1 Biological Synthesis and Transformation of Nitriles

Biological Synthesis and Transformation of Nitriles

Item Type text; Electronic Dissertation

Authors Nelp, Micah

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Download date 10/10/2021 06:33:53

Link to Item http://hdl.handle.net/10150/621560

BIOLOGICAL SYNTHESIS AND TRANSFORMATION OF NITRILES

Micah Taylor Nelp

______

A Dissertation Submitted to the Faculty of the

DEPARTMENT OF CHEMISTRY AND BIOCHEMISTRY

In Partial Fulfillment of the Requirements for the Degree of

DOCTOR OF PHILOSOPHY

WITH A MAJOR IN CHEMISTRY

In the Graduate College

THE UNIVERSITY OF ARIZONA

2016

As members of the Dissertation Committee, we certify that we have read the dissertation prepared by Micah Taylor Nelp entitled Biological Synthesis and Transformation of Nitriles and recommend that it be accepted as fulfilling the dissertation requirement for the Degree of Doctor of Philosophy

______Date: April 1, 2016 Vahe Bandarian

______Date: April 1, 2016 Matthew Cordes

______Date: April 1, 2016 Indraneel Ghosh

______Date: April 1, 2016 John Jewett

______Date: April 1, 2016 Elisa Tomat

Final approval and acceptance of this dissertation is contingent upon the candidate’s submission of the final copies of the dissertation to the Graduate College.

I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement.

______Date: April 1, 2016 Dissertation Director: Indraneel Ghosh

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

SIGNED: Micah Taylor Nelp

TABLE OF CONTENTS

LIST OF FIGURES ...... 5

ABSTRACT ...... 6

CHAPTER I: A Review of Nitrile Biosynthesis and Degradation...... 8

CHAPTER 2: Summary of the Present Study...... 32

CHAPTER 3: Characterizing the Post-Translational Synthesis of the Active Site Complex of Nitrile Hydratase ...... 43

CHAPTER 4: Development of an Assay to Determine the Means of Metal Specificity in Nitrile Hydratase ...... 66

REFERENCES ...... 74

Appendix A: A Single Enzyme Transforms a Carboxylic Acid into a Nitrile through and Amide Intermediate ...... 83

Appendix B: The Alpha Subunit of Nitrile Hydratase is Sufficient for Catalytic Activity and Post-Translational Modification ...... 101

Appendix C: A Protein Derived Oxygen is the Source of the Amide Oxygen of Nitrile Hydratase ...... 119

List of Figures

Figure 1.1 A summary of the enzymatic pathways herein described for the synthesis and transformation of nitriles. 9

Figure 1.2 The biosynthetic pathway for the production of cyanogenic glycosides. 10

Figure 1.3 Mechanism of aldoxime dehydratase. 13

Figure 1.4 Mechanism of nitrile synthetase, ToyM. 16

Figure 1.5 Mechanism of nitrilase. 19

Figure 1.6 (A) Structure of the active site of cobalt-type nitrile hydratase (PDB 1IRE). (B) Mechanism of nitrile hydratase. 22

Figure 1.7 Mechanism of rhodanese. 25

Figure 1.8 Mechanism of cyanase. 28

Figure 1.9 Mechanism of nitrile reductase. 30

Figure 2.1 The structure of cobalt type nitrile hydratase. 35

Figure 2.2 Possible mechanisms of nitrile hydratase. 39

Figure 3.1 Possible series of intermediates in the route from apoTNH to holoTNH. 48

Figure 3.2 A comparison of the increase in the A) enzyme activity, B) ratio of abundance of holo enzyme to apo enzyme, and C) absorbance at 300 nm over the course of the activation assays. 56

Figure 3.3 Mass spectrum of TNH showing multiple charge states for the subunit of interest, ToyJ, as well as ToyK. 59

Figure 3.4 Mass spectra of TNH at various times throughout the activation process. 60

Figure 3.5 UV-visible spectra of apo-TNH activation assay over time. 62

Figure 3.6 EPR spectra of apo-TNH activated in vitro compared to that of enzyme expressed in cobalt-containing media and purified in holo form. 64

Figure 4.1 Structures of A) iron and B) cobalt type nitrile hydratases with the predictive amino acid pairs highlighted in red. 67

Abstract

Nitrile-containing natural products are rare in Nature, and there have been very few studies on the mechanisms by which they are synthesized and utilized. The biosynthesis of 7-deazapurine containing natural products is a unique case whereby both formation of a nitrile and its conversion to an amide are documented. The overall theme of this work is to interrogate the biosynthesis of the nitrile intermediate in the pathway and its subsequent hydration to an amide.

The biosynthesis 7-cyano-7-deazaguanine (preQ0), the key intermediate in the biosynthesis of the hypermodified base queuosine and the toyocamycin natural product, is accomplished by preQ0 synthetase through a series of unprecedented reactions whereby the carboxylate moiety of the substrate, 7-carboxy-7-deazaguanine (CDG), is successively activated by adenylation, reacted with ammonia, and dehydrated to produce the nitrile. This one-enzyme synthesis of a nitrile is unique as the only other known route to nitriles proceeds through at least two enzymes.

Nitrile hydratases are metalloenzymes that selectively hydrate nitriles to the amide and are used industrially to produce acrylamide and nicotinamide. These enzymes use a trivalent iron or cobalt complex comprised of two backbone amidate ligands and three cysteine thiolate ligands of which two are modified to the sulfenato and sulfinato form. This work describes aspects of a particular nitrile hydratase, toyocamycin nitrile hydratase (TNH). Whereas most nitrile hydratases are heterodimeric, TNH is heterotrimeric, and yet what was discovered is that only the subunit containing the active site metal complex is required for activity. This single subunit analog of the protein was used for single turnover assays in 18O-labeled water to show with high resolution mass

spectrometry that the source of the product amide oxygen is actually the enzyme itself and likely the sulfenato ligand oxygen acting as a nucleophile. The mechanism of the active site complex synthesis is described showing that this is self-catalytic in the presence of cobalt(II) and molecular oxygen.

CHAPTER I

A Review of Nitrile Biosynthesis and Degradation

Abstract - The rarity of nitriles in biology is belied by the rich chemistry life has evolved to create and transform them. Nitrile containing compounds represent a small fraction of all those published, but reports of newly discovered nitrile containing compounds suggest the unique chemistry of this functional group has not gone unnoticed by evolution and may prove to have undiscovered importance. 1 Indeed the use of nitriles in pharmaceuticals for their incredibly small steric bulk and polarity is, like most advanced chemical inventions, probably already being used by life, and that importance is underscored by the presence of this functional group in essential modified RNA bases used in all kingdoms of life. 2,3,4 This chapter will focus on the varied means by which enzymes are utilized in biology to produce and transform nitriles.

Nitrile Biosynthesis - Of the known nitrile containing natural products, there have only been two routes to their synthesis that have been described in mechanistic detail.

The aldoxime pathway is the best characterized of those and is present in at least 2500 species of plants as well as some arthropods. 1, 5 It involves multiple enzymes and produces cyanogenic glycosides that can act as a defense mechanism. 6 These cyanogenic glycosides are normally stored in vacuoles but are released into the cytoplasm upon herbivory where they encounter glycosidases that convert them into α-hydroxy nitriles that can then spontaneously form aldehydes and toxic cyanide (Fig. 1.2). 7

Figure 1.1 A summary of the enzymatic pathways herein described for the synthesis and transformation of nitriles.

Figure 1.2. The biosynthetic pathway for the production of cyanogenic glycosides.

NH2 N N Glycosyltransferase N OH R Glu R HO R O R O Glucosidase Amino Acid Nitrile alpha-Hydroxy Nitrile Cyanogenic Glycoside

Aldoxime P450 P450 spontaneous aa Dehydratase ox

HO OH N spontaneous HO N N OH O R + R HC R O N,N-Dihydroxy Amino Acid Aldoxime Aldehyde Cyanide

These cyanogenic compounds derive from amino acids in a two-enzyme pathway in which they are hydroxylated twice to form N,N-dihydroxy-amino acids which are degraded to aldoximes and then dehydrated to the nitrile by a second enzyme. Each step in this pathway is catalyzed by cytochrome P450 type enzymes (Fig. 1.2). 8 9 In typical fashion, these cytochrome P450 enzymes hydroxylate their substrate amino acids using molecular oxygen and NADPH as a reducing equivalent.10 The first step of dihydroxylation is accomplished by P450aa which when purified and reconstituted is shown to produce the singly and doubly N-hydroxylated products. It is then thought this can spontaneously decompose to the aldoxime. 8 From plants, the enzyme required for dehydration of the aldoxime to the nitrile, called P540ox, has been purified and shown to have activity, but its reaction mechanism has not been characterized. 11

Enzymatic nitrile synthesis from aldoxime substrates, however, has been characterized with bacterial enzymes. This is accomplished by a unique heme enzyme, aldoxime dehydratase. 12, 13 This has been found among many bacteria and fungi many of which are also capable of nitrile hydration, and the pathway from amino acid to nitrile using the plant amino acid hydroxylase and bacterial aldoxime dehydratase has even been reconstituted in E. coli. 14, 15, 16

Using a unique mechanism, this histidine ligated heme iron bacterial aldoxime dehydratase binds the substrate hydroxylated nitrogen directly through the ferrous heme iron. 17 The metal then donates electrons to the nitrogen to eliminate the hydroxyl group as water. The carbon nitrogen triple bond is then formed when the α-proton is abstracted

and the nitrogen iron bond is resolved leaving the iron in the resting ferrous state (Fig.

1.3). 15

Figure 1.3. Mechanism of aldoxime dehydratase.

Bacterial pathways further metabolize these nitriles to the amide using nitrile hydratases, and they also typically possess amidases to finally produce ammonia and the corresponding carboxylic acid. 18 Some bacteria have even been observed to be able to survive with this aldoxime pathway using aldoximes as the sole source of nitrogen. 13

Plant pathways on the other hand continue in producing cyanogenic glycosides by hydroxylation at the α-position followed with protection against decomposition to cyanide and aldehyde by glycosylation of the new hydroxyl, usually with glucose (Fig.

1.2). 19 The important plant hormone indole 3-acetic acid is a derivative of indole 3- nitrile, which may be synthesized from indole 3-acetaldoxime in similar fashion as the aldoxime pathway though direct in vitro evidence is lacking. 20

The aldoxime pathway was the only known biosynthetic route to nitriles until the discovery of a nitrile synthetase used in the production of 7-dezapurines in S. rimosus. 21

This enzyme is an essential part of many pathways that produce modified RNA bases and antibiotics, all of which share the ring structure of a 7-deazapurine and the nitrile

3 intermediate, preQ0, which is derived from GTP. The core steps in these pathways leading to the synthesis of the nitrile-containing preQ0 are conserved, and it is from this point that they diverge making use of three unique nitrile transforming enzymes to modify this nitrile group into an amine, amide, or amidino groups, all of which will be discussed in detail later.

The nitrile synthetase, ToyM or QueC, catalyzes the conversion of 7-carboxy-7- deazaguanine (CDG) to 7-cyano-7-deazaguanine (preQ0), using an ammonia-derived

nitrogen, and ATP to activate the carboxylate. 21 ToyM activates the acid through adenylation allowing ammonia to attack to form the amide, 7-amido-7-deazaguanine. The nitrile is then formed through an unprecedented second activation of this new amide substrate, poising the oxygen to leave when the lone pair from the amide nitrogen forms the triple bond (Fig. 1.4). 22

Figure 1.4. Mechanism of nitrile synthetase, ToyM.

The evolution of a single enzyme capable of this dual substrate activation likely occurred through an amide synthetase gaining the ability to act upon its product amide.

This is supported by the homology of nitrile synthetase to amide synthetases such as asparagine synthase. Additionally the first half reaction from carboxylic acid to amide proceeds much more quickly than the second, from amide to nitrile, suggesting the enzyme was under pressure to optimize the amide synthetase reaction before it gained the nitrile forming function. 22

The synthesis of nitriles has only been mechanistically described in detail in two pathways, but it is known from studies of other nitrile producing organisms that these cannot account for all the known nitrile-containing natural products. Borrelidin and cyanosporoside are both nitrile-containing products derived from marine Streptomyces, and their biosynthesis suggests there are potentially many more routes to nitriles. 23, 24

The elucidation of these pathways will thus double the known mechanistic paradigms for nitrile synthesis along with those of the aldoxime and nitrile synthetase pathways to significantly advance the understanding of the origin of many nitrile products.

The operons corresponding to the biosynthetic routes to borrelidin and cyanosporoside suggest a common amine intermediate produced by pyridoxal phosphate dependent amino transferases acting upon a carboxylic acid or aldehyde carbonyl groups.

It is then hypothesized that these routes diverge in the oxidation to the nitrile; in the case of cyanosporoside through the use of a flavin dependent oxidoreductase, and in the case

of borrelidin through the use of a cytochrome P450 hydroxylase acting similarly to those that produce aldoximes in the aldoxime pathway. 23, 24

Nitrile Transformations - Nitrile compounds are typically toxic, such as the latent sources of cyanide already mentioned, but they also include many antibiotics, anticancer agents, and herbicides. 25, 26 Industrially produced nitriles pose a new challenge to biological defenses as they are usually neurotoxic, and though not lethal to bacteria, they do inhibit growth. 27, 28 Despite this, nitriles are quite resistant to biochemical transformation, passing through the body as pharmaceuticals usually unchanged. 2 This second part will focus on how organisms have evolved a variety of ways to convert nitriles.

Nitriles are often hydrolyzed to their corresponding acids through the use of nitrilases, which are widespread from bacteria, fungi, animals, to plants. 29, 30 An active site cysteine thiolate is used to nucleophilicly attack at the nitrile carbon creating a thioimidate intermediate. Ammonia is lost by addition of water, and a second addition of water resolves the enzyme bound thioester regenerating the active site thiolate and freeing the acid (Fig. 1.5). 31

Figure 1.5. Mechanism of nitrilase.

This mechanism is supported by the ability of nitrilases to accept amides as substrates. 32 Indeed some nitrilases preferentially produce the amide and can thus be considered nitrile hydratases. 33 This is presumed to be the result of abortive release after the addition of the first water, but curiously for one such nitrilase, the amide product did not act as a substrate. Nevertheless the same conserved active site cysteine is required for catalysis to produce both acid and amide products supporting the overall mechanistic paradigm of nitrilases. 34

For the majority of nitrilases and in synthetic chemistry, the hydrolysis of nitriles solely to the amide is hindered by the fact that amides are more susceptible to hydrolysis than the starting nitrile. Thus synthetically it is difficult to achieve nitrile hydration to the amide without also creating a carboxylic acid byproduct. A fascinating biological solution to this problem is the family of enzymes known as nitrile hydratases which are capable of converting a wide range of nitriles to their corresponding amides without any acid byproduct. 35, 36

Nitrile hydratases employ a metal cofactor, either trivalent iron or cobalt, bound directly to the enzyme through three cysteine thiolate sulfurs and two backbone amidate nitrogens. These enzymes make a highly unique active site complex by post- translationally modifying two of the equatorial thiolato ligands to the sufinato and sulfenato ligands, both of which are required for catalysis (Fig. 1.6A). 37, 38 This is an oxygen dependent modification, and the mechanism is still not understood. 39

The mechanism of these nitrile hydratases has been an active area of research for decades since their discovery and especially since the beginning of their use as biocatalysts on an industrial scale. 40 Recent computational and crystallographic studies have shown these enzymes likely bind the nitrile directly through the active site metal and use the sulfenato oxygen to nucleophilicly attack at the nitrile carbon. This metallocycle is then resolved by attack of water at the sulfenato oxygen allowing the original sulfenato oxygen to leave forming a bound imidate which can protonate and rearrange to the amide (Fig. 1.6B). 41 42 43 Recently this was tested using isotopically labeled water to show that after a single turnover, the amide oxygen does not derive from solvent water but rather the protein itself and thus likely the sulfenato nucleophile

(Appendix C). 44

Figure 1.6. (A) Structure of the active site of cobalt-type nitrile hydratase (PDB 1IRE).

(B) Mechanism of nitrile hydratase.

Thiocyanate hydrolases are closely related to nitrile hydratases, and they employ the same unique active site cobalt complex including the cysteine sulfurs modified to the sufinato and sulfenato ligands to hydrolyze thiocyanate into carbonyl sulfide and ammonia. 45–47 Organisms with this enzyme appear to be able to utilize the resulting carbonyl sulfide as a source of sulfide for eventual anaerobic respiration to sulfate. 45

The striking similarities between thiocyanate hydrolase and nitrile hydratase including the necessity of both the sulfenato and sulfinato ligands for catalysis, where oxidation to two sulfinato ligands abrogates activity, and the ability of thiocyanate to bind directly to the metal of thiocyante mimic complexes suggest thiocyanate hydrolases employ a mechanism very similar to nitrile hydratases. 48,49 Thiocyanate hydrolases hydrate thiocyante presumably first to the S-thiocarbamate which can then rearrange to carbonyl sulfide and ammonia.

Thiocyanate hydrolases have not been shown to be capable of hydration of nitriles, but there is evidence that nitrile hydratases can accept thiocyanates to form a different set of products, the thiolate and presumably the formamide. 50, 51 This has even been used as a colorimetric test for a cobalt-type nitrile hydratase with the substrate 2- nitro-5-thiocyanato benzoic acid where the resultant thiolate is highly colored and absorbs at 412 nm. 52 There is, however, contradictory evidence that not all nitrile hydratases are capable of such reactions. 53 The substrate binding pocket of thiocyanate hydrolase is notably different than those of nitrile hydratases, which could account for its

ability to selectively catalyze thiocyanate hydrolysis, but further experimental evidence will be required to fully describe the mechanism of these interesting enzymes. 46

Though members of the nitrile hydratase family are able to catalyze the hydration of many different nitriles, the simplest nitrile, cyanide, is notably not a substrate for nitrile hydratases but rather a potent competitive inhibitor. 54 55 56 The enzymatic degradation of cyanide is accomplished mainly by two types of enzymes in the nitrilase family. 57 Cyanide hydratases, found in fungi, convert hydrogen cyanide into formamide.

58 The mechanism appears to be that of other nitrilases based upon the presence of a strictly conserved active site cysteine residue that is essential for catalysis as well as their ability to accept longer chain nitriles as substrates. 59, 60 The other type is called cyanide dihydratase or cyanidase, which catalyzes the production of formate and ammonia. 61

Interestingly from a mechanistic standpoint, though both types are homologous to members of the nitrilase family, cyanide dihydratase is distinct from cyanide hydratase as it does not produce detectable formamide intermediate and is incapable of accepting formamide as a substrate. 62

Rhodanese is a widely distributed enzyme present in bacteria and humans and is thus likely a very ancient enzyme capable of detoxifying cyanide. 63, 64 It adds a sulfide to cyanide forming less toxic thiocyanate. An essential cysteine thiolate reacts with a sulfide donor, usually thiosulfite, to create an enzyme-bound persulfide with sulfite as a co- product. 65, 66, 67 Cyanide can then attack the newly installed sulfur atom generating thiocyanic acid and regenerating the active site thiolate (Fig. 1.7). 68

Figure 1.7. Mechanism of rhodanese.

Two other mechanistic routes to cyanide metabolism take advantage of the unique chemical properties of cyanide distinct from other nitriles. The nucleophilicity of cyanide is used to create β-cyano alanine from cysteine by the plant enzyme β-cyanoalanine synthase, which is pyridoxal phosphate dependent. 69 This enzyme displaces the thiol group of cysteine with the cyanide producing hydrogen sulfide as a co-product. 70 The resultant nitrile is then amenable to metabolism by so called β-cyanoalanine nitrilases, though one of the major products is actually the amide, asparagine. 71, 72 Cyanide is also distinct from other nitriles in its ability to act as a substrate for nitrogenase. 73 Ammonia and methane are produced as the six electron reduced product as well as methylamine reduced with four electrons. 74 Though this may be thought of as a side reaction for the enzyme named for nitrogen reduction, organisms are able to survive using this nitrogenase activity with cyanide as the only nitrogen source, and bioremediation efforts to remove cyanide have even been tested using bacterial strains that have particularly high levels of nitrogenase production and cyanide resistance. 75. 76

The preceding descriptions of cyanide metabolism may just be a fraction of what is used in biology to mitigate the toxicity of cyanide. From studies on bacteria using whole cell lysate, there is strong evidence of cyanide oxygenase and dioxygenase activity. 77, 78 Details on which enzymes are catalyzing these reactions and their mechanisms should prove to be fertile ground for research.

Cyanate degradation is yet another example of unique chemistry employed by enzymes to degrade a carbon-nitrogen triple bond. These plant and bacterial cyanases

produce carbamate which breaks down spontaneously to ammonia and carbon dioxide. 79,

80 But curiously, this reaction does not proceed by simple hydration. Instead the enzyme makes use of bicarbonate as a substrate to attack the cyanate creating carbon dioxide and carbamate, which degrades to a second carbon dioxide and ammonia (Fig. 1.8). 81

Figure 1.8. Mechanism of cyanase.

With the exception of enzymatic degradations of cyanide and cyanate, the nitrile transforming reactions described thus far have only been that of hydration or hydrolysis.

The repertoire of biochemistry is not likely to be so limited, and two examples have been discovered to date that expand this to reduction to the amine and the addition of ammonia to form the amidino functional group. Both of these, nitrile reductase and archaeosine synthase, are found in the 7-deazapurine pathway and utilize the nitrile, preQ0, produced by nitrile synthetase.

In eubacteria, the nitrile substrate preQ0 is reduced to the primary amine, preQ1, and used with further modification to become the hypermodified RNA base queuosine. 3

This novel transformation was found to use NADPH as the reducing co-substrate. An essential active site cysteine residue is thought to add to the nitrile to form an enzyme- bound thiol imine, which is the direct substrate for reduction to the tetrahedral enzyme bound amine that is further reduced to the primary amine in a second NADPH reduction that frees the cysteine thiol (Fig. 1.9). 82, 83, 84 This mechanism bears a striking resemblance to that of nitrilases wherein a cysteine thiol attacks the nitrile carbon to make an enzyme bound thiol imine that is more susceptible to attack by water than the nitrile, but in this case the enzyme is able to use this more reactive intermediate to catalyze reduction. The similarity in mechanism appears to be convergent as nitrile reductase is not related to other nitrilases but rather to the nucleotide binding family of

GTP cyclohydrolases. 85

Figure 1.9. Mechanism of nitrile reductase.

Archaeosine is a modified RNA base found in all archaebacteria. 86 It is synthesized from the nitrile-containing RNA base, preQ0, that is already inserted into tRNA using a tRNA guanine transglycosylase. 87,88 Careful mechanistic studies were done to show that ammonia; from glutamine, asparagine, or simply authentic ammonia; is added into the nitrile without first hydrolyzing it to the amide and that this could be accomplished without any activating molecules such as ATP. 89

CHAPTER 2

Summary of the Present Study

In accordance with the University of Arizona Graduate College dissertation formatting guidelines, the appendices include three published papers, which represent the major original work and discoveries in this dissertation, and their summaries including the contributions of co-authors are included here.

Appendix A is a description of studies done on nitrile synthetase. Before this work it was only known that nitrile synthetase uses a carboxylic acid, ammonia, and adenosine triphosphate as substrates to produce the product nitrile. To characterize this further, the substrate, 7-carboxy -7-deazaguanine (CDG) was synthesized and incubated with the enzyme, also known as ToyM. Using LC-MS, an intermediate amide was identified, and this 7-amido-7-deazaguanine, or ADG, was synthesized and proven to be competent for the production of the nitrile without the addition of ammonia. The nature of activation was explored where adenosine monophosphate was identified as a co- product suggesting adenylation of the substrate carboxylic acid. This was supported with a pyrophosphate detection assay that showed both ADG and CDG incubated with ToyM produce pyrophosphate as a co-product, which showed both the substrate acid and intermediate amide are activated in the same way with adenylation. The adenylated acid intermediate was isolated and characterized in the absence of added ammonium salts to the assays. This was done using LC-MS and LC-MSMS with labeled natural abundance and uniformly labeled 13C adenosine triphosphate that produced the expected mass shifts.

