation of the aminoglycoside antibiotic butirosin B, which is a two-electron oxidation of the C3 secondary alcohol of 2-deoxy-scyllo-inosamine (DOIA) to the corresponding ketone. The enzyme is a member of the radical S-adenosylmethionine (SAM) superfamily of enzymes, and has been shown to use a 5’-deoxyadenosyl 5’-radical (5’-dA•) to abstract the C3 hydrogen atom of DOIA, a one-electron oxidation of the substrate, which initiates turnover and results in production of a substrate-radical intermediate. Previous studies on

BtrN suggested that it contains one [4Fe–4S] cluster, a requisite cofactor found in all radical SAM (RS) family members, which serves to deliver the electron necessary for the reductive cleavage of SAM to generate the 5’-dA•. In this work, we show using

Mössbauer spectroscopy in combination with site-directed mutagenesis and quantitative analyses for labile iron and sulfide that BtrN contains two [4Fe–4S]2+ clusters, each of which is ligated by only three cysteinyl residues. We suggest that one of the two iron– sulfur clusters binds in contact with the C3 hydroxyl group of DOIA, and acts as the immediate acceptor of an electron from the substrate-radical intermediate via inner- sphere electron transfer, affording the product, amino-2-deoxy-scyllo-inosose, and a reduced, [4Fe–4S]+, iron–sulfur cluster. This finding places BtrN in accord with AtsB, another RS dehydrogenase that contains multiple [4Fe–4S] clusters, and suggests a common mechanism for RS dehydrogenation.

292 7.2 Introduction

BtrN catalyzes the third step in the biosynthetic pathway of the 2-deoxystreptamine

(DOS) (1)-containing aminoglycoside antibiotic butirosin B. The reaction involves a seemingly simple two-electron oxidation of the C3 secondary alcohol of 2-deoxy-scyllo- inosamine (DOIA) to a ketone, affording amino-2-deoxy-scyllo-inosose (amino-DOI)

(Figure 7-1) (2). Interestingly, BtrN does not contain or employ any of the usual suspects that act as cofactors or cosubstrates in these types of oxidations, such as flavin-, pyridine-

, or quinone-containing nucleotide metabolites. In fact, a thorough in vitro analysis of

BtrN and its reaction indicates that the protein is a member of the radical S-adenosyl-L- methionine (SAM) superfamily, and that it catalyzes this two-electron oxidation via a radical mechanism (2, 3).

Figure 7-1. Reaction catalyzed by BtrN Enzymes within the radical SAM (RS) superfamily use SAM as a precursor to a 5’- deoxyadenosyl 5’-radical (5’-dA•), an obligate intermediate, which initiates turnover by abstracting a key hydrogen atom from the appropriate substrate (4, 5). All known RS enzymes contain at least one [4Fe–4S] cluster that is ligated by three cysteinyl residues lying in a canonical CxxxCxxC motif (6, 7). The sole proven variation is the RS enzyme

ThiC, in which a [4Fe–4S] cluster is ligated by cysteinyl residues lying in a CxxCxxxxC motif (8, 9). SAM binds bidentate via its -amino and -carboxylate groups to the non-

293 cysteinyl-coordinated Fe site of this cluster (10-12). In its reduced state, the [4Fe–4S]+ cluster injects an electron into the sulfonium group of SAM, inducing fragmentation of the molecule into L-methionine and a 5’-dA• (13, 14). Spectroscopic, biochemical, and analytical characterization of BtrN suggested that the protein contains one [4Fe–4S] cluster per polypeptide, and that the 5’-dA• initiates turnover by abstracting the hydrogen atom from C3 of the substrate (2, 3). It was postulated that the remaining electron is returned to this [4Fe–4S]2+ cluster concomitant with ketone formation to allow for subsequent rounds of SAM cleavage by the [4Fe-4S]+ cluster in the absence of additional reducing equivalents (2).

The reaction catalyzed by BtrN is similar to those of the RS-dependent sulfatase maturation enzymes, which catalyze the oxidation of a seryl or cysteinyl residue to a formylglycyl (FGly) residue on a cognate protein (15-18). This FGly residue is then used as a cofactor in the hydrolysis of organosulfate monoesters (19). Recent characterization of AtsB, a RS-dependent sulfatase maturation enzyme, showed that it contains three

[4Fe–4S] clusters, only one of which—Cluster A—is housed in the CxxxCxxC motif

(17). A working hypothesis was advanced in which one of the two remaining clusters is coordinated by the target seryl or cysteinyl residue of the substrate to facilitate loss of the second electron from a radical-containing substrate intermediate via inner-sphere electron transfer to afford the oxidized product (17). The finding that BtrN contains only one

[4Fe-4S] cluster (2) would suggest that this mechanism is not operative, or that it is not universally conserved among RS dehydrogenases. However, a close inspection of the primary structure of the protein shows that it contains eight cysteinyl residues, a quantity sufficient to coordinate two [4Fe–4S] clusters. In this work, we demonstrate using a

294 combination of spectroscopic and analytical methods that BtrN harbors two [4Fe-4S] clusters, and is therefore in accord with the mechanism that we proposed for RS- dependent dehydrogenation, specifically for the reaction catalyzed by AtsB (17). The results suggest that the presence of at least two [4Fe-4S] clusters is a general feature of

RS dehydrogenases.

Materials and Methods

Materials. All DNA-modifying enzymes and reagents were purchased from New

England Biolabs (Beverly, MA). DNA sequencing was carried out at the Pennsylvania

State University Nucleic Acid Facility. 57Fe (97-98%) metal was purchased from

Pennwood Chemicals (Great Neck, NY). It was dissolved with heating in an anaerobic

57 57 solution of 2 N H2SO4 (1.5 mol of H2SO4 per mole of Fe). The Fe solution was used as is for supplementation in E. coli culture media, and was titrated to pH 6.5 with an anaerobic solution of saturated sodium bicarbonate for in vitro reconstitution.

Coupled Assay of BtrN Wt and Variants. BtrN Wt (or variants) was added to the following reaction mixture to a final concentration of 20 µM in a total of 150 µL: 50 mM

HEPES pH 7.5, 2 mM SAM, 5mM 2-DOIA, 3 mM Dithionite, 10 mM Glutamine, 0.5 mM PLP, and 20 µM BtrR. The reaction was performed at 37 C, with 10 µL aliquots quenched at 0, 1, 5, 10, and 15 minute intervals into 10 µL of 100 mM H2SO4 and 100

µM Tryptophan (Internal Standard). The samples were subsequently centrifuged at

10,000 rpm and analyzed with the below LC-MS method. Standard curves were generated with purchased compounds of the highest quality for 5’-dAdenosine and 2-

DOS. To make sure that the BtrR was not rate limiting, a reaction was performed with

40 µM BtrR. This reaction yielded a similar rate.

295 General Methods: High-performance liquid chromatography (HPLC) with detection by mass spectrometry (LC-MS) was conducted on an Agilent Technologies (Santa Clara,

CA) 1200 system, which was fitted with an autosampler for sample injection and coupled to an Agilent Technologies 6410 QQQ mass spectrometer. The system was operated with the associated MassHunter software package, which was also used for data collection and analysis. Assay mixtures were separated on an Agilent Technologies Zorbax Rapid

Resolution SB-C18 column (2.4 mm x 35 mm, 3.5 µm particle size), which was equilibrated in 90% solvent A (5 mM perfluoroheptanoic acid–6 mM ammonium formate

(aqueous), pH 3) and 10% acetonitrile. A gradient of 10–30% acetonitrile was applied from 0 to 8 min, and then from 30–10% acetonitrile from 8 to 8.5 min to restore the system to the initial conditions. The column was allowed to re-equilibrate for 3.5 min under initial conditions before subsequent sample injections. Detection of substrates and products (Table 1) was performed using electrospray ionization in positive mode (ESI+) with multiple reaction monitoring.

Table 7-1. Retention times and monitored m/z values for products and substrates Retention Time Parent Ion* Product Ion 1† Product Ion 2† DOIA 1.2 min 164.1 (90) 146.1 (9) 110.1 (11) 5’-dA 4.7 min 252.1 (90) 136 (13) 119 (50) Tryp (IS) 6.2 min 188 (130) 146.1 (10) 118 (21) DOS 9.5 min 163.1 (90) 102 (10) 145 (9) *Respective fragmentor voltages are in parenthesis. †Respective collision energies are in parenthesis.

Spectroscopic Methods. UV-visible spectra were recorded on a Cary 300 spectrometer

(Varian, Walnut Creek, CA), employing the associated WinUV software package for operating the instrument and manipulating the data. Mössbauer spectra were recorded on

296 a spectrometer from WEB Research (Edina, MN), which was equipped with an SVT-400 cryostat from Janis Research Co (Wilmington, MA). Spectra were collected in constant acceleration mode in transmission geometry. Isomer shifts are quoted relative to the centroid of α-Fe at room temperature. Spectra were analyzed with the program WMOSS from WEB Research.

Cloning of the Bacillus circulans btrN Gene. The btrN gene (Q8G907), codon- optimized for expression in E. coli, was obtained from GeneArt (Burlingame, CA), supplied in plasmid pMA. It was excised by digestion with NdeI and EcoRI, and cloned into similarly digested expression vector pET-26b, allowing for the protein to be produced with a C-terminal hexahistidine tag. The resulting plasmid was denoted pBtrNWt. The sequence of the codon-optimized btrN gene is displayed below. Bases in upper-case letters are those corresponding to codons that encode cysteinyl residues, and those that are underlined correspond to codons that encode the cysteinyl residues of the

CxxxCxxC motif

Gene sequence from GeneArt for btrN.

5’atggataaactgttcagcatgatcgaagtggaagtgaacagccagTGCaaccgtaccTGCtggtatTGCccgaacag cgtgagcaaacgtaaagaaaccggcgaaatggacccggcgctgtataaaaccctgatggaacaactgtctagcctggattttgc gggccgtattagctttcatttttacggcgaaccgctgctgTGCaaaaacctggacctgtttgtgggcatgaccaccgaatatattc cgcgtgcgcgtccgattatttataccaacggcgatttcctgaccgaaaaacgtctgcagaccctgaccgaactgggcattcagaa atttattgtgacccagcatgcgggtgcgaaacataaatttcgcggcgtgtatgatcagctggccggtgcggataaagaaaaagtg gtgtacctggatcatagcgatctggttctgagcaaccgtggcggcattctggataacattccgcaggcgagcaaagcgaacatg agcTGCatggtgccgagcaacctggccgtggtgaccgtgctgggcaatgtgctgccgTGCtttgaagatttcaaccagaaa atggtgatgggcaacattggcgaacagcatattagcgatatctggcataacgataaatttaccagcttccgcaaaatgctgaaaga

297 aggccaccgcggcaaaagcgatctgTGCaaaaatTGCaacaacgtgagcgtgcagaccgaagaacagtatgattacgtg ctgaattcgagctccgtcgacaagcttg -3’

Construction of Cys→Ala Variants of BtrN. Cys→Ala variants of BtrN were constructed by site-directed mutagenesis using the Stratagene QuikChange II site directed mutagenesis kit. The following table lists the primers used for constructing the single variants of BtrN. The C16A/C20A/C23A triple variant was prepared as previously described for the C35A/C39A/C42A triple variant for AtsB (17).

Table 7-2. Primers for constructing BtrN variants Primer name Sequence Forward 5’-CGAACCGCTGCTGgccAAAAACCTGGACCTG-3’ C69A Reverse 5’-CAGGTCCAGGTTTTTggcCAGCAGCGGTTCG-3’ Forward 5’-GCGAACATGAGCgccATGGTGCCGAGC-3’ C169A Reverse 5’-GCTCGGCACCATggcGCTCATGTTCGC-3’ Forward 5’-GCAATGTGCTGCCGgccTTTGAAGATTTCAACC-3’ C187A Reverse 5’-GGTTGAAATCTTCAAAggcCGGCAGCACATTGC-3’ Forward 5’-GCAAAAGCGATCTGgccAAAAATtgcAACAACG-3’ C232A Reverse 5’-CGTTGTTgcaATTTTTggcCAGATCGCTTTTGC-3’ Forward 5’-CGATCTGtgcAAAAATgccAACAACGTGAGCG-3’ C235A Reverse 5’-CGCTCACGTTGTTggcATTTTTgcaCAGATCG-3’ 5’-GGATAAACTGTTCAGCATGATCGAAGTGGAAGTGAA Forward CAGCCAGgccAACCGTACCgccTGGTATgccCCG-3’ C16A/C20A/C23A 5’-CCATTTCGCCGGTTTCTTTACGTTTGCTCACGCTG Reverse TTCGGggcATACCAggcGGTACGGTTggcCTGGC-3’

Expression and Purification of the btrN Gene. The btrN gene, or constructs encoding variants of the protein, was transformed into E. coli BL21(DE3)/pDB1282 by standard methods. The protein was expressed and purified as previously described (17).

Reconstitution of the Fe/S clusters were carried out as previously described (17, 20).

Fe/S sulfur analysis was performed as previously described (21, 22).

Amino Acid Analysis of BtrN. Amino acid analysis of BtrN was carried out at the

Molecular Structure Facility at the University of California (Davis, CA). The protein was

298 exchanged by gel filtration (NICK pre-poured column) into 50 mM HEPES buffer (pH

7.5) containing 100 mM NaCl. The eluate was divided into 150 μL fractions, which were lyophilized to dryness using a Savant SpeedVac concentrator (Thermo Scientific,

Waltham, MA). One fraction was used to determine the protein concentration by the procedure of Bradford (23) before lyophilization. The remaining fractions were shipped for amino acid analysis, which was performed in triplicate.

7.3 Results and Discussion

BtrN containing a C-terminal hexahistidine tag was purified by immobilized metal affinity chromatography under anaerobic conditions (as described in Supporting

Information) to ≥95% homogeneity as judged by SDS-PAGE. It displays a brown color, and its UV-vis spectrum is consistent with that of an iron–sulfur (Fe/S) protein (Figure 7-

2) and similar to that previously published (3). Analysis of the protein for Fe and S2- reveals that it contains 4.8 ± 0.4 equiv of the former and 6.1 ± 0.3 equiv of the latter per polypeptide (Table S3). The 4.2-K/53-mT Mössbauer spectrum of as-isolated (AI), hexahistidine tagged, wild-type (wt) BtrN is shown in Figure 7-3A as vertical bars. 98% of its intensity can be simulated to a quadrupole doublet with parameters (isomer shift (δ) of 0.44 mm/s; and quadrupole splitting parameter (ΔEQ) of 1.13 mm/s) typical of [4Fe-

4S]2+ clusters. In addition, the electron paramagnetic resonance (EPR) spectrum of this sample did not reveal any signals. These observations, in combination with Fe and protein analyses (4.8 Fe per BtrN), demonstrate that AI wt BtrN harbors 1.2 [4Fe-4S]2+ clusters. When AI wt BtrN is further reconstituted with iron and sulfide (RCN wt BtrN),

299 it is found to contain 7.7 ± 0.1 equiv of the former and 8.8 ± 0.8 equiv of the latter per polypeptide (Table S3).

The corresponding Mössbauer spectrum (Figure 7-3B, vertical bars) reveals a large fraction of total Fe (87%) that is identical to AI wt BtrN (solid line in Figure 7-3B).

Given that RCN wt BtrN contains 7.7 Fe per polypeptide, analysis by Mössbauer spectroscopy indicates the presence of 1.7 [4Fe-4S]2+ clusters per polypeptide. Thus, reconstitution increases the amount of [4Fe-4S]2+ clusters significantly (1.4 fold), suggesting that BtrN harbors two [4Fe-4S] clusters.

300

Figure 7-3. 4.2-K/53-mT Mössbauer spectra of wt BtrN (A and B) and

To quantify cluster content more rigorously, a triple variant, in which the cysteinyl residues of the canonical radical SAM motif (C16, C20, and C23) are replaced with alanyl residues, was constructed and analyzed. The AI triple variant contains 3.7 ± 0.1 Fe and 4.6 ± 0.3 S2- per polypeptide, slightly less than that observed for the wt protein

(Table 7-3). The 4.2-K/53-mT Mössbauer spectrum of the AI triple variant (Figure 7-

3C) is similar, albeit not identical to that of AI wt BtrN, because it contains pronounced shoulders, which are indicated by the arrows. The spectrum can be analyzed with two quadrupole doublets [δ1 = 0.44 mm/s, ΔEQ,1 = 1.24 mm/s (70% of total intensity) and δ2 =

0.44 mm/s, ΔEQ,2 = 0.78 mm/s (27% of total intensity)]. In conjunction with the finding of 3.7 Fe per polypeptide, we conclude that the AI BtrN C16A/C20A/C23A triple variant harbors 0.9 [4Fe-4S]2+ clusters. Reconstitution of this protein with additional iron and sulfide results in uptake of additional iron (6.2 Fe per polypeptide), but the associated

4.2-K/53-mT Mössbauer spectrum (Figure 7-3D, vertical bars) demonstrates that only

301 ~57% of the total intensity is attributable to [4Fe-4S]2+ clusters (solid line), which corresponds to a stoichiometry of 0.9 [4Fe-4S]2+ per triple variant. The remainder of the spectrum is attributed to adventitiously bound iron, because an identical EPR sample did not reveal the presence of Fe/S clusters with S = 1/2 ground states. Collectively, these results indicate that BtrN harbors two [4Fe-4S] clusters.

Table 7-3. Activity, Fe, and S of Wt and Variants of BtrN Iron per sulfide per polypeptide polypeptide SR Activitya Activityb AI RC AI RC BtrN Wt 3.7 0.22 4.8±0.4 7.7±0.1 6.1±0.3 8.8±0.8 BtrN C16/20/23A ND ND 3.7±0.1 6.2±0.1 4.6±0.3 6.0±0.2 BtrN C69A 3.7 0.29 4.7±0.2 11.0±0.2 3.2±0.1 7.9±0.6 BtrN C235A 0.25 0.02 1.8±0.04 17.2±0.2 1.7±0.1 8.2±0.6 a -1 : defined as Vmax/[ET] min for production of 5’-dAdenosine from SAM b -1 : defined as Vmax/[ET] min for production of 2-DOS

The model proposed for radical SAM dehydrogenation indicates that at least two [4Fe–

4S] clusters are coordinated by only three cysteinyl ligands, which would present open coordination sites for binding of SAM to one cluster and the appropriate substrate (DOIA in the case of BtrN) to the other. In order to investigate which cysteines in BtrN act as ligands to this second Fe/S cluster, single CA variants of the five cysteinyl residues outside of the CxxxCxxC motif were constructed with the intention of assessing the ability of each to catalyze the reaction. Three of the variants (C169A, C187A, and

C232A) were produced as insoluble aggregates, and were not analyzed further. The remaining two variants (C69A and C235A) were soluble, and displayed UV-vis spectra

(Figures 7-5, 7-6) and quantities of Fe and S2- (Table 7-3) that are consistent with the presence of Fe/S clusters.

302

The activities of wt BtrN, the C16A/C20A/C23A triple variant, and both the C69A and

C235A single variants were determined by high-performance liquid chromatography

(HPLC) with detection by mass spectrometry using an Agilent Technologies (Santa

Clara, CA) QQQ spectrometer as described in Materials and Methods. Assays included the pyridoxal 5’-phosphate-containing enzyme BtrR (19), which converts the product of the BtrN reaction (amino-DOI) into 2-deoxystreptamine (DOS), a commercially available compound. The time-dependent quantification of 5’-dA and DOS was accomplished using known concentrations of these two metabolites to generate a standard curve.

303

As shown in Table S3, wt BtrN displays significant SAM reductase activity (3.7 min-

1)—which is the production of 5’-dA without concomitant substrate turnover—while the formation of DOS occurred with a rate constant of 0.22 min-1, similar to that observed for

AtsB (17), as well as that reported by Yokoyama et al (0.2 min-1) for assays in which the production of DOS was monitored (2). As expected, neither SAM reductase activity nor

DOIA reduction was detected for the C12A/C20A/C23A triple variant. The BtrN C69A variant behaved similarly to wt BtrN, while the C235A variant displayed SAM reductase

304 and substrate turnover activities that were 7% and 9%, respectively, of that displayed by wt hexahistidine-tagged BtrN.

CxxxCxxC RS motif and ligate the Fe/S cluster involved in the reductive cleavage of

SAM. The finding that three of the remaining cysteinyl residues (C169, C187, and C232) afford insoluble proteins when they are changed to alanyl residues, while CA substitutions at two of the remaining cysteinyl residues afford proteins that are active, suggests strongly that the second Fe/S cluster that we detect is ligated by C169, C187,

305 and C232. In addition, our finding that this cluster is ligated by only three protein-based cysteinyl residues is consistent with a model in which an open coordination site on the second cluster allows the substrate C3 hydroxyl group to coordinate to the cluster

(Figure 7-7).

Figure 7-7. Proposed mechanism for BtrN. Binding of (co)substrate and dissociation of (co)products are labeled with ① and ②, respectively. Electron flow is indicated by red arrows.

306 This cluster coordination should facilitate loss of the proton of the C3 hydroxyl group upon coordination to the Fe/S cluster, hydrogen-atom abstraction at C3 by the 5’-dA•, and inner-sphere electron transfer from the ensuing substrate radical to the [4Fe-4S]2+ cluster to afford the product, amino-DOI. It is conceivable that the electron transferred upon generation of amino-DOI to the second Fe/S cluster is subsequently transferred to the RS-[4Fe-4S] cluster, which would allow the latter to reductively cleave SAM and generate another 5’-dA• for the next reaction cycle, as was originally proposed by

Yokoyama et al (2).

7.4 Conclusions

There appears to be one significant difference between the BtrN described herein and that previously characterized by Yokoyama et al (2). BtrN described herein displays significant SAM reductase activity (~3.7 min-1), while that previously described was reported to couple SAM cleavage with DOS formation efficiently. The authors of that report then used formation of 5’-dA as the standard method for further kinetic

-1 characterization of BtrN, reporting kcat values of 1.2 to 2.3 min , somewhat less than what we observe for SAM reductase activity, but 5 to 10-fold greater than what they (0.2 min-1) and we (0.22 min-1) report for DOS formation. Despite this small discrepancy, it is clear from the spectroscopic and biochemical studies reported herein that BtrN, a radical

SAM dehydrogenase, contains two [4Fe–4S] clusters rather than one, and that the model originally proposed for the RS enzyme AtsB may in fact be a common mechanistic feature of RS dehydrogenases.

307 7.5 Acknowledgments

We are indebted to Dr. Jonathan B. Spencer and Dr. Fanglu Huang for their generous gift of 2-deoxy-scyllo-inosamine and the plasmids for production of BtrR and BtrC.

308 7.6 References

1. Abbreviations. AI, as-isolated; 5'-dA, 5'-deoxyadenosine, 5'-dA•, 5'-

deoxyadenosyl 5'-radical; amino-DOI, 2-deoxy-scyllo-inosose; DOIA, 2-deoxy-

scyllo-inosamine; DOS, 2-deoxystreptamine; RCN, reconstituted; RS, radical

SAM; SAM, S-adenosyl-L-methionine; wt, wild-type.

2. Yokoyama, K., Numakura, M., Kudo, F., Ohmori, D., and Eguchi, T. (2007)

Characterization and mechanistic study of a radical SAM dehydrogenase in the

biosynthesis of butirosin, J. Am. Chem. Soc. 129, 15147-15155.

3. Yokoyama, K., Ohmori, D., Kudo, F., and Eguchi, T. (2008) Mechanistic study

on the reaction of a radical SAM dehydrogenase BtrN by electron paramagnetic

resonance spectroscopy, Biochemistry 47, 8950-8960.

4. Booker, S. J. (2009) Anaerobic functionalization of unactivated C–H bonds, Curr.

Opin. Chem. Biol. 13, 58–73.

5. Frey, P. A., and Booker, S. J. (2001) Radical mechanisms of S-

adenosylmethionine-dependent enzymes, Adv. Protein Chem. 58, 1-45.

6. Sofia, H. J., Chen, G., Hetzler, B. G., Reyes-Spindola, J. F., and Miller, N. E.

(2001) Radical SAM, a novel protein superfamily linking unresolved steps in

familiar biosynthetic pathways with radical mechanisms: functional

characterization using new analysis and information visualization methods,

Nucleic Acids. Res. 29, 1097-1106.

7. Walsby, C. J., Ortillo, D., Yang, J., Nnyepi, M. R., Broderick, W. E., Hoffman, B.

M., and Broderick, J. B. (2005) Spectroscopic approaches to elucidating novel

309 iron-sulfur chemistry in the "radical-SAM" protein superfamily, Inorg. Chem. 44,

727-741.

8. Chatterjee, A., Li, Y., Zhang, Y., Grove, T. L., Lee, M., Krebs, C., Booker, S. J.,

Begley, T. P., and Ealick, S. E. (2008) Reconstitution of ThiC in thiamine

pyrimidine biosynthesis expands the radical SAM superfamily, Nat. Chem. Biol.

4, 758–765.

9. Martinez–Gomez, N. C., and Downs, D. M. (2008) ThiC is an [Fe–S] cluster

protein that requires AdoMet to generate the 4-amino-5-hydroxymethyl-2-

methylpyrimidine moiety in thiamin synthesis, Biochemistry 47, 9054–9056.

10. Chen, D., Walsby, C., Hoffman, B. M., and Frey, P. A. (2003) Coordination and

mechanism of reversible cleavage of S-adenosylmethionine by the [4Fe-4S]

center in lysine 2,3-aminomutase, J. Am. Chem. Soc. 125, 11788-11789.

11. Layer, G., Heinz, D. W., Jahn, D., and Schubert, W. D. (2004) Structure and

function of radical SAM enzymes, Curr. Opin. Chem. Biol. 8, 468-476.

12. Walsby, C. J., Ortillo, D., E., B. W., Broderick, J. B., and Hoffman, B. M. (2002)

An anchoring role for FeS clusters: chelation of the amino acid moiety of S-

adenosylmethionine to the unique iron site of the [4Fe–4S] cluster of pyruvate

formate–lyase activating enzyme, J. Am. Chem. Soc. 124, 11270-11271.

13. Cheek, J., and Broderick, J. B. (2001) Adenosylmethionine-dependent iron-sulfur

enzymes: versatile clusters in a radical new role, J. Biol. Inorg. Chem. 6, 209-226.

14. Frey, P. A., Hegeman, A. D., and Reed, G. H. (2006) Free radical mechanisms in

enzymology, Chem. Rev. 106, 3302-3316.

310 15. Benjdia, A., Leprince, J., Guillot, A., Vaudry, H., Rabot, S., and Berteau, O.

(2007) Anaerobic sulfatase-maturating enzymes: radical SAM enzymes able to

catalyze in vitro sulfatase post-translational modification, J. Am. Chem. Soc. 129,

3462.

16. Fang, Q., Peng, J., and Dierks, T. (2004) Post-translational formylglycine

modification of bacterial sulfatases by the radical S-adenosylmethionine protein

AtsB, J. Biol. Chem. 279, 14570-14578.

17. Grove, T. L., Lee, K. H., St Clair, J., Krebs, C., and Booker, S. J. (2008) In vitro

characterization of AtsB, a radical SAM formylglycine-generating enzyme that

contains three [4Fe-4S] clusters, Biochemistry 47, 7523-7538.

18. Szameit, C., Miech, C., Balleininger, M., Schmidt, B., von Figura, K., and Dierks,

T. (1999) The iron sulfur protein AtsB is required for posttranslational formation

of formylglycine in the Klebsiella sulfatase, J. Biol. Chem. 274, 15375-15381.

19. Schmidt, B., Selmer, T., Ingendoh, A., and von Figura, K. (1995) A novel amino

acid modification in sulfatases that is defective in multiple sulfatase deficiency,

Cell 82, 271–278.

20. Cicchillo, R. M., Lee, K. H., Baleanu-Gogonea, C., Nesbitt, N. M., Krebs, C., and

Booker, S. J. (2004) Escherichia coli lipoyl synthase binds two distinct [4Fe-4S]

clusters per polypeptide, Biochemistry 43, 11770-11781.

21. Beinert, H. (1978) Micro methods for the quantitative determination of iron and

copper in biological material, Methods Enzymol. 54, 435-445.

311 22. Beinert, H. (1983) Semi-micro methods for analysis of labile sulfide and of labile

sulfide plus sulfane sulfur in unusually stable iron-sulfur proteins, Anal. Biochem.

131, 373-378.

23. Bradford, 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.

312 Chapter 8

Further Characterization of Cys-Type and Ser-Type Anaerobic Sulfatase Maturating Enzymes Suggests a Commonality in Mechanism of Catalysis†

This chapter was reproduced from “Grove, T.L., Ahlum, J.H., Qin, R.M., Lanz, N.D., Radle, M.I., Krebs, C., Booker, S.J. Further Characterization of Cys-Type and Ser-Type Anaerobic Sulfatase Maturating Enzymes Suggests a Commonality in Mechanism of Catalysis. Biochemistry. 2013 Mar 11. [Epub ahead of print]

313 8.1 Abstract

The anaerobic sulfatase maturating enzyme from Clostridium perfringens (anSMEcpe) catalyzes the two-electron oxidation of a cysteinyl residue on a cognate protein to a formyglycyl residue (FGly) using a mechanism that involves organic radicals. The FGly residue plays a unique role as a cofactor in a class of enzymes termed arylsulfatases, which catalyze the hydrolysis of various organosulfate monoesters. anSMEcpe has been shown to be a member of the radical S-adenosylmethionine (SAM) family of enzymes,

[4Fe–4S] cluster–requiring proteins that use a 5’-deoxyadenosyl 5’-radical (5’-dA•) generated from a reductive cleavage of SAM to initiate radical-based catalysis. Herein, we show that anSMEcpe contains in addition to the [4Fe–4S] cluster harbored by all radical SAM (RS) enzymes, two additional [4Fe–4S] clusters, similar to the radical SAM protein AtsB, which catalyzes the two-electron oxidation of a seryl residue to a FGly residue. We show by size-exclusion chromatography that both AtsB and anSMEcpe are monomeric proteins, and site-directed mutagenesis studies on AtsB reveal that individual

CysAla substitutions at seven conserved positions result in insoluble protein, consistent with those residues acting as ligands to the two additional [4Fe–4S] clusters. Ala substitutions at an additional conserved Cys residue (C291 in AtsB; C276 in anSMEcpe) afford proteins that display intermediate behavior. These proteins exhibit reduced solubility and drastically reduced activity, behavior that is conspicuously similar to that of a critical Cys residue in BtrN, another radical SAM dehydrogenase [Grove, T. L., et al

(2010) Biochemistry, 49, 3783–3785]. We also show that wild-type anSMEcpe acts on peptides containing other oxidizeable amino acids at the target position. Moreover, we

314 show that the enzyme will convert threonyl peptides to the corresponding ketone product, and also allo-threonyl peptides, but with a significantly reduced efficiency, suggesting that the proS hydrogen atom of the normal cysteinyl substrate is stereoselectively removed during turnover. Lastly, we show that the electron generated during catalysis by

AtsB and anSMEcpe can utilized for multiple turnovers, albeit through a reduced flavodoxin-mediated pathway.