Finally, evidence from the rates of the two reactions, using CDG and ammonia or ADG

alone, was used to speculate that this enzyme may have evolved from an amide synthetase that gained the ability to carry out a similar reaction on its product amide, poising it for dehydration to the nitrile. This was thought to be so because the enzyme catalyzes the more complex condensation of ammonia, CDG, and ATP more quickly than the relatively simpler activation of ADG with ATP suggesting the enzyme has been optimized for the first reaction.

Since their discovery nearly forty years ago, nitrile hydratases have been used as one of the only biocatalysts in industry where they produce kilotons of acrylamide and methacrylamide each year from acylonitrile and methacrylonitrile respectively. 40, 36 The utility of nitrile hydratases lies in their ability to hydrolyze nitriles under neutral, aqueous conditions to the corresponding amides using a relatively simple iron or cobalt cofactor.

As a result, both the mechanism of hydration and that of post-translational active site synthesis have been the subject of a great deal of inquiry.

The hydrolysis of nitriles to amides and then to carboxylic acids is often achieved using acid or base catalysts under reflux. Other methods using hydrogen peroxide have also been widely used. 90 The hydrolysis solely to the amide, however, is made difficult by the similar reactivities of amides and nitriles towards hydrolysis resulting in a mixture of amide and acid products. The enzymatic conversion of nitriles to amides without acid byproducts has thus been of enormous benefit. The main biocatalysts used for this are nitrile hydratases, which were discovered in bacteria for their ability to hydrate

acetonitrile. 40 And they were quickly thereafter used industrially to produce acrylamide and nicotinamide. 91

Nitrile hydratases employ a low spin non-corrin cobalt or non-heme iron active site complex comprised of two peptide backbone amidate ligands and two cysteine sulfur ligands oxygenated to the sulfenic and sulfinic acids which together serve as the equatorial ligands. A third unmodified cysteine thiol acts as an axial ligand with the other site occupied by a solvent derived water (Fig. 2.1). 37, 38 This unprecedented complex is shared only with the closely related thiocyanate hydrolase. 46

Figure 2.1 The structure of cobalt type nitrile hydratase is shown with the full protein, a zoomed in region near the active site complex, and a schematic view of that same complex.

These enzymes are typically multimeric containing an α and β subunit that come together just around the active site metal complex. Appendix B describes work done on a single-subunit nitrile hydratase. Toyocamycin nitrile hydratase (TNH) is a unique three- subunit member of the nitrile hydratase family. The α subunits contain the active site complex, and the β subunits reside just above that complex with multiple conserved residues within hydrogen bonding distance of the modified cysteine ligands suggesting a key role in catalysis and possibly active site complex formation. Within TNH, the α subunit, ToyJ, is related to other nitrile hydratase α subunits, and the β and γ subunits are related to the N- and C- terminal halves of what is the single β subunit in other nitrile hydratases. It was discovered in the course of various experiments to add a hexahistidine tags to these three subunits that the α subunit can be expressed in soluble form without co-expression of the other two subunits, and remarkably this protein was the same orange color as the subunit-replete enzyme showing the active site complex was likely present.

Metals analysis with inductively coupled plasma emission spectroscopy showed this single subunit, ToyJ, possessed a stoichiometric amount of cobalt. This was confirmed to be the case using high resolution mass spectrometry that showed the addition of one cobalt and three oxygen atoms consistent with a fully mature post-translationally modified complex. Electron paramagnetic resonance experiments showed that this cobalt was in the low spin d6 state consistent with subunit replete nitrile hydratases. Having proved that ToyJ is sufficient for active site synthesis, the activity of this enzyme in hydrating the substrate toyocamycin was tested. Remarkably ToyJ is active without the β and γ subunits. The role of these two additional subunits, without which the enzyme is clearly capable of nitrile hydration, was then analyzed. Comparing the activity of ToyJ

and ToyJKL with the natural substrate toyocamycin and a new substrate 3-cyanopyridine showed that these additional subunits, ToyKL, impart substrate specificity and catalytic efficiency. For this, Dr. Linda Breci worked on the mass spectrometry on ToyJ, and Dr.

Andrei Astashkin obtained the EPR spectra on the samples of ToyJ and ToyJKL.

The mechanism of hydration is only now being detailed experimentally, but two discoveries made major advances towards uncovering it. The first is that the enzyme is inactivated when an exogenous ligand is bound to the axial position, nitric oxide in the case of iron-type nitrile hydratases and possibly carbon monoxide in the case of cobalt nitrile hydratases, which can be reversed with visible light that cleaves the metal axial- ligand bond allowing for an exchangeable axial coordination site. 92, 93, 94 This suggests the substrate nitrile binds directly to the metal in the course of catalysis or that the axial hydroxo ligand acts as the nucleophile or that the axial hydroxo acts as a general base to activate a water molecule as the nucleophile (Fig 2.2A). The hydroxo ligand acting as a general base seemed to be most likely since the complex was thought to be exchange inert, a reasonable expectation for a low spin d6 complex, in the case of the cobalt type nitrile hydratases. However, crystallographic studies showed that isonitriles bind to the metal center, stopped flow experiments showed that addition of nitriles to nitrile hydratases caused spectroscopic changes similar to that of azide-bound nitrile hydratase mimics, and some mimic compounds were synthesized that are capable of reversibly binding nitriles directly to the low spin metal. 95, 96, 97 Additionally, electron paramagnetic resonance experiments proved that water derived oxygens can quickly exchange, and

computational studies showed this was likely the result of strong sigma donation from the axial thialoato causing an increase in the lability of the axial site trans to this. 98, 99, 100

Together these suggested an initially unexpected role for the redox inactive metal centers, that of a Lewis acid to poise a bound nitrile for nucleophilic attack (Fig. 2.2B-D).

Figure 2.2 Possible mechanisms of nitrile hydratase. Bond from ligands to the rest of the enzyme are omitted for clarity. Representative references include A) Mitra and Holz,

2007 126; B) Hashimoto et al., 2008 95; C) Hopmann, 2014 41; and D) Light et al. 2015 42.

The second major advance was the discovery that the post-translationally modified sulfenato and sulfinato ligands are required for activity. 39 The additional oxygens on the modified cysteine residues may be acting to withdraw electron density from the active site, increasing the Lewis acidity of the metal ion. 101 Studies on a series of (cysteinato-N,S)bis(ethylenediamine)cobalt(III) complexes where the sulfur is oxidized to the sulfenato and sulfinato forms demonstrated the higher electronegativity of the oxidized cysteine ligands and their higher field strength relative to thiolato ligands. 102 In nitrile hydratases, these oxygens are within hydrogen bonding distance of two conserved arginine side chains that further allow electron withdrawal from the metal. 103

A more important role for the sulfenato was revealed when it was discovered that if it is oxidized to a sulfinato ligand activity is abolished. 104 This was shown in exquisite detail in crystal structures of thiocyanate hydrolase where the enzyme was only active with oxidation of the correct thiolato ligand to the sulfenato form, and it also showed the enzyme lost activity when this ligand was further oxidized to the sulfinato form. 48 Thus the sulfenate was proposed to act as a general base catalyst in activating a water molecule to attack the nitrile (Fig. 2.2B).

It was overlooked that the sulfenato oxygen itself could be acting as the nucleophile. Early cobalt-sulfenato complexes demonstrated the nucleophilicity of this type of ligand, and computational studies on nitrile hydratases showed the sulfenato sulfur was likely to have a high positive charge able to support a strongly negatively charged sulfenato oxygen. 105, 106, 107 In one of the few known nitrile hydratase mimic compounds that hydrates nitriles, a sulfenato ligand is essential for this activity, and further, using labeled water, Heinrich et al. observed that exchange into this sulfenato

ligand was turnover dependent, clearly implicating the sulfenato as the nucleophile which first attacks the nitrile. 108 This was followed by evidence that the sulfenato ligand reacts with electrophilic boronic acids in an actual nitrile hydratase. 109

Computational work has recently been published that supports both of these mechanistic insights: that the nitrile substrates bind directly to the active site metal in the axial position and that the sulfenate oxygen then attacks the nitrile carbon to form a metallocycle that is then resolved by attack of water at the sulfenato sulfur or, in a more circuitous route, by the axial thiolato sulfur attacking the sulfenato sulfur after which this disulfide is resolved by attack of water on what was originally the sulfenato sulfur ligand

(Fig. 2.2C-D). 41, 42

A clearer picture emerged with a time resolved crystallographic study in which important arginine residues had been mutated to slow the catalytic turnover of the nitrile hydratase used. What this demonstrated was an unambiguous structure of a metal bound nitrile that formed a covalent bond with the sulfenato oxygen, and this was confirmed using 18O-water and FTIR spectroscopy to show that the unique sulfenato stretch was perturbed by turnover dependent exchange. 43

Concurrently, high resolution mass spectrometry experiments were done as part of this work described in Appendix C that characterized not only the enzyme in solution but also the product amide to show that an enzyme bound oxygen is the source of the amide oxygen, definitively ruling out an initial activated water nucleophile. This was done using a single subunit version of nitrile hydratase that was slow enough to allow for single turnover conditions in small volumes of 18O-labeled water. The enzyme and

substrate were mixed in 97% 18O-labeled water at various concentrations of substrate from sub-stoichiometric amount to four-hundred times that of enzyme. Mass spectrometry on the resultant products showed that after a single turnover, the vast majority of product did not contain labeled water and that only after multiple turnovers did the product incorporate the labeled oxygen from bulk solvent. Mirroring this, the protein was also characterized and showed a turnover dependent exchange of a single oxygen, all of which is consistent with the sulfenato acting as the initial nucleophile in catalysis. 44

What remains to be resolved experimentally is the way in which water is finally incorporated into the catalytic cycle, either by attack directly on the sulfenato-sulfur or whether a disulfide is formed first after which water can attack the equatorial thiolato ligand to regenerate the sulfenate. This may be attempted by the capture of the disulfide species using mass spectrometry in which substrate and enzyme are mixed just prior to infusion, though not observing this potential intermediate will not lend support to either mechanism as the metal-bound disulfide may just be too reactive to observe.

The synthesis of the active site complex in nitrile hydratases is a much less explored aspect of these enzymes. This entails the correct binding of iron or cobalt and the asymmetric oxidation of two equatorial cysteine thiols to the sulfenic and sulfinic acids. The source of the additional oxygen atoms and the mechanism of their addition to the cysteine sulfurs remains unknown, and experiments on these problems are detailed in chapter 3.

Chapter 3

Characterizing the Post-Translational Synthesis of the Active Site Complex

Abstract

The active site complex of nitrile hydratases is unique in a variety of respects: it contains two post-translationally modified cysteine thiolate ligands oxidized to the sulfinato and sulfenato forms, it involves direct ligation of the metal to two deprotonated amide nitrogens derived from the backbone of the protein, and it contains cobalt or iron in their low spin trivalent state. 37, 38 Perhaps most significantly, the synthesis of such an unprecedented complex is accomplished in a self-catalyzed series of reactions. 39, 116

Beyond this, very little is known about the mechanism of complex formation. In order to elucidate the mechanism of activation, the source of the oxygens on the modified cysteine sulfenato and sulfinato ligands must first be determined, and this chapter will described the design of experiments that will allow mass spectrometric characterization of in vitro activated nitrile hydratase to determine the source of the modified cysteine oxygen atoms using 18O-labeled water and molecular oxygen.

Introduction

The activation of nitrile hydratases has been shown in many cases to be dependent on separate activator proteins, which are specific to cobalt or iron type nitrile hydratase and are typically associated with nitrile hydratase in the same operons. 110, 111 These

activators in some cases appears to dissemble the subunits of nitrile hydratase to access the α subunits to insert the correct metal ion. 112, 113 Though these activator proteins are clearly useful in many cases in producing active, recombinant nitrile hydratases, they are not required uniformly. 114 Multiple nitrile hydratases, including both iron and cobalt types, have been reported that do not require an activator to be co-expressed heterologously in order to achieve activity. 115, 53 The possibility that native enzymes in the E. coli expression host are able to catalyze this reaction has not been excluded, but reports of purified apo-nitrile hydratases being activated by the addition of cobaltous or ferric salts supports the notion that this is a self-catalytic process. 39, 116

Somewhat ironically, the first evidence of self-catalytic synthesis of the active site came from experiments on heterologously expressed apo-enzyme attained by not coexpressing the corresponding activator. The iron-type nitrile hydratase from

Rhodococcus species N-771 was shown to be insoluble as initially purified, but after treatment with guanidinium hydrochloride and reconstitution with 10 µM ferric citrate in the presence of oxygen, the enzyme gained activity. Mass spectrometric analysis of tryptic digests of this enzyme removed during the activation process demonstrated that the presence of sulfinic acid modification was correlated with the increase in activity, and though not directly observed, the sulfenic acid modification too was associated with activity. The presence of a sulfenic acid, which is chemically unstable under these conditions, was inferred by the formation of a disulfide during the denaturation of the enzyme at low pH. Under these conditions, two thiolates could not form a disulfide, but a disulfide could readily result from reaction of the Cys-S-OH moiety with a free thiol. 39

Similar oxygen-dependent activation of an apo thermophilic nitrile hydratase in the

presence of cobalt(II) has also been observed. 37 Collectively, these experiments show conclusively that nitrile hydratases are capable of installing the active site complex without additional enzymes in an oxygen-dependent manner.

The order of synthesis of the modified amino acids was suggested by crystal structures of the closely related thiocyante hydrolase wherein the cobalt metal and sulfinic acid modification were present in crystals of the enzyme that were inactive. The protein from these crystals, however, gained activity after storage and was shown to have incorporated the sulfenic acid modification concomitantly. 48 Thus it seems likely that the metal ion binds first to three unmodified cysteine thiolate ligands and two amidate ligands, the sulfinic acid modification then forms, and lastly the sulfenic acid is formed and the metal becomes oxidized to the +3 state.

The role of residues outside the coordination sphere of the fully mature active site complex remains largely unknown. As is discussed in Chapter 4, the identity of conserved amino acid pairs just outside the active site sequence correlates with the metal ion found in the protein as follows. Threonine and tyrosine are found in cobalt-type enzymes, whereas serine and threonine are present in the iron nitrile hydratases in the sequence C(T/S)LCSC(Y/T). The interchange of these residues, however, is not sufficient to redirect metal preference suggesting that a more intricate mechanism is at play in the incorporation of the metal. 117

Nitrile hydratases typically comprise two subunits, where the residues that form the ligands for the active site derive solely from the α-subunit. 118 The active site complex of nitrile hydratases exists at the interface of the α- and β-subunits with two conserved β-

subunit-derived arginines side chains that hydrogen bond to the modified sulfenato and sulfinato ligands. 103 These and other conserved residues in the β-subunit that sit just above the active site may reasonably be expected to play a role in activation, but a study in this work proved that the β-subunit residues are not required for full activation. The α- subunit of the cobalt-type toyocamycin nitrile hydratase, when expressed in E. coli in the presence of cobalt(II) chloride, leads to an active nitrile hydratase, suggesting that neither an activator nor the β- or γ -subunits are required for activity. 53 Therefore, the sequence determinants that are required to direct the posttranslational activation of this nitrile hydratase are present in the α subunit.

Inorganic complexes that mimic the active sites of nitrile hydratases go further in showing the minimal requirements for metal binding and cysteine oxidation. The

Chottard group was able to synthesize an N2S3 iron complex that oxidized under air from the tristhiolato form to one containing a sulfinato ligand that became O-bonded. 119 This same group was also able to synthesize a cobalt complex with N2S2 ligands in the equatorial plane much like that of nitrile hydratases, and this formed the S-bonded sulfenato ligand upon addition of molecular oxygen. Remarkably this complex is catalytically active and has been shown to hydrate up to 50 nitriles to the amide per complex. 108

A similar equatorial N2S2 six-amino acid peptide mimic is also capable of binding to cobalt and oxidizing a thiolato ligand to the sulfinic acid in the presence of molecular oxygen. Studies on this process showed that the cobalt binds as cobalt(II), the sulfinato is formed, and then slowly the cobalt is oxidized to cobalt (III), as it is in the resting state of mature nitrile hydratases. 120 This is consistent with cobalt(II) activation of nitrile

hydratases, but the previously described example of activation of apo-enzyme with ferric iron seems to contradict this. It is possible however that this iron(III) is reduced to an active form upon coordination to three thiolate and two amidate ligands in the highly anionic environment of an apo-nitrile hydratase active site.

Together these data point to a mechanism involving the binding of the metal in the +2 state to the apo-enzyme; oxygen dependent oxidation of one of the equatorial thiolato ligands to the sulfinato; and then oxygen dependent oxidation of the metal to the

+3 state and the oxidation of the second thiolato ligand to the sulfenato form, the order of the last two of which is not clear (Fig. 3.1).

Figure 3.1 Possible series of intermediates in the route from apo-TNH to holo-TNH.

The nature of oxygen’s role in this process is not clear. One may plausibly invoke a role in oxidation of the active site thiolato residues, either indirectly or directly. Indeed, the sulfenato and sulfinato oxygen atoms may be derived from molecular oxygen. A well studied model for the posttranslational mechanism is the reaction catalyzed by cysteine dioxygenase (CDO). This enzyme has a non-heme iron atom coordinated by three histidine residues and two exchangeable solvent water molecules. The substrate, cysteine, can coordinate in place of the solvent ligands and undergo an oxygen-dependent conversion to cysteine sulfinate. 121 Studies with labeled oxygen have demonstrated that that both oxygen atoms in the sulfinate moiety of the product of CDO are derive from molecular oxygen. 122

As in the proposed mechanism of nitrile hydratase activation, the metal center in

CDO is in the high spin +2 oxidation state where it is then thought to bind to molecular oxygen to form the iron(III) superoxo intermediate. The distal oxygen can then add to the thiolato ligand forming an iron(IV) oxo which in turn adds in to the newly formed sulfenato ligand to form the sulfinato and iron(II). 121, 123 Thus this serves as a likely route to the intermediate sulfinato complex in nitrile hydratases observed in the peptide mimic and thiocyanate hydrolase crystal structures, but it does not explain the ultimate formation of the singly oxygenated thiolato ligand to the sulfenato.

As mentioned before, cobaltous salts can be introduced to apo-nitrile hydratases that can then self-catalytically form the active cobaltic complex only after the sulfinato is present. This suggests a one electron oxidation produces a radical co-product. One can imagine a scenario in which a second oxygen molecule binds and forms the sulfenato ligand on the second equatorial thiolato but is stopped there before proceeding to the

sulfinato at which point the remaining metal bound oxygen atom radical may be quenched leaving the metal in the +3 state.

The main objective of the studies presented here was to develop a system in which the posttranslational activation of the nitrile hydratase could be monitored. The studies in this chapter show that nitrile hydratase undergoes an oxygen-dependent process culminating in appearance of toyocamycin nitrile hydratase activity and posttranslational activation of the protein. The system can be used in the future to identify the source of the oxygen atoms using labeled molecular oxygen to probe the mechanism underlying this fascinating transformation.

Materials and Methods

Expression and Purification. Toyocamycin nitrile hydratase (TNH) was expressed and purified in the apo form by expression in the absence of cobalt(II) chloride and in the presence ethylenediaminetetraacetic acid (1 mM) to complex any trace metal ions. 53 All purification steps were carried out under anaerobic conditions in a Coy chamber (~95% nitrogen and 5% hydrogen) using the same method as previously described but with 0.5

M sodium chloride added to each buffer. 53 The enzyme eluted from the HisTrap column was concentrated with an Amicon pressurized concentrator with a 10 kDa filter and injected onto a HiPrep 16/60 Sephacryl S-200 size exclusion column (GE Healthcare) pre-equilibrated with 50 mM piperazine-N,N’-(2-ethanesulfonic acid) (PIPES) pH 7.4 and 0.5 M sodium chloride. The fractions containing the heterohexameric TNH complex were pooled and concentrated, frozen with liquid nitrogen, and stored at -80 °C.

The enzyme was purified under anaerobic conditions because the active site thiolates have been known to form disulfides in the absence of a metal ion. 117 The enzyme was exchanged into a PIPES buffer because this does not chelate metal ions well as opposed to phosphate or Tris containing buffers that might interfere with activation by binding cobalt(II).

Activation Assays. Apo-enzyme (50 µL of 200 µM, final concentration 100 µM), the concentration of which was determined by a standard Bradford assay using bovine serum albumin as a standard, was added to aerobic water (40 µL) and cobalt(II) chloride

(10 µL of 1 mM, final concentration of 100 µM). Above 100 µM CoCl2 enzyme begins to precipitate, but at 100 µM it is stable. These solutions were allowed to incubate for 2 h.

An aliquot (10 µL) was withdrawn at 0, 2.5, 5, 10, 20, 40, 80, and 120 min and analyzed by mass spectrometry for protein modification and HPLC for enzymatic activity.

Enzyme Mass Spectrometry. For mass spectrometric characterization, 10 µL of the aliquots that were removed at each time point were added to 0.5 mL of water (Fisher

Optima grade) containing 0.1% acetic acid (v/v). Buffer and salt were removed by exchanging the buffer two times using a centrifugal concentrator with a polyethersulfone support and 10 kDa cutoff. This was repeated three times, with the final resuspension occurring with 200 µL of the buffer for a final enzyme concentration of ~ 5 µM. The enzyme solution was directly infused into a Thermo Fisher LTQ Orbitrap XL mass spectrometer with typical settings as follows: injection flow rate 100 µL/min, m/z range

900 – 2000, resolution 100,000, injection time 200 ms, sheath gas flow rate 11, auxiliary

gas flow rate 3, sweep gas flow rate 38, spray voltage 3.00 kV, capillary temperature 275

°C, capillary voltage 49.00 V, tube lens 150 V, multipole offset -1 V, Lens 0 -19.50 V, multipole 0 offset -5.25 V, lens 1 -37.00 V, gate lens -56.00 V, multipole 1 offset -12.00

V, mutipole RF amplitude 400 V p-p, and front lens -6.00 V. The Xtract settings were: generate MH+ Masses Mode, resolution 100,000 at 200, S/N Threshold 10, fit factor

44%, remainder 25%, AveragineLowSulfur model, max charge 20.

Activity Assays. The activity of the enzyme during activation assays was monitored by removing 1 µL of the activation reaction and diluting it to 100 µL with 50 mM potassium phosphate pH 6.5 (1 µM enzyme). An aliquot of this dilution (5 µL) was added to a 45 µL assay solution containing 1 mM toyocamycin (Berry and Associates),

50 mM potassium phosphate pH 6.5, and a final concentration of 0.1 mM TNH. The reactions were quenched at 1 and 3 min by removing 20 µL and mixing with 4 µL 30%

(w/v) trichloroacetic acid. Conversion of toyocamycin to sangivamycin was monitored by

HPLC as described previously. 44

UV-Visible Spectroscopy of Activation of TNH. To determine the spectral changes that accompany activation of TNH, activation reactions were set up essentially as described above except that the final enzyme concentration was 25 µM in a total volume of 0.1 mL. Spectral changes were monitored from 30 s - 80 min in an HP Agilent 8453 diode array spectrophotometer.

EPR Spectroscopy of TNH. Electron paramagnetic spectra (EPR) were acquired of apo-enzyme after incubation with CoCl2. An activation solution prepared as described above was frozen with liquid nitrogen after 2 h and lyophilized to remove oxygenated solvent and so that enzyme could be resuspended in water containing 10% glycerol in the manner in which TNH was prepared for EPR analysis as previously described. 53

Determination of Disulfide Content of Enzyme. The concentration of free thiols in the apo-enzyme was quantified using the colorimetric Ellman’s reagent, 5,5’-dithiobis-(2- nitrobenzoic acid) (DTNB). A denaturing solution of anaerobic 6 M guanidinium hydrochloride, 0.1 M sodium chloride, 2.5 mM EDTA, and 50 mM Tris HCl pH 8.0 was prepared, and to 990 µL of this were added 10 µL 200 µM apo-TNH in a quartz cuvette.