315

8.2 Introduction

Radical SAM (RS)1 dehydrogenases are a burgeoning class of S-adenosylmethionine

(SAM)-requiring enzymes that catalyze the two-electron oxidation of organic substrates via intermediates containing unpaired electrons (1-7). These enzymes, as do all RS proteins, contain a [4Fe–4S] cluster cofactor that is absolutely required for turnover (1-3,

8). The [4Fe–4S] cluster is coordinated by the -amino and -carboxylate groups of

SAM, and in its reduced state, provides the essential electron for the reductive cleavage of SAM into methionine and a 5’-deoxyadenosyl 5’-radical (5’-dA•) (9, 10). The 5’-dA•, in turn, initiates turnover by abstracting a hydrogen atom (H•) from a strategic position, often cleaving unactivated or weakly activated C–H bonds (11-15). Three RS dehydrogenases spanning two distinct classes have been the subject of detailed in vitro mechanistic investigation. One, BtrN, catalyzes the third step in the biosynthetic pathway of the 2-deoxystreptamine (DOS)-containing aminoglycoside antibiotic, butirosin B, which entails the two-electron oxidation of the C3 secondary alcohol of 2-deoxy-scyllo- inosamine (DOIA) to a ketone, affording amino-2-deoxy-scyllo-inosose (amino-DOI)

(Figure 8-1A) (3).

316

Figure 8-1. Reactions catalyzed by BtrN (A) and anSMEs (B).

The remaining two, AtsB and anSMEcpe, are anaerobic sulfatase modifying enzymes

(anSMEs), which catalyze the two-electron oxidation of a target seryl or cysteinyl residue on their cognate arylsulfatases to a formylglycyl (FGly) residue (Figure 8-1B) (2, 4, 16,

17). The FGly residue serves as an obligate cofactor in the cleavage of a variety of sulfate monoesters by this class of enzymes (18-21). Crystallographic and mechanistic studies have shown that the FGly residue exists as a hydrate, wherein one oxygen acts as a nucleophile in the attack on the sulfur atom of the sulfate monoester. Release of sulfate is concomitant with collapse of the sulfated geminal diol to the aldehyde (22-24). This mechanism for generating the FGly cofactor is distinguished from another non-RS mechanism found in higher eukaryotes and some bacteria, which requires reducing equivalents and the intervention of dioxygen (25-28); however, in both cases the FGly cofactor is typically found in the conserved sequence motif C/S-X-P/A-S/X-R-X-X-X-

L/X-T/X-G/X-R/X, with the C/S highlighted in bold type as the site of modification (16,

29).

Characterization of AtsB, BtrN, and anSMEcpe verified their membership in the RS superfamily of enzymes (1-3, 17, 30). In addition to their canonical CxxxCxxC motifs,

317 which bear the Cys ligands that coordinate the iron–sulfur (Fe/S) cluster involved intimately in the cleavage of SAM, they were all shown to contain [4Fe–4S] clusters and to cleave SAM reductively to 5’-deoxyadenosine (5’-dA) and methionine during catalysis. However, the number of Fe/S clusters on these enzymes has been a subject of disagreement. In the initial characterization of BtrN, Yokoyama, et al. used quantitative analyses for iron and sulfide after reconstitution of the Fe/S cluster to demonstrate the presence of only one [4Fe–4S] cluster (presumed to be the RS Fe/S cluster) per polypeptide (8). By contrast, Grove, et al. used a combination of analytical (quantitative

Fe, S2-, and protein analyses) and spectroscopic (UV-vis and Mössbauer) methods to demonstrate that BtrN harbors two [4Fe–4S] clusters (31). Using the same experimental methodology, it was also demonstrated that AtsB harbors three [4Fe–4S] clusters (2). It was suggested that one of the remaining two non-RS [4Fe–4S] clusters might coordinate to the substrate to facilitate the two-electron oxidation. For the related enzyme anSMEcpe, Benjdia, et al. reported that their reconstituted protein contained 5.7 ± 0.5 equiv of iron (sulfide not quantified). This stoichiometry in concert with characterization of the protein by UV–vis, resonance Raman, and electron paramagnetic resonance (EPR) spectroscopy led the authors to suggest that the protein most likely contained one [4Fe–

4S] cluster, although they left open the possibility that it might contain two, and suggested that further studies would be required to determine this conclusively (1).

The Cys-type anSME from Clostridium perfringens (anSMEcpe) shares 48% sequence similarity with the Ser-type anSME from Klebsiella pneumoniae (AtsB). It is slightly smaller in size (370 aa vs 395 aa), but contains 18 Cys residues per polypeptide as opposed to 13 Cys residues on AtsB. Eleven Cys residues are common between the two

318 proteins and are conserved throughout anSMEs. In light of the differences in cluster content observed between these two proteins using different strategies for protein overproduction and spectroscopic methods for Fe/S cluster characterization, we set out to characterize anSMEcpe in a quantitative manner with respect to cluster stoichiometry as well as turnover with various peptide substrates. Herein, we show that anSMEcpe harbors three [4Fe–4S]2+ clusters in its fully active form, as was found for AtsB. Thus, these results further corroborate our proposal that all natural RS-dehydrogenases require at least two [4Fe–4S] clusters for turnover (31). Moreover, we show via site-directed mutagenesis that seven Cys residues in addition to the three that coordinate the RS cluster are absolutely required, and their substitution with Ala residues affords completely insoluble proteins. Similar to findings by Grove, et al. on BtrN, one Cys residue, when substituted with Ala, affords a soluble protein that can be characterized; however, its activity is greatly diminished, supporting a key role for this residue in catalysis. Last, we show that anSMEcpe is capable of converting Cys, Ser, and SeCys residues to FGly residues, as well as threonyl residues to the corresponding keto product, while the reaction of the corresponding allo-threonyl-containing substrate does not lead to substantial formation of the keto product. Collectively these results suggest that the key step in catalysis by anSMEs is abstraction of the 3-proS H• from the substrate by the 5’- dA• intermediate. Also discussed is the fate of the second electron removed from the target Ser or Cys residue during the two-electron oxidation.

319 8.3 Materials and Methods

Materials. All DNA-modifying enzymes and reagents were purchased from New England

Biolabs (Ipswich, MA), as were Vent polymerase and its associated 10 buffer. Oligonucleotide primers were obtained from Integrated DNA Technologies (Coralville, IA). C. perfringens (strain

NCTC 8237) genomic DNA (ATCC 13124D-5) was purchased from American Type Culture

Collection (Manassas, VA). 5’-Deoxyadenosine (5’-dA), sodium sulfide (nonahydrate), sodium dithionite (DT), -mercaptoethanol, L-tryptophan, L-(+)-arabinose, and ferric chloride were purchased from Sigma–Aldrich Chemicals (St. Louis, MO). N-(2-hydroxyethyl)piperazine-N'-(2- ethanesulfonic acid) (HEPES) was purchased from Fisher Scientific (Pittsburgh, PA), and imidazole was purchased from J. T. Baker Chemical Co (Phillipsburg, NJ). Potassium chloride and glycerol were purchased from EMD Chemicals (Gibbstown, NJ), while dithiothreitol (DTT) was purchased from Gold Biotechnology (St. Louis, MO). Coomassie blue dye-binding reagent for protein concentration determination was purchased from Pierce (Rockford, IL), as was the bovine serum albumin (BSA) standard (2 mg/mL). Talon metal affinity resin was purchased from

Clontech (Mountain View, CA). Sephadex G-25 resin and NICK and NAP prepoured gel filtration columns were purchased from GE Biosciences (Piscataway, NJ). Fmoc-Thr(tBu)-OH

(99%), Fmoc-allo-Thr(tBu)-OH (99%), and Fmoc-Se-4-methoxybenzyl selenocysteine (99%) were purchased from Chem-Impex International. All other chemicals were of the highest grade available.

S-Adenosyl-L-methionine (SAM) was synthesized enzymatically and purified as described previously (32). Flavodoxin (Flv) and flavodoxin reductase (Flx) were purified from E. coli

BL21(DE3) containing plasmids pTYB1-Flv and pTYB1-Flx as described previously (33, 34).

Fmoc-formylglycine (dimethylacetal) was kindly provided by Professor Carolyn Bertozzi and Dr.

Jason Rush (UC Berkeley).

320 DNA sequencing was carried out at the Pennsylvania State University Nucleic Acid

Facility. Analyses for iron and sulfide were performed by the procedures of Beinert (35-

37). SPEX CertiPrep (Metuchen, NJ) Cläritas PPT single element Fe (1000 mg/L in 2%

HNO3) was used to prepare iron standards for quantitative iron analysis. Protein concentration was measured by the procedure of Bradford using bovine serum albumin

(Fraction V) as a standard (38).

Spectroscopic Methods. UV-visible spectra were recorded on a Cary 50 spectrometer (Varian,

Walnut Creek, CA) using the associated WinUV software package for operating the instrument and manipulating the data. Mössbauer spectra were recorded on a spectrometer from WEB

Research (Edina, MN), which was equipped with an SVT-400 cryostat from Janis Research Co

(Wilmington, MA). Spectra were collected in constant acceleration mode in transmission geometry. Isomer shifts are quoted relative to the centroid of α-Fe at room temperature. Spectra were analyzed with the program WMOSS from WEB Research. 57Fe (97-98%) metal for

Mössbauer spectroscopy was purchased from Isoflex USA (San Francisco, CA). For preparation

57 of a FeSO4 solution, the solid was dissolved with heating in an anaerobic solution of 2 N H2SO4

57 57 (1.5 mol of H2SO4 per mole of Fe). The Fe solution was used as is for supplementation in E. coli culture media, or was titrated to pH 6.5 with an anaerobic solution of saturated sodium bicarbonate for in vitro reconstitution. X-band (~9.5 GHz) electron paramagnetic resonance

(EPR) spectroscopy was conducted on a Bruker ESP 300 spectrometer equipped with an Oxford

Instruments Model ESP 900 continuous flow cryostat. EPR parameters for various samples are provided in the appropriate figure legends.

Cloning of the cpe0635 gene from Clostridium perfringens. The gene corresponding to anSMEcpe (cpe0635) was amplified from C. perfringens genomic DNA (ATCC# 13124D-5) using the polymerase chain reaction (PCR) in combination with a forward primer containing an

321 NdeI restriction site (underlined) (5’-CGC-GCC-CGC-ATA-TGC-CAC-CAT-TAA-GTT-TGC-

TTA-TTA-AGC-3’) and a reverse primer containing a BamHI restriction site (underlined) (5’-

CCG-GAT-CCG-ATT-TAA-TAT-TGT-TGG-CAA-CAT-TTA-TTA-ACC-3’). The reverse primer was designed to remove the stop codon from the C-terminus of the gene, which affords addition of a 22-amino acid C-terminal extension containing a hexahistidine tag. The PCR was conducted using a Stratagene (La Jolla, CA) Robocycler thermocyler as described previously

(39), and the amplified gene was isolated and cloned into expression vector pET-26b by standard procedures. Several constructs were analyzed by DNA sequencing, which revealed that they all had identical sequences. The chosen construct was designated pCpe0635Wt.

Construction of the C15A/C19A/C22A anSMEcpe triple variant. The C15A/C19A/C22A anSMEcpe triple variant was constructed using the Stratagene QuikChange II site-directed mutagenesis kit as described previously (2). The forward primer used was 5’-CCA-TTA-AGT-

TTG-CTT-ATT-AAG-CCA-GCT-TCT-AGT-GGA-GCT-AAT-TTA-AAA-GCC-ACT-TAT-

GCT-3’, while the reverse primer used was 5’-CTT-AAC-ATT-TCT-ATT-ATC-ACT-TAA-

AGA-ATG-ATA-AAA-AGC-ATA-AGT-GGC-TTT-TAA-ATT-AGC -3’. The underlined letters represent the altered codons.

Expression of the Cpe0635 gene and purification of anSMEcpe. Plasmid pCpe0635Wt, or constructs encoding variants of anSMEcpe, was transformed into E. coli BL21(DE3)/pDB1282 by standard methods, and the encoded Cpe0635 gene expressed as described previously for overproduction of AtsB (2). The protein was also purified as previously described. Reconstitution of the Fe/S clusters of anSMEcpe was conducted as described previously (2, 33).

Construction of CysAla variants of AtsB and anSMEcpe. Single CysAla substitutions in anSMEcpe (Cys276) and AtsB (Cys residues 127, 245, 270, 276, 291, 331, 334, 340, 344, and

357) were engineered using the Stratagene QuikChange II site-directed mutagenesis kit with

322 primers listed in Table 8-1 as described above. Expression of the variant constructs and purification of the encoded proteins were done exactly as described previously (2).

Table 8-1. List of primers for site-directed mutagenesis of AtsB Primer Sequence

AtsB C127A Forward 5’-gctgatcaacgacgcatggGCCcgactgttccgcg-3’

AtsB C127A Reverse 5’-cgcggaacagtcgGGCccatgcgtcgttgatcagc-3’

AtsB C245A Forward 5’-ggcggaagcgcGCCgatagagggcg-3’

AtsB C245A Reverse 5’-cgccctctatcGGCgcgcttccgcc-3’

AtsB C270A Forward 5’-ccagcggcagcGCCgtgcacagcg-3’

AtsB C270A Reverse 5’-cgctgtgcacGGCgctgccgctgg-3’

AtsB C276A Forward 5’-cgtgcacagcgcccgcGCCggcagcaacctgg-3’

AtsB C276S Reverse 5’-ccaggttgctgccGGCgcgggcgctgtgcacg-3’

AtsB C276S Forward 5’-cgtgcacagcgcccgcTCCggcagcaacctgg-3’

AtsB C276A Reverse 5’-ccaggttgctgccGGAgcgggcgctgtgcacg-3’

AtsB C291A Forward 5’-ggacagctctacgccGCCgaccacctgatcaacg-3’

AtsB C291A Reverse 5’-cgttgatcaggtggtcGGCggcgtagagctgtcc-3’

AtsB C331A Forward 5’-gcgccgcgaaGCCcagacttgctcgg-3’

AtsB C331A Reverse 5’-ccgagcaagtctgGGCttcgcggcgc-3’

AtsB C334A Forward 5’-ccgcgaatgccagactGCCtcggtaaaaatgg-3’

AtsB C334A Reverse 5’-ccatttttaccgaGGCagtctggcattcgcgg-3’

AtsB C340A Forward 5’-cggtaaaaatggtcGCCcagggcggctgccc-3’

AtsB C340A Reverse 5’-gggcagccgccctgGGCgaccatttttaccg-3’

AtsB C344A Forward 5’-ggtctgccagggcggcGCCccggcgcatctcaacgccg-3’

AtsB C344A Reverse 5’-cggcgttgagatgcgccggGGCgccgccctggcagacc-3’

323 AtsB C357A Forward 5’-ggcaacaaccgcctcGCCggaggctactaccgc-3’

AtsB C357A Reverse 5’-gcggtagtagcctccGGCgaggcggttgttgcc-3’ anSMEcpe C276 Forward 5’-ggagtgtttatcctGCTgatttttatgttttagataaatgg-3’ anSMEcpe C276 Reverse 5’-ccatttatctaaaacataaaaatcAGCaggataaacactcc-3’

Amino acid analysis of anSMEcpe. Amino acid analysis of anSMEcpe was carried out at the

Molecular Structure Facility at the University of California–Davis (Davis, CA). The protein was exchanged by gel filtration (NICK pre-poured column) into 50 mM HEPES buffer (pH 7.5) containing 100 mM NaCl. The eluate was divided into 50 L fractions, which were lyophilized to dryness using a Savant SpeedVac concentrator (Thermo Scientific; Waltham, MA). One fraction was used to determine the protein concentration by the procedure of Bradford before lyophilization. The remaining fractions were shipped for amino acid analysis, which was performed in quadruplicate. It was found that the concentration determined by the procedure of

Bradford is an overestimate and therefore must be multiplied by 0.69 to achieve the true anSMEcpe concentration.