A standard Bradford assay using bovine serum albumin as a standard was used to determine the concentration of enzyme. To this 20 µL of 100 mM DTNB dissolved in dimethylsulfoxide were added with pipetting to mix. After two minutes of incubation, the absorbance at 412 nm was recorded to quantify TNB, thio-(2-nitrobenzoic acid) (ε412 nm =

13,700 M-1 cm-1 in 6 M guanidinium hydrochloride). 124

Results and Discussion

In order to characterize the synthesis of the active site complex of nitrile hydratases, this process has to be carried out under controlled conditions in vitro so that labeled oxygen and water can be added selectively to determine the source of the modified cysteine ligand oxygens in the enzyme. To this end, apo-TNH was obtained in

soluble form using by a slight modification to earlier protocols for the expression and purification of holo-TNH. Specifically, the addition of EDTA to the expression culture to bind free cobalt from the Luria-Bertani media and, more obviously, simply not adding additional cobalt chloride was sufficient to prevent the in vivo synthesis of the cobalt- containing holo-TNH complex. Apo-TNH, however, is less stable during purification than holo-TNH where it was found to precipitate in the course of purification. The addition of 0.5 M sodium chloride was sufficient to prevent the apo-enzyme from precipitating under assay conditions, which is consistent with earlier studies that showed this heterohexameric enzyme is stabilized for the most part through non-polar interactions. 25

In the reconstitution experiments, apo-TNH was combined with cobalt chloride under aerobic conditions. The enzyme solution remained stable up to 0.1 mM CoCl2, but precipitation was observed at higher concentrations. Therefore, all the experiments were carried out at this limiting concentration of the metal ion. The activity of the protein during the incubation was measured by assaying TNH with toyocamycin as a substrate.

A substrate concentration of 1 mM was chosen as the holo-TNH appears to become saturated at this concentration above which (5 mM) it shows more complex behavior perhaps due to cooperativity (see Appendix B). As shown in Fig. 3.2A, activity of activated apo-TNH reaches a plateau by ~ 1 h, suggesting that the activation is complete.

The turnover number of holo-TNH under these conditions is 25 sec-1, which was determined within the linear range of activity. This can be compared to the previously reported turnover rate of the holo-TNH (175 sec-1) under the same buffer conditions, which shows this activated TNH acquires roughly 15% of the activity of holo-TNH.

Further controls using holo-TNH purified the same way in which apo-TNH was in this case and treated the same way as under activation conditions may show this difference in activity is not the result of incomplete or incorrect synthesis of the active site complex but rather some sort of inactivation due to those differing purification and activation conditions.

To rule out that the lower activity observed with the activated apo-TNH is due to formation of disulfides in the as purified enzyme, Ellman’s reagent was used to quantify the free thiols in apo-TNH under denaturing conditions. The experiments show that apo-

TNH retains ~4.9+0.4 free thiols per heterotrimer. The ToyJKL heterotrimer has 5 cysteine residues. Therefore, the lower level of activation is likely not due to oxidation of the thiol residues in the as purified enzyme.

Figure 3.2. A comparison of the increase in the A) enzyme activity, B) ratio of abundance of holo enzyme to apo enzyme, and C) absorbance at 300 nm over the course of the activation assays.

The requirement of oxygen for this process was examined by carrying out the reactions in the anaerobic chamber. In these experiments, protein was combined with cobalt (II) using all anaerobic buffers. The as-isolated protein has activity of 0.9 sec-1 at the beginning of incubation, which increases to 1.8 sec-1 after 80 min. This is in contrast

to the observation in the presence of O2, where 13-fold activation is observed in a similar timeframe. The residual activity observed at the beginning of the incubation likely reflects small amount of protein activation in the cell, and while we made every effort to exclude oxygen, it is possible that residual oxygen is responsible for the ~ 2-fold increase in activity during the incubation.

The mass spectra of the enzyme were also obtained at various time points during activation. These show ToyJ and ToyK subunits above noise at multiple charge states.

ToyK is detectable, but the optimized ionization conditions used to maximize the signal for ToyJ have reduced the ionization efficiency significantly for the ToyL subunit (Fig.

3.3). These spectra were charge deconvolved to produce the spectra in Figure 3.3, which show species corresponding to apo-ToyJ at 21,084.08 Da. This mass is consistent with post-translational processing of the Met residue in the recombinant protein obtained from

E. coli. Both the raw spectra and the deconvolutions reveal a number of other low abundance species. These may represent the binding of other metals to the active site such as iron that has carried over from the cell. These additional peaks may be the intermediates in the synthesis of the active site, but as they overlap, it is difficult to assign them to reasonable intermediates such as the complex containing just two additional oxygens or perhaps four additional oxygens that may occur as molecular oxygen binds to the sulfinato complex that has not yet formed the sulfenato ligand. While these do not

interfere with the analysis, collectively, they may explain why the activation reactions do not yield activity closer to those that are observed with holo-TNH. The Xtract program, which was used to obtain the deconvoluted spectrum, treats the charges derived from the cobalt(III) as protons and so assigns the peak in deconvolution lower by the mass of three protons or 3.0229 Da from the theoretical.

Over the course of the activation assay, the disappearance of apo-enzyme at

21,084.08 Da is concurrent with the increase in holo-enzyme at 21,187.97 Da that contains three additional oxygen atoms and a cobalt that displaces three protons. The holo peak becomes clear after about 10 min and becomes the base peak in this protein envelope (Fig. 3.4). The ionization efficiencies of each of these forms of ToyJ are not identical and so direct measurement of concentration is not possible with this technique.

Instead the increase in holo enzyme was quantified using the ratio of the abundance of the apo-TNH peak to that of the holo-TNH peak and plotted in Fig. 3.2B.

Figure 3.3 Mass spectrum of TNH showing multiple charge states for the subunit of interest, ToyJ, as well as ToyK.

ToyJ ToyK

1000 1200 1400 1600 1800 2000 m/z

Figure 3.4 Mass spectra of TNH at various times throughout the activation process. From the top are the 0, 5, 10, 20, and 80 min time points.

To obtain additional evidence for the formation of the correct holo-enzyme active site complex UV-visible spectra of the enzyme were obtained during the reconstitution.

The spectra reveal a shoulder from 300 – 430 nm, consistent with a similar feature observed in the UV-visible spectra of holo-ToyJKL and other known cobalt-type nitrile hydratases (Fig. 3.5). 125 This time-dependence of the appearance of this feature mirrors the increase in enzyme activity and the relative abundance of holo-enzyme in mass spectra (Fig. 3.2C).

Figure 3.5. UV-visible spectra of apo-TNH activation assay over time. The inset is of holo-TNH as purified for comparison.

The correct synthesis of the active site complex may also be probed using EPR spectroscopy. The holo-enzyme contains a diamagnetic low spin d6 ion which can be reduced with sodium dithionite to produce a low spin d7 complex that is EPR active. The final product of the activation assay was reduced to the Co(II) species with sodium dithionite and shows a weak signal matching that observed for holo-enzyme treated similarly, showing conclusively that the correct active site complex is formed in vitro as in vivo (Fig. 3.6). The signal to noise of the in vitro activated product is likely low as a result of the lower efficiency of activation, about 15% based upon activity assays.

This is a particularly promising result in that it shows the possibility of trapping intermediates in the synthesis of the complex which can be characterized with EPR. No such intermediate has been identified to date, but since the proffered metal is a high spin divalent cobalt and the apo-enzyme active site is presumably not yet oxidized before the metal binds and thus provides a relatively weaker field set of ligands, an early high spin cobalt complex is to be expected.

Figure 3.6. EPR spectra of apo-TNH activated in vitro compared to that of enzyme expressed in cobalt-containing media and purified in holo form. The spectrum of holo-

TNH is from the previously published EPR of this enzyme (Appendix B). Experimental conditions: microwave frequency, 9.335 GHz; microwave power, 2 mW; magnetic field modulation amplitude, 0.5 mT; temperature, 30 K.

Activated ToyJKL

g = 2

Holo ToyJKL

2000 2500 3000 3500 4000 B (Gauss) 0

These results presented in this chapter establish a series of conditions and assays that can be used to monitor the conversion of apo-TNH to holo-TNH in vitro. Based upon the activity, UV-vis, EPR, and mass spectrum data, active site complex synthesis is occurring, albeit at lower levels per enzyme than is achieved in vivo. The ability to acquire isotopically resolved mass spectra of activated holo-TNH as in these assays affords a means of testing the source of the oxygen atoms in the posttranslationally modified complex. Using labeled water and molecular oxygen, this holo-TNH envelope will have a readily detectable shift to indicate which of these oxygen sources is contributing the modified ligand oxygens. This will provide the basis for proposed mechanisms which can then be further probed in future studies.

CHAPTER 4

Towards the Development of an Assay to Determine the Means of Metal Specificity in

Nitrile Hydratase

Nitrile hydratases can be grouped into two different classes based upon their binding of either cobalt or iron in their active site complexes that are otherwise indistinguishable in the first coordination sphere. This metal binding difference is based apparently on the identity of just two active site residues that do not bind directly to the metals in the mature post-translationally modified complex. The threonine/serine and tyrosine/threonine pairs in the sequence VC(T/S)LCSC(Y/T) indicate a cobalt binding enzyme with threonine and tyrosine and an iron enzyme with the serine and threonine version (Fig. 4.1). 126 This second coordination sphere environment is remarkably predictive of the metal binding for every nitrile hydratase thus far observed, but these residues have not proven to be sufficient to control that metal binding discrimination when mutated. A site directed mutant of a cobalt-type nitrile hydratase from

Psuedomanas putida in which the cobalt pair T/Y was mutated to the iron predictive S/T pair did not result in an iron binding enzyme but rather one that was overall poor at binding any metal and that instead had a disulfide formed between two of the active site cysteine thiols that normally bind to the metal. 117 Understanding the ability of enzymes, a third of which are metalloenzymes, to insert the correct metal despite vast differences in binding affinities between various biologically relevant metal ions is an important subject and one to which nitrile hydratases and their well defined metal preferences may contribute significantly. 127

Figure 4.1 Structures of A) iron and B) cobalt type nitrile hydratases with the predictive

- amino acid pairs highlighted in red. The sequence of the iron-type (A) is CSLC(SO2

- - )SC(SO )TA from PDB 4FM4. The sequence of the cobalt-type (B) is CTLC(SO2

)SC(SO-)YP from PDB 1IRE.

Since most nitrile hydratases have been assumed to require an activator protein for metal insertion and post-translational modification, it could be hypothesized these predictive residues are not necessary for correct metal insertion and discrimination directly but rather that they are necessary for correct binding of the apo-nitrile hydratase to its corresponding activase, which differ significantly between the cobalt and iron inserting types. 113, 128 But as more evidence emerged, it appeared not all nitrile hydratases require an activator accessory protein and that these nitrile hydratases still incorporated the correct metal based on their active site sequences. 115, 116, 53 In two of these cases, the metal is cobalt, so it is not persuasive that this enzyme simply inserted the only metal ion option that was present as iron would necessarily be present at micromolar concentrations since it is necessary for the cells to survive. 129 Another report of an iron-type nitrile hydratase expressed in E. coli in the absence of its corresponding activator but with supplemental cobalt(II) chloride in the growth medium did show some incorporation and activity but only about 2.5% of that with the iron replete enzyme. 130

Thus the fidelity of each nitrile hydratase in choosing the correct metal ion for incorporation must be based on an as yet undiscovered feature of these enzymes that at least appears to require the correct predictive pair. This would suggest other residues are involved, but comparisons of residues amongst the various iron and cobalt types have not uncovered conserved differences precluding an obvious route to mutagenic studies that might switch one nitrile hydratase’s metal preference for the other.

If no model can be made a priori about what residues are responsible for metal discrimination, an alternative is to allow evolutionary selection to create this change for us, potentially uncovering a route through sequence space to accomplish this switch which may then be used to as a guide to uncover the evolutionary path these enzymes utilized in nature to achieve the same thing.

In a nitrogen limiting environment, the activity of nitrile hydratase can be used to convert otherwise inert nitrile substrates to amides which can then be hydrolyzed by more common amidases to afford the necessary ammonia which can be immediately used or sequestered thus providing a selective advantage to those organisms bearing this activity.

131 This has been demonstrated by the growth of bacteria utilizing nitrile hydratases with nitriles as the sole source of nitrogen. 132 The activity of nitrile hydratases, which depends upon metal insertion and correct post-translational modification of the active site cysteine ligands, then could be used to select from a library of random mutants of a cobalt-type nitrile hydratase to bind iron instead.

Two major obstacles exist to this. It is likely necessary to have a high concentration of nitrile substrate as whatever mutant that can switch metal preference to be an active iron-type nitrile hydratase may as a result of those mutations have a greatly reduced catalytic efficiency, and high concentrations of nitriles are known to have bacteriostatic effects. 28 The other is that common expression strains such as E. coli posses genes for amidases, but these may need to be expressed at a much higher level so

that the limiting factor for growth and thus selection of those transformants is the nitrile hydrating activity of whatever new iron-type nitrile hydratase that may be expressed. 133

A library of genes encoding ToyJ, the subunit containing the active site complex, in the pACYC-DUET expression vector was constructed by error prone PCR using primers corresponding only to the protein open reading frame. This was done with a standard protocol using Taq DNA polymerase, high concentrations of each primer (10

µM), a high concentration magnesium chloride (7 mM), deoxythymidine triphosphate and deoxycytidine triphosphate (1 mM), deoxyadenosine triphosphate and deoxyguanosine triphosphate (0.2 mM), manganese chloride (0.5 mM), potassium chloride (50 mM), and Tris HCl pH 8.3 (10 mM). 134 The genes for toyocamycin nitrile hydratase, toyJKL, in expression plasmids constructed as previously described or in the mutated pACYC-Duet plasmid for toyJ, were transformed into electrocompetent E. coli

BL21 (DE3). These were then grown on M9 minimal media agar plates with less than the usual 18 mM ammonium chloride (0-5 mM ammonium chloride instead), with 34 µg/mL each of the selection antibiotics kanamycin, streptomycin, and chloramphenicol. 3-

Cyanopyridine, a known substrate for cobalt containing ToyJKL, was added as the source of nitrogen at various concentrations from 2 – 10 mM. These were grown without addition of cobalt salts and incubated at 25 °C, the optimal temperature for soluble expression of ToyJ. 53

Expression strains were also constructed using the same conditions as above but with toyJKL in expression plasmids used for native expression so that a fourth gene

product could also be expressed from a pet28a vector. 135 This additional protein expressed was the E. coli nicotinamidase, PNCA, which is known to hydrolyze the product of ToyJKL and 3-cyanopyridine, nictotinamide, to the nicotinic acid and ammonia. 136 This was amplified from genomic DNA of E. coli strain W3110 using the primers described in the primer table. For these cultures, ampicillin at 100 µg/mL was used in addition to the other three antibiotics. Zinc chloride (50 µM) was also added to these four-protein cultures as this is an essential cofactor for PNCA. 137 This is the extent to which these assays have been performed to date.

In each round, controls can also be done expressing wild type ToyJKL, and both the mutated and wild type were also grown on plates with cobalt(II) chloride added to a final concentration of 50 µM. This allows a minimal effort to discover a metal-switched variant as only when colonies are present on the plates containing mutated TNH but not the wild type TNH containing plates will a selective advantage from those mutations be apparent and worth pursuing with sequencing and perhaps further rounds of mutation.

Finding the optimal conditions for a successfully metal switched mutant to grow faster than those without active nitrile hydratase may require trying various other conditions including gradients of nitrile substrate to avoid the bacteriostatic effects (this could be done for instance by pouring plates in layers with successively lower concentrations of substrate nitrile so that the nitrile can diffuse slowly towards the bacteria at the agar surface and remain at a low concentration).

Another possibility is to use a cobalt-type nitrile hydratase from a thermophilic organism as the starting point of mutation. The nitrile from Pseudonocardia thermophila has been described that can withstand temperatures of 50 °C. 109 Thermophilic proteins offer the advantage of greater structural durability towards mutations and so open more possibilities to explore beneficial mutations in finding a metal switching variant. 138

Finding any clues as to how such similar enzymes can bind similar metals so selectively will provide a deeper understanding of how metalloenzymes generally are able to selectively incorporate their correct metal ions.

Table of Primers

Sequence 5’-3’

ToyJ ATCATCACAGCAGCGAGAACCTGTACTTCCAGGGCCATATGAG

Forward TACCGAACACGTCAGCC

ToyJ Reverse AGTGGTGGTGGTGGTGGTGCTCGAGTGCGGCCGCAAGCTTTCA

CACAGGAGCCGTGCC

PNCA CGGCCTGGTGCCGCGCGGCAGCCATATGCCCCCTCGCGCCCTG

Forward TTAC

PNCA CAGTGGTGGTGGTGGTGGTGCTCGAGTTACCCCTGTGTCTCTT

Reverse CCCAGTCTGCCAG

References

(1) Fleming, F.; Fleming, F. Nitrile-Containing Natural Products. Natural Product Reports 1999, 16, 597–606. (2) Fleming, F. F.; Yao, L.; Ravikumar, P. C.; Funk, L.; Shook, B. C. Nitrile- Containing Pharmaceuticals: Efficacious Roles of the Nitrile Pharmacophore. J. Med. Chem. 2010, 53, 7902–17. (3) McCarty, R.; Bandarian, V. Biosynthesis of Pyrrolopyrimidines. Top Curr Chem 2012, 43, 15–25. (4) Mandal, D.; Köhrer, C.; Su, D.; Babu, I. R.; Chan, C. T.; Liu, Y.; Söll, D.; Blum, P.; Kuwahara, M.; Dedon, P. C.; et al. Identification and Codon Reading Properties of 5- Cyanomethyl Uridine, a New Modified Nucleoside Found in the Anticodon Wobble Position of Mutant Haloarchaeal Isoleucine tRNAs. RNA 2014, 20, 177–88. (5) Zagrobelny, M.; Bak, S.; Møller, B. L. Cyanogenesis in Plants and Arthropods. Phytochemistry 2008, 69, 1457–68. (6) Vetter, J. Plant Cyanogenic Glycosides. Toxicon 2000, 38, 11–36. (7) Møller, B. Functional Diversifications of Cyanogenic Glucosides. Current Opinion in Plant Biology 2010, 13, 337–346. (8) Sibbesen, O.; Koch, B.; Halkier, B. A.; Møller, B. L. Cytochrome P-450TYR Is a Multifunctional Heme-Thiolate Enzyme Catalyzing the Conversion of L-Tyrosine to P- Hydroxyphenylacetaldehyde Oxime in the Biosynthesis of the Cyanogenic Glucoside Dhurrin in Sorghum Bicolor (L.) Moench. J. Biol. Chem. 1995, 270, 3506–11. (9) Bak, S.; Kahn, R.; Nielsen, H.; Møller, B.; Halkier, B. Cloning of Three A-Type Cytochromes P450, CYP71E1, CYP98, and CYP99 from Sorghum Bicolor (L.) Moench by a PCR Approach and Identification by Expression in Escherichia Coli of CYP71E1 as a Multifunctional Cytochrome P450 in the Biosynthesis of the Cyanogenic Glucoside Dhurrin. Plant Mol Biol 1998, 36, 393–405. (10) Kahn, R. A.; Bak, S.; Halkier, B. A.; Moller, B. L. Isolation and Reconstitution of Cytochrome P450ox and in Vitro REconstitution of the Entire Biosyntehtic Pathway of the Cyanogenic Glucoside Dhhurrin from Sorgum. Plant Physiology 1997, 115, 1661– 1670. (11) Kahn, R. A.; Bak, S.; Svendsen, I.; Halkier, B. A.; Møller, B. L. Isolation and Reconstitution of Cytochrome P450ox and in Vitro Reconstitution of the Entire Biosynthetic Pathway of the Cyanogenic Glucoside Dhurrin from Sorghum. Plant Physiol. 1997, 115, 1661–70. (12) Kato, Y.; Nakamura, K.; Sakiyama, H.; Mayhew, S.; Asano, Y. Novel Heme- Containing Lyase, Phenylacetaldoxime Dehydratase from Bacillus Sp. Strain OxB-1: Purification, Characterization, and Molecular Cloning of the Gene. Biochemistry-us 2000, 39, 800–9. (13) Oinuma, K.-I.; Hashimoto, Y.; Konishi, K.; Goda, M.; Noguchi, T.; Higashibata, H.; Kobayashi, M. Novel Aldoxime Dehydratase Involved in Carbon-Nitrogen Triple Bond Synthesis of Pseudomonas Chlororaphis B23. Sequencing, Gene Expression, Purification, and Characterization. J. Biol. Chem. 2003, 278, 29600–8. (14) Kato, Y.; Ooi, R.; Asano, Y. Distribution of Aldoxime Dehydratase in Microorganisms. Appl. Environ. Microbiol. 2000, 66, 2290–6.

(15) Hashimoto, Y.; Hosaka, H.; Oinuma, K.-I.; Goda, M.; Higashibata, H.; Kobayashi, M. Nitrile Pathway Involving Acyl-CoA Synthetase: Overall Metabolic Gene Organization and Purification and Characterization of the Enzyme. J. Biol. Chem. 2005, 280, 8660–7. (16) Miki, Y.; Asano, Y. Biosynthetic Pathway for the Cyanide-Free Production of Phenylacetonitrile in Escherichia Coli by Utilizing Plant Cytochrome P450 79A2 and Bacterial Aldoxime Dehydratase. Appl. Environ. Microbiol. 2014, 80, 6828–36. (17) Konishi, K.; Ohta, T.; Oinuma, K.-I.; Hashimoto, Y.; Kitagawa, T.; Kobayashi, M. Discovery of a Reaction Intermediate of Aliphatic Aldoxime Dehydratase Involving Heme as an Active Center. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 564–8. (18) Xie, S.-X. X.; Kato, Y.; Komeda, H.; Yoshida, S.; Asano, Y. A Gene Cluster Responsible for Alkylaldoxime Metabolism Coexisting with Nitrile Hydratase and Amidase in Rhodococcus Globerulus A-4. Biochemistry 2003, 42, 12056–66. (19) Gleadow, R. M.; Møller, B. L. Cyanogenic Glycosides: Synthesis, Physiology, and Phenotypic Plasticity. Annu Rev Plant Biol 2014, 65, 155–85. (20) Sugawara, S.; Hishiyama, S.; Jikumaru, Y.; Hanada, A.; Nishimura, T.; Koshiba, T.; Zhao, Y.; Kamiya, Y.; Kasahara, H. Biochemical Analyses of Indole-3- Acetaldoxime-Dependent Auxin Biosynthesis in Arabidopsis. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 5430–5. (21) McCarty, R. M.; Somogyi, A.; Lin, G.; Jacobsen, N. E.; Bandarian, V. The Deazapurine Biosynthetic Pathway Revealed: In Vitro Enzymatic Synthesis of PreQ(0) from Guanosine 5’-Triphosphate in Four Steps. Biochemistry 2009, 48, 3847–52. (22) Nelp, M.; Bandarian, V. A Single Enzyme Transforms a Carboxylic Acid into a Nitrile through an Amide Intermediate. Angewandte Chemie Int Ed 2015, 54, 10627– 10629. (23) Schulze, CJ; Bray, WM; Loganzo, F; Lam, MH Borrelidin B: Isolation, Biological Activity, and Implications for Nitrile Biosynthesis. Journal of natural … 2014. (24) Lane, AL; Nam, SJ; Fukuda, T; Yamanaka, K Structures and Comparative Characterization of Biosynthetic Gene Clusters for Cyanosporasides, Enediyne-Derived Natural Products from Marine Actinomycetes. Journal of the … 2013. (25) Uematsu, T.; Suhadolnik, R. J. Nucleoside Antibiotics. VI. Biosynthesis of the Pyrrolopyrimidine Nucleoside Antibiotic Toyocamycin by Streptomyces Rimosus. Biochemistry 1970, 9, 1260–6. (26) Veselá, A. B.; Pelantová, H.; Sulc, M.; Macková, M.; Lovecká, P.; Thimová, M.; Pasquarelli, F.; Pičmanová, M.; Pátek, M.; Bhalla, T. C.; et al. Biotransformation of Benzonitrile Herbicides via the Nitrile Hydratase-Amidase Pathway in Rhodococci. J. Ind. Microbiol. Biotechnol. 2012, 39, 1811–9. (27) Tarskikh, M. M.; Klimatskaya, L. G.; Kolesnikov, S. I. Pathogenesis of Neurotoxicity of Acrylates Acrylonitrile and Acrylamide: From Cell to Organism. Bull. Exp. Biol. Med. 2013, 155, 451–3. (28) Unknown Cyanide Compounds in Biology. Ciba Found. Symp. 1988, 140, 1–261. (29) Kobayashi, M.; Shimizu, S. Versatile Nitrilases: Nitrile-Hydrolysing Enzymes. FEMS Microbiology Letters 1994, 120, 217–223. (30) Pace, H. C.; Brenner, C. The Nitrilase Superfamily: Classification, Structure and Function. Genome Biol. 2001, 2, REVIEWS0001.