Synthesis and purification of substrate peptides. The following peptide substrates, each containing an N-terminal acetyl (Ac) group (except the IS peptide Kp9Ser), were synthesized at the Penn State Core Research Facilities utilizing standard Fmoc chemistry (bold and underlined type indicates target location of modification): Cp18Cys, Ac-NH2-YTAVPSCIPSRASILTGM-

COOH; Kp18Cys, Ac-NH2-YYTSPMCAPARSMLLTGN-COOH; Kp18Ser, Ac-NH2-

YYTSPMSAPARSMLLTGN-COOH; Kp18SeCys, Ac-NH2-YYTSPMSeCAPARSMLLTGN-

COOH; Kp18Thr, Ac-NH2-YYTSPMTAPARSMLLTGN-COOH; Kp18alloThr, Ac-NH2-

YYTSPMaTAPARSMLLTGN-COOH; Kp18FGly, Ac-NH2-YYTSPMfGAPARSMLLTGN-

COOH; and Kp9Ser, NH2-PMSAPARSM. The first two letters of each peptide name correspond to the organism (Clostridium perfringens or Klebsiella pneumoniae) from which the peptide

324 sequence is derived; the number corresponds to the length; and the amino acid abbreviation corresponds to the amino acid in the target position. Fmoc-S-4-methoxybenzyl selenocysteine, used in the synthesis of Kp18SeCys, was purchased from Chem-Impex International (Wood

Dale, IL) and used as received. Subsequent to synthesis, the peptide (0.035 mmol, ~278 mg) was cleaved from the resin in a solution of 2% triisopropylsilane (100 µL), 100 µL water, and 2.5 % thioanisole (125 µL) in neat TFA (5 mL) containing 1.3 equiv 2,2’-dithiobis(5-nitropyridine) (14 mg) at room temperature for 2 h, after which the cleaved resin was removed by filtration. The crude peptides were then precipitated by addition of ice-cold diethyl ether (1:10 dilution). The peptide mixture was redissolved in a 50% acetonitrile solution (v/v in water) and the appropriate full-length peptide was purified by reverse-phase HPLC (Agilent 1100 System; Santa Clara, CA) using an Agilent Zorbax SB-C18 (9.4  250 mm) semi-preparative column. A three-solvent system was employed in the separation: 0.1% trifluoroacetic acid (TFA) in water (Solvent A);

0.1% TFA in acetonitrile (Solvent B); and methanol (Solvent C). The column was equilibrated in a solution consisting of 85% Solvent A, 10% Solvent B, and 5% Solvent C. Upon injection of the crude peptide mixture, a gradient of 10-50% Solvent B was applied over 29 min, after which

Solvent B was increased to 80% over 1 min. Finally, Solvent B was returned to 10% (initial conditions) over 1 min and the column was allowed to re-equilibrate for 10 min. Throughout the run Solvent C was maintained constant, the flow rate was maintained at 4 mL min-1, and detection of the peptide was monitored by UV-vis spectroscopy at 275 nm. The peak corresponding to the deprotected full-length peptide was collected and lyophilized to dryness to obtain the final product as a white solid. The peptide was then re-dissolved in water and its concentration was determined using a molar absorptivity at 274 nm of 1405 M-1 cm-1 (one Tyr residue) for Cp18Cys and 2810 M-1 cm-1 (two Tyr residues) for the remaining peptides, except for Kp9Ser. The IS peptide Kp9Ser was purified as described above with monitoring at 220 nm. Its final concentration was determined by dissolving a weighed amount in an appropriate volume of water.

325 The purified peptides were analyzed by LC-MS using an Agilent 6410 Triple Quadrupole (QQQ)

ESI-MS instrument in positive mode with an MS2 scan width of 500 – 2000 m/z to verify their masses.

Activity determination of anSMEcpe. Reactions contained in a total volume of 150 µL: 50 mM

HEPES, pH 7.5, 150 mM KCl, 1 mM SAM, 3 mM DT, 1 mM peptide substrate, and either 4 µM

(DT assays) or 40 µM (Flv/Flx/NADPH assays) WT anSMEcpe. Reaction mixtures lacking DT were incubated for 5 min at 37 °C, and 10 µL aliquots were removed (t=0) and added to 10 µL of a solution containing 100 mM H2SO4, 100 µM Kp9Ser (IS), and 100 µM L-tryptophan (IS) to yield final IS concentrations of 50 µM. Reactions were initiated by the addition of DT and incubated for appropriate times before being quenched as described above. The samples were subjected to centrifugation at 18,000  g in a bench-top microcentrifuge and analyzed by LC-MS using Method 1 or Method 2 as described below. Standard curves were generated with 5’-dA or the appropriate purified peptides. All final concentrations were multiplied by a dilution factor of

2 to determine original concentrations in the assay mixtures. When the Flv/Flx/NADPH reducing system replaced DT, their concentrations were 50 µM, 15 µM, and 2 mM, respectively. When reactions were carried out with Kp18Thr or Kp18alloThr, each peptide was present at a concentration of 500 µM, and the concentrations of AtsB or anSMEcpe were adjusted to 200 µM or 100 µM, respectively. Products were analyzed as described above, as well as by MALDI MS using dinitrophenylhydrazine (DNPH) as a derivatizing agent as previously described (2).

LC-MS Method 1. HPLC with detection by mass spectrometry (LC-MS) was conducted on an

Agilent Technologies (Santa Clara, CA) 1200 system, which was fitted with an autosampler for sample injection and coupled to an Agilent Technologies 6410 QQQ mass spectrometer. The system was operated with the associated MassHunter software package, which was also used for data collection and analysis. Assay mixtures were separated on an Agilent Technologies Zorbax

Rapid Resolution SB-C18 column (2.4 mm  35 mm, 3.5 µm particle size), which was

326 equilibrated in 80% Solvent A (5 mM perfluoroheptanoic acid–6 mM ammonium formate in water, pH 3) and 20% acetonitrile at a flow rate of 0.4 mL min-1. A gradient of 20–30% acetonitrile was applied from 0 to 2 min, and then from 30 to 20% acetonitrile from 2 to 2.5 min to restore the system to initial conditions. The column was allowed to re-equilibrate for 1.5 min under initial conditions before subsequent sample injections. Detection of 5’-dA and tryptophan was performed using electrospray ionization in positive mode (ESI+) with multiple reaction monitoring. Relevant retention times and ions monitored are given in Table 8-2.

Table 8-2. Retention times and monitored m/z values for Detection Method 1 Retention Time Parent Ion* Product Ion 1† Product Ion 2†

5’-dA 4.7 min 252.1 (90) 136 (13) 119 (50) Tryptophan (IS) 6.2 min 188 (130) 146.1 (10) 118 (21)

*Respective fragmentor voltages are in parenthesis.

†Respective collision energies are in parenthesis.

LC-MS Method 2. Data collection and analysis was carried out as in Method 1 with the following modifications: the column was equilibrated in 92% Solvent A (0.1% formate in water, pH 3.0) and 8% acetonitrile at a flow rate of 0.5 mL min-1. A gradient of 8–26% acetonitrile was applied from 0.5 to 2 min, and then from 26–28% acetonitrile from 2 min to 4 min. The column was restored to initial conditions from 4 min to 4.5 min and allowed to equilibrate for another 2 min before subsequent sample injections. Detection of substrates and products (Table S3) was performed using electrospray ionization in positive mode (ESI+) with MRM. Relevant retention times and ions monitored are given in Table 8-3.

327

Table 8-3. Retention times and monitored m/z values for Detection Method 2 Substrate Retention Time Parent Ion* Product Ion 1† Product Ion 2†

Kp9Ser (IS) 1.4 min 474.4 (180) 719.3 (15) 561.3 (11) Kp18FGly 3.9 min 1000.7 (180) 905.9 (12) 404.2 (20) Kp18Ser 4.0 min 1001.7 (180) 906.9 (12) 404.2 (12) Kp18Cys 4.4 min 1009.9 (180) 1727.8 (12) 914.9 (12) Kp18SeCys 4.8 min 1033.1 (180) 1414.6 (0) 291.1 (24) Cp18Cys 4.8 min 955.3 (180) 477.2 (16) 421.2 (8) Kp18Thr 4.1 min 1008.7 (180) 1059.6 (18) 914 (14) Kp18alloThr 3.9 min 1008.7 (180) 1059.6 (18) 914 (14)

*Respective fragmentor voltages are in parenthesis.

†Respective collision energies are in parenthesis.

Molecular sieve chromatography of anSMEcpe and AtsB. Molecular sieve chromatography of anSMEcpe and AtsB was performed with slight modifications of a previously described procedure (40) using an ÄKTA (GE Healthcare, Piscataway, NJ) liquid chromatography system, which was maintained inside a Coy anaerobic chamber. A HiPrep 16/60 Sephacryl S-200 HR column (GE Healthcare) column was equilibrated in a buffer composed of 10 mM HEPES pH

7.5, 500 mM KCl, 5 mM DTT, and 10% glycerol at a flow rate of 0.3 ml min-1. WT RCN anSMEcpe (100 µL of a 737 µM solution) or WT RCN AtsB (100 µL of a 568 µM solution) and standards (500 µL of 0.1 – 1 mg ml-1 solutions) were injected on the column, which was maintained at a flow rate of 0.3 mL min-1 throughout the chromatographic procedure (470 min).

Adenosine (267 Da), cytochrome c (12.4 kDa), Coir albumin (75 kDa) and β-amylase (200 kDa) were used to generate a standard curve of known molecular masses, while the void volume (V0) of the column was determined using blue dextran (2,000 kDa). The elution volumes (Ve) of the

-1 standards were obtained, and the ratios of Ve V0 were plotted as a function of the log of their

328 respective molecular masses. The standard curve was then used to extrapolate the apparent molecular mass of Wt RCN anSMEcpe or AtsB from their corresponding elution volumes. In some analyses, 100 nmol AtsB was combined with 125 nmol AtsA or 2 µmol Kp18Ser before injection.

Fate of the second reducing equivalent upon abstraction of a H• by the 5’-dA•. An anaerobic solution of DT was prepared in 1 M HEPES buffer, pH 7.5, and its concentration was determined

-1 -1 spectrophotometrically using potassium ferricyanide (420 = 1020 M cm ) as a standard and assuming that 1 mol of DT reduces 2 mol of ferricyanide. Flavodoxin semiquinone (Flv•) was generated by adding 0.5 equiv of DT to 1.05 equiv of oxidized Flv (Flvox) and then incubating at

37 °C for 1 h, and its concentration was subsequently determined spectrophotometrically ( =

4570 M-1 cm-1) (41). The anSMEcpe reaction was initiated by adding Flv• (204 µM final concentration) to a reaction mixture containing the following components in a final volume of 1 mL: 100 µM anSMEcpe, 50 mM HEPES, pH 7.5, 200 mM KCl, 2 mM SAM, and 2 mM

Kp18Cys. The mixture was incubated at 37 °C, and at designated times, 250 µL aliquots were removed and loaded into EPR tubes, which were subsequently submerged in cryogenic isopentane (-130 °C) to rapidly freeze the solution. Monitoring of Flv• was performed by EPR at

77 K under nonsaturating conditions (see appropriate figure legends), and spin quantification was determined by comparison of the double integral of the signal to that of a 1 mM Cu(II)-EDTA standard collected under identical (nonsaturating) conditions. Low-temperature spectra were also collected at 13 K to monitor reduction of the Fe/S clusters. Product analysis was conducted in parallel by removing 10 µL aliquots of each reaction before freezing and quenching in acid as described above. The data were fitted to Equation 1, which describes a burst phase followed by a linear steady-state phase.

329

Figure 8-2. Equation describing a burst phase followed by a steady-state phase.

8.4 Results

Overproduction of anSMEcpe. The cpe0635 gene from C. perfringens was cloned into a pET-26b expression vector to yield a construct that overproduces anSMEcpe containing a C-terminal hexahistidine tag separated from the last native amino acid (aa) by a spacer of 16 aa. Sequencing of the cloned gene revealed a number of aa alterations from the sequence reported in the database (42). Subsequent recloning and resequencing of the gene indicated that these changes did not result from cloning artifacts, but were indeed authentic for this particular strain of C. perfringens. These changes include the following substitutions: D56E, I69T, R78K, I177V, R179K, Q212K, L224F,

S309L, K324R, and D341A. We employed a strategy for overproducing soluble anSMEcpe in Escherichia coli (Ec), in which the cpe0635 gene on plasmid pCpe0635Wt was coexpressed with genes from plasmid pDB1282 (33, 34, 43), which derive from an operon encoding proteins known to be involved in Fe/S cluster biosynthesis in

Azotobacter vinelandii. This strategy was used successfully to overproduce sufficient amounts of soluble AtsB for biochemical and spectroscopic characterization (2).

330

Figure 8-3. SDS–PAGE analysis of anSMEcpe. Lane 1, molecular mass markers. Lane 2, purified anSMEcpe (45,740 Da). The gel was stained with Coomassie Brilliant Blue.

In addition, overproduction was conducted in M9 minimal medium to allow for efficient incorporation of 57Fe into the protein for analysis by Mössbauer spectroscopy.

Figure 8-3 depicts an SDS–PAGE analysis of the purified protein, which displays migratory properties that are consistent with its molecular mass (45,740 Da) as calculated from its aa sequence. From 16 L of M9 culture, >250 mg of protein are routinely obtained. This yield is a significant improvement over that observed by Benjdia, et al.

(~5 mg from 12 L of culture) (1), as well as for the previous overproduction of AtsB (2).

Amino acid analysis of anSMEcpe indicates that the Bradford (38) method for protein concentration determination overestimates its concentration by a factor of 1.45 when using BSA (Fraction V) as a standard. Therefore, a correction factor of 0.69 (i.e., 1/1.45) is multiplied by the protein concentration determined by the Bradford method to yield the true protein concentration.

331

Spectroscopic and analytical characterization of wild-type anSMEcpe. The as-isolated

(AI) UV–vis spectrum of anSMEcpe is shown in Figure 8-4A (solid line). The spectrum is consistent with the presence of [4Fe–4S] clusters, showing a broad absorption that extends beyond 700 nm and a distinct feature at 397 nm. In contrast to the spectrum of the AI enzyme recorded by Benjdia, et al., there is very little evidence of [2Fe–2S] clusters (1). The ratio of the absorbance at 397 nm to that at 279 nm, which gives a qualitative assessment of cluster content, is 0.35, significantly greater than the ratio observed by Benjdia et al. (0.19), even for their reconstituted enzyme (0.29), suggesting that anSMEcpe used in this study is of significantly better quality and may be suitable for quantitative cluster analyses and rigorous biochemical characterization (34). Analytical determinations of iron and sulfide associated with AI anSMEcpe indicates 9.6 ± 0.1 of the former and 10.0 ± 0.2 of the latter, suggestive of more than one [4Fe–4S] cluster.

332 Figure 8-4A also indicates that the absorbance at 397 nm is 0.207 for a 5.0 µM sample of anSMEcpe, resulting in a molar absorptivity of ~41,400 M-1 cm-1 at 397 nm. Given that average molar absorptivities in this region for inorganic model peptide-ligated [4Fe–4S] clusters in organic solvents range from 12,100 to 17,500 M-1 cm-1 (44), this analysis strongly suggests that AI anSMEcpe contains more than one [4Fe–4S] cluster, consistent with results from Fe and S2- analysis.

Reconstitution of AI anSMEcpe results in an increase in the stoichiometry of Fe (14.1

± 0.3) and S2- (12.8 ± 0.7) associated with the protein and increased intensity in its UV- vis spectrum (Figure 8-4A, dashed line). Although this behavior is suggestive of increased cluster incorporation, analysis by more definitive spectroscopic techniques is required, because adventitiously bound Fe/S species derived from the reconstitution procedure can also produce similar spectra (34, 39, 45).

333

Mössbauer-spectroscopic characterization of wild-type anSMEcpe. To determine the type and stoichiometry of Fe/S clusters more definitively, AI and reconstituted (RCN) samples of 57Fe-enriched WT anSMEcpe were analyzed by Mössbauer spectroscopy. The

4.2-K/53-mT Mössbauer spectrum of AI anSMEcpe (523 µM; 9.6 Fe per polypeptide) is shown in Figure 8-5A, and is dominated by an intense quadrupole doublet. The EPR

334 spectrum of an identical sample revealed the presence of a small amount of [3Fe–4S]+ clusters (14 µM spin, 42 µM Fe, 0.8% of total Fe) (Figure 8-6, red trace), corresponding to 0.8% of the total Fe (i.e. [3 Fe  14 µM]/[5.02 mM total Fe]). Such a small amount of a paramagnetic cluster with three distinct Fe subsites is beyond the detection limit of

Mössbauer spectroscopy (46). The Mössbauer spectrum can be analyzed with one broad quadrupole doublet (95% of total Fe) with parameters typical of [4Fe-4S]2+ clusters: isomer shift () of 0.44 mm/s and quadrupole splitting parameter (EQ) of 1.14 mm/s

(solid line in Figure 8-5A). The weak absorption at 0.6 mm/s (see arrow) is at a position typical of the high-energy line of spectra of [2Fe-2S]2+ clusters and is most likely associated with a small amount (3%) of this cluster type, which is often observed as the degradation product of [4Fe-4S] clusters (46). The nature of the weak shoulder (<2% of total Fe) at 1.7 mm/s (see arrow) is not clear. Mössbauer analysis, in addition to the stoichiometry of 9.6 Fe ions per polypeptide, therefore reveals that AI WT anSMEcpe harbors 2.3 [4Fe–4S] clusters.