(31) Gong, J.-S. S.; Lu, Z.-M. M.; Li, H.; Shi, J.-S. S.; Zhou, Z.-M. M.; Xu, Z.-H. H. Nitrilases in Nitrile Biocatalysis: Recent Progress and Forthcoming Research. Microb. Cell Fact. 2012, 11, 142. (32) Kobayashi, M.; Goda, M.; Shimizu, S. Nitrilase Catalyzes Amide Hydrolysis as Well as Nitrile Hydrolysis. Biochem. Biophys. Res. Commun. 1998, 253, 662–6. (33) Piotrowski, M.; Volmer, J. J. Cyanide Metabolism in Higher Plants: Cyanoalanine Hydratase Is a NIT4 Homolog. Plant Mol. Biol. 2006, 61, 111–22. (34) Piotrowski, M.; Schönfelder, S.; Weiler, E. W. The Arabidopsis Thaliana Isogene NIT4 and Its Orthologs in Tobacco Encode Beta-Cyano-L-Alanine Hydratase/nitrilase. J. Biol. Chem. 2001, 276, 2616–21. (35) Prasad, S.; Bhalla, T. C. Nitrile Hydratases (NHases): At the Interface of Academia and Industry. Biotechnol. Adv. 2010, 28, 725–41. (36) Kobayashi, M.; Shimizu, S. Metalloenzyme Nitrile Hydratase: Structure, Regulation, and Application to Biotechnology. Nat. Biotechnol. 1998, 16, 733–6. (37) Nagashima, S.; Nakasako, M.; Dohmae, N.; Tsujimura, M.; Takio, K.; Odaka, M.; Yohda, M.; Kamiya, N.; Endo, I. Novel Non-Heme Iron Center of Nitrile Hydratase with a Claw Setting of Oxygen Atoms. Nat. Struct. Biol. 1998, 5, 347–51. (38) Miyanaga, A.; Fushinobu, S.; Ito, K.; Wakagi, T. Crystal Structure of Cobalt- Containing Nitrile Hydratase. Biochem. Biophys. Res. Commun. 2001, 288, 1169–74. (39) Murakami, T.; Nojiri, M.; Nakayama, H.; Odaka, M.; Yohda, M.; Dohmae, N.; Takio, K.; Nagamune, T.; Endo, I. Post-Translational Modification Is Essential for Catalytic Activity of Nitrile Hydratase. Protein Sci. 2000, 9, 1024–30. (40) Asano, Y.; Tani, Y.; Yamada, H. A New Enzyme “Nitrile Hydratase” Which Degrades Acetonitrile in Combination with Amidase. Agricultural and Biological Chemistry 1980, 44, 2251–2252. (41) Hopmann, K. H. Full Reaction Mechanism of Nitrile Hydratase: A Cyclic Intermediate and an Unexpected Disulfide Switch. Inorg Chem 2014, 53, 2760–2. (42) Light, K. M.; Yamanaka, Y.; Odaka, M.; Solomon, E. I. Spectroscopic and Computational Studies of Nitrile Hydratase: Insights into Geometric and Electronic Structure and the Mechanism of Amide Synthesis. Chem Sci 2015, 6, 6280–6294. (43) Yamanaka, Y.; Kato, Y.; Hashimoto, K.; Iida, K.; Nagasawa, K.; Nakayama, H.; Dohmae, N.; Noguchi, K.; Noguchi, T.; Yohda, M.; et al. Time-Resolved Crystallography of the Reaction Intermediate of Nitrile Hydratase: Revealing a Role for the Cysteinesulfenic Acid Ligand as a Catalytic Nucleophile. Angew. Chem. Int. Ed. Engl. 2015, 54, 10763–7. (44) Nelp, M.; Song, Y.; Wysocki, V.; Bandarian, V. A Protein-Derived Oxygen Is the Source of the Amide Oxygen of Nitrile Hydratases. J Biol Chem 2016, jbc.M115.704791. (45) Katayama, Y.; Narahara, Y.; Inoue, Y.; Amano, F.; Kanagawa, T.; Kuraishi, H. A Thiocyanate Hydrolase of Thiobacillus Thioparus. A Novel Enzyme Catalyzing the Formation of Carbonyl Sulfide from Thiocyanate. J. Biol. Chem. 1992, 267, 9170–5. (46) Arakawa, T.; Kawano, Y.; Kataoka, S.; Katayama, Y.; Kamiya, N.; Yohda, M.; Odaka, M. Structure of Thiocyanate Hydrolase: A New Nitrile Hydratase Family Protein with a Novel Five-Coordinate cobalt(III) Center. J. Mol. Biol. 2007, 366, 1497–509. (47) Katayama, Y.; Matsushita, Y.; Kaneko, M.; Kondo, M.; Mizuno, T.; Nyunoya, H. Cloning of Genes Coding for the Three Subunits of Thiocyanate Hydrolase of Thiobacillus Thioparus THI 115 and Their Evolutionary Relationships to Nitrile

Hydratase. J. Bacteriol. 1998, 180, 2583–9. (48) Arakawa, T.; Kawano, Y.; Katayama, Y.; Nakayama, H.; Dohmae, N.; Yohda, M.; Odaka, M. Structural Basis for Catalytic Activation of Thiocyanate Hydrolase Involving Metal-Ligated Cysteine Modification. J. Am. Chem. Soc. 2009, 131, 14838–43. (49) Shearer, J.; Kung, I. Y.; Lovell, S.; Kaminsky, W.; Kovacs, J. A. Why Is There an “Inert” Metal Center in the Active Site of Nitrile Hydratase? Reactivity and Ligand Dissociation from a Five-Coordinate Co(III) Nitrile Hydratase Model. J. Am. Chem. Soc. 2001, 123, 463–8. (50) Katayama, Y.; Hashimoto, K.; Nakayama, H.; Mino, H.; Nojiri, M.; Ono, T. A.; Nyunoya, H.; Yohda, M.; Takio, K.; Odaka, M. Thiocyanate Hydrolase Is a Cobalt- Containing Metalloenzyme with a Cysteine-Sulfinic Acid Ligand. J. Am. Chem. Soc. 2006, 128, 728–9. (51) Hjort, C.; Godtfredsen, S.; Emborg, C. Isolation and Characterization of a Nitrile Hydratase from a Rhodococcus Sp. J Chem Technol Biot 1990, 48, 217–226. (52) Stevens, J. M.; Rao Saroja, N.; Jaouen, M.; Belghazi, M.; Schmitter, J. M.; Mansuy, D.; Artaud, I.; Sari, M. A. Chaperone-Assisted Expression, Purification, and Characterization of Recombinant Nitrile Hydratase NI1 from Comamonas Testosteroni. Protein Expr. Purif. 2003, 29, 70–6. (53) Nelp, M.; Astashkin, A.; Breci, L.; McCarty, R.; Bandarian, V. The Alpha Subunit of Nitrile Hydratase Is Sufficient for Catalytic Activity and Post-Translational Modification. Biochemistry-us 2014, 53, 3990–3994. (54) Asano, Y.; Fujishiro, K.; Yamada, H. Aliphatic Nitrile Hydratase from Arthobacter Sp. J-1 Purification and Characterization. Agricultural and Biological Chemistry 46, 1165–1174. (55) Nagasawa, T.; Nanba, H.; Ryuno, K.; Takeuchi, K.; Yamada, H. Nitrile Hydratase of Pseudomonas Chlororaphis B23. Purification and Characterization. Eur. J. Biochem. 1987, 162, 691–8. (56) Cramp, R. A.; Cowan, D. A. Molecular Characterisation of a Novel Thermophilic Nitrile Hydratase. Biochim. Biophys. Acta 1999, 1431, 249–60. (57) O’Reilly, C.; Turner, P. D. The Nitrilase Family of CN Hydrolysing Enzymes - a Comparative Study. J. Appl. Microbiol. 2003, 95, 1161–74. (58) Fry, W. E.; Munch, D. C. Hydrogen Cyanide Detoxification by Gloeocercospora Sorghi. Physiol Plant Pathol 1975, 7, 23–33. (59) Brown, D. T.; Turner, P. D.; O’Reilly, C. Expression of the Cyanide Hydratase Enzyme from Fusarium Lateritium in Escherichia Coli and Identification of an Essential Cysteine Residue. FEMS Microbiol. Lett. 1995, 134, 143–6. (60) Nolan, L. M.; Harnedy, P. A.; Turner, P.; Hearne, A. B.; O’Reilly, C. The Cyanide Hydratase Enzyme of Fusarium Lateritium Also Has Nitrilase Activity. FEMS Microbiol. Lett. 2003, 221, 161–5. (61) Ingvorsen, K.; Højer-Pedersen, B.; Godtfredsen, S. E. Novel Cyanide- Hydrolyzing Enzyme from Alcaligenes Xylosoxidans Subsp. Denitrificans. Appl. Environ. Microbiol. 1991, 57, 1783–9. (62) Meyers, P. R.; Rawlings, D. E.; Woods, D. R.; Lindsey, G. G. Isolation and Characterization of a Cyanide Dihydratase from Bacillus Pumilus C1. J. Bacteriol. 1993, 175, 6105–12. (63) Jarabak, R.; Westley, J. Human Liver Rhodanese. Nonlinear Kinetic Behavior in a

Double Displacement Mechanism. Biochemistry 1974, 13, 3233–3236. (64) Ray, W. K.; Zeng, G.; Potters, M. B.; Mansuri, A. M.; Larson, T. J. Characterization of a 12-Kilodalton Rhodanese Encoded by glpE of Escherichia Coli and Its Interaction with Thioredoxin. J. Bacteriol. 2000, 182, 2277–84. (65) Ploegman, J. H.; Drent, G.; Kalk, K. H.; Hol, W. G. The Structure of Bovine Liver Rhodanese. II. The Active Site in the Sulfur-Substituted and the Sulfur-Free Enzyme. J. Mol. Biol. 1979, 127, 149–62. (66) Gliubich, F.; Gazerro, M.; Zanotti, G.; Delbono, S.; Bombieri, G.; Berni, R. Active Site Structural Features for Chemically Modified Forms of Rhodanese. J. Biol. Chem. 1996, 271, 21054–61. (67) Miller-Martini, D. M.; Chirgwin, J. M.; Horowitz, P. M. Mutations of Noncatalytic Sulfhydryl Groups Influence the Stability, Folding, and Oxidative Susceptibility of Rhodanese. J. Biol. Chem. 1994, 269, 3423–8. (68) Miller-Martini, D. M.; Hua, S.; Horowitz, P. M. Cysteine 254 Can Cooperate with Active Site Cysteine 247 in Reactivation of 5,5’-dithiobis(2-Nitrobenzoic Acid)- Inactivated Rhodanese as Determined by Site-Directed Mutagenesis. J. Biol. Chem. 1994, 269, 12414–8. (69) Akopyan, T. N.; Braunstein, A. E.; Goryachenkova, E. V. Beta-Cyanoalanine Synthase: Purification and Characterization. Proc. Natl. Acad. Sci. U.S.A. 1975, 72, 1617–21. (70) Yi, H.; Juergens, M.; Jez, J. M. Structure of Soybean Β-Cyanoalanine Synthase and the Molecular Basis for Cyanide Detoxification in Plants. Plant Cell 2012, 24, 2696– 706. (71) O’Leary, B.; Preston, G. M.; Sweetlove, L. J. Increased Β-Cyanoalanine Nitrilase Activity Improves Cyanide Tolerance and Assimilation in Arabidopsis. Mol Plant 2014, 7, 231–43. (72) Jenrich, R.; Trompetter, I.; Bak, S.; Olsen, C. E.; Møller, B. L.; Piotrowski, M. Evolution of Heteromeric Nitrilase Complexes in Poaceae with New Functions in Nitrile Metabolism. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 18848–53. (73) Kelly, M.; Postgate, J. R.; Richards, R. L. Reduction of Cyanide and Isocyanide by Nitrogenase of Azotobacter Chroococcum. Biochem. J. 1967, 102, 1C–3C. (74) Li, J.; Burgess, B. K.; Corbin, J. L. Nitrogenase Reactivity: Cyanide as Substrate and Inhibitor. Biochemistry 1982, 21, 4393–402. (75) Koksunan, S.; Vichitphan, S.; Laopaiboon, L.; Vichitphan, K.; Han, J. Growth and Cyanide Degradation of Azotobacter Vinelandii in Cyanide-Containing Wastewater System. J. Microbiol. Biotechnol. 2013, 23, 572–8. (76) Kao, C. M.; Liu, J. K.; Lou, H. R.; Lin, C. S.; Chen, S. C. Biotransformation of Cyanide to Methane and Ammonia by Klebsiella Oxytoca. Chemosphere 2003, 50, 1055– 61. (77) Ebbs, S. Biological Degradation of Cyanide Compounds. Curr. Opin. Biotechnol. 2004, 15, 231–6. (78) Quesada, A.; Guijo, M. I.; Merchán, F.; Blázquez, B.; Igeño, M. I.; Blasco, R. Essential Role of Cytochrome Bd-Related Oxidase in Cyanide Resistance of Pseudomonas Pseudoalcaligenes CECT5344. Appl. Environ. Microbiol. 2007, 73, 5118– 24. (79) Anderson, P. M. Purification and Properties of the Inducible Enzyme Cyanase.

Biochemistry 1980, 19, 2882–8. (80) Sung, Y. C.; Fuchs, J. A. Characterization of the Cyn Operon in Escherichia Coli K12. J. Biol. Chem. 1988, 263, 14769–75. (81) Johnson, W. V.; Anderson, P. M. Bicarbonate Is a Recycling Substrate for Cyanase. J. Biol. Chem. 1987, 262, 9021–5. (82) Van Lanen, S. G.; Reader, J. S.; Swairjo, M. A.; de Crécy-Lagard, V.; Lee, B.; Iwata-Reuyl, D. From Cyclohydrolase to Oxidoreductase: Discovery of Nitrile Reductase Activity in a Common Fold. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 4264–9. (83) Lee, B. W.; Van Lanen, S. G.; Iwata-Reuyl, D. Mechanistic Studies of Bacillus Subtilis QueF, the Nitrile Oxidoreductase Involved in Queuosine Biosynthesis. Biochemistry 2007, 46, 12844–54. (84) Ribeiro, A.; Yang, L.; Ramos, M.; Fernandes, P.; Liang, Z.-X.; Hirao, H. Insight into Enzymatic Nitrile Reduction: QM/MM Study of the Catalytic Mechanism of QueF Nitrile Reductase. ACS Catalysis 2015, 5, 3740–3751. (85) Kim, Y.; Zhou, M.; Moy, S.; Morales, J.; Cunningham, M. A.; Joachimiak, A. High-Resolution Structure of the Nitrile Reductase QueF Combined with Molecular Simulations Provide Insight into Enzyme Mechanism. J. Mol. Biol. 2010, 404, 127–37. (86) Gregson, J. M.; Crain, P. F.; Edmonds, C. G.; Gupta, R.; Hashizume, T.; Phillipson, D. W.; McCloskey, J. A. Structure of the Archaeal Transfer RNA Nucleoside G*-15 (2-Amino-4,7-Dihydro- 4-Oxo-7-Beta-D-Ribofuranosyl-1H-pyrrolo[2,3- D]pyrimidine-5-Carboximi Dam Ide (archaeosine)). J. Biol. Chem. 1993, 268, 10076–86. (87) Watanabe, M.; Matsuo, M.; Tanaka, S.; Akimoto, H.; Asahi, S.; Nishimura, S.; Katze, J. R.; Hashizume, T.; Crain, P. F.; McCloskey, J. A.; et al. Biosynthesis of Archaeosine, a Novel Derivative of 7-Deazaguanosine Specific to Archaeal tRNA, Proceeds via a Pathway Involving Base Replacement on the tRNA Polynucleotide Chain. J. Biol. Chem. 1997, 272, 20146–51. (88) Bai, Y.; Fox, D. T.; Lacy, J. A.; Van Lanen, S. G.; Iwata-Reuyl, D. Hypermodification of tRNA in Thermophilic Archaea. Cloning, Overexpression, and Characterization of tRNA-Guanine Transglycosylase from Methanococcus Jannaschii. J. Biol. Chem. 2000, 275, 28731–8. (89) Phillips, G.; Chikwana, V. M.; Maxwell, A.; El-Yacoubi, B.; Swairjo, M. A.; Iwata-Reuyl, D.; de Crécy-Lagard, V. Discovery and Characterization of an Amidinotransferase Involved in the Modification of Archaeal tRNA. J. Biol. Chem. 2010, 285, 12706–13. (90) Hall, J.; Gisler, M. A Simple Method for Converting Nitriles to Amides. Hydrolysis with Potassium Hydroxide in Tert-Butyl Alcohol. The Journal of Organic Chemistry 1976, 41, 3769–3770. (91) Asano, Y.; Tachibana, M.; Tani, Y.; Yamada, H. Purification and Chracterization of Amidase Which Participates in Nitrile Degradation. Agricultural and Biological Chesmistry 1982, 46, 1175–1181. (92) Noguchi, T.; Honda, J.; Nagamune, T.; Sasabe, H.; Inoue, Y.; Endo, I. Photosensitive Nitrile Hydratase Intrinsically Possesses Nitric Oxide Bound to the Non- Heme Iron Center: Evidence by Fourier Transform Infrared Spectroscopy. FEBS Lett. 1995, 358, 9–12. (93) Odaka, M.; Fujii, K.; Hoshino, M.; Noguchi, T.; Tsujimura, M.; Nagashima, S.; Yohda, M.; Nagamune, T.; Inoue, Y.; Endo, I. Activity Regulation of Photoreactive

Nitrile Hydratase by Nitric Oxide. Journal of the American Chemical Society 1997, 119, 3785–3791. (94) Sari, M.-A. A.; Jaouen, M.; Saroja, N. R.; Artaud, I. Influence of Cobalt Substitution on the Activity of Iron-Type Nitrile Hydratase: Are Cobalt Type Nitrile Hydratases Regulated by Carbon Monoxide? J. Inorg. Biochem. 2007, 101, 614–22. (95) Hashimoto, K.; Suzuki, H.; Taniguchi, K.; Noguchi, T.; Yohda, M.; Odaka, M. Catalytic Mechanism of Nitrile Hydratase Proposed by Time-Resolved X-Ray Crystallography Using a Novel Substrate, Tert-Butylisonitrile. J. Biol. Chem. 2008, 283, 36617–23. (96) Gumataotao, N.; Kuhn, M. L.; Hajnas, N.; Holz, R. C. Identification of an Active Site-Bound Nitrile Hydratase Intermediate through Single Turnover Stopped-Flow Spectroscopy. J. Biol. Chem. 2013, 288, 15532–6. (97) Shearer, J.; Jackson, H. L.; Schweitzer, D.; Rittenberg, D. K.; Leavy, T. M.; Kaminsky, W.; Scarrow, R. C.; Kovacs, J. A. The First Example of a Nitrile Hydratase Model Complex That Reversibly Binds Nitriles. J. Am. Chem. Soc. 2002, 124, 11417–28. (98) Sugiura, Y.; Kuwahara, J.; Nagasawa, T.; Yamada, H. Significant Interaction between Low-Spin iron(III) Site and Pyrroloquinoline Quinone in Active Center of Nitrile Hydratase. Biochem. Biophys. Res. Commun. 1988, 154, 522–8. (99) Jin, H.; Turner, I.; Nelson, M.; Gurbiel, R.; Doan, P.; Hoffman, B. Coordination Sphere of the Ferric Ion in Nitrile Hydratase. Journal of the American Chemical Society 1993, 115, 5290–5291. (100) Lugo-Mas, P.; Dey, A.; Xu, L.; Davin, S. D.; Benedict, J.; Kaminsky, W.; Hodgson, K. O.; Hedman, B.; Solomon, E. I.; Kovacs, J. A. How Does Single Oxygen Atom Addition Affect the Properties of an Fe-Nitrile Hydratase Analogue? The Compensatory Role of the Unmodified Thiolate. J. Am. Chem. Soc. 2006, 128, 11211–21. (101) Tyler, L.; Noveron, J.; Olmstead, M.; Mascharak, P. Modulation of the P K a of Metal-Bound Water via Oxidation of Thiolato Sulfur in Model Complexes of Co(III) Containing Nitrile Hydratase: Insight into Possible Effect of Cysteine Oxidation in Co−Nitrile Hydratase. Inorganic Chemistry 2003, 42, 5751–5761. (102) Sloan, C.; Krueger, J. Oxidation of Coordinated Cysteine. Formation of Sulfinato- and sulfenatocobalt(III) Complexes. Inorganic Chemistry 1975, 14, 1481–1485. (103) Piersma, S. R.; Nojiri, M.; Tsujimura, M.; Noguchi, T.; Odaka, M.; Yohda, M.; Inoue, Y.; Endo, I. Arginine 56 Mutation in the Beta Subunit of Nitrile Hydratase: Importance of Hydrogen Bonding to the Non-Heme Iron Center. J. Inorg. Biochem. 2000, 80, 283–8. (104) Tsujimura, M.; Odaka, M.; Nakayama, H.; Dohmae, N.; Koshino, H.; Asami, T.; Hoshino, M.; Takio, K.; Yoshida, S.; Maeda, M.; et al. A Novel Inhibitor for Fe-Type Nitrile Hydratase: 2-Cyano-2-Propyl Hydroperoxide. J. Am. Chem. Soc. 2003, 125, 11532–8. (105) Adzamli, K.; Libson, K.; Lydon; Elder; Deutsch, E. Synthesis, Characterization, and Reactivity of Coordinated Sulfenic Acids. Inorganic Chemistry 1979, 18, 303–311. (106) Lydon; Deutsch, E. Chemistry and Reactivity of Sulfur-Bonded Sulfenato- cobalt(III) Complexes. Inorganic Chemistry 1982, 21, 3180–3185. (107) Smentek, L.; Hess, A.; Nowak; Ohtsuka; Hasegawa; Nakatsuji Density Functional Study on Geometry and Electronic Structure of Nitrile Hydratase Active Site Model. International Journal of Quantum Chemistry 2002, 90, 1174–1187.

(108) Heinrich, L.; Mary‐Verla, A.; Li, Y.; Vaissermann, J.; Chottard, J. Cobalt(III) Complexes with Carboxamido‐N and Sulfenato‐S or Sulfinato‐S Ligands Suggest That a Coordinated Sulfenate‐S Is Essential for the Catalytic Activity of Nitrile Hydratases. Eur J Inorg Chem 2001, 2001, 2203–2206. (109) Martinez, S.; Wu, R.; Sanishvili, R.; Liu, D.; Holz, R. The Active Site Sulfenic Acid Ligand in Nitrile Hydratases Can Function as a Nucleophile. J. Am. Chem. Soc. 2014, 136, 1186–9. (110) Lu, J.; Zheng, Y.; Yamagishi, H.; Odaka, M.; Tsujimura, M.; Maeda, M.; Endo, I. Motif CXCC in Nitrile Hydratase Activator Is Critical for NHase Biogenesis in Vivo. FEBS Letters 2003, 553, 391–396. (111) Liu, Y.; Cui, W.; Fang, Y.; Yu, Y.; Cui, Y.; Xia, Y.; Kobayashi, M.; Zhou, Z. Strategy for Successful Expression of the Pseudomonas Putida Nitrile Hydratase Activator P14K in Escherichia Coli. BMC Biotechnology 2013, 13, 48. (112) Zhou; Hashimoto; Shiraki; Kobayashi Discovery of Posttranslational Maturation by Self-Subunit Swapping. Proceedings of the National Academy of Sciences 2008, 105, 14849–14854. (113) Liu, Y.; Cui, W.; Xia, Y.; Cui, Y.; Kobayashi, M.; Zhou, Z. Self-Subunit Swapping Occurs in Another Gene Type of Cobalt Nitrile Hydratase. PLoS ONE 2012, 7, e50829. (114) Nojiri, M.; Yohda, M.; Odaka, M.; Matsushita, Y.; Tsujimura, M.; Yoshida, T.; Dohmae, N.; Takio, K.; Endo, I. Functional Expression of Nitrile Hydratase in Escherichia Coli: Requirement of a Nitrile Hydratase Activator and Post-Translational Modification of a Ligand Cysteine. J. Biochem. 1999, 125, 696–704. (115) Kuhn, M. L.; Martinez, S.; Gumataotao, N.; Bornscheuer, U.; Liu, D.; Holz, R. C. The Fe-Type Nitrile Hydratase from Comamonas Testosteroni Ni1 Does Not Require an Activator Accessory Protein for Expression in Escherichia Coli. Biochem. Biophys. Res. Commun. 2012, 424, 365–70. (116) Kim, S. H.; Padmakumar, R.; Oriel, P. Cobalt Activation of Bacillus BR449 Thermostable Nitrile Hydratase Expressed in Escherichia Coli. Appl. Biochem. Biotechnol. 2001, 91-93, 597–603. (117) Miyanaga, A.; Fushinobu, S.; Ito, K.; Shoun, H.; Wakagi, T. Mutational and Structural Analysis of Cobalt-Containing Nitrile Hydratase on Substrate and Metal Binding. Eur. J. Biochem. 2004, 271, 429–38. (118) Hourai, S.; Miki, M.; Takashima, Y.; Mitsuda, S.; Yanagi, K. Crystal Structure of Nitrile Hydratase from a Thermophilic Bacillus Smithii. Biochem. Biophys. Res. Commun. 2003, 312, 340–5. (119) Heinrich; Li; Vaissermann; Chottard; Chottard, JC A Pentacoordinated Di-N- Carboxamido-Dithiolato-O-Sulfinato-iron(III) Complex Related to the Metal Site of Nitrile Hydratase. Angew. Chem. Int. Ed. Engl. 1999, 38, 3526–3528. (120) Dutta, A.; Flores, M.; Roy, S.; Schmitt, J.; Hamilton, A.; Hartnett, H.; Shearer, J.; Jones, A. Sequential Oxidations of Thiolates and the Cobalt Metallocenter in a Synthetic Metallopeptide: Implications for the Biosynthesis of Nitrile Hydratase. Inorganic Chemistry 2013, 52, 5236–5245. (121) McCoy; Bailey; Bitto; Bingman; Aceti; Fox; Phillips Structure and Mechanism of Mouse Cysteine Dioxygenase. Proceedings of the National Academy of Sciences 2006, 103, 3084–3089.

(122) Lombardini, J. B.; Singer, T. P.; Boyer, P. D. Cystein Oxygenase. II. Studies on the Mechanism of the Reaction with 18oxygen. J. Biol. Chem. 1969, 244, 1172–5. (123) Kumar, D.; Thiel, W.; de Visser, S. P. Theoretical Study on the Mechanism of the Oxygen Activation Process in Cysteine Dioxygenase Enzymes. J. Am. Chem. Soc. 2011, 133, 3869–82. (124) Riddles, P.; Blakeley, R.; Zerner, B. [8] Reassessment of Ellman’s Reagent; 1983; Vol. 91, pp. 49–60. (125) Nagasawa, T.; Takeuchi, K.; Yamada, H. Characterization of a New Cobalt- Containing Nitrile Hydratase Purified from Urea-Induced Cells of Rhodococcus Rhodochrous J1. Eur. J. Biochem. 1991, 196, 581–9. (126) Mitra, S.; Holz, R. C. Unraveling the Catalytic Mechanism of Nitrile Hydratases. J. Biol. Chem. 2007, 282, 7397–404. (127) Waldron, K. J.; Robinson, N. J. How Do Bacterial Cells Ensure That Metalloproteins Get the Correct Metal? Nat. Rev. Microbiol. 2009, 7, 25–35. (128) Lu, J.; Zheng, Y.; Yamagishi, H.; Odaka, M.; Tsujimura, M.; Maeda, M.; Endo, I. Motif CXCC in Nitrile Hydratase Activator Is Critical for NHase Biogenesis in Vivo. FEBS Lett. 2003, 553, 391–6. (129) Andrews, S. C.; Robinson, A. K.; Rodríguez-Quiñones, F. Bacterial Iron Homeostasis. FEMS Microbiol. Rev. 2003, 27, 215–37. (130) Nojiri, M.; Nakayama, H.; Odaka, M.; Yohda, M.; Takio, K.; Endo, I. Cobalt- Substituted Fe-Type Nitrile Hydratase of Rhodococcus Sp. N-771. FEBS Lett. 2000, 465, 173–7. (131) Kim, M.; Zhang, Z.; Okano, H.; Yan, D.; Groisman, A.; Hwa, T. Need-Based Activation of Ammonium Uptake in Escherichia Coli. Mol. Syst. Biol. 2012, 8, 616. (132) Asano, Y.; Yasuda, T.; Tani, Y.; Yamada, H. A New Enzymatic Method of Acrylamide Production. Agricultural and Biological Chemistry 1982, 46, 1183–1189. (133) Pardee, A. B.; Benz, E. J.; St Peter, D. A.; Krieger, J. N.; Meuth, M.; Trieshmann, H. W. Hyperproduction and Purification of Nicotinamide Deamidase, a Microconstitutive Enzyme of Escherichia Coli. J. Biol. Chem. 1971, 246, 6792–6. (134) Wilson, D. S.; Keefe, A. D. Random Mutagenesis by PCR. Curr Protoc Mol Biol 2001, Chapter 8, Unit8.3. (135) Song, Y.; Nelp, M. T.; Bandarian, V.; Wysocki, V. H. Refining the Structural Model of a Heterohexameric Protein Complex: Surface Induced Dissociation and Ion Mobility Provide Key Connectivity and Topology Information. ACS Cent Sci 2015, 1, 477–487. (136) Frothingham, R.; Meeker-O’Connell, W. A.; Talbot, E. A.; George, J. W.; Kreuzer, K. N. Identification, Cloning, and Expression of the Escherichia Coli Pyrazinamidase and Nicotinamidase Gene, pncA. Antimicrob. Agents Chemother. 1996, 40, 1426–31. (137) Du, X.; Wang, W.; Kim, R.; Yakota, H.; Nguyen, H.; Kim, S. H. Crystal Structure and Mechanism of Catalysis of a Pyrazinamidase from Pyrococcus Horikoshii. Biochemistry 2001, 40, 14166–72. (138) Bloom, J. D.; Labthavikul, S. T.; Otey, C. R.; Arnold, F. H. Protein Stability Promotes Evolvability. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 5869–74.

Appendix A:

A Single Enzyme Transforms a Carboxylic Acid into a Nitrile through an Amide Intermediate

This research was originally published in Angewandte Chemie. Micah T. Nelp and Vahe Bandarian. A Single Enzyme Transforms a Carboxylic Acid into a Nitrile through an Amide Intermediate. Angewandte Chemie 2015, 127, 10773-10775 © Wiley.

DOI: 10.1002/ange.201504505

83 Angewandte Chemie

Deutsche Ausgabe:DOI:10.1002/ange.201504505 Biosynthesis Hot Paper Internationale Ausgabe:DOI:10.1002/anie.201504505 ASingle Enzyme Transforms aCarboxylic Acid into aNitrile through an Amide Intermediate Micah T. Nelp and Vahe Bandarian*

Abstract: The biosynthesis of nitriles is knowntooccur through specialized pathways involving multiple enzymes; however,inbacterial and archeal biosynthesis of 7-deazapur- ines,asingle enzyme,ToyM, catalyzesthe conversion of the carboxylic acid containing 7-carboxy-7-deazaguanine (CDG) into its corresponding nitrile,7-cyano-7-deazaguanine

(preQ0). The mechanism of this unusual direct transformation was shown to proceed via the adenylation of CDG,which activates it to form the newly discoveredamide intermediate 7- amido-7-deazaguanine (ADG). This is subsequently dehydrated to form the nitrile in aprocess that consumes asecond equivalent of ATP. The authentic amide intermediate is shown to be chemically and kinetically competent. The ability of ToyM to activate two different substrates,anacid and an amide,accounts for this unprecedented one-enzyme catalysis of nitrile synthesis,and the differential rates of these two half reactions suggest that this catalytic ability is derived from an amide synthetase that gained anew function. Figure 1. Twobiosynthetic pathways to nitrile secondary moieties in Nitrile-containing products are widespread in nature and nature. are found in secondary metabolites such as cyanoglucosides and alkanenitriles.[1] Although these are increasingly being recognized as important next-generation antibiotics and amine nitrogens.[10, 11] Thecrystal structure of ToyM, solved cytotoxic agents with pharmaceutical potential,[2,3] thebio- before its enzymatic function was known, showed the synthesis of these compounds is very poorly understood. Only presence of astructural zinc divalent cation and revealed one biosynthetic route,which utilizes at least three separate that it adopts aPPi loop structure characteristic of ATP- enzymes,has been described in mechanistic detail, and it utilizing enzymes.[12] This suggests that the mechanistic involves aldoxime formation[4] and asubsequent dehydra- strategy employed by ToyM may involve adenylation. The tion[5,6] (Figure 1). Thediscovery of the 7-deazapurine bio- simplest mechanism for ToyM (Figure 2) would be adenyla- synthetic operon of Streptomyces rimosus revealed the first tion to activate 7-carboxy-7-deazaguanine (CDG), followed enzyme capable of acting on acarboxylic acid to form by the addition of ammonia to generate 7-amido-7-deaza- anitrile[7] (Figure 1). This nitrile synthetase,ToyM, was shown guanine (ADG). Activation of ADG by asecond equivalent to be ATPdependent and to use ammonia as the source of the of ATPwould precede collapse to form the nitrile product 7- [8] nitrile nitrogen. Homologues of ToyM are widely distrib- cyano-7-deazaguanine (preQ0). Such amechanism would be uted since this nitrilation is essential to the biosynthesis of the arare example of asingle enzyme activating two functional modified nucleosides archaeosine and the ubiquitous queuosine,aswell as many pyrrolopyrimidine nucleoside antibiotics,including toyocamycin and sangivamycin.[9] ToyM bears evolutionary similarity to glutamine-dependent asparagine synthetases and the lysidine-forming enzyme TilS,both of which employ ATPinthe adenylation of substrate molecules,thus poising them for condensation with

[*] M. T. Nelp, Prof. V. Bandarian Department of Chemistry and Biochemistry,University of Arizona 1041 East Lowell Street, Biological Sciences West, Tucson AZ 85721-0088 (USA) E-mail:[email protected] Supportinginformation for this article is available on the WWW under http://dx.doi.org/10.1002/anie.201504505. Figure 2. Proposed mechanism of ToyM.

Angew.Chem. 2015, 127,10773 –10775 ⌫ 2015 Wiley-VCH Verlag84GmbH &Co. KGaA, Weinheim 10773 Angewandte. Zuschriften

groups,both the acid as well as the resultant amide nesium was required to observe maximal quantities of the intermediate,and presents anew paradigm for the biosyn- intermediate (Figure S2Binthe Supporting Information). thesis of nitriles from carboxylic acids. Theidentity of this intermediate was confirmed by To examine the nitrilation catalyzed by ToyM, the comparison with the UV/Vis spectra and mass spectra of recombinant protein was produced in an E. coli expression authentic ADG,which was synthesized enzymatically from

host. Protein was quantified by amino acid analysis,and preQ0 by using ToyJKL, anitrile hydratase from the same metals analysis confirmed the presence of 1.1 0.2 equiv- pathway as ToyM.[13] To determine whether ADG is agenuine ⌃ alents of Zn2+ per monomer.Totest the mechanistic proposal intermediate,itwas incubated with ToyM and ATP. In these

in Figure 2, we incubated ToyM with CDG,ATP,and experiments (Figure 3B), preQ0 was formed at the expense of

magnesium, and monitored the reaction by LC–MS (Fig- ADG with nearly identical t1/2 values (ca. 4min). In addition, ure S1 in the Supporting Information). We found that AMP formed with aprofile that nearly mirrored the

magnesium supplementation significantly increased product formation of preQ0,thus suggesting that adenylation of formation (Figure S2Ainthe Supporting Information), and it ADG is the likely mode of activation. Therefore,the

was routinely included in the assays.The HPLC traces showed formation of preQ0 from CDG appears to occur in two

that preQ0 was produced as previously shown, as well as an ATP-dependent half reactions,the first of which leads to the

intermediate that increased as CDG was consumed and formation of ADG,which is converted into preQ0 in the

decreased as preQ0 formed (Figure 3A). This intermediate second step. exhibited aUV/Vis spectrum highly similar to that of CDG Direct evidence of the nature of the activated intermedi- (Figure S3 in the Supporting Information) and amass con- ate was obtained by combining ToyM, CDG,and ATPinthe sistent with 7-amido-7-deazaguanine (ADG;[M + H]+ m/z of absence of ammonia. This reaction was filtered to remove the

194.0670;theoretical 194.0678). As with preQ0,added mag- enzyme and analyzed by ion pairing LC–MS in positive-ion mode.Inthese experiments,wedetected the formation of an intermediate that elutes at ca. 25 min. Examination of the mass spectrum suggested that the peak appears at m/z 524.1030, which is consistent with adenylated CDG (theoretical m/z 524.1044;Figure 4). Theappearance of this new peak is dependent on the presence of ToyM, CDG,and ATP, but it is abolished in the presence of ammonia (Figure 4A).

Figure 4. A) Extracted ion traces from ToyM reactions with CDG and ATPinthe absence of ammonium ions, showing the adenylated intermediateforms only when all three are present. B) MS/MS analysis of the adenylatedintermediate showing fragments of CDG and AMP.

Theidentity of adenylated CDG was confirmed by using 13 U- C10-labeled ATP. Theresulting covalent adduct exhibited the expected mass shift from m/z 524.1030 to m/z 534.1376 (theoretical m/z 534.1379), which is consistent with the incorporation of ten 13Catoms from the labeled ATP. MS/ MS analysis of the intermediate further confirmed its identity by revealing fragments consistent with expulsion of [CDG + H]+ (m/z 195.0511) and aketene-like fragment of [CDG]+ (m/ z 177.0406). We also observed fragments corresponding to [adenosine phosphate + H]+ (m/z 348.0701) and [adenine + H]+ (m/z 136.0615), both of which shift when 13C-lableled 13 ATPisused. With the labeled substrate,[C10-adenosine + 13 Figure 3. A) Reaction of ToyM with CDG, ammonium chloride, and phosphate + H] is observed at m/z 358.1035 and [ C5- ATP. B) Reaction of ToyM with ADG and ATP. adenine + H]+ at m/z 141.0782 (Figure 4B and Figure S4 in

10774 www.angewandte.de ⌫ 2015 Wiley-VCH Verlag GmbH &C85o. KGaA,Weinheim Angew.Chem. 2015, 127,10773 –10775 Angewandte Chemie

the Supporting Information). We note that the presence of is an elegant example of finding new chemical capability magnesium abolished the buildup of the adenylated inter- through the promiscuity of an amide synthetase enzyme to mediate,presumably as aresult of magnesium catalyzing the achieve what requires at least three enzymes in other known hydrolysis of the adenylated intermediate in solution. There- synthetic pathways. fore,these assays were all carried out in the absence of added In conclusion, ToyM is an amide synthetase,aswell as divalent cation. anitrile synthetase.Its ability to accept both the acid and To further examine whether the acid and amide substrates amide forms of CDG enables sequential amidation and are activated in asimilar manner,ToyM was incubated with dehydration to the nitrile.This provides anew paradigm for ATP, CDG,and ammonia or ADG for 8.5 h, after which these studying the origin of the many nitrile-containing secondary reaction mixtures were tested by using acoupled enzyme metabolites,the biosynthetic provenance of which is currently pyrophosphate detection assay.[14] Reactions with both sub- unknown, and it gives molecular detail to the synthesis of an strates showed the characteristic signal for the presence of important intermediate in many pathways,the nitrile-con- pyrophosphate in at least the same stoichiometry as that of taining compound preQ0. the preQ0 product (Figure 5). Taken together with the formation of AMP with both substrates,this result demon- Keywords: bioorganic chemistry ·biosynthesis ·enzymes · strates conclusively that ToyM is capable of catalyzing the evolution ·nitriles same adenylation reaction on two substrates,CDG and ADG, to eventually form the nitrile preQ0. Howtocite: Angew.Chem. Int. Ed. 2015, 54,10627–10629 Angew.Chem. 2015, 127,10773–10775

[1] F. F. Fleming, Nat. Prod. Rep. 1999, 16,597 –606. [2] J. C. Schulze,W.M.Bray,F.Loganzo, M. Lam, T. Szal, A. Villalobos,F.E.Koehn, R. G. Linington, J. Nat. Prod. 2014, 77, 2570 –2574. [3] A. L. Lane,S.Nam, T. Fukuda, K. Yamanaka, C. A. Kaufmman, P. R. Jensen, W. Fenical, B. S. Moore, J. Am. Chem. Soc. 2013, 135,4171 –4174. [4] M. F. Oldfield,R.N.Bennet,G.Kiddle,R.M.Wallsgrove,N.P. Botting, Plant Physiol. Biochem. 1999, 37,99–108. [5] Y. Kato,T.Tsuda, Y. Asano, BBAProteins and Proteomics 2007, 1774,856 –865. [6] K. I. Oinuma, Y. Hashimoto,K.Konishi, G. Masahiko,T. Noguchi, H. Higashibata, M. Kobayashi, J. Biol. Chem. 2003, 278,29600 –29608. [7] R. M. McCarty,V.Bandarian, Chem. Biol. 2008, 15,790 –798. [8] R. M. McCarty,V.Bandarian, Biochemistry 2009, 48,3847 – 3852. [9] R. M. McCarty, V. Bandarian, Bioorg.Chem. 2012, 43,15–25. [10] V. Fresquet, J. B. Thoden, H. M. Holden, F. M. Raushel, Bioorg. Figure 5. Pyrophosphate-dependent consumption of NADH through Chem. 2004, 32,63–75. acoupled enzyme assay showing that with both substrates (CDG and [11] T. Suzuki, K. Miyauchi, FEBS Lett. 2010, 584,272 –277. ADG), ToyM produces pyrophosphate, whereas with either no sub- [12] N. Cicmil, R. H. Huang, Proteins Struct. Funct. Bioinf. 2008, 72, strate or no enzyme, only backgroundconsumption of NADH is 1084 –1088. observed. [13] M. T. Nelp,A.V.Astashkin, L. A. Breci, V. Bandarian, Bio- chemistry 2014, 53,3990 –3994. [14] W. E. O’Brien, Anal. Biochem. 1976, 76,423 –3994. Enzyme promiscuity has been integral to the formation of [15] M. Z. Schimdt, E. C. Mundorff,M.Dojka, E. Bermudez, J. E. [15,16] the immense biosynthetic repertoire found in nature. Ness,S.Govindarajan, P. C. Babbit, J. Minshull, J. A. Gerlt, According to this model, the efficiency of enzymes for their Biochemistry 2003, 42,8387 –8393. ancestral primary reaction is likely higher than that for the [16] P. J. O’Brien, D. Herschlag, Chem. Biol. 1999, 6,R91 –R105. reactions for which they gain catalytic ability over time.[17] [17] S. D. Copley, Curr.Opin. Chem. Biol. 2003, 7,265 –272. ToyM bears tantalizing clues to its evolution in that the first half reaction, the amidation, proceeds more rapidly than dehydration to the nitrile in the second half reaction. It is thus Received:May 18, 2015 likely that this unprecedented and simplified nitrile synthesis Published online: July 17, 2015

Angew.Chem. 2015, 127,10773 –10775 ⌫ 2015 Wiley-VCH Verlag86GmbH &Co. KGaA, Weinheim www.angewandte.de 10775

Supporting Information

A Single Enzyme Transforms a Carboxylic Acid into a Nitrile through an Amide Intermediate Micah T. Nelp and Vahe Bandarian* ange_201504505_sm_miscellaneous_information.pdf

Table of Contents

S1. Experimental Procedures

1.1 Expression and Purification of ToyM 1.2 Determination of protein concentration and metal content 1.3 Synthesis of CDG and ADG 1.4 Activity assays

S2. Supporting Figures

Figure S1. HPLC elution profile of the reaction of ToyM with CDG, ammonium chloride, and ATP.

Figure S2. Activity of ToyM at various magnesium sulfate concentrations.

Figure S3. Comparison of the UV-visible spectra of ADG and CDG.

Figure S4. Fragment ions of the adenylated intermediate observed using MS/MS.

Figure S5. Synthesis Scheme for CDG and ADG.

Figure S6. HPLC-MS analysis of purified CDG methyl ester.

Figure S7. HPLC-MS analysis of purified CDG.

Figure S8. HPLC-MS analysis of purified ADG.

S3. References

S1. Experimental Procedures

1.1 Expression and Purification of ToyM

ToyM was expressed and purified as described previously[1].

1.2 Determination of protein concentration

Routine protein concentration was determined with UV-visible spectroscopy using the extinction coefficient of ToyM at 280 nm calculated to be 25,603 M-1 cm-1 with the Expasy ProtParam tool. The concentration as determined by UV-visible spectroscopy was compared to protein concentration determined by amino acid analysis that was conducted by the Molecular Structure Facility at the University of California, Davis. This resulted in a corrected extinction coefficient at 280 nm of 22,288 M-1 cm-1.

ICP-OES to determine zinc metal content was performed at the University of Arizona Department of Hydrology and Water Resources. This showed there to be 1.1 ± 0.2 mol zinc per mol ToyM

1.3 Synthesis of CDG and ADG

CDG was synthesized as described Gangjee et al. modified as follows (Supporting Figure S8).[2] The presence and purity of each compound were confirmed with HPLC-FT-MS on an LTQ Orbitrap XL using the elution program described by Young et al. [3]

(1) Methyl 2-amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3-d]pyrimidine-5- carboxylate. Methyl formate (10.6 g, 175 mmol) was added to THF (50 mL) and brought to 0 °C in an ice bath with stirring. To this was added sodium methoxide (2.16 g, 40 mmol). Over 30 min, methyl chloroacetate (3.26 g, 30 mmol) was added dropwise from a small separatory funnel. The resulting solution was kept at 0 °C for 2 h and allowed to come to room temperature and allowed to stand 1 h. The excess sodium methoxide was quenched by addition of water (50 mL). Excess methyl formate and THF were removed with heat. The remaining aqueous solution was brought to a boil, and then 2,4-diamino- 6-hydroxypyrimidine (1.8 g, 15 mmol) was added. The white precipitate that formed after 20 min of refluxing was removed and dried with vacuum filtration. HPLC RT: 41 + + min. FTMS m/z of [M+H] calculated for (C8H9N4O3) 209.0675, found 209.0668 (Supporting Figure S3).