The 4.2-K/53-mT Mössbauer spectrum of RCN WT anSMEcpe (173 µM; 14.2 Fe per polypeptide) (Figure 8-5B) is also dominated by the same intense quadrupole doublet associated with the [4Fe-4S]2+ clusters of AI WT anSMEcpe. Approximately 75% of the total Fe can be attributed to the [4Fe-4S]2+ clusters of AI anSMEcpe (Figure 8-5B, solid line), resulting in a stoichiometry of 2.7 [i.e. (14.2 Fe)  (0.75)/(4 Fe per cluster)] [4Fe-

4S]2+ clusters per polypeptide. The remaining 25% of Fe gives rise to a broad absorption, which is attributed to unspecifically bound Fe, because the EPR spectrum of an identical sample reveals only a small amount of [3Fe-4S]+ clusters (7 μM spin, 21 μM Fe, 0.9% of

335 total Fe) (Figure 8-6, black trace) and no other signals attributable to Fe/S clusters with spin state S = ½ are observed. Thus, the combination of Mössbauer spectroscopy and analytical methods strongly suggests the presence of three [4Fe-4S] clusters on anSMEcpe, as was reported for the related enzyme, AtsB, from Klebsiella pneumoniae

(2).

Characterization of AI and RCN C15A/C19A/C22A triple variant anSMEcpe by

Mössbauer spectroscopy. To verify the stoichiometry of three [4Fe–4S] clusters per WT anSMEcpe polypeptide, a triple variant, in which the Cys residues that ligate the RS Fe/S cluster are changed to Ala residues, was constructed (anSMEcpeC15A/C19A/C22A). This substitution of all coordinating residues to the RS Fe/S cluster with noncoordinating residues should lead to its complete elimination, resulting in a stoichiometry of two [4Fe–

4S] clusters per polypeptide. anSMEcpeC15A/C19A/C22A was noticeably less stable than the

WT protein, which is in contrast to that observed for AtsB, wherein the corresponding triple variant was more stable than the WT protein (2). Nonetheless, ~15 mg of 57Fe- labeled protein was isolated from 8 L of culture, considerably less than that obtained for the WT protein, but sufficient for appropriate characterization.

The UV-vis spectrum of the AI anSMEcpeC15A/C19A/C22A is still consistent with the presence of [4Fe–4S] clusters, exhibiting a pronounced feature at 397 nm and an

A397/A279 ratio of 0.24 (Figure 8-4B, solid line), consistent with the finding of 3.2 ± 0.1

Fe and 7.5 ± 0.1 S2- per polypeptide. Reconstitution of the triple variant results in an increase in the A397/A279 ratio (0.42) (Figure 8-4B, dashed line) as well as iron and sulfide associated with the protein (8.8 ± 0.4 and 15.1 ± 0.9, respectively). However, the

336 spectral features between 550 and 700 nm suggest the presence of adventitiously bound iron in this sample.

The 4.2-K/53-mT Mössbauer spectrum of AI anSMEcpeC15A/C19A/C22A (472 µM; 3.2 Fe per polypeptide) (Figure 8-5C) is dominated by a quadrupole doublet associated with

2+ [4Fe-4S] clusters:  = 0.44 mm/s, EQ = 1.16 mm/s, 80% intensity (dotted line). In addition, the peak at 0.6 mm/s suggests the presence of [2Fe-2S]2+ clusters ( = 0.31 mm/s, EQ = 0.51 mm/s, 17% intensity, dashed line). The greater relative fraction of

2+ [2Fe-2S] clusters in anSMEcpeC15A/C19A/C22A compared to that in WT anSMEcpe suggests a greater instability of the remaining [4Fe–4S] clusters in the triple variant. In addition, an identical EPR sample does not show signals of Fe/S clusters with half-integer spin ground states (Figure 8-6, green trace). The Mössbauer data, in concert with the observed stoichiometry of 3.2 Fe per polypeptide, indicates that AI

2+ 2+ anSMEcpeC15A/C19A/C22A contains 0.6 [4Fe-4S] and 0.3 [2Fe-2S] clusters per polypeptide.

337

Reconstitution of anSMEcpeC15A/C19A/C22A with additional Fe and sulfide leads to greater incorporation of Fe/S clusters. The 4.2-K/53-mT Mössbauer spectrum of RCN anSMEcpeC15A/C19A/C22A (281 µM; 8.8 Fe per polypeptide) (Figure 8-5D) is dominated by

2+ a quadrupole doublet associated with [4Fe-4S] clusters ( = 0.44 mm/s, EQ = 1.16 mm/s, 70% intensity), while the remainder is associated with unspecifically bound Fe, given that an identical EPR sample does not show signals of Fe/S clusters with half- integer spin ground states (Figure 8-6, blue trace). Given the stoichiometry of 8.8 Fe per polypeptide, it is concluded that RCN anSMEcpeC15A/C19A/C22A harbors 1.5 [4Fe-4S] clusters. This stoichiometry clearly indicates that the triple variant harbors more than one

338 [4Fe-4S] cluster. The fact that it does not contain a full complement of two [4Fe-4S] clusters is rationalized by the greater instability of the protein.

Gel-filtration analysis of anSMEcpe. To assess the quaternary structure of WT anSMEcpe, the RCN protein was subjected to molecular-sieve chromatography on a

Sephacryl S-200 HR gel-filtration column connected to an ÄKTA preparative liquid chromatography system housed in a Coy anaerobic chamber. A series of protein standards was used to generate a plot of log molecular mass of a given standard versus Ve

-1 V0 , wherein Ve is the elution volume of the standard and V0 is the void volume of the column. This plot was then used to extrapolate the molecular mass of anSMEcpe from its

-1 determined Ve V0 value. Hexahistidine-tagged anSMEcpe migrates as a symmetrical single peak of molecular mass 37,500 Da under the conditions described in Materials and

Methods (Figure 8-7A).

339

Its calculated molecular mass of 45,740 Da would therefore be most consistent with a monomeric quaternary structure. A similar experiment was also conducted for hexahistidine-tagged AtsB, which migrates as a symmetrical peak of molecular mass

33,500 Da (Figure 8-7B, blue line). Its calculated molecular mass of 46,432 Da would suggest that the protein also exhibits a monomeric quaternary structure, although the possibility of a dimeric structure exists. Interestingly, when AtsB is mixed with its peptide substrate (Kp18Ser, MW 2,001 Da) before being applied to the column, it migrates as a protein of 35,800 Da, consistent with a protein/peptide complex (Figure 8-

7B, black line). By contrast, when it is mixed with its natural protein substrate (Kp AtsA), it migrates still as a protein of 33,500 Da (Figure 8-7B, red line), consistent with

340 previous suggestions that AtsB acts on AtsA before it is folded into its native tertiary structure (17). The absence of a peak for AtsA in the chromatogram is due to monitoring at 395 nm, which allows for the selective monitoring of AtsB migration. The observation that the protein/peptide complex migrates almost exactly as the sum of the masses of the protein (33,500 Da) and peptide (2,001 Da) determined from molecular-sieve chromatography argues for a monomeric structure over a dimeric structure. Unless the protein exhibits half-of-the-sites reactivity, the protein/peptide complex for dimeric AtsB would be expected to exhibit a molecular mass of ~37,502 Da (33,500 + 4,002 Da).

341 Activity determination of anSMEcpe. Sulfatase maturating enzymes (SMEs) act on protein substrates, installing the required FGly cofactor in arylsulfatases (18-22, 26, 47).

There is a consensus sequence motif C/S-X-P-S/X-R-X-X-X-L/X-T/X-G/A-R/X found among the various protein substrates irrespective of the mechanism used to generate the

FGly cofactor, in which an invariant Arg residue is separated from the Cys or Ser residue to be modified by three amino acids, the second of which is typically Pro, but which can also be Ala (16, 48). Initial activity determinations in this work were conducted with peptides used to study AtsB rather than those that mimic the natural protein substrate for anSMEcpe, given that these were on hand. The FGly modification was quantified by

HPLC with detection by QQQ mass spectrometry (LC/MS) using a peptide standard of the same sequence but containing an authentic FGly residue at the target position. Figure

8-8 displays LC-MS data used to quantify FGly production in a typical assay, which reveals that the FGly-containing product forms at the expense of the substrate. Although the peak corresponding to the FGly product is irregular, due to the highly electrophilic nature of the aldehyde, all regions of the peak correspond to the expected m/z value for the peptide containing the FGly modification. Moreover, the FGly product migrates exactly—both with respect to retention time and shape—as a standard peptide synthesized with an FGly residue at the target position.

342

Figure 8-9. Turnover of WT RCN anSMEcpe with Kp18Cys. A) Time-dependent formation of 5′-dA (closed triangles) and Kp18FGly (open squares), and depletion of Kp18Cys (closed squares) using DT as the requisite electron donor. Reaction mixtures contained 4 μM anSMEcpe, 1 mM SAM, 0.5 mM Kp18Cys, and 3 mM dithionite. The data are the averages of three independent trials, and error bars denote one standard deviation. Vmax/[ET] values for 5’-dA and Kp18FGly formation are 2.98 ± 0.071 min-1 and 2.30 ± 0.100 min-1, respectively, while the -1 Vmax/[ET] value for consumption of Kp18Cys is 2.37 ± 0.017 min . B) Time-dependent formation of 5′-dA (closed triangles) and Kp18FGly (open squares) using the Flv/Flx/NADPH reducing system. Reaction mixtures contained 40 μM anSMEcpe, 1 mM SAM, 1 mM Kp18Cys, 50 μM Flv, 15 μM Flx, and 2 mM NADPH. The data are the averages of three independent trials, and error bars denote one standard deviation. Vmax/[ET] values for 5’-dA and Kp18Cys are 0.28 ± 0.022 min-1 and 0.26 ± 0.022 min-1, respectively. In Figure 8-9a, the activity of anSMEcpe (4 µM) using Kp18Cys (500 µM) as the substrate and DT as the reductant is displayed. Formation of the FGly product (open

-1 squares) occurs with a Vmax/[ET] of 2.31 ± 0.10 min , while formation of 5’-dA (closed

-1 triangles) occurs with a Vmax/[ET] of 2.98 ± 0.07 min . In addition, ~100 turnovers take place within the 30 min span of the assay. Figure 8-9b depicts activity profiles of anSMEcpe (40 µM) using Kp18Cys as the substrate and the Flv/Flx/NADPH reducing system as the source of the requisite electron. Similarly to that observed for AtsB, the reaction is significantly slower under these conditions, displaying Vmax/[ET] values of

0.28 ± 0.022 min-1 and 0.26 ± 0.022 min-1 for 5’-dA (closed triangles) and FGly (open

343 squares) formation, respectively. Importantly, for each of these assays product formation is stoichiometric with substrate consumption. In addition, these Vmax/[ET] values are significantly higher than those observed for AtsB under similar conditions (2).

Figure 8-10. Turnover of WT RCN anSMEcpe with Cp18Cys. A) Time-dependent formation of 5’-dA (black triangles) and depletion of Cp18Cys (red squares) in the presence of dithionite. Reaction mixtures contained 4 μM anSMEcpe, 1 mM SAM, 1 mM Cp18Cys, and 3 mM dithionite. The data are the averages of two independent trials, and error bars denote one standard -1 deviation. Vmax/[ET] values for 5’-dA formation and peptide consumption are 4.50 ± 0.052 min and 1.91 ± 0.259 min-1, respectively. B) Time-dependent formation of 5’-dA (black triangles) and depletion of Cp18Cys (red squares) in the presence of the Flv/Flx/NADPH reducing system. Reaction mixtures contained 40 μM anSMEcpe, 1 mM SAM, 1 mM Cp18Cys, 50 μM Flv, 15 μM Flx, and 2 mM NADPH. The data are the averages of two independent trials, and error bars denote one standard deviation. Vmax/[ET] values for 5’-dA formation and peptide consumption are 0.22 ± 0.003 min-1 and 0.21 ± 0.032 min-1, respectively.

Activity determinations were also conducted with a peptide substrate that corresponds to the sequence of the natural substrate for anSMEcpe. Only substrate consumption was monitored in these assays because of lack of an FGly-containing peptide standard.

However, using multiple different assays we have never observed formation of significant amounts of any intermediate species; loss of substrate peptide is always concomitant with formation of product peptide. The Vmax/[ET] for 5’-dA formation and

344 consumption of Cp18Cys are 4.50 ± 0.052 min-1 and 1.91 ± 0.259 min-1, respectively, using DT as reductant, indicating that a significant amount of abortive cleavage of SAM occurs in the presence of this substrate (Figure 8-10A). In the presence of the

Flv/Flx/NADPH reducing system the rates are 0.224 ± 0.003 min-1 and 0.213 ± 0.032 min-1, respectively, similar to those obtained with the Kp18Cys substrate and indicating tight coupling of SAM cleavage and FGly formation (Figure 8-10B).

Figure 8-11. Turnover of WT RCN anSMEcpe with Kp18Ser. A) Time-dependent formation of 5′-dA (closed triangles), and Kp18FGly (open squares) in the presence of DT. Reaction mixtures contained 4 μM anSMEcpe, 1 mM SAM, 1 mM Kp18Ser, and 3 mM DT. The data are the averages of two independent trials, and error bars denote one standard deviation. Vmax/[ET] values for 5’-dA and Kp18FGly are 1.00 ± 0.029 min-1 and 0.85 ± 0.001 min-1 , respectively. B) Time-dependent formation of 5′-dA (closed diamonds) and Kp18FGly (open squares) in the presence of the Flv/Flx/NADPH reducing system. Reaction mixtures contained 40 μM anSMEcpe, 1 mM SAM, 1 mM Kp18Ser, 50 μM Flv, 15 μM Flx, and 2 mM NADPH. The data are the averages of three independent trials, and error bars denote one standard deviation. -1 -1 Vmax/[ET] values for 5’-dA and Kp18FGly are 0.074 ± 0.009 min and 0.073 ± 0.004 min , respectively. Our previous studies indicated that AtsB can act as a Cys-type anSME, although its natural substrate bears a Ser residue at the target position. Studies by Benjdia, et al. showed that anSMEcpe can indeed oxidize Ser-containing substrates; however, the

345 experiments were qualitative in nature and did not permit direct comparison of rates. In

Figure 8-11, turnover of anSMEcpe with Kp18Ser is shown. As can be observed, the rates are significantly lower than that in the presence of Kp18Cys. When using DT as the

-1 reductant, Vmax/[ET] is 1.00 ± 0.029 and 0.85 ± 0.001 min for formation of 5’-dA and the FGly product, respectively. When using the Flv/Flx/NADPH reducing system,

-1 Vmax/[ET] is 0.074 ± 0.009 and 0.073 ± 0.004 min for formation of 5’-dA and the FGly product, respectively. These rates are approximately three-fold lower with either reductant when Kp18Ser is substituted for Kp18Cys.