(2) 2-amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3-d]pyrimidine-5-carboxylic acid (2). Methyl 2-amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3-d]pyrimidine-5-carboxylate (0.1 g, 0.5 mmol) was added to water (5 mL) and DMSO (1 mL). Sodium hydroxide (0.06 mL 5 M) was added, and this was refluxed for 1 hr. After cooling, a white precipitate formed and was removed and dried with vacuum filtration. HPLC RT: 49 min. + + FTMS m/z of [M+H] calculated for (C7H7N4O3) 195.0518, found 195.0514 (Supporting Figure S4).

(3) 2-amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3-d]pyrimidine-5-carboxamide (3). 2-Amino-4-oxo-4,7-dihydro-3H-pyrrolo[2,3-d]pyrimidine-5-carbonitrile, preQ0, (0.1 g, 0.57 mmol) synthesized as previously described4 was dissolved in DMSO (1.5 mL) and added to 50 mM potassium phosphate buffer pH 7.5 with 15% methanol (6 mL).[4] His6-ToyJKL purified as previously described was added to this suspension to a final concentration of 0.2 mM. [5] This was stirred at room temperature for 2 days. Insoluble product was removed with centrifugation. Protein was removed from the resultant supernatant using VWR centrifugal filters (modified polyethersulfone 10K membrane). DMSO was added to the solid product pellet to denature residual ToyJKL, and the eluent from the centrifugal concentrators was used to resuspend and wash this solid product. This process was repeated three times to afford a white precipitate devoid of residual + + ToyJKL activity. HPLC RT: 36 min. FTMS m/z of [M+H] calculated for (C7H8N5O2) 194.0678, found 194.0670 (Supporting Figure S5).

1.4 Activity assays

Reactions with ToyM were carried out with 0.05 M PIPES·NaOH pH 7.5, 0.025 M MgCl2, 0.1 M NaCl, 1 mM dithiothreitol, 2 U New England Biolabs inorganic pyrophosphatase, 5 mM ATP, and 0.15 mM CDG or ADG. With CDG, 5 mM ammonium chloride was also added. The reactions were initiated by addition of 0.02 mM ToyM and quenched at various times by addition of 30% (w/v) trichloroacetic acid to a final concentration of 1.5% (w/v). PreQ0, AMP, ADP, ADG, and residual CDG were quantified by HPLC analysis using a Thermo Scientific Hypersil gold column (4.6 x 250 mm). These were separated at 0.3 mL/min using an initial buffer of 0.05 M ammonium acetate pH 5.5 for 1.5 min, then by a gradient to 2% acetonitrile over 3.5 min, followed by an increased gradient to 10% acetonitrile over 15 min, and a final gradient to 18% acetonitrile over 22 min. This elution program allowed all the components of the mixture to be resolved as follows. ADP elutes at 12.2 min, AMP at 18.1 min, ADG at 31.6 min, preQ0 at 35.8 min, and CDG at 37.4 min.

Pyrophosphate was detected using the same reaction conditions as above excluding inorganic pyrophosphatase. These were run for 8.5 h after which 0.05 mL of each were added to 0.33 mL pyrophosphate reagent (Sigma) and 0.62 mL water. The loss of NADH was monitored at 340 nm for 2 h in temperature controlled cuvettes at 30 °C with time points taken every 13 sec. Negative controls without ToyM, CDG, or ADG and a positive control including 0.08 mM sodium pyrophosphate were run as well.

The adenylated intermediate was detected in reactions that contained all components of the assay except ammonia. The reactions contained 0.05 M HEPES·NaOH pH 7.0, 0.08 mM CDG, and 5 mM ATP. The reactions were incubated for 30 min. Control assays where ammonia (5 mM) was added, or ToyM, CDG or ATP were not included were carried out as above. Protein was removed with VWR centrifugal filters (modified polyethersulfone 10K membrane). The resultant eluent (0.01 mL) was diluted into 0.1 mL of 15 mM tert-butyl ammonium bromide. This was injected onto an Agilent Zorbax Eclipse XDB-C-18 (4.6 X 250 mM) which had been equilibrated with 5 mM ammonium acetate pH 4.8 at a flow rate of 0.3 mL/min. An aliquot of the sample was injected and the column was washed for 3 min with the starting buffer, after

which a gradient to 0.2% acetonitrile over 1.75 min, to 0.8% acetonitrile over 1.25 min, to 1.8% acetonitrile over 0.25 min, to 3.2% over 1.75 min, to 5% acetonitrile over 0.75 min, to 25% over 15 min, and to 50% acetonitrile over 15 min, was used to separate the components. Under these conditions the adenylated intermediate elutes at 25 min. To further assign the intermediate, ESI-MS/MS was performed using the HCD cell of the LTQ Orbitrap XL instrument. The parent ions at m/z of 524 (or 534 for 13C-labeled ATP) were isolated with a 1.0 m/z width and fragmented by collision with helium gas at collision energy of 35% for 0.1 msec.

S2. Supporting Figures

Figure S1. HPLC elution profile of the reaction of ToyM with CDG, ammonium chloride, and ATP.

Figure S2. Activity of ToyM at various magnesium sulfate concentrations. These reactions were carried out as described for activity assays without addition of yeast inorganic pyrophosphatase as that enzyme is itself magnesium dependent. Panel A shows the production of the final product nitrile, preQ0. B shows the production of the amide intermediate, ADG, that is formed quickly in the dead time of the reaction with 5 and 25 mM magnesium followed by a decline corresponding to the increase in final product, preQ0. This shows that without detectable levels of magnesium, ToyM is still active in forming the amide intermediate, ADG.

Figure S3. Comparison of the UV-visible spectra of ADG (in blue) and CDG (in red).

NH2 N N N H NH N N H HN H3N O H H3N O H O O O N N HO N N O O O P O O P O O O OH HO m/z 348.07095 m/z 524.10439

NH2 N NH2 N NH N N H HN H2N O H HN NH O O N N O O O O P C O O O HO H m/z 177.04127 m/z 524.10439

NH3 N N NH3 N H N H2N O H HN NH HN NH O N N O O O O O P O O O O HO H m/z 195.05183 m/z 524.10439

NH2 N N H N H N N HN NH NH 3 O H H3N O N N O O N N O P O O O OH m/z 136.06232 m/z 524.10439

Figure S4. Fragment ions of the adenylated intermediate observed using MS/MS.

Figure S5. Synthesis Scheme for CDG-ester (1), CDG (2), and ADG (3).

Figure S6. HPLC-MS analysis of purified CDG methyl ester (1). A shows the UV absorbance over the entire elution. B shows the relative intensity of the ions corresponding to the mass of CDG methyl ester over time, and C shows mass spectrum averaged over the time at which CDG methyl ester elutes

Figure S7. HPLC-MS analysis of purified CDG (2). A shows the UV absorbance over the entire elution. B shows the relative intensity of the ions corresponding to the mass of CDG over time, and C shows mass spectrum averaged over the time at which CDG elutes

Figure S8. HPLC-MS analysis of purified ADG (3). A shows the UV absorbance over the entire elution. B shows the relative intensity of the ions corresponding to the mass of ADG over time, and C shows mass spectrum averaged over the time at which ADG elutes

S3. References

[1] R. M. McCarty, V. Bandarian, Biochemistry. 2009, 48, 3847-3852.

[2] A. Gangjee, A. Vidwans, E. Elzein, J. J. McGuire, S. F. Queener, R. L. Kisliuk, J. Med. Chem. 2001, 44, 1993-2003.

[3] A. P. Young, V. Bandarian, Biochemistry. 2011, 50, 10573-10575. [4] D. Quaranta, R. M. McCarty, V. Bandarian, C. Rensing, J. Bacteriol. 2007, 189,

5361-5371.

[5] M. T. Nelp, A. V. Astashkin, L. A. Breci, R. M. McCarty, V. Bandarian, Biochemistry. 2014, 53, 3990-3994.

100

Appendix B:

The Alpha Subunit of Nitrile Hydratase is Sufficient for Catalytic Activity and Post- Translational Modification

This research was originally published in the Biochemistry. Micah T. Nelp, Andrei V. Astashkin, Linda A. Breci, Reid M. McCarty, and Vahe Bandarian. The Alpha Subunit of Nitrile Hydratase is Sufficient for Catalytic Activity and Post-Translational Modification. Biochemistry 2014, 53, 3990-3994. © American Chemical Society.

DOI: 10.1021/bi500260j

101 This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Article

pubs.acs.org/biochemistry

The Alpha Subunit of Nitrile Hydratase Is Suﬃcient for Catalytic Activity and Post-Translational Modiﬁcation Micah T. Nelp, Andrei V. Astashkin, Linda A. Breci, Reid M. McCarty,† and Vahe Bandarian* Department of Chemistry and Biochemistry, University of Arizona, 1041 E. Lowell Street, Biological Sciences West 540, Tucson, Arizona 85721-0088, United States *S Supporting Information

ABSTRACT: Nitrile hydratases (NHases) possess a mononuclear iron or cobalt cofactor whose coordination environment includes rare post-translationally oxidized cysteine sulfenic and sulfinic acid ligands. This cofactor is located in the α-subunit at the interfacial active site of the heterodimeric enzyme. Unlike canonical NHases, toyocamycin nitrile hydratase (TNHase) from Streptomyces rimosus is a unique three-subunit member of this family involved in the biosynthesis of pyrrolopyrimidine antibiotics. The subunits of TNHase are homologous to the α- and β-subunits of prototypical NHases. Herein we report the expression, purification, and characterization of the α- subunit of TNHase. The UV−visible, EPR, and mass spectra of the α-subunit TNHase provide evidence that this subunit alone is capable of synthesizing the active site complex with full post-translational modifications. Remarkably, the isolated post- translationally modified α-subunit is also catalytically active with the natural substrate, toyocamycin, as well as the niacin precursor 3-cyanopyridine. Comparisons of the steady state kinetic parameters of the single subunit variant to the heterotrimeric protein clearly show that the additional subunits impart substrate specificity and catalytic efficiency. We conclude that the α- subunit is the minimal sequence needed for nitrile hydration providing a simplified scaffold to study the mechanism and post- translational modification of this important class of catalysts.

itrile hydratases (NHases) catalyze the industrially antibiotics.8 Sequence comparisons revealed that the TNHase important conversion of nitriles to amides under mild encoded by the toyJ, toyK, and toyL genes is homologous to the conditionsN and are a rare example of metalloenzymes that ligate family of nitrile hydratases. ToyJ is homologous to the α- trivalent cobalt or iron in noncorrin or nonheme environ- subunit of NHase and ToyK and ToyL to the C- and N- ments.1,2 The active site complexes of both metal forms are terminal halves, respectively, of the β-subunit of NHase. similar in many respects: they are low spin with low redox However, unlike all previously reported NHase proteins that potentials, and they share the same set of ligands, including have two subunits, the Streptomyces protein is heterotrimeric three cysteine sulfurs, two of which are post-translationally (see Figure 1).9 ToyJKL is also homologous to thiocyanate oxidized, one to cysteine sulfenic acid and the other to cysteine hydrolase, which shares with ToyJKL a composition of three sulfinic acid. The modified cysteine residues are necessary for different subunits.10 activity.3 Along with the closely related thiocyanate hydrolase, Herein we report the expression and purification of NHases are the only known enzymes to employ these unusual recombinant ToyJKL and ToyJ and demonstrate that ToyJ modifications in metal binding and possibly catalysis.4 alone is sufficient for the formation of the post-translationally The mechanisms of post-translational cysteine modification modified, catalytically active NHase. and nitrile hydration remain unknown, but the importance of the β-subunit in each of these processes is implied by the fact ■ EXPERIMENTAL PROCEDURES that all known NHases possess a β-subunit or homologous subunits that are intimately involved with the active site Cloning of ToyJ and ToyJKL. Plasmids containing all three subunits were prepared from those constructed and complex. A highly conserved hydrogen bonding network links 8 the two subunits,5 including bonds between the modified described previously. For expression of His6-ToyJ, toyJ was cysteine residues in the active site of the α-subunit and two β cloned into the NdeI and XhoI sites of pET28a. ToyJKL, where arginine residues, which when mutated drastically reduce ToyL was His6-tagged was prepared by coexpression of toyJ, activity or abolish it altogether.6 toyK, and toyL in Escherichia coli as follows. ToyJ was expressed Uematsu and Suhadolnik first described the enzymatic by cloning toyJ into the NdeI/XhoI sites of pACYCDuet-1. activity of toyocamycin nitrile hydratase (TNHase) in ToyK was expressed by introducing toyK into the NdeI/XhoI Streptomyces rimosus, where it is involved in catalyzing the conversion of the nucleoside antibiotic toyocamycin to Received: March 1, 2014 sangivamycin.7 We cloned this NHase in the course of studies Revised: May 19, 2014 involving the biosynthesis of pyrrolopyrimidine nucleoside 102Published: June 10, 2014

Biochemistry Article Callahan (Burlingame, CA) performed ICP-OES to determine metal content. Amino acid analysis was conducted by the Molecular Structure Facility at the University of California, Davis. Empirically determined correction factors of 0.3 for ToyJ and 0.55 for ToyJKL were found to correlate the Bradford assay results to the amino acid analysis. Activity Assays. The concentration of active enzyme was estimated to be that of cobalt content of the sample as determined by ICP-OES. Reactions to determine protein activity contained 0.05 M potassium phosphate (pH 6.5), varying concentrations of toyocamycin (Berry and Associates) from 0.1 to 2.5 mM and varying concentrations of 3- cyanopyridine from 0.05 to 0.75 M. The reactions were initiated by addition of 10 μM ToyJ or 10 nM ToyJKL and quenched at various times by addition of 30% (w/v) trichloroacetic acid to a final concentration of 5% (w/v). The products sangivamycin or nicotinamide and residual toyoca- Figure 1. Structure of the two-subunit cobalt nitrile hydratase from fi 20 mycin or 3-cyanopyridine were quanti ed by HPLC using an Pseudonocardia thermophila (PDB 1IRE) showing the active site Agilent Zorbax Eclipse XDB-C-18 (4.6 × 50 mm) with an metal closely associated with both subunits. The single β-subunit is isocratic elution of 80% 0.1 M triethylammonium acetate (pH depicted in both red and green showing where it is homologous to fl ToyK and ToyL. 6.8) and 20% methanol at a ow rate of 0.75 mL/min. Sangivamycin elutes at 2.3 min, and toyocamycin elutes at 3.4 min. Nicotinamide elutes at 1.3 min, and 3-cyanopyridine sites of pCDFDuet-1. ToyL was expressed from a plasmid eluted at 2.3 min. containing toyL in the NdeI/XhoI sites of pET28a. Thiocyanate degradation was assayed using the colorimetric 11 Growth and Expression of ToyJ and ToyJKL. Expression method of Bowler modified as follows. Potassium thiocyanate plasmids were introduced into E. coli BL21(DE3) by (1.5 mM) in 0.05 M HEPES·NaOH pH 7.0 was incubated with electroporation and plated on Lennox broth (LB) agar 0.25 mM ToyJ or 0.25 mM ToyJKL for 24 h. Control reactions containing 34 μg/mL kanamycin for ToyJ alone or 34 μg/ contained buffer only. Samples were quenched by diluting 10 mL kanamycin, 34 μg/mL streptomycin, and 34 μg/mL μL of the reaction mixture into 90 μL of 0.5 M FeNO3 and 0.5 chloramphenicol for expression of ToyJKL. All fermentations M HNO3. Precipitated protein was removed by centrifugation. contained these same concentrations of antibiotics. A single Residual thiocyanate concentration in the ± enzyme was colony was used to inoculate 0.1 L LB medium which was compared using the absorbance of the iron thiocyanate grown overnight at 37 °C. The overnight culture was complex at 455 nm. Solutions of potassium thiocyanate were distributed evenly among six Fernbach flasks containing 1 L used as a calibration standard. To ensure that the protein had LB medium and grown to an OD600 nm ≈ 0.5 at which point the not lost activity over the course of the reaction, an aliquot of temperature was lowered to 30 °C, and protein expression was the overnight mixture containing enzyme was also assayed with induced by addition of isopropyl-β-D-thiogalactopyranoside to a toyocamycin, and the rate of formation of sangivamycin was final concentration of 0.1 mM. The cultures were supplemented compared to an aliquot of freshly thawed enzyme to ensure that at the time of induction to 0.05 mM CoCl2. Cells were pelleted the protein was active. at 5000g after 8 h. The pellets were frozen in liquid nitrogen UV−Vis Spectroscopy. UV−visible spectra were obtained and stored at −80 °C. using an Agilent 8453 UV−visible system. Purification of ToyJ and ToyJKL. All steps were carried Mass Spectrometry of ToyJ. Mass spectra were obtained out at 4 °C. Cell paste (∼15 g) was resuspended in 0.05 M using an LTQ Velos Orbitrap in the positive mode. ToyJ was potassium phosphate buffer containing 0.05 M imidazole and 1 prepared by exchanging 0.1 mL aliquots into deionized water mM phenylmethylsulfonylfluoride at pH 7.4. Cells were lysed using a nick column (GE Healthcare). This was lyophilized and with a Branson sonifier for a total of 15 min with 3 s bursts at resuspended in a 7:3 (v/v) solution of 0.1% trifluoroacetic acid 50% amplitude interspersed with 6 s pauses to equilibrate and methanol to a final concentration of 0.1 mM. The solution temperature. Insoluble components were removed by cen- was infused directly using an Advion Nanomate in LC Coupler trifugation at 18000g. Clarified lysate was loaded onto a 5 mL at a flow rate of 200 nL/min with an applied voltage of 2.0 kV. HisTrapHP column (GE Healthcare). Protein was eluted with Resolution was set to 100000 from 400 to 1800 m/z with 4.361 a linear gradient from 0.05 M potassium phosphate buffer ms ion injection time with two microscans. containing 0.05 M imidazole at pH 7.4 to one containing 0.5 M Chemical Reduction and EPR of ToyJKL and ToyJ. imidazole over 25 mL. Fractions containing TNHase were ToyJKL in 0.05 M potassium phosphate (pH 7.4) was identified by an orange color and by SDS-PAGE, pooled, and lyophilized and resuspended in anaerobic water and 20% concentrated using an Amicon Vivaspin Turbo centrifugal glycerol in a Coy anaerobic chamber. It was reduced by concentrator (PES 10K membrane). The protein was addition of sodium dithionite to a final concentration of 10 exchanged into 0.05 M HEPES·NaOH (pH 7.5) using a Bio- mM; the final concentration of cobalt was 0.5 mM as Rad 10DG column. Protein aliquots were frozen in liquid N2 determined by ICP-OES. EPR samples of this were frozen in and stored at −80 °C. liquid nitrogen anaerobically. Determination of Protein Concentration and Metal Controls containing as isolated enzyme contained 0.15 mM Content. Protein concentration was determined using the 103ToyJ or 0.22 mM ToyJKL in 0.05 M HEPES·NaOH (pH 7.4) Bradford method (BioRad) with BSA as a standard. Garratt with 20% glycerol (v/v).

3991 dx.doi.org/10.1021/bi500260j | Biochemistry 2014, 53, 3990−3994

Biochemistry Article EPR spectra were obtained using a Bruker Elexsys E500 To our knowledge, this is the first report of successful spectrometer at the microwave frequency of 9.335 GHz, expression and purification of a holo α-subunit of a nitrile microwave power of 2 mW, and the magnetic field modulation hydratase. Therefore, ESI-MS was used to further confirm the amplitude of 0.5 mT. The measurement temperature was 30 K. presence of the post-translational modification, which was Determination of Oligomeric State with Size Ex- inferred from the similarities between the UV−visible spectra of clusion Chromatography. The oligomeric states of ToyJ and ToyJ and ToyJKL. An Orbitrap Velos mass spectrometer was ToyJKL were determined using analytical size exclusion used because it affords the sensitivity and resolution to clearly chromatography with a HiPrep 16/60 Sephacryl S200 column differentiate between protein that is fully or partially modified. (GE Healthcare) and gel filtration protein standard (BioRad Deconvolution of the mass envelope (Supporting Figure S3, #151-1901) as indicated in Figures S1 and S2. The enzymes Supporting Information) shows that the most abundant species and standards were injected into the column pre-equilibrated is consistent with the mass of ToyJ that lacks the starting with 0.02 M HEPES·NaOH (pH 7.4) and 0.15 M NaCl and methionine, as expected for a bacterial recombinant protein. eluted over 0.15 L at 0.1 mL/min. Elution volumes were Remarkably, the protein contains three additional oxygens and determined using UV traces and SDS-PAGE analysis. a cobalt. The mass of the most abundant species obtained from global deconvolution of the m/z envelope shown is 23353.086 ■ RESULTS AND DISCUSSION amu for [M + H+]. The theoretical mass corresponding to fully fi + Recombinant ToyJKL with ToyL N-terminally His6-tagged was modi ed ToyJKL is 23352.9724 for [M + H ] as predicted purified from E. coli grown in LB medium supplemented with using the QualBrowser Xtract program that is part of the fi CoCl2 (50 μM). Puri ed ToyJKL is orange and has a UV− Xcalibur 2.1.0 SP1.1160 operating software for the Orbitrap visible spectrum that is typical of other cobalt-containing nitrile instrument. We note the molecular formula used to calculate hydratases, which also exhibit low intensity shoulders at 314 the theoretical mass of modified ToyJ is that of the neutral and 429 nm (blue trace in Figure 2).12 The protein molecule wherein three protons have been removed to accommodate the charge of the trivalent cobalt atom. The difference between the observed and calculated values is 0.114 amu, which is within the 10 ppm error of the instrument. The mass spectra of some preparations of ToyJ occasionally exhibit an additional m/z envelope corresponding to ToyJ with only two oxygens (see Figure 3). This is consistent with partial modification or loss of the sulfenic acid oxygen.13

Figure 2. Comparison of UV−visible spectra of ToyJ and ToyJKL. concentration and metal content of several preparations were determined by amino acid analysis and by ICP-OES, the results of which revealed there are 0.6 ± 0.2 cobalt atoms per heterotrimer (Supporting Figure S1, Supporting Information) in the recombinant heterologously expressed protein. Figure 3. MS of ToyJ. The +25 charge state was selected and isolated To delineate the role of each subunit in metal insertion, in an Orbitrap Velos with a mass width of 1.6 and resolution of cysteine oxidation, and nitrile hydration, we explored if the α- 100000. subunit, which is the site of cobalt insertion and post- translational modification, could be expressed in a soluble Since several residues in the β-subunit of NHases have been and active form in the absence of the ToyK and ToyL subunits. previously shown to be intimately involved with the active site 14 Indeed, N-terminal His6-tagged α-subunit, ToyJ, was readily complex, we used electron paramagnetic resonance (EPR) to expressed in the absence of the other two subunits. Gel explore whether the absence of the β-subunit could affect the filtration chromatography shows that ToyJ is monomeric charge and spin state of the active site cobalt. EPR of ToyJKL (Supporting Figure S2, Supporting Information). The UV− and ToyJ as purified did not show any paramagnetic species visible spectrum of purified ToyJ is very similar to that of holo (red trace in Figure 4), which suggests that the cobalt in each ToyJKL (orange trace in Figure 2). Amino acid analysis and protein is in the low spin trivalent state. ToyJKL chemically ICP-OES revealed that the protein contains 1.1 ± 0.1 equiv of reduced with dithionite (black trace in Figure 4) exhibits an cobalt per monomer. Iron was not detected in any of the EPR spectrum consistent with a low spin Co(II) center. The preparations. In the context of the cobalt content of ToyJKL, EPR spectrum of the reduced protein bears a striking similarity we propose that the metal preference of ToyJKL, ToyJ, and to that of Co(II) in corrin complexes, revealing the remarkable possibly the other homologous NHases derives solely from 104ability of the unique noncorrin set of ligands in NHases to interactions with the α-subunit. enforce a low spin state.15 It was not possible to reduce ToyJ to