Figure 8-12. Time-dependent formation of 5′-dA (black triangles) and Kp18FGly (red squares) in the presence of Kp18SeCys. Reaction mixtures contained 40 μM anSMEcpe, 1 mM SAM, 0.5 mM Kp18SeCys, 50 μM Flv, 15 μM Flx, and 2 mM NADPH. Vmax/[ET] values for 5’-dA and Kp18FGly formation are 0.053 min-1 and 0.032, respectively. The target Cys residue was also replaced with a SeCys residue, which has a number of properties that are similar to those of Cys. In addition, a substrate containing a SeCys residue would permit investigation of substrate coordination to an Fe/S cluster by

346 selenium X-ray absorption spectroscopy (49-51). Figure 8-12 displays turnover of anSMEcpe in the presence of Kp18SeCys and the Flv/Flx/NADPH reducing system. The reaction is linear for the first 10 min, but becomes uncoupled at longer incubation times, which is different from that observed for substrates containing Cys or Ser at the target

-1 position. A fit to the first three time points affords Vmax/[ET] values of ~0.053 min for both 5’-dA and Kp18FGly.

Fate of the second reducing equivalent upon abstraction of a H• by the 5’-dA•. All RS enzymes require the input of an electron to initiate reductive cleavage of SAM to a 5’- dA•, which is used most often to oxidize substrates by one electron via H• abstraction. In the reactions catalyzed by AtsB and anSMEcpe, the substrate is oxidized further by one electron, wherein the presumed radical intermediate transfers an electron to an undetermined acceptor. It has been suggested that the electron is returned to the RS [4Fe–

4S] cluster after each turnover, implying that the introduction of one electron can prime the system for multiple turnovers as has been shown for the RS enzyme, DesII, in a reaction with a substrate analog (52). To address the fate of the remaining electron, Flv• was generated by treatment of 0.5 equiv of DT with 1.05 eq. of Flv and then added to a reaction mixture containing the following components after quantification of the Flv• concentration: anSMEcpe (100 µM), SAM (2 mM), and Kp18Cys (2 mM), and Flv• (204

µM). At designated times (1, 5, and 15 min), aliquots were removed and added to EPR tubes, which were subsequently immersed in cryogenic isopentane (~-130 °C) to quench the reaction by rapid freezing.

347

EPR at 77 K as described in Materials and Methods, while parallel aliquots were removed from the reaction to quantify product formation by LC/MS. As can be seen in

Figure 8-13A, the concentration of Flv• is essentially unchanged throughout the 15 min incubation. By contrast, Figure 8-13B shows that greater than 200 µM product is formed

(~2 turnovers) during the same time period, and that FGly formation (open squares) is tightly coupled to SAM cleavage (5’-dA, closed triangles). The open circles in Figure 8-

13B correspond to the Flv• concentrations in Figure 8-13A; the slight change in concentration of Flv• during the 15 min period most likely derives from slight O2 contamination. If the sole role of Flv is to prime the reaction such that the emitted electron from the substrate radical intermediate is returned to the RS cluster to be used in

348 a subsequent round of SAM cleavage, it would be expected that the concentration of Flv• should decrease by 50% (from 200 µM to 100 µM) within the first 3 min of the reaction, which corresponds to the time required for one full turnover.

Figure 8-14. Low-temperature X-Band EPR of Flv• and anSMEcpe during turnover. Reaction mixtures contained 100 μM anSMEcpe, 2 mM SAM, 2 mM Kp18Cys and 204 μM Flv• -1 -1 (580=4.57 mM cm ) at 13 K at 1 min (green), 15 min (red), and 30 min (black). Spectra were collected on a Bruker ESP-300 X-Band EPR spectrometer under the following conditions: frequency, 9.51 GHz; temperature, 13 K; power, 0.101 mW; and modulation amplitude, 10 Gauss. The observation that the concentration of Flv• does not change significantly over the course of multiple turnovers suggests that the ejected electron is ultimately returned to

Flvox at the end of each turnover event. Consistent with this observation, parallel EPR spectra recorded at 13 K do not show evidence of a reduced [4Fe-4S] cluster (Figure 8-

14), which would argue against recycling of the ejected electron by storing it internally

349 on an Fe/S cluster. Whether reduction of Flvox occurs through a reduced RS [4Fe-4S] cluster intermediate or a reduced auxiliary cluster intermediate is not yet clear. Of note is the bi-phasic nature of the appearance of 5’-dA and Kp18FGly, indicating that a burst phase is followed by a steady-state phase. A fit of the data to an appropriate equation results in the following kinetic parameters: burst amplitude, 113 µM; kburst, 0.32 ± 0.078

-1 -1 min ; kss, 0.059 ± 0.011 min . The burst phase, which corresponds to ~1.1 equiv of enzyme may arise from rate-limiting product release; however, we have not rigorously characterized this aspect of the reaction. A similar experiment carried out with AtsB (150

µM), SAM (1 mM), Kp18Ser (1 mM), and 75 µM Flv• showed essentially identical

-1 results, albeit with a smaller burst phase (burst amplitude, 10.6 µM; kburst, 2.0 min ; kss,

0.015 min-1) (Figure 8-15).

Figure 8-15. Correlation of spectral changes and product formation during AtsB turnover with Kp18Cys. A) X-Band EPR (77 K) spectra of a reaction mixture containing 150 μM AtsB, 1 mM SAM, 1 mM Kp18Ser, 75 µM Flvox, and 75 μM Flv• at 1 min (red), 15 min (green), and 30 min (black). Spectra were recorded as described in Materials and Methods. B) Time-dependent quantification of Flv• (open circles) and 5′-dA (closed triangles). The black line is a fit of the 5’- dA data to an equation describing a burst phase followed by a steady-state linear phase.

350

Stereochemistry of AtsB and anSMEcpe. Recent studies of Benjdia, et al. verified the hypothesis that the role of the 5’-dA• in RS dehydrogenases is to abstract a hydrogen atom from the carbon undergoing oxidation, which was initially demonstrated by

Yokoyama et al for BtrN (3, 53). Using a peptide containing a target Cys residue isotopically substituted at C3 with deuterium, they provided evidence via mass spectrometry and NMR for transfer of deuterium to 5’-dA. However, the C3 hydrogens of cysteine are prochiral, and it would be expected that an enzyme would act stereoselectively in the removal of an H• from this position. Given that seryl residues are oxidized to FGly both by AtsB and anSMEcpe, we assessed whether threonyl and allo- threonyl residues, which are chiral at C3, are converted into the corresponding ketone product. As shown in Figure 8-16, the configuration of L-threonine at its two chiral carbons is 2S,3R, while the configuration of L-allo-threonine is 2S,3S.

351

Figure 8-16. Stereochemical designation of threonine and allo-threonine.

Therefore, conversion of substrate containing a threonyl residue at the target position would require abstraction of the proS hydrogen, while conversion of a substrate containing an allo-threonyl residue at the target position would require abstraction of the proR hydrogen. Figure 8-17 displays the results of activity determinations with

Kp18Thr and Kp18alloThr, containing L-threonyl, and L-allo-threonyl residues, respectively, at the target position. As can be observed, (Figure 8-17A, closed squares)

~130 µM Kp18Thr is consumed in ten min in a reaction containing 100 µM anSMEcpe and DT as the requisite reductant, and MALDI-TOF analysis of the DPNH-derivatized product (m/z = 2195.4) is consistent with its assignment as the corresponding ketone derivative (Figure 8-18A). By contrast, only ~20 µM Kp18alloThr is consumed under identical conditions before the reaction levels off (Figure 8-17B, closed squares). This amount of substrate consumption could derive from L-Thr contamination at the target

352 position, especially given that the reaction stops abruptly. MALDI-TOF analysis of the

DPNH-derivatized product (m/z = 2195.4) verifies that there is a much smaller, but observable, amount of the corresponding ketone product (Figure 8-18b). AtsB was also able to use Kp18Thr as a substrate, but to a lesser extent, as judged by the relative intensities of the substrates with respect to the derivatized products (Figure 8-19).

Figure 8-17. Turnover of WT RCN anSMEcpe with Kp18Thr or Kp18alloThr. A) Time- dependent formation of 5′-dA (closed black triangles) and disappearance of Kp18Thr (closed red squares). Reaction mixtures contained 100 μM anSMEcpe, 1 mM SAM, 0.5 mM Kp18Thr, and -1 3 mM DT. Vmax/[ET] values for 5’-dA formation and Kp18Thr disappearance are 0.29 min and 0.11 min-1 , respectively. B) Time-dependent formation of 5′-dA (closed triangles) and disappearance of Kp18alloThr (closed red squares). Reaction mixtures contained 100 μM anSMEcpe, 1 mM SAM, 0.5 mM Kp18alloThr, and 3 mM DT. Vmax/[ET] values for 5’-dA formation and Kp18alloThr disappearance are 0.07 min-1 and 0.007 min-1, respectively.

353

354 Determination of cysteinyl residues that ligate the [4Fe–4S] clusters in anSMEs. AtsB contains 13 Cys residues, three of which lie in the canonical CxxxCxxC motif. Site- directed mutagenesis of the remaining ten Cys residues was conducted to determine which might coordinate the auxiliary clusters. Seven of the CysAla variants (C270A,

C276A, C331A, C334A, C340A, C344A, and C357A) were produced in a completely insoluble form and not studied further.

Two of the variants, C127A and C245A, were freely soluble and behaved like WT AtsB in both purification and activity. The UV-vis spectra for both of these AI variants are displayed in Figure 8-20 (solid and dashed lines, respectively), and reveal spectral

355 envelopes that are similar to that of the WT protein. Moreover, their A395/A280 values of

0.38 are also similar to that of the AI WT protein (0.42) (2). The AI C127A variant contained 9.8 ± 0.1 and 9.6 ± 0.5 iron and sulfide ions, respectively, per polypeptide, while the AI C245A variant contained 12.0 ± 1.1 and 15.0 ± 0.3 iron and sulfide ions, respectively, per polypeptide.

Figure 8-21. UV-vis spectrum of AI AtsB C291A. The protein (6.4 µM), contained 6.7 ± 0.1 irons and 5.6 ± 0.6 sulfides per polypeptide. The A405/A280 ratio was 0.39.

Activity determinations on both of these AI proteins were conducted using the

-1 Flv/Flx/NADPH reducing system, yielding Vmax/[ET] values of 0.061 and 0.052 min ,

356 respectively, for the C127A and C245A variants, similar to the Vmax/[ET] value for the

WT protein under similar conditions (0.040 min-1).

The C291A variant could be isolated, but was sparingly soluble and poorly behaved.

The UV-vis spectrum for this variant is shown in Figure 8-21, and reveals a spectral envelope that is similar to that of the WT protein. Its A405/A280 value of 0.39 would indicate high cluster incorporation; however, it contained only 6.7 ± 0.1 irons and 5.6 ±

0.6 sulfides per polypeptide. Efforts to reconstitute this protein resulted in its

357 precipitation from solution. The activity determination of this variant was not above the limit of detection of the assay when the Kp18Ser peptide was used as substrate, suggesting that this Cys residue is either structurally or functionally important in the reaction of anSMEs. When activity determinations were conducted in the presence of DT and the Kp18Cys peptide approximately 10 µM 5’-dA and 2 µM FGly product were observed. The equivalent CysAla variant was constructed for anSMEcpe (C276A) and found, in contrast to AtsB C291A, to be stable and readily soluble. The UV-vis spectrum of this protein shows an almost identical spectral envelope as WT anSMEcpe (Figure 8-

22). In a similar fashion as AtsB C291A, this protein was almost completely inactive in production of Kp18FGly. After a 30-min incubation in the presence of 200 µM anSMEcpe C276A, approximately 6 µM FGly was produced (Figure 8-23). On the other hand, SAM reductase activity was higher than that of the AtsB C291A variant, albeit less than one-half an equivalent of 5’-dA was produced after 30 min after 60 min of incubation with 70 µM protein.

358

Figure 8-23. Time-dependent formation of 5′-dA (closed triangles) and Kp18FGly (open squares) with RCN anSMEcpe C276A and Kp18Cys. Reaction mixtures contained 200 μM anSMEcpe C276A, 1 mM SAM, 1 mM Kp18Cys, and 3 mM DT. The data are the averages of two independent trials, and error bars denote one standard deviation. Vmax/[ET] values for 5’-dA and Kp18FGly are 0.14 ± 0.001 min-1 and 0.001 ± 0.0001 min-1, respectively.

8.5 Discussion

anSMEcpe shares 48% sequence similarity with Kp AtsB; however, it is a Cys-type anSME, and therefore its in vivo role is to catalyze the oxidation of a target Cys residue to

FGly. Its mechanism of catalysis is predicted to be identical to that proposed for AtsB, except that the presumed thioaldehyde product is then hydrolyzed to the aldehyde with elimination of H2S. anSMEcpe has been characterized previously using a number of

359 spectroscopic techniques, including UV-vis, resonance Raman, and EPR spectroscopies

(1). Although the previous studies were consistent with the presence of Fe/S clusters on the enzyme, cluster content was not rigorously determined. The protein studied by

Benjdia, et al. contained an N-terminal hexahistidine tag, and was overproduced largely in inclusion bodies, yielding ~5 mg of soluble protein from 12 L of growth medium. In contrast, it has been our strategy to include an accessory plasmid that harbors genes that encode proteins that are involved in Fe/S cluster biosynthesis in A. vinelandii, the homologs of which are known to encode proteins involved in Fe/S cluster biosynthesis in other organisms (34). This strategy allowed purification to near-homogeneity of >250 mg of anSMEcpe containing a C-terminal hexahistidine tag from 16 L of minimal medium.

This yield is considerably higher than that reported by Benjida, et al. as well as that for

AtsB (30 mg from 16 L of medium). Indeed, we find that WT anSMEcpe is a much better behaved than WT AtsB, and therefore better suited for detailed mechanistic and structural investigations.

In the work presented herein, Mössbauer spectroscopy was used in concert with analytical determinations of 57Fe content to establish not only the configuration of Fe/S clusters associated with anSMEcpe, but also the stoichiometry of each particular cluster type per anSMEcpe polypeptide. When anSMEcpe is overproduced along with proteins encoded by plasmid pDB1282, the AI enzyme contains 2.3 [4Fe–4S] clusters (95% of all

57Fe), with ~3% of all 57Fe occurring as [2Fe–2S] clusters and ~2% occurring as an undefined cluster type. Upon reconstitution of AI anSMEcpe, the protein contains ~2.7

[4Fe–4S] clusters (75% of all 57Fe), with the remaining 25% of all 57Fe existing as

360 unspecifically bound iron. Analysis of a triple variant of anSMEcpe, in which the Cys ligands to the RS [4Fe–4S] clusters were changed to Ala residues—a state that should not allow cluster ligation—showed that the AI protein contained 0.6 [4Fe–4S] clusters and

0.3 [2Fe–2S] clusters, while the RCN triple variant contained 1.5 [4Fe–4S]2+ clusters.

Our model of three [4Fe–4S] clusters per polypeptide for anSMEcpe would predict that the triple variant would harbor two [4Fe–4S] clusters. In contrast to AtsB, in which the analogous triple variant is more soluble than the WT protein, we find that the anSMEcpe triple variant is less stable and less soluble than its corresponding WT protein. We believe that the increased heterogeneity in the AI triple variant and the significantly reduced cluster content derives from the instability of this protein.

Previous site-directed mutagenesis studies on AtsB revealed, as expected, that one of the clusters is ligated by C35, C39, and C42, which are found in the canonical CxxxCxxC

RS signature sequence (2). However, the large number of Cys residues (13) in the primary structure of AtsB did not readily allow determination of the ligands to the two remaining clusters, or determination of which Cys residues were partnered in the ligation of any given cluster. Given the presence of two auxiliary Fe/S clusters, our original working hypothesis was that one would be the immediate acceptor of an electron from the substrate-radical intermediate generated via H• abstraction by the 5’-dA•, and that the other cluster would act as a conduit through which the ejected electron would be transferred to an acceptor, presumed to be Flvox. This hypothesis suggested the possibility of two phenotypes for CysAla variants of the cysteines coordinating the two auxiliary clusters: (1) variants that are completely inactive due to an inability to transfer an electron from the substrate radical intermediate, and (2) variants that are inactive with Flv but

361 active with DT, presuming that oxidized DT (i.e. bisulfite) can accept an electron from the reduced auxiliary cluster (54). In an effort to determine the ligands that ligate the auxiliary clusters and perhaps provide evidence for the role(s) of these clusters, we created single CysAla substitutions at the ten cysteines outside of the CxxxCxxC motif, with the intent of purifying and characterizing the corresponding proteins. We found that the behavior of the resulting variants could be grouped into three categories: those that afforded proteins that behaved essentially like WT AtsB (C127A and C245A); those that afforded completely insoluble proteins (C270A, C276A, C331A, C334A,

C340A, C344A, and C357A); and one that afforded a sparingly soluble protein exhibiting measureable, but very poor, activity (C291A). Based on these observations, we feel confident that C127 and C245 play no major role in catalysis, while C270, C276, C331,

C334, C340, C344, and C357 contribute ligands to the two auxiliary [4Fe-4S] clusters.

The role of C291 is more difficult to assign because of its intermediate behavior. The significantly reduced activity of the C291A variant might suggest a role such as the general base to which the substrate proton is donated during the dehydrogenation reaction; however, its significantly reduced solubility might suggest that it serves as a ligand to one of the auxiliary [4Fe-4S] clusters, implying that both of these clusters are fully ligated. We note that C276 in anSMEcpe, the equivalent residue to C291 in AtsB, behaved similarly. Consistent with two fully ligated auxiliary clusters, our efforts to establish substrate ligation to an auxiliary cluster using selenium X-ray absorption spectroscopy and Kp18SeCys were unsuccessful (unpublished results). It should be mentioned that we observed a similar outcome with variants of BtrN, a RS dehydrogenase that has only one auxiliary cluster (31). This enzyme contains eight Cys

362 residues, three of which (C16, C20, and C23) coordinate the RS cluster, and one of which behaves like the WT protein. Three additional Cys residues, which when substituted with

Ala, were produced completely as insoluble aggregates, suggesting that they coordinate the auxiliary [4Fe–4S] cluster. One Cys residue, C235, behaved similarly to C291 of

AtsB and C276 of anSMEcpe. Although the C235A variant of BtrN could be purified, it was poorly soluble, and exhibited a Vmax/[ET] that was less than 10% of that of the WT enzyme. If indeed both auxiliary clusters in AtsB are fully ligated by Cys residues, it is highly likely that the two auxiliary clusters in anSMEcpe and the one auxiliary cluster in

BtrN are similarly ligated.

Our current studies do not allow us to deduce the role(s) of the auxiliary clusters in RS dehydrogenases. In fact, it is conceivable that they simply maintain the structural integrity of the protein. Interestingly, a subclass of the glycyl radical enzyme (GRE) activases, proteins that catalyze formation of glycyl radical cofactors on cognate enzymes, are also believed to harbor three [4Fe–4S] clusters, although the stoichiometry has not been rigorously determined (7, 55). It has been speculated that the two auxiliary clusters in the GRE activases might act as a conduit for reduction of the RS Fe/S cluster

(56). This role is unlikely in AtsB and anSMEcpe, however, given that these enzymes catalyze their reactions in the presence of flavodoxin with rate constants that are equal to or better than those exhibited by many other RS enzymes that do not contain auxiliary clusters but are also activated by flavodoxin. Our studies herein, however, suggest that after each turnover, the ejected electron is returned ultimately to Flvox, given that the concentration of Flv• does not change significantly during catalysis. If the ejected electron were returned to the RS cluster as its final destination, we would expect that (i)

363 the reaction should exhibit a lag phase (corresponding to slow reduction of the RS [4Fe–

4S] cluster) followed by a faster phase (return of the ejected electron to the RS [4Fe–4S] for use in subsequent rounds of SAM cleavage) that approaches the steady-state rate of the reaction in the presence of dithionite; and (ii) the concentration of the Flv• should have been reduced by the concentration of enzyme in the assay (50%), given the burst of product corresponding to one equiv of enzyme, which suggests that all active sites are functional. Whether the electron is returned to Flvox via the auxiliary clusters or the RS cluster is currently unknown.

The RS enzyme, DesII, catalyzes a key step in the biosynthesis of D-desosamine, a deoxysugar found in a number of macrolide antibiotics. This reaction is the conversion of thymidine diphosphate (TDP)-4-amino-6-deoxy-D-glucose to TDP-3-keto-4,6-dideoxy-D- glucose, which is somewhat similar to the reaction catalyzed by the coenzyme B12- dependent enzyme, ethanolamine ammonia lyase (57). This reaction, with respect to the substrate, is redox-neutral; however, DesII catalyzes stoichiometric production of 5’-dA with respect to product rather than regeneration of SAM after each turnover, therefore requiring the input of two electrons during turnover (52). Interestingly, DesII will also catalyze a two-electron oxidation of the nonphysiological substrate, TDP-D-quinovose (4- hydroxy-6-deoxy-D-glucose), converting it to TDP-3-keto-6-deoxy-D-glucose. In this instance, although the ratio of 5’-dA to product remains 1:1, the reaction does not require external reducing equivalents once primed, suggesting that the ejected electron is returned to the RS [4Fe–4S] — the sole Fe/S cluster on the protein — after each turnover

(52).

364 anSMEcpe and AtsB each harbor a CxxCxxxxxCxxxC motif, which our studies herein indicate contains cysteines that contribute ligands to auxiliary [4Fe–4S] clusters.

Interestingly, this motif is highly conserved in a newly designated subclass of RS enzymes, TIGR04085, which are those that contain SPASM domains. The acronym

SPASM derives from the finding that the founding members of this family catalyze key steps in the maturation of subtilosin, PQQ, anaerobic sulfatases, and mycofactin. In addition, the conserved cysteine-containing motif that each member shares is always C- terminal to the RS cysteine-containing motif (58, 59). Only in the anSMEs has the cluster stoichiometry been rigorously established in this subclass of RS enzymes (2), and the roles of the auxiliary cluster(s) have not been delineated in any SPASM domain- containing protein. Nevertheless, these enzymes share the characteristic of catalyzing reactions on protein or peptide substrates.

Our results with peptide substrates containing threonyl residues at the target position suggest the following working hypothesis for catalysis by AtsB and anSMEcpe. After reductive cleavage of SAM, the 5’-dA• abstracts the 3-proS H• of the substrate, yielding a substrate radical. Subsequent to electron transfer to an auxiliary cluster and loss of a substrate proton — in an order that has not been established — the ejected electron is transferred to Flvox via an auxiliary cluster. These studies herein, and future studies, will provide much needed insight into a growing class of RS enzymes — including those containing SPASM domains — that use multiple Fe/S clusters to catalyze their reactions(7).

365 8.6 Acknowledgement

We thank Professor Carolyn Bertozzi and Dr. Jason Rush for authentic formylglycine.

366 8.7 References

1. Benjdia, A., Subramanian, S., Leprince, J., Vaudry, H., Johnson, M. K., and

Berteau, O. (2008) Anaerobic sulfatase-maturating enzymes – first dual substrate

radical S-adenosylmethionine enzymes, J. Biol. Chem. 283, 17815–17826.

2. Grove, T. L., Lee, K. H., St Clair, J., Krebs, C., and Booker, S. J. (2008) In vitro

characterization of AtsB, a radical SAM formylglycine-generating enzyme that

contains three [4Fe-4S] clusters, Biochemistry 47, 7523-7538.

3. Yokoyama, K., Numakura, M., Kudo, F., Ohmori, D., and Eguchi, T. (2007)

Characterization and mechanistic study of a radical SAM dehydrogenase in the

biosynthesis of butirosin, J. Am. Chem. Soc. 129, 15147-15155.

4. Benjdia, A., Leprince, J., Guillot, A., Vaudry, H., Rabot, S., and Berteau, O.

(2007) Anaerobic sulfatase-maturating enzymes: radical SAM enzymes able to

catalyze in vitro sulfatase post-translational modification, J. Am. Chem. Soc. 129,

3462.

5. Benjdia, A., Subramanian, S., Leprince, J., Vaudry, H., Johnson, M. K., and

Berteau, O. (2010) Anaerobic sulfatase-maturating enzyme--a mechanistic link

with glycyl radical-activating enzymes?, FEBS J 277, 1906-1920.

6. Booker, S. J., and Grove, T. L. (2010) Mechanistic and functional versatility of

radical SAM enzymes, F1000 Biol. Rep. 2:52.

7. Lanz, N. D., and Booker, S. J. (2012) Identification and function of auxiliary iron-

sulfur clusters in radical SAM enzymes, Biochim Biophys Acta 1824, 1196-1212.

367 8. Yokoyama, K., Ohmori, D., Kudo, F., and Eguchi, T. (2008) Mechanistic study

on the reaction of a radical SAM dehydrogenase BtrN by electron paramagnetic

resonance spectroscopy, Biochemistry 47, 8950-8960.

9. Walsby, C. J., Ortillo, D., Yang, J., Nnyepi, M. R., Broderick, W. E., Hoffman, B.

M., and Broderick, J. B. (2005) Spectroscopic approaches to elucidating novel

iron-sulfur chemistry in the "radical-SAM" protein superfamily, Inorg. Chem. 44,

727-741.

10. Dowling, D. P., Vey, J. L., Croft, A. K., and Drennan, C. L. (2012) Structural

diversity in the AdoMet radical enzyme superfamily, Biochim. Biophys. Acta

1824, 1178-1195.

11. Booker, S. J. (2009) Anaerobic functionalization of unactivated C–H bonds, Curr.

Opin. Chem. Biol. 13, 58–73.

12. Fontecave, M., Mulliez, E., and Ollagnier-de Choudens, S. (2001)

Adenosylmethionine as a source of 5'-deoxyadenosyl radicals, Curr. Opin. Chem.

Biol. 5, 506-511.

13. Frey, P. A., and Magnusson, O. T. (2003) S-Adenosylmethionine: a wolf in

sheep's clothing, or a rich man's adenosylcobalamin?, Chem. Rev. 103, 2129-

2148.

14. Challand, M. R., Driesener, R. C., and Roach, P. L. (2011) Radical S-

adenosylmethionine enzymes: mechanism, control and function, Nat. Prod. Rep.

28, 1696–1721.

368 15. Hiscox, M. J., Driesener, R. C., and Roach, P. L. (2012) Enzyme catalyzed

formation of radicals from S-adenosylmethionine and inhibition of enzyme

activity by the cleavage products, Biochim Biophys Acta 1824, 1165-1177.

16. Berteau, O., Guillot, A., Benjdia, A., and Rabot, S. (2006) A new type of bacterial

sulfatase reveals a novel maturation pathway in prokaryotes, J. Biol. Chem. 281,

22464-22470.

17. Fang, Q., Peng, J., and Dierks, T. (2004) Post-translational formylglycine

modification of bacterial sulfatases by the radical S-adenosylmethionine protein

AtsB, J. Biol. Chem. 279, 14570-14578.

18. Hanson, S. R., Best, M. D., and Wong, C.-H. (2004) Sulfatases: structure,

mechanism biological activity, inhibition, and synthetic utility, Angew. Chem. Int.

Ed. 43, 5736-5763.

19. Parenti, G., Meroni, G., and Ballabio, A. (1997) The sulfatase gene family, Curr.

Opin. Genet. Dev. 7, 386-391.

20. Schirmer, A., and Kolter, R. (1998) Computational analysis of bacterial sulfatases

and their modifying enzymes, Chem. Biol. 5, R181-R186.

21. von Figura, K., Schmidt, B., Selmer, T., and Dierks, T. (1998) A novel protein

modification generating an aldehyde group in sulfatases: its role in catalysis and

disease, Bioessays 20, 505-510.

22. Boltes, I., Czapinska, H., Kahnert, A., von Bülow, R., Dierks, T., Schmidt, B.,

von Figura, K., Kertesz, M. A., and Usón, I. (2001) 1.3 Å structure of

arylsulfatase from Pseudomonas aeruginosa establishes the catalytic mechanism

of sulfate ester cleavage in the sulfatase family, Structure 9, 483-491.

369 23. Lukatela, G., Krauss, N., Theis, K., Selmer, T., Gieselmann, V., von Figura, K.,

and Saenger, W. (1998) Crystal structure of human arylsulfatase A: the aldehyde

function and the metal ion at the active site suggest a novel mechanism for sulfate

ester hydrolysis, Biochemistry 37, 3654-3664.

24. Recksiek, M., Selmer, T., Dierks, T., Schmidt, B., and von Figura, K. (1998)

Sulfatases, trapping of the sulfated enzyme intermediate by substituting the active

site formylglycine, J. Biol. Chem. 273, 6096-6103.

25. Cosma, M. P., Pepe, S., Annunziata, I., Newbold, R. F., Grompe, M., Parenti, G.,

and Ballabio, A. (2003) The multiple sulfatase deficiency gene encodes an

essential and limiting factor for the activity of sulfatases, Cell 113, 445-456.

26. Dierks, T., Dickmanns, A., Preusser-Kunze, A., Schmidt, B., Mariappan, M., von

Figura, K., Ficner, R., and Rudolph, M. G. (2005) Molecular basis for multiple

sulfatase deficiency and mechanism for formylglycine generation of the human

formylglycine-generating enzyme, Cell 121, 541-552.

27. Dierks, T., Schmidt, B., Borissenko, L. V., Peng, J., Preusser, A., Mariappan, M.,

and von Figura, K. (2003) Multiple sulfatase deficiency is caused by mutations in

the gene encoding the human Cα-formylglycine generating enzyme, Cell 113,

435-444.

28. Carlson, B. L., Ballister, E. R., Skordalakes, E., King, D. S., Breidenbach, M. A.,

Gilmore, S. A., Berger, J. M., and Bertozzi, C. R. (2008) Function and structure

of a prokaryotic formylglycine-generating enzyme, J. Biol. Chem. 283, 20117-

20125.

370 29. Schmidt, B., Selmer, T., Ingendoh, A., and von Figura, K. (1995) A novel amino

acid modification in sulfatases that is defective in multiple sulfatase deficiency,

Cell 82, 271–278.

30. Benjdia, A., Deho, G., Rabot, S., and Berteau, O. (2007) First evidences for a

third sulfatase maturation system in prokaryotes from E-coli aslB and ydeM

deletion mutants, FEBS Lett. 581, 1009-1014.

31. Grove, T. L., Ahlum, J. H., Sharma, P., Krebs, C., and Booker, S. J. (2010) A

consensus mechanism for radical SAM-dependent dehydrogenation? BtrN

contains two [4Fe–4S] clusters, Biochemistry 49, 3783–3785.

32. Iwig, D. F., and Booker, S. J. (2004) Insight into the polar reactivity of the onium

chalcogen analogues of S-adenosyl-L-methionine, Biochemistry 43, 13496-13509.

33. Cicchillo, R. M., Iwig, D. F., Jones, A. D., Nesbitt, N. M., Baleanu-Gogonea, C.,

Souder, M. G., Tu, L., and Booker, S. J. (2004) Lipoyl synthase requires two

equivalents of S-adenosyl-L-methionine to synthesize one equivalent of lipoic

acid, Biochemistry 43, 6378-6386.

34. Lanz, N. D., Grove, T. L., Gogonea, C. B., Lee, K. H., Krebs, C., and Booker, S.

J. (2012) RlmN and AtsB as models for the overproduction and characterization

of radical SAM proteins, Methods Enzymol. 516, 125-152.

35. Beinert, H. (1978) Micro methods for the quantitative determination of iron and

copper in biological material, Methods Enzymol. 54, 435-445.

36. Beinert, H. (1983) Semi-micro methods for analysis of labile sulfide and of labile

sulfide plus sulfane sulfur in unusually stable iron-sulfur proteins, Anal. Biochem.

131, 373-378.

371 37. Kennedy, M. C., Kent, T. A., Emptage, M., Merkle, H., Beinert, H., and Münck,

E. (1984) Evidence for the formation of a linear [3Fe-4S] cluster in partially

unfolded aconitase, J. Biol. Chem. 259, 14463-14471.