3992 dx.doi.org/10.1021/bi500260j | Biochemistry 2014, 53, 3990−3994

Biochemistry Article Table 1. Kinetic Parameters of ToyJKL and ToyJ Catalyzed Hydrations of Toyocamycin and 3-Cyanopyridine

kcat/KM −1 a −1 −1 protein substrate kcat (s ) KM (mM) (mM s ) ToyJKL toyocamycin 159 ± 2 2.8 × 10−2 ± 5.7 × 103 1 × 10−3 ToyJ toyocamycin 0.44 ± 0.04 15 ± 2 3.0 × 10−2 ToyJKL 3- 79 ± 5 99 ± 2 8.0 × 10−1 cyanopyridine ToyJ 3- 35 ± 3 1.1 × 103 ± 3.2 × 10−2 cyanopyridine 0.1 × 103 aAll reported activities are corrected for the cobalt content of the protein as determined by ICP-OES.

concentrations, the activity of ToyJKL deviates from the ideal fi Figure 4. EPR of ToyJKL. EPR of ToyJKL. The EPR spectrum of Michaelis−Menten t (Supporting Figure S7, Supporting ToyJKL (0.5 mM) chemically with sodium dithionite (10 mM) to the Information). Controls show that this is not the result of Co(II) state is shown in black. By comparison, no EPR signal is DMSO used to solubilize the substrate. At this point we cannot observed in the absence of dithionite as shown in the red trace with rule out that the deviation results from other factors. The three- 0.22 mM ToyJKL. Experimental conditions: microwave frequency, dimensional structure of ToyJKL is not known, but all fi 9.335 GHz; microwave power, 2 mW; magnetic eld modulation published NHase homologues are dimeric, and one cannot amplitude, 0.5 mT; temperature, 30 K. exclude the potential for cooperativity. To determine if TNHase is a promiscuous enzyme as with obtain the corresponding spectra for comparison as the protein other nitrile hydratases, we compared the turnover of ToyJ and quickly precipitated upon the addition of dithionite, presumably ToyJKL with 3-cyanopyridine, which is a common substrate as a result of degradation of the active site complex, and only and is turned over to nicotinamide.16 Both ToyJKL and ToyJ ff high spin Co(II) was detected in such samples. accept 3-cyanopyridine as substrate, the di erence in kcat/Km Having established that ToyJ is sufficient for the biogenesis values being only 25-fold (Supporting Figures S8 and S9, of the active site complex, we tested whether it is active against Supporting Information, and Table 1). This difference is orders the natural substrate, toyocamycin. ToyJ catalyzes the time- and of magnitude smaller than that with toyocamycin, which enzyme-dependent conversion of toyocamycin to the amide supports the notion that the ToyKL subunits are responsible sangivamycin, which are readily separated via HPLC (Support- for substrate specificity. ing Figures S4 and S5, Supporting Information). We carried out Yamanaka et al. recently reported that the closely related a steady-state kinetic comparison of ToyJ and ToyJKL with thiocyanate hydrolase could be engineered to hydrate nitriles toyocamycin as substrate (shown in Figure 5 and summarized by mutating arginine residues in the γ- and β-subunits to in Table 1). Maximal activity of ToyJ with toyocamycin is 360- phenylalanine and tryptophan, respectively.17 We tested ToyJ fold lower than the wild type, and the KM for the substrate is alone for activity with thiocyanate. Even in the presence of increased at least 500-fold. Since ToyJ does not saturate under substantial concentrations of enzyme and thiocyanate (0.25 the conditions of the assay, the fit provides a lower limit for this mM ToyJ or ToyJKL and 1.5 mM thiocyanate), we were value. Nevertheless the data clearly show that the ToyK and unable to detect any thiocyanate degradation after 24 h ToyL subunits impart ≥105 to the overall catalytic efficiency. (Supporting Table S1, Supporting Information). We note that Control experiments demonstrate that the activity of ToyJ under these conditions, the protein retained 80−90% of the requires intact protein, as boiled ToyJ is inactive (Supporting starting activity with toyocamycin, indicating that absence of Figure S6, Supporting Information). At high substrate activity was not due to protein denaturation. This observation

105 Figure 5. Michaelis−Menten plots of ToyJKL- and ToyJ-catalyzed conversion of toyocamycin to sangivamycin.

3993 dx.doi.org/10.1021/bi500260j | Biochemistry 2014, 53, 3990−3994

Biochemistry Article suggests that the substrate preference of TNHase for nitriles (4) Arakawa, T., Kawano, T., Kataoka, S., Katayama, Y., Kamiya, N., derives at least in part from the α-subunit Yohda, M., and Odaka, M. (2007) Structure of Thiocyanate The activity of the α-subunit of a nitrile hydratase Hydrolase: A New Nitrile Hydratase Family Protein with a Novel distinguishes among possible mechanisms of nitrile hydration Five-Coordinate Cobalt(III) Center. J. Mol. Biol. 366, 1497−1509. by adding significant weight to those that rely solely on residues (5) Nakasako, M., Odaka, M., Yohda, M., Dohmae, N., Takio, K., Kamiya, N., and Endo, I. (1999) Tertiary and Quaternary Structures of derived from the α-subunit. The mechanisms proposed by 18 19 Photoreactive Fe-Type Nitrile Hydratase from Rhodococcus sp. N-771: Hashimoto et al. and Martinez et al. involve mainly α- Roles of Hydration Water Molecules in Stabilizing the Structures and subunit residues and particularly the cysteine-sulfenate oxygen, Structural Origin of the Substrate Specificity of the Enzyme. known to be necessary for activity, to act as a general base in Biochemistry. 38, 9887−9898. activating a water molecule to attack the metal bound nitrile or (6) Piersma, S. R., Nojiri, M., Tsujimura, M., Noguchi, T., Odaka, M., as the nucleophile itself, respectively. Yohda, M., Inoue, Y., and Endo, I. (2000) Arginine 56 Mutation in the This study provides insight into the vast differences in the β Subunit of Nitrile Hydratase: Importance of the Hydrogen Bonding catalytic efficiency between model complexes and NHases. The to the Non-Heme Iron Center. J. Inorg. Biochem. 80, 283−288. activity of ToyJ demonstrates what is possible in a single (7) Uematsu, T., and Suhadolnik, R. L. (1974) In Vivo and Enzymatic molecule with a significantly truncated hydrogen bonding Conversion of Toycamycin to Sangivamycin by Streptomyces rimosus. Arch. Biochem. Biophys. 162, 614−619. network relative to the complete enzyme. Further studies on (8) McCarty, R. M., and Bandarian, V. (2008) Deciphering the contributions of each subunit will help to elucidate and fi ffi Deazapurine Biosynthesis: Pathway for Pyrrolypyrimidine Nucleosides model the signi cant catalytic e ciency found in NHases. Toyocamycin and Sangivamycin. Chem. Biol. 15, 790−798. (9) Blackwell, E. A., Dodds, E. D., Bandarian, V., and Wysocki, V. H. ■ ASSOCIATED CONTENT (2011) Revealing the Quaternary Structure of a Heterogeneous Noncovalent Protein Complex through Surface-Induced Dissociation. S Supporting Information * Anal. Chem. 83, 2862−2865. Figures of analytical size exclusion with ToyJKL and ToyJ; ESI- (10) Katayama, Y., Narahara, Y., Inoue, Y., Amano, F., Kanagawa, T., MS of ToyJ; HPLC trace of sangivamycin and toyocamycin; and Kuraishi, H. (1992) A Thiocyanate Hydrolase of Thiobacillus ToyJ activity at various enzyme concentrations; comparison of thioparus. J. Biol. Chem. 267, 9170−9175. activity with ToyJ and boiled ToyJ; ToyJKL activity in (11) Bowler, R. G. (1944) Determination of Thiocyanate in Blood hydrating toyocamycin and in hydrating 3-cyanopyridine and Serum. Biochem. J. 38, 385−388. ToyJ activity in hydrating 3-cyanopyridine. Table of turnover of (12) Nojiri, M., Nakayama, H., Odaka, M., Yohda, M., Takio, K., and ToyJ and ToyJKL with thiocyanate. This material is available Endo, I. (2000) Cobalt-Substituted Fe-type Nitrile Hydratase of free of charge via the Internet at http://pubs.acs.org. Rhodococcus sp. N771. FEBS Lett. 465, 173−177. (13) Nagashima, S., Nakasako, M., Dohmae, N., Tsujimura, M., Takio, K., Odaka, M., Yohda, M., Kamiya, N., and Endo, I. (1998) ■ AUTHOR INFORMATION Novel Non-Heme Iron Center of Nitrile Hydratase with a Claw Corresponding Author Setting of Oxygen Atoms. Nat. Struct. Biol. 5, 347−351. *Telephone: (520) 626-0389. Fax: (520) 626-9204. E-mail: (14) Endo, I., Nojiri, M., Tsujimura, M., Nakasako, M., Nagashima, [email protected]. S., Yohda, M., and Odaka, M. (2001) Fe-Type Nitrile Hydratase. J. Inorg. Biochem. 83, 247−253. Present Address (15) Pilbrow, J. R. (1982) EPR of B12-Dependent Enzyme Reactions † (R.M.M.) College of Pharmacy, and Department of Chemistry and Related Systems, in B12 Vol. 1 Chemistry (Dolphin, D., Ed.) pp and Biochemistry, University of Texas at Austin, Austin, Texas 431−462, John Wiley & Sons, New York. 78712. (16) Nagasawa, T., Mathew, C. D., Mauger, J., and Yamada, H. Funding (1988) Nitrile Hydratase-Catalyzed Production of Nicotinamide from 3-Cyanopyridine in Rhodococcus rhodochrous J1. Appl. Environ. This research was supported by the National Institute of Microbiol. 54, 1766−1769. General Medical Sciences of the National Institutes of Health (17) Yamanaka, Y., Arakawa, T., Watanabe, T., Namima, S., Sato, M., via grants to V.B. R01 GM72623 and a Career Award in Hori, S., Ohtaki, A., Noguchi, K., Katayama, Y., Yohda, M., and Odaka, Biomedical Sciences from the Burroughs Wellcome Fund. The M. (2013) Two Arginine Residues in the Substrate Pocket acquisition of the EPR spectrometer used in this work was Predominantly Control the Substrate Selectivity of Thiocyanate funded by NIH/NCRR Grant 1S10RR026416-01 to A.V.A. Hydrolase. J. Biol. Sci. Bioeng. 116, 22−27. Orbitrap funding provided by NIH/NCRR 1 S10 RR028868- (18) Hashimoto, K., Suzuki, H., Taniguchi, K., Noguchi, T., Yohda, 01. R.M.M. acknowledges support from a Biological Chemistry M., and Odaka, M. (2008) Catalytic Mechanism of Nitrile Hydratase Training grant from the NIH (T32 GM008804). Proposed by Time-resolved X-ray Crystallography Using a Novel Substrate tert-Butylisonitrile. J. Biol. Chem. 283, 36617−36623. Notes (19) Martinez, S., Wu, R., Sanishvili, R., Liu, D., and Holz, C. H. The authors declare no competing financial interest. (2014) The Active Site Sulfenic Acid Ligand in Nitrile Hydratases Can Function as a Nucleophile. J. Am. Chem. Soc. 136, 1186−1189. ■ REFERENCES (20) Miyanaga, A., Fushinobu, S., Ito, K., and Wakagi, T. (2001) Crystal Structure of Cobalt-Containing Nitrile Hydratase. Biochem. (1) Kobayashi, M., and Shimizu, S. (1996) Metalloenzyme Nitrile Biophys. Res. Commun. 288, 1169−1174. Hydratase: Structure, Regulation, and Application to Biotechnology. Nat. Biotechnol. 16, 733−736. (2) Kovacs, J. A. (2004) Synthetic Analogues of Cysteine-Ligated Non-Heme Iron and Non-Corrinoid Cobalt Enzymes. Chem. Rev. 104, 825−848. (3) Murakami, T., Nojiri, M., Hiroshi, N., Odaka, M., Masafumi, Y., Naoshi, D., Takio, K., Nagamune, T., and Endo, I. (2000) Post- Translational Modification is Essential for Catalytic Activity of Nitrile 106 Hydratase. Protein Sci. 9, 1024−1030.

3994 dx.doi.org/10.1021/bi500260j | Biochemistry 2014, 53, 3990−3994

Supporting Information

The Alpha Subunit of Nitrile Hydratase Is

Sufficient for Catalytic Activity and Post-

Translational Modification

Micah T. Nelp, Andrei V. Astashkin, Linda A. Breci, Reid M. McCarty and Vahe Bandarian

Department of Chemistry and Biochemistry, University of Arizona, 1041 E. Lowell Street, Tucson, AZ 85721, USA

Table of Contents

S1. Supporting Figures and Tables

Figure S1. Analytical size exclusion with ToyJKL and standards.

Figure S2. Analytical size exclusion with ToyJ and standards.

Figure S3. ESI-MS of ToyJ showing full envelope.

Figure S4. HPLC trace of sangivamycin and toyocamycin.

Figure S5. ToyJ activity at various enzyme concentrations.

Figure S6. Comparison of activity with ToyJ and boiled ToyJ.

Figure S7. ToyJKL activity in hydrating toyocamycin to 4.5 mM.

107

Figure S8. ToyJKL activity in hydrating 3-cyanopyridine.

Figure S9. ToyJ activity in hydrating 3-cyanopyridine.

Table 1. Turnover of ToyJ and ToyJKL with thiocyanate.

S1. Supporting Figures

Figure S1. Analytical size exclusion with ToyJKL and standards.

108

Figure S2. Analytical size exclusion with ToyJ and standards.

109

Figure S3. ESI-MS of ToyJ showing full envelope. The +27, +25, and +23 charge state of the fully modified ToyJ species are highlighted.

110

Figure S4. HPLC trace of the product sangivamycin and substrate toyocamycin.

111

Figure S5. ToyJ activity at various enzyme concentrations.

112

Figure S6. Comparison of activity with ToyJ and heat denatured ToyJ. All assays contained 0.5 mM toyocamycin and 10 µM protein ToyJ.

113

Figure S7. ToyJKL activity in hydrating toyocamycin to 4.5 mM.

114

Figure S8. ToyJKL activity in hydrating 3-cyanopyridine.

115

Figure S9. ToyJ activity in hydrating 3-cyanopyridine.

116

Table S1. Turnover of ToyJ and ToyJKL with thiocyanate.

%Activity of TNHase [Thiocyanate] after 24 hr. with toyocamycin after Experiment incubation (µM) overnight incubation with thiocyanate + ToyJ (0.25 mM) 154 ± 4 83% + ToyJKL (0.25 mM) 149 ± 4 90% – Enzyme control 151 ± 4 – KSCN 0

117

References (1) McCarty, R. M.; Bandarian, V. Deciphering Deazapurine Biosynthesis: Pathway for Pyrrolypyrimidine Nucleosides Toyocamycin and Sangivamycin. Chem. Biol 2008 . , 15, 790-798.

(2) Bowler, R.G. Determination of Thiocyanate in Blood Serum. Biochem. J. 1944, 38, 385-388.

118

Appendix C:

A Protein Derived Oxygen is the Source of the Amide Oxygen of Nitrile Hydratase

This research was originally published in the Journal of Biological Chemistry. Micah T. Nelp, Yang Song, Vicki H. Wysocki, and Vahe Bandarian. A Protein Derived Oxygen is the Source of the Amide Oxygen of Nitrile Hydratase. Journal of Biological Chemistry. 2016 © American Society for Biochemistry and Molecular Biology.

DOI: jbc.M115.704791.

119

A Protein-Derived Oxygen is the Source of the Amide Oxygen of Nitrile Hydratases

Micah T. Nelp+, Yang Song#, Vicki H. Wysocki#, and Vahe Bandarian+*

+From the Department of Chemistry, University of Utah, Salt Lake City, Utah, 84112. # From the Department of Chemistry and Biochemistry, The Ohio State University, Columbus, Ohio, 43210.

Running title: Source of the Amide Oxygen of Nitrile Hydratase

To whom correspondence should be addressed: Prof. Vahe Bandarian, Department of Chemistry, University of Utah, 315 South 1400 East, TBBC 3627, Salt Lake City, Utah, 84112-0850. Phone 801-581-6366. E-mail [email protected].

Keywords: Nitrile hydratase, enzyme mechanism, metalloenzyme, enzyme catalysis, enzyme, protein chemical modification bacterial enzymes utilize trivalent iron or ABSTRACT: Nitrile hydratase cobalt metal cations in this non-redox metalloenzymes are unique and important reaction, and, uniquely, these metal ions are biocatalysts that are used industrially to ligated directly to the protein backbone produce high value amides from their through amidate ligands and three cysteine corresponding nitriles. After more than three sulfur ligands, of which two are post- decades since their discovery, the translationally modified to the sulfinic and mechanism of this class of enzymes is sulfenic acid (Fig. 1A) (3). The mechanism becoming clear with evidence from multiple of hydration is slowly emerging through recent studies that the cysteine-derived multiple studies and appears to involve the sulfenato ligand of the active site metal sulfenic acid oxygen as the nucleophile. serves as the nucleophile that initially This sulfenate oxygen is proposed to attacks the nitrile. Herein we describe the initially attack the metal-bound nitrile first direct evidence from solution phase carbon atom creating a cyclic intermediate catalysis that the source of the product that is then resolved through attack at the carboxamido oxygen is the protein. Using sulfenic sulfur either directly by water or 18O-labeled water under single turnover from the neighboring axial thiolato sulfur conditions and native high resolution protein (Fig. 1B) (4-6). Alternative mechanisms mass spectrometry we show that the have also found support from time-resolved incorporation of labeled oxygen into both crystallographic and kinetic studies that product and protein is turnover dependent involve an activated water molecule and that only a single oxygen is exchanged attacking the metal-bound nitrile (Fig. 1C) into the protein even under multiple turnover (7-8). conditions, lending significant support to Early studies on sulfenato ligands in proposals that the post-translationally cobalt complexes demonstrated the modified sulfenato group serves as the nucleophilicity of the sulfenato oxygen (9- nucleophile to initiate hydration of nitriles. 10), and recently this same characteristic ______was observed in NHases wherein the ___ sulfenato oxygen reacted with electrophilic boronic acids (11). Quite tellingly, the

activity of NHase, the related thiocyanate Nitrile hydratases (NHases) catalyze the hydrolase, and NHase mimics were found to formation of amides from nitriles in depend on the presence of these sulfenato aqueous, pH-neutral environments without ligands where oxidation to the sulfinate any carboxylic acid byproducts (1-2). These abrogated activity (12-14).

120

To find direct evidence of either a protein- halves of what is the single β-subunit in or solvent-based nucleophile, we became other known NHases (19). interested in the observations made by The active site of NHases is at the Heinrich et al. with an iron NHase mimic interface of the subunits with β-subunit bearing two sulfenato ligands that is capable residues within hydrogen bonding distance of hydrating nitriles with up to 17 turnovers of the modified cysteine ligands of the (14). They discovered that labeled oxygen active site metal complex (4, 15). Despite from solvent water was exchanged into one the importance of the β-subunit for catalytic of these sulfenato groups only if the nitrile efficiency (21), the recombinantly produced hydration reaction occurred. This provides a α-subunit of TNH (ToyJ) is catalytically way to test the mechanistic paradigms active in the absence of the β- and γ- shown in Fig. 1B and 1C with actual NHase subunits (16). Moreover, as purified, the α- 18 enzymes. In O-labeled water, the subunit alone contains one equivalent of mechanism shown in Fig. 1B predicts a cobalt(III) and three additional oxygen single turnover should result in a product atoms, as expected for a fully post- 16 with no isotopic enrichment as the O- translationally modified NHase α-subunit. oxygen originally from the sulfenato ligand The main difference between ToyJ and the is used and not the bulk solvent. This full complement of ToyJKL is ToyJ alone mechanism also predicts that after a single has lower catalytic efficiency (k /K ) and 18 cat M turnover in O-labeled water the enzyme specificity for toyocamycin, suggesting that should then possess the labeled oxygen the β-subunit generally, or the β- and γ- derived from solvent in the sulfenato group. subunits specifically in TNHase, are By contrast, the mechanism in Fig. 1C responsible for the substrate specificity and predicts that solvent-derived oxygen would increased catalytic efficiency of these be incorporated on the first and all enzymes (16). subsequent turnovers and that the protein ToyJ is a convenient model system with would never become enriched with solvent- which to explore the mechanism of NHases derived oxygen. Single turnover as the slow turnover number (k = 0.44 + experiments to distinguish between these cat 0.04 s-1) and high K (15 + 2 mM) allow for possibilities are challenging because of the m easily accessed single turnover conditions relatively high catalytic efficiency of this that would be challenging to achieve with class of enzyme with turnover rates for the subunit-replete enzyme, ToyJKL (k = aromatic nitrile substrates ranging from 90 cat 159 + 2 s-1 and K = 2.8x10-2 + 1x10-3 mM) sec-1 to 159 sec-1 and turnover rates for m (16). Foremost among the advantages of aliphatic nitrile substrates approaching 1,000 ToyJ is that it provides a means of testing sec-1 and even 4,500 sec-1 for a thermophilic the reaction without the need for large, and nitrile hydratase (15-18). consequently costly, quantities of labeled Toyocamycin nitrile hydratase (TNH) is a water that would be required for stopped cobalt-type nitrile hydratase, encoded by the flow experiments. In this manuscript we toyJKL genes of Streptomyces rimosus, have carried out 18O incorporation which catalyzes the hydration of 18 experiments using H2 O and ToyJ, which toyocamycin to sangivamycin (19). unambiguously show the source of the Interestingly, while homologous to other product amide oxygen is a non- NHase proteins, TNH is a three-subunit exchangeable ToyJ-bound oxygen that, hexameric protein complex (20) where the when taken in the context of other studies of α-subunit is homologous to the α-subunit of NHases (4-6), is likely the nucleophilic other nitrile hydratases, and the β- and γ- oxygen of the cysteine-sulfenate. subunits (ToyK and ToyL, respectively) are homologous to the N- and the C-terminal Experimental Procedures

121

ToyJ was purified as previously described were incubated 2 h, after which 10 µL were (16) modified to include 50 mM sodium removed and flash frozen for chloride in each purification buffer. The characterization of ToyJ by mass concentration of active ToyJ, as determined spectrometry. The remaining samples were by cobalt metal content, was quantified by applied to a 10 kDa centrifugal filter with inductively coupled plasma mass modified polyethersulfone support to spectrometry performed by the Arizona remove protein for analysis of the product. Laboratory for Emerging Contaminants at The eluent containing the sangivamycin the University of Arizona. The concentration product was analyzed by HPLC using a of active sites in each assay was determined gradient described above. The HPLC was solely by the cobalt concentration. coupled to a Thermo Fisher LTQ Orbitrap Monitoring turnover by ToyJ. Turnover XL instrument for in-line mass analysis. The of ToyJ was monitored as follows. MS was set to focus on m/z 200 to 400 in the Reactions were initiated by mixing 0.1 mM positive mode at 100,000 resolution. toyocamycin (Berry and Associates, Dexter To characterize 18O incorporation into the MI) in 50 mM potassium phosphate pH 6.8 protein, samples of ToyJ taken from the with 30 µM ToyJ in the same buffer, which assays were first buffer exchanged into 0.1 had been lyophilized and resuspended in M ammonium acetate via Amicon ultra-0.5 water. An aliquot of the reaction mixture centrifugal filter devices with 10 kDa cutoff (20 µL) was withdrawn at various times (EMD Millipore, Billerica, MA). The final (~10 s - 2 h) and mixed with 4 µL of 30% concentration of the protein was estimated to (v/v) trichloroacetic acid. The precipitated be ~5 µM. Samples were infused directly protein was removed by centrifugation. The into an Exactive EMR mass spectrometer residual substrate toyocamycin and resultant (Thermo Fisher Scientific, Waltham, MA). product sangivamycin were quantified using Glass capillaries (approximate O.D. of 1.5- reverse phase HPLC analysis as described 1.8mm and wall thickness of 0.2 mm) for previously (22) modified to use a lower nano electrospray ionization were pulled on concentration of ammonium acetate of 50 a P-97 micropipet puller (Sutter Instruments, mM in the mobile phase. Hercules, CA) and filled with sample Incorporation of 18O into sangivamycin solution. A 0.9-1.1kV ionization voltage was and ToyJ. The turnover-dependent isotopic applied to a platinum wire inserted into the content of both ToyJ and the product amide, back of the capillary. Typical instrument sangivamycin were determined as follows. settings for native mass spectrometry Substrate (at 2 µM, 5 µM, 10 µM, 20 µM, experiments were: m/z range 400-8000, in 50 µM, 0.2 mM, 0. 5 mM, 2 mM, 10 mM, source CID 20 V, HCD 20V, resolution and 20 mM) and enzyme (60 µM) solutions 140,000, microscan 10, AGC is fixed and were prepared separately, each in half the injection time 200 ms, capillary temperature final volume, flash frozen with liquid 275 °C, S-lens RF 101, source DC offset 25 nitrogen, and lyophilized to remove natural V, injection flatpole DC 12 V, inter flatpole abundance water. All protein aliquots were lens 10 V, bent flatpole DC 8 V, transfer flash frozen just once after purification to multipole DC tune offset -20 V, C-trap maintain the highest possible activity. For entrance lens tune offset 2 V, trapping gas each assay, the lyophilized protein and pressure 5. Protein mass spectra in the m/z substrate were re-suspended in equal range 1500-3200 were charge-deconvolved volumes of 97% enriched 18O-water and deisotoped by Thermo Xtract software. (Cambridge Isotopes, Tewksbury, MA) and The Xtract settings were: generate MH+ were incubated separately for 1 h to ensure Masses Mode, resolution 100,000 at 200, complete exchange. The reactions were S/N Threshold 10, fit factor 44%, remainder initiated by pipetting the protein solution 25%, AveragineLowSulfur model, max into the substrate solution and the mixtures charge 30.

122

The relative abundances of unlabeled and of the amide oxygen in the product. To this singly-18O-labeled-ToyJ populations were end, we tested whether it is feasible to carry estimated as follows. Theoretical isotopic out the reaction under single turnover envelopes were simulated by combining conditions with ToyJ in a timescale allowing theoretical spectra of 1% - 99% singly-18O- for hand mixing after being lyophilized and labeled-ToyJ with unlabeled ToyJ. The reconstituted in small volumes of 97% 18O- isotope distributions of 100% unlabeled water. In these experiments, 30 µM ToyJ ToyJ and the 100% singly-18O-labeled-ToyJ was incubated with 0.1 mM of the substrate, were obtained from the Bruker Isotope toyocamycin, and formation of the product, Pattern software from the theoretical protein sangivamycin, was monitored by LC-MS. formula. The experimental profile data was As expected given the sub-saturating processed to centroid data, noise subtracted, concentration of substrate (Km = 15 + 2 and normalized. The processed experimental mM) (16), there is little turnover (4%) data were compared to the family of occurring in the reaction after 10 s, and theoretical data, and the best fits were quantitative conversion of toyocamycin to estimated by calculating the sum of squared sangivamycin is achieved in ~120 min (Fig. deviation of the intensity for each of the 3). Additionally we tested the effect of peaks. The entire process is illustrated in lyophilization on the activity of the protein Fig. 2 and coded in Matlab script (see and determined ToyJ retains ~82% of the Supporting Information). starting activity after lyophilization. The turnover dependent-exchange of The source of the amide oxygen of labeled water from bulk solvent to enzyme sangivamycin was explored by carrying out and eventually to substrate was modeled the incubations under single and multiple with KinTek Explorer software (23) using turnover conditions by using a 24-fold the following scheme: excess of ToyJ active sites to toyocamycin to achieve single turnover, and up to a 400- fold excess of toyocamycin relative to ToyJ ToyJ-16O + toyocamycin à ToyJ-18O + for multiple turnovers. The products were sangivamycin-16O analyzed by LC-MS. The mass spectrum of natural abundance sangivamycin control ToyJ-18O + toyocamycin à ToyJ-18O + sample (Fig. 4A) exhibits a base peak at m/z Sangivamycin-18O of 310.1146 m/z, exactly the theoretical m/z value for sangivamycin with the molecular + 16 18 formula [C12H15N5O5 + H] . The peak The concentrations of ToyJ- O, ToyJ- O, 18 16 18 corresponding to the presence of a single O sangivamycin- O and sangivamycin- O 18 + were computed as a function of constant [ O-M+H] appears at m/z 312.1189, exactly matching the theoretical m/z value [ToyJ] and 0.001-10 mM toyocamycin. In 18 + this model, we neglected the contribution to for [C12H15N5O4 O1 + H] . This peak is 0.99 16 + 0.02% as abundant as the base peak, labeling of ToyJ- O by exchange of bulk which is consistent with the natural solvent ± sangivamycin because controls 18 described above showed this to be abundance of O in a molecule containing five oxygens (5 x 0.2% = 1%). The high negligible. resolution Orbitrap mass analyzer can resolve the 18O-containing peak from other Results and Discussion relatively abundant isotope combinations The key to distinguishing between giving a +2 peak, such as the species 13 13 + mechanisms of nitrile hydratase proposed in containing two C [C10 C2H15N5O5 + H] at the literature (4-8), those invoking an m/z 312.1213 (expected abundance: 0.8%; activated water molecule or sulfenate acting observed abundance: 0.77 + 0.01%), a single 13 15 13 15 as the nucleophile, is to determine the source C and a single N [C11 C1H15N4 N1O5 +

123

H]+ at m/z 312.1150 (expected abundance: A water molecule tightly bound in the 0.2%; observed abundance: 0.20 + 0.01%), active site is unlikely to be the source of this and other lower abundance isotope product carboxamido oxygen. The protein in combinations (highlighted in Fig. 4A) that these assays underwent lyophilization to are expected to be at or above a 0.02% remove natural abundance water and was abundance relative to the base peak at m/z also incubated for at least an hour after 310.1146. being resuspended in the labeled 18O-water When ToyJ is incubated with 10 mM before being added to substrate. Studies 17 toyocamycin in 18O water, the base peak of using O-labeled water and electron the product sangivamycin shifts to m/z paramagnetic spectroscopy have shown that 312.1189 (Fig. 4B), indicating that turnover water molecules in the active site, in 18O water leads to incorporation of bulk specifically the active site metal-bound aquo solvent into the product. In contrast to this or hydroxo ligand, are easily exchangeable multiple turnover experiment, when a sub- (24-25). Additionally the active sites of stoichiometric concentration of substrate is NHases exist at the interface of the α- and β- used, such as 1 µM sangivamycin (Fig. 4C), subunits (26-27), and since ToyJ in this the base peak for [M+H]+ of sangivamycin experiment is monomeric (16), the active appears at m/z 310.1146, as it does with the site is unlikely to harbor tightly non- natural abundance product. The singly 18O- exchangeable water molecules as it is labeled peak is slightly enriched compared probably exposed to the solvent. We note to that observed in the natural abundance that no crystal structure of this specific sangivamycin control, 3.5 ± 0.8% compared NHase, TNH or ToyJKL, has been solved, to 0.99 ± 0.02%. This may arise because a but a recent study using multiple mass small percentage of enzymes may undergo spectrometry techniques has shown the multiple turnovers, and this effect can be structure of TNH is highly similar to other amplified in the course of mixing. known structures of nitrile hydratases (28). Nevertheless, the fact that nearly all the This series of experiments is most product contains 16O, when the reaction is consistent with a mechanism whereby ToyJ carried out in 97% 18O-labeled water, shows is obligatorily labeled on a single turnover that the newly installed oxygen in the from bulk 18O-water and transfers the product does not derive directly from bulk resulting 18O to product on subsequent solvent. turnovers. The solvent having been ruled out Control reactions were carried out where as the source of the product amide oxygen, the product amide sangivamycin was treated we interrogated the protein using high- under assay conditions with 30 µM ToyJ in resolution mass spectrometry to determine if 97% enriched 18O-water. No exchange of there is indeed a turnover dependent 18O was observed, as the ratio of the 18O- exchange of oxygen from solvent to enzyme containing peak relative to the base peak of after one turnover as would be expected if a unlabeled sangivamycin was single oxygen is transferred to product from indistinguishable from that of sangivamycin enzyme. 18 produced and incubated in natural ToyJ incubated in 97% H2 O in the abundance water (compare Fig. 4A and absence of substrate shows a mass envelope 4D). Indeed, the relative intensities of the at the +8 charge state within 13 ppm of the various natural abundance isotopes at M+2 theoretical natural abundance envelope for are identical to those observed with the the fully post-translationally mature enzyme unlabeled sangivamycin control. Therefore, containing the cobalt complex with the incorporation of labeled water at neutral pH sulfinic and sulfenic acid residues, which does not occur independent of catalysis on shows there is no exchange of oxygen into the time scale of the reactions. the enzyme from labeled water in the time scale of the assays (Fig. 5A). However,

124

ToyJ incubated in this manner but with the the data detailed in Methods and addition of 400-fold excess of the nitrile summarized in Fig. 2. The incorporation of substrate, toyocamycin, shows an envelope 18O into the product was monitored by LC- shifted 0.25 m/z higher at the +8 charge MS as described above. The resulting data state, which corresponds to a single 16O in from both product amide and enzyme are the protein being exchanged for 18O (8 x overlaid in Fig. 8. Remarkably, the 0.25 = 2 Da). This mass increase is within incorporation of bulk solvent into the protein 13 ppm error of the theoretical values based (shown in purple squares) parallels the data on the isotopic distribution. The exchange of on the appearance of 18O in sangivamycin one oxygen upon multiple turnovers is (shown in red circles). These incorporation shown clearly in the deconvolved spectra in profiles were analyzed quantitatively using Fig. 5B where ToyJ having undergone KinTek Explorer software (23) and a model multiple turnovers is assigned a mass of that is depicted in Fig. 8 invoking labeling 23519.30 Da, which is 2.01 Da higher than of ToyJ on each turnover to produce ToyJ- ToyJ incubated in the absence of substrate at 18O. The data from the model (solid lines in 23517.29 Da. In accordance with the Fig. 8) show that the incorporation of 18O labeling of product amide sangivamycin, from bulk solvent into sangivamycin mirrors ToyJ is clearly becoming labeled by bulk the experimental observations for solvent only upon turnover. incorporation of 18O into both sangivamycin Intriguingly though ToyJ does not and ToyJ. We note that in the course of undergo exchange of oxygen with solvent some of the experiments, after 2 h of water in the time scale of the assays, when incubation at room temperature, proteolytic ToyJ is incubated with natural abundance fragments of ToyJ still containing the active product amide, sangivamycin, in labeled site region corresponding to amino acids 32- water, a slight increase in the average mass 214 are detectable in the mass spectra. Close of this ToyJ is observed on the order of 10% examination of the mass envelopes for these of that which occurs with substrate nitrile. proteins also showed the same turnover- 18 This suggests that the amide is able to bind linked increase in O content. in a catalytically similar fashion to poise the As we were preparing this manuscript, an nucleophilic oxygen of ToyJ for exchange elegant study using time-resolved (Fig. 6). crystallography and FT-IR on the To further probe the source of the amide Rhodococcus iron-containing NHase was oxygen we analyzed both sangivamycin and reported that showed the sulfenato oxygen is ToyJ under conditions where ToyJ was likely this protein-based nucleophile (6). mixed with various concentrations of Earlier time-resolved studies using an substrate nitrile ranging from sub- isonitrile substrate analog suggested the stoichiometric (1 µM) to excess (10 mM) sulfenic oxygen may be acting as a general relative to ToyJ. The mass spectra of ToyJ base to activate a water molecule to attack 18 from these experiments are shown in Fig. 7, the isonitrile carbon (7). The O which also details the full m/z range used to incorporation data in the present study (Fig. acquire these data (Fig. 7A) as well as the 8) rules out a solvent-derived nucleophile zoomed in range used for deconvolution mechanism (Fig. 1C) and thus strongly (Fig. 7B) and a further zoomed in region supports the mechanism involving the showing the isotopically resolved envelope enzyme-based oxygen nucleophile in the of ToyJ in the +8 charge state (Fig. 7C). form of a sulfenate-nucleophile (Fig. 1B), Remarkably, bulk 18O appears in the product which is supported by recent computational and enzyme but only when ~one turnover studies suggesting the sulfenato nucleophile has been completed by the enzyme. mechanism (4-5). This work provides the first direct evidence from the solution phase The proportion of unlabeled to singly 18O- that a protein-based nucleophilic oxygen is labeled ToyJ was extracted from analysis of

125

the source of the carboxamido oxygen in first evidence acquired from both product product amides of NHases, representing the and enzyme directly.

Conflict of interest: The authors declare they have no conflicts of interest with the contents of this article.

Author contributions: MN, YS, VW, and VB conceived of the experiments and wrote this paper. MN and YS performed the experiments.

126

REFERENCES

1. Nagasawa, T., Nanba, H. Ryuno, K. Takeuchi, K., Yamada, H. (1987) Nitrile Hydratase of Pseudomonas chlororaphis B23. Purification and Characterization. Eur. J. Biochem. 2, 691-698. 2. Kobayashi, M., Shimizu, S. (1998) Metalloenzyme Nitrile Hydratase: Structure, Regulation, and Application to Biotechnology. Nat. Biotechnol. 16, 733-736. 3. Kovacs, J.A. (2004) Synthetic Analogues of Cysteinate-Ligated Non-Heme Iron and Non-Corrin Cobalt Enzymes. Chem Rev. 104, 825-848. 4. Hopmann, K.H. (2014) Full Reaction Mechanism of Nitrile Hydratase: A Cyclic Intermediate and an Unexpected Disulfide Switch. Inorg. Chem. 53, 2760-2762. 5. Light, K.M., Yamanaka, Y., Odaka, M., Solomon, E.I. (2015) Spectroscopic and Computational Studies of Nitrile Hydratase: Insights into Geometric and Electronic Structure and the Mechanism of Amide Synthesis. Chem. Sci. 6, 6280-6294. 6. Yamanaka, Y., Kato, Y., Hashimoto, K., Iida, K., Nagasawa, K., Nakayama, H., Dohmae, N., Noguchi, K., Noguchi, T., Yohda, M., Odaka, M. (2015) Time-Resolved Crystallography of the Reaction Intermediate of Nitrile Hydratase: Revealing a Role for the Cysteine Sulfenic Acid Ligand as a Catalytic Nucleophile. Angew. Chem. Int. Ed. Engl. 54, 10763-10767. 7. Hashimoto, K., Suzuki, H., Taniguchi, K., Noguchi, T., Yohda, M., Odaka, M. (2008) Catalytic Mechanism of Nitrile Hydratase Proposed by Time-resolved X-ray Crystallography Using a Novel Substrate, tert-butylisonitrile. J. Biol. Chem. 283, 36617- 36623. 8. Mitra, S., Holz, R. C. (2007) Unraveling the Catalytic Mechanism of Nitrile Hydratases. J. Biol. Chem. 282, 7396-7404. 9. Adzamli, K.I., Libson K., Lydon J.D. Elder R.C., Deutsch E. (1979) Synthesis, Characterization, and Reactivity of Coordinated Sulfenic Acids. Inorg. Chem. 18, 303- 311. 10. Lydon, J.D., Deutsch, E. (1981) Chemistry and Reactivity of S-Bonded Sulfenato- Cobalt(III) Complexes. Inorg. Chem. 21, 3180-3185. 11. Martinez, S., Wu, R., Sanishvili, R., Liu, D. Holz, R. (2014) The Active Site Sulfenic Acid Ligand in Nitrile Hydratases Can Function as a Nucleophile. J. Am. Chem. Soc. 136, 1186-1189. 12. Tsujimura, M., Odaka, M., Nakayama, H., Dohmae, N., Koshino, H., Asami, T., Hoshino, M., Takio, K., Yoshida, S., Maeda, M., Endo, I. (2003) A Novel, Inhibitor for Fe-type Nitrile Hydratase: 2-Cyano-2-Propyl Hydroperoxide. J. Am. Chem. Soc., 125, 11532- 11538. 13. Arakawa, T., Kawano, Y., Katayama, Y., Nakayama H., Dohmae, N., Yohda, M., Odaka, M. (2009) Structural Basis for Catalytic Activation of Thiocyanate Hydrolase Involving Metal-Ligated Cysteine Modification. J. Am. Chem. Soc. 131, 14838-1484. 14. Heinrich, L., Mary-Verla, A., Li, Y., Vaissermann, J., Chottard, J. (2001) Cobalt(III) Complexes with Carboxamido-N and Sulfenato-S of Sulfinato-S Ligands Suggest that Coordinated Sulfenate-S is Essential for the Catalytic Activity of Nitrile Hydratases. Eur. J. Inorg. Chem. 2203-2206.

127

15. Miyanaga, A., Fushinobu, S., Ito, K., Shoun, H., Wakagi, T. (2004) Mutational and Structural Analysis of Cobalt-Containing Nitrile Hydratase on Substrate and Metal Binding. Eur. J. Biochem. 271, 429-438. 16. Nelp, M.T., Astashkin, A.V., Breci, L.A., McCarty, R.M., Bandarian, V. (2014) The Alpha Subunit of Nitrile Hydratase is Sufficient for Catalytic Activity and Post- Translational Modification. Biochem. 53, 3990-3994. 17. Takarada, H., Kawano, Y., Hashimoto, K., Nakayama, H., Ueda, S., Yohda, M., Kamiya, N., Dohmae, N., Maeda, M., Odaka, M. (2006) Mutational Study on αGln90 of Fe-Type Nitrile Hydratase from Rhodococcus sp. N771. Biosc. Biotechnol. Biochem. 70, 881-889. 18. Pereira, R.A., Graham, D., Rainey, F.A., Cowan, D.A. (1998) A Novel Thermostable Nitrile Hydratase. Extremophiles 2, 347-357. 19. McCarty, R. M., Bandarian, V. (2008) Deciphering Deazapurine Biosynthesis: Pathway for Pyrrolopyrimidine Nucleosides Toyocamycin and Sangivamycin. Chem. Biol., 15, 790-798. 20. Blackwell, A.E., Dodds, E.D., Bandarian, V., Wysocki, V.H. (2011) Surface-Induced Dissociation Reveals the Quaternary Substructure of a Heterogenous Non-Covalent Protein Complex. Anal. Chem. 83, 2862-2865. 21. Piersma, S.R., Nojiri, M., Tsujimura, M., Odaka, M., Yohda, M., Inoue, Y., Endo, I. (2000) Ariginine 56 Mutation in the Beta Subunit of Nitrile Hydratase: Importance of the Hydrogen Bonding to the Non-Heme Iron Center. J. Inorg. Biochem. 80, 283-288. 22. Pomerantz, S.C., McCloskey, J.A. (1990) Analysis of RNA Hydrolyzates by Liquid Chromatography. Methods Enzymol. 193, 796-824. 23. Johnson, K. A., Simpson, Z. B., and Blom, T. (2009) Global Kinetic Explorer: A New Computer Program for Dynamic Simulation and Fitting of Kinetic Data. Anal. Biochem. 387, 20-29. 24. Sugiura, Y., Kuwahara, J. (1987) Nitrile Hydratase: The First Non-Heme Iron Enzyme with a Typical Low-Spin Fe(III)-Active Center. J. Am. Chem. Soc. 109, 5848-5850. 25. Jin, H., Turner, I.M., Nelson, M.J., Gurbiel, R.J., Doan, P.E., Hoffman, B.M. (1993) Coordination Sphere of the Ferric Ion in Nitrile Hydratase. J. Am. Chem. Soc. 115, 5290- 5291. 26. Nagashima, S., Nakasako, M., Naoshi, D., Tsujimura, M., Takio, K., Odaka, M., Yohda, M., Kamiya, N., Endo, I. (1998) Novel Non-Heme Iron Center of Nitrile Hydratase with a Claw Setting of Oxygen Atoms. Nat. Struct. Mol. Biol. 5, 347-351. 27. Miyanaga, A., Fushinobu, S., Ito, K., Wakagi, T. (2001) Crystal Structure of Cobalt- Containing Nitrile Hydratase. Biochem. Biophys. Res. Comm. 288, 1169-1174. 28. Song. Y., Nelp, M.T., Bandarian, V., Wysocki, V.W. (2015) Refining the Structural Model of Heterohexameric Protein Complex: Surface Induced Dissociation and Ion Mobility Provide Key Connectivity and Topology Information. ACS Cent. Sci. DOI: 10.1021/acscentsci.5b00251

128

Footnotes This research was supported by the National Institute of General Medical Sciences of the National Institutes of Health via grants to V.B. R01 GM72623 and V.W. R01 GM 113658, and a Career Award in Biomedical Sciences to V. B. from the Burroughs Wellcome Fund.

129

Figure 1. Active site structure and mechanism of NHase. A) The X-ray crystal structure of the cobalt-type NHase homolog from Pseudonocardia thermophila JCM 3095 (PDB: 1IRE) shows the active site cobalt complex including two modified cysteine sulfur ligands oxidized to the sulfinic and sulfenic acids. The sequence shown is CTLCSC. B) Potential mechanism of nitrile hydratase in which the sulfenate oxygen acts as the initial nucleophile to attack the nitrile carbon. C) Potential mechanism of nitrile hydratase wherein a solvent water molecule serves as the initial nucleophile. (In B and C, bonds from enzyme to modified-cysteine and amidate ligands are omitted for clarity).

130

Figure 2. Analysis of ToyJ mass spectrometry data for incorporation of 18O. The steps used to analyze the mass spectra of ToyJ are depicted. In step 1, the experimental profile data were processed to centroid data, noise subtracted, and normalized. In step 2, a library of theoretical envelopes corresponding to 1% - 99% singly-18O-labeled-ToyJ with unlabeled ToyJ in the +8 charge state were constructed using Bruker Isotope Pattern software. In step 2, the processed experimental data were finally compared to the family of theoretical data, and the best fits were estimated by calculating the sum of squared deviation of the intensity for each of the peaks.

131

Figure 3. Time dependence of conversion of toyocamycin to sangivamycin. Quantitative conversion of 0.1 mM toyocamycin to sangivamycin is achieved in the presence of 30 µM ToyJ.

100 3.3 [Sangivamycin Product] / [Enzyme]

80 2.7

60 2.0

40 1.4

4% Maximal Turnover 20 10 s after mixing 0.7

% Toyocamycin converted to Sangivamycin Toyocamycin % 0 0 20 40 60 80 100 120 140 Time (min)

132

Figure 4. Exchange of 18O from bulk solvent into sangivamycin is turnover dependent. Shown are the mass spectra of sangivamycin in the +1 charge state. The natural abundance peak is at m/z 310.1146 and that of the 18O-sangivamycin peak is at m/z 312.1189. The traces shown are for A) unlabeled sangivamycin, B) sangivamycin produced from 10 mM toyocamyin in 97% 18O-labeled water, C) sangivamycin produced from 1 µM toyocamycin in 97% 18O-labeled water, and D) control incubation of unlabeled sangivamycin with ToyJ in 97% 18O-labeled water. The high resolution spectra were obtained using an LTQ Orbitrap XL mass spectrometer.

133

Figure 5. Exchange of 18O from bulk solvent into ToyJ is turnover dependent. A) The deconvoluted mass spectra shown are of the +8 charge state of ToyJ. The envelopes shift 0.25 m/z units after multiple turnovers corresponding to the incorporation of just one labeled 18O- oxygen from solvent (8 x 0.25 = 2 Da). In black lines are the theoretical envelopes corresponding to ToyJ and singly 18O-labeled ToyJ. B) Shown are the deconvolved and deisotoped spectra of ToyJ from the 0 and 10 mM toyocamycin samples. The ToyJ-sodium adducts are denoted with an asterisk. The pattern of sodium adducts is influenced by the varying