38. Bradford, 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.

39. Cicchillo, R. M., Lee, K. H., Baleanu-Gogonea, C., Nesbitt, N. M., Krebs, C., and

Booker, S. J. (2004) Escherichia coli lipoyl synthase binds two distinct [4Fe-4S]

clusters per polypeptide, Biochemistry 43, 11770-11781.

40. Cicchillo, R. M., Baker, M. A., Schnitzer, E. J., Newman, E. B., Krebs, C., and

Booker, S. J. (2004) Escherichia coli L-serine deaminase requires a [4Fe–4S]

cluster in catalysis, J. Biol. Chem. 279, 32418-32425.

41. Jenkins, C. M., and Waterman, M. R. (1998) NADPH–flavodoxin redutase and

flavodoxin from Escherichia coli: Characteristics as a soluble microsomal P450

reductase, Biochemistry 37, 6106–6113.

42. The_UniProt_Consortium. (2007) The Universal Protein Resource (UniProt),

Nucleic Acids Res. 35, D193-D197.

43. Johnson, D. C., Unciuleac, M. C., and Dean, D. R. (2006) Controlled expression

and functional analysis of iron-sulfur cluster biosynthetic components within

Azotobacter vinelandii, J. Bacteriol. 188, 7551–7561.

44. Holm, R. H., and Ibers, J. A. (1977) Synthetic analogues of the active sites of

iron–sulfur proteins, In Iron-Sulfur Proteins (Lovenberg, W., Ed.), Academic

Press, New York.

372 45. Cosper, M. M., Jameson, G. N. L., Hernández, H. L., Krebs, C., Huynh, B. H.,

and Johnson, M. K. (2004) Characterization of the cofactor composition of

Escherichia coli biotin synthase, Biochemistry 43, 2007-2021.

46. Krebs, C., Henshaw, T. F., Cheek, J., Huynh, B. H., and Broderick, J. B. (2000)

Conversion of 3Fe–4S to 4Fe–4S clusters in native pyruvate formate-lyase

activating enzyme: Mössbauer characterization and implications for mechanism,

J. Am. Chem. Soc. 122, 12497-12506.

47. Ghosh, D. (2007) Human sulfatases: a structural perspective to catalysis, Cell.

Mol. Life Sci. 64, 2013-2022.

48. Dierks, T., Lecca, M. R., Schlotterhose, P., Schmidt, B., and von Figura, K.

(1999) Sequence determinants directing conversion of cysteine to formylglycine

in eukaryotic sulfatases, EMBO J 18, 2084–2091.

49. Cosper, N. J., Booker, S. J., Ruzicka, F., Frey, P. A., and Scott, R. A. (2000)

Direct FeS cluster involvement in generation of a radical in lysine 2,3-

aminomutase, Biochemistry 39, 15668-15673.

50. Eidsness, M. A., Scott, R. A., Prickril, B. C., DerVartanian, D. V., Legall, J.,

Moura, I., Moura, J. J. G., and Peck, H. D. J. (1989) Evidence for selenocysteine

coordination to the active site nickel in the [NiFeSe]hydrogenases from

Desulfovibrio baculatus, Proc. Natl. Acad. Sci. U S A 86, 147–151.

51. Shokes, J. E., Duin, E. C., Bauer, C., Jaun, B. H., R., Koch, J., and Scott, R. A.

(2005) Direct interaction of coenzyme M with the active-site Fe–S cluster of

heterodisulfide reductase, FEBS Lett. 579, 1741–1744.

373 52. Ruszczycky, M. W., Choi, S. H., and Liu, H. W. (2010) Stoichiometry of the

redox neutral deamination and oxidative dehydrogenation reactions catalyzed by

the radical SAM enzyme DesII, J. Am. Chem. Soc. 132, 2359-2369.

53. Benjdia, A., Leprince, J., Sandström, C., Vaudry, H., and Berteau, O. (2009)

Mechanistic investigations of anaerobic sulfatase-maturating enzyme: Direct Cβ

H-atom abstraction catalyzed by a radical AdoMet enzyme, J. Am. Chem. Soc.

131, 8348–8349.

– 54. Mayhew, S. G. (1978) The redox potential of dithionite and SO2 from

equilibrium reactions with flavodoxins, methyl viologen and hydrogen plus

hydrogenase, Eur. J. Biochem. 85, 535–547.

55. Demick, J. M., and Lanzilotta, W. N. (2011) Radical SAM activation of the B12-

independent glycerol dehydratase results in formation of 5'-deoxy-5'-

(methylthio)adenosine and not 5'-deoxyadenosine, Biochemistry 50, 440–442.

56. Yu, L., Blaser, M., Andrei, P. I., Pierik, A. J., and Selmer, T. (2006) 4-

Hydroxyphenylacetate decarboxylases: Properties of a novel subclass of glycyl

radical enzymes systems, Biochemistry 45, 9584–9592.

57. Banerjee, R., and Ragsdale, S. W. (2003) The many faces of vitamin B12:

catalysis by cobalamin-dependent enzymes, Annu. Rev. Biochem. 72, 209-247.

58. Haft, D. H., and Basu, M. K. (2011) Biological systems discovery in silico:

radical S-adenosylmethionine protein families and their target peptides for

posttranslational modification, J. Bacteriol. 193, 2745-2755.

374 59. Haft, D. H., Selengut, J. D., Richter, R. A., Harkins, D., Basu, M. K., and Beck,

E. (2013) TIGRFAMS and genome properties in 2013, Nuc. Acids Res. 41, D387–

D395.

Appendix A

EPR Characterization of Radical Species Trapped in the Radical SAM dehydrogenases anSMEcpe and AtsB

376 A.1 Abstract

This section will outline the radical species that have been isolated in the reaction of two anaerobic Sulfatase maturating enzymes (anSMEs), anSMEcpe and AtsB. AtsB and anSMEcpe catalyze the oxidation of a target seryl or cysteinyl residue within the active site of their respective sulfatases (1-3). These reactions are believed to occur, in vivo, during transitional of the sulfatase as the final folded form of the protein is not a substrate for these enzymes. However, the reactions can be reconstituted in vitro with the use of small 18mer peptide substrates. Under normal reaction conditions oxidation reactions do not build up radical intermediates. This section will detail the strategies used to trap various radical intermediate in the reactions of AtsB and anSMEcpe.

377 A.2 Introduction

Since Schirmer et al first proposed that anSMEs contain multiple Fe/S clusters by sequence comparison, the question has remained what is the role of these additional clusters (4). In the foregoing chapters of this dissertation, AtsB and anSMEcpe have been unequivocally shown to contain three [Fe–4S] clusters that are indispensable for reactivity (3, 5). Any disruption of the clusters results in essentially no reaction.

Grove et al proposed that the additional clusters were required to remove the electron out of the active site, as with the two extra clusters acting as an obligatory sink

(3, 5). This was proposed to also facilitate electron transfer to an oxidant such as

Flavodoxin. A recent crystal structure of anSMEcpe has provided additional support for this function, but direct interaction of the clusters with chemistry has remained elusive

(6). A pathway to provide support for this mechanism would utilize radical intermediates coupled with isotopic labeling of the protein clusters, but as no radical under turnover conditions is found and perturbation of the cluster results in loss of activity this route has not been pursued.

Herein, we employ several peptide substrate analogues that act as substrates in the reactions of AtsB and, even more so, anSMEcpe. The radical species that are trapped have been characterized to different levels. These radical intermediates may help interrogate the role of the additional clusters.

378 A.3 Materials and Methods

Wt anSMEcpe and AtsB were purified as described in chapters 6 and 8. Peptides were synthesized and purified as described in chapter 8. The following peptides were used as substrates to trap radical species in the reactions:

Kp18allylgly

H H H

H H Ac-YYTSPM--APARSMLLTGN Kp18MeCys

CH3 H S H Ac-YYTSPM--APARSMLLTGN Kp18Thr

HO CH3 H Ac-YYTSPM--APARSMLLTGN

379 A.4 Results

M 1 2

Figure A-2. SDS-PAGE anlaysis of AtsB reaction with Kp18Allylgly before (lane 1) after 20 min (lane 2). The band appears to shift upward indicating a cross-linked intermediate may form (red arrows).

380

tlg022608_05 AtsBWt Allyl Radical 5mW tlg011810_28 CpWt Allyl Radical 5mW

0.2

0.0 dX"/dH

-0.2

3320 3340 3360 3380 3400 Magnetic field, G

Figure A-3. EPR of AtsB and anSMEcpe with Kp18Allyl peptide after 1 min. The spectra were recorded a 77K.

381

tlg122309_16 Cp0635 Wt 57Fe w/ Allyl peptide (5mW) tlg011810_28 Cp0635 Wt 56Fe w/ Allyl peptide (5mW) 0.4

0.2

0.0 Intensity

-0.2

-0.4

3340 3360 3380 3400 3420G Field (G)

Figure A-4. Peptides EPR of anSMEcpe labeled with 56Fe or 57Fe after reaction with Kp18Allyl peptide for 1 min. The spectra were recorded a 77K.

382

0.20 tlg122309_21 anSMEcpe, SAM, MeCys peptide tlg122309_23 anSMEcpe, SAM, Cp peptide 0.15

0.10

0.05

0.00

-0.05

-0.10

-0.15

3320 3340 3360 3380 3400 3420 G

Figure A-5. EPR of anSMEcpe with Kp18MeCys or Cp18Cys peptide after 30 s. The spectra were recorded a 77K.

383

100 < 46 Da 30

25 80

20 60

Intensity 40 10

20 5

0 0 980 1000 1020 1040 1060 1080 1100 m/z

384

44.0 Cp0635 Wt w/ SAM, Kp18Thr

43.5 ] 3 43.0

42.5

42.0 Intensity [10 Intensity

41.5 80 K 41.0 3340 3360 3380 3400 3420 Gauss

Figure A-7. EPR of anSMEcpe with Kp18Thr peptide after 1 min. The spectra were recorded at 80K.

385 A.5 Conclusions

As is evident, several substrates are capable of producing radical species in the reaction with AtsB and anSMEcpe. The allyl-radical that is generated in AtsB seems to react with the protein to generate a covalent species, but this is yet to be proven. In addition, labeling of the proteins with 57Fe does not have a large effect on the line width of the radical. This indicates that the radicals are not in very close proximity to the cluster species as originally thought. Interestingly, the Kp18MeCys substrate produces the largest concentration of radical but also acquires a modification that increases its mass by 46 Da. This increase corresponds to a methyl-thio group. Lastly, the Kp18Thr peptide forms a radical intermediate as it was shown in chapter 8 this substrate forms the keto-product. These results lay the foundation for exploration of the anSMEs mechanism.

386 A.6 References

1. Benjdia, A., Leprince, J., Guillot, A., Vaudry, H., Rabot, S., and Berteau, O.

(2007) Anaerobic sulfatase-maturating enzymes: radical SAM enzymes able to

catalyze in vitro sulfatase post-translational modification, J Am Chem Soc 129,

3462-3463.

2. Benjdia, A., Subramanian, S., Leprince, J., Vaudry, H., Johnson, M. K., and

Berteau, O. (2008) Anaerobic sulfatase-maturating enzymes – first dual substrate

radical S-adenosylmethionine enzymes, J. Biol. Chem. 283, 17815–17826.

3. Grove, T. L., Lee, K. H., St Clair, J., Krebs, C., and Booker, S. J. (2008) In vitro

characterization of AtsB, a radical SAM formylglycine-generating enzyme that

contains three [4Fe-4S] clusters, Biochemistry 47, 7523-7538.

4. Schirmer, A., and Kolter, R. (1998) Computational analysis of bacterial sulfatases

and their modifying enzymes, Chemistry and Biology (London) 5, R181-R186.

5. Grove, T. L., Ahlum, J. H., Sharma, P., Krebs, C., and Booker, S. J. (2010) A

consensus mechanism for radical SAM-dependent dehydrogenation? BtrN

contains two [4Fe–4S] clusters, Biochemistry 49, 3783–3785.

6. Goldman, P. J., Grove, T. L., McLaughlina, M. M., Sites, L. A., .Booker, S. J.,

and Drennan, C. L. (2013) X-ray structure of an AdoMet radical activase reveals

an anaerobic solution for formylglycine posttranslational modification, Proc Natl

Acad Sci U S A xx, xx.

VITA

Tyler L. Grove

Education PhD, 2013 Department of Chemistry, The Pennsylvania State University, University Park, PA Advisor: Dr. Squire J. Booker Bachelor of Science, 2006 in Chemistry, Shepherd University, Shepherdstown, WV Advisors: Professors. Eugene Volker and Dan DiLella Publications 11. Peter J. Goldman, Tyler L. Grove, Martin M. McLaughlin, Lauren A. Site, Squire J .Booker, and Catherine L. Drennan. X-ray structure of an AdoMet radical activase reveals an anaerobic solution for formylglycine posttranslational modification. PNAS. Accepted April 2013 10. Tyler L. Grove, Jessica A. Ahlum, Rosie M. Qin, Nicholas D. Lanz, Matthew I. Radle, Carsten Krebs, and Squire J. Booker. Further Characterization of Cys-Type and Ser-Type Anaerobic Sulfatase Maturating Enzymes Suggests a Commonality in Mechanism of Catalysis. Biochemistry. Accepted March 2013 9. Tyler L. Grove, Jovan Livada, Michael T. Green, Squire J. Booker, and Alexey Silakov. A Kinetically Competent Substrate Radical Intermediate in Catalysis by the Antibiotic Resistance Protein Cfr. Nature Chemical Biology. Accepted January 2013 8. Nicholas D. Lanz, Tyler L. Grove, Camelia Baleanu Gogonea, Kyung-Hoon Lee, Carsten Krebs, Squire J. Booker. RlmN and AtsB as Models for the Overproduction and Characterization of Radical SAM Proteins. Methods in Enzymology. Vol 516, 2012, Pages 125–152 7. Tyler L. Grove, Radle Matthew I., Krebs Carsten, Booker Squire J. Cfr and RlmN Contain a Single [4Fe-4S] Cluster, which Directs Two Distinct Reactivities for S-Adenosylmethionine: Methyl Transfer by SN2 Displacement and Radical Generation. Journal of the American Chemical Society. 2011 Sep 14. 6. Boal Amie K., Tyler L. Grove, McLaughlin Monica I., Yennawar Neela H., Booker Squire J., Rosenzweig Amy C.. Structural basis for methyl transfer by a radical SAM enzyme. Science. 2011 May 27;332(6033):1089-92 5. Tyler L. Grove, Benner Jack S., Raddle, Matthew, Ahlum Jessica H., Landgraff, Bradley, Krebs Carsten, Booker Squire J. A Radically Different Mechanism for S-adenosylmethionine-dependent Methyltransferases Exhibited by the Antibiotic Resistance Protein Cfr and its Homologue RlmN. Science. 2011 Apr 29;332(6029):604-7. 4. Booker Squire J., Tyler L. Grove Mechanistic and functional versatility of radical SAM enzymes. F1000 Biology Reports 2010, 2:52 3. Grove Tyler L., Ahlum Jessica H., Sharma Priya, Krebs Carsten and Booker Squire J. A General Cofactor Requirement for Radical S adenoyslmethionine-dependent Dehydrogenases? BtrN contains two [4Fe–4S] Clusters. Biochemistry 2010,49, 3783-3785. 2. Tyler L. Grove; Lee, Kyung-Hoon; St. Clair, Jennifer; Krebs, Carsten; Booker, Squire J. In Vitro Characterization of AtsB, a Radical SAM Formylglycine-Generating Enzyme That Contains Three [4Fe-4S] Clusters. Biochemistry. 2008 Jul 15;47(28):7523-38. 1. Booker, Squire J.; Cicchillo, Robert M.; Tyler L. Grove Self-sacrifice in radical S-adenosylmethionine proteins. Current Opinion in Chemical Biology. 2007 Oct;11(5):543-52. Awards PSU Roberts Fellowship 2005-2006 PSU Braucher Fellowship 2008-2009 PSU Millers Fellowship 2009-2010 PSU Braucher Fellowship 2011-2012 PSU Alumni Association Dissertation Award 2012-2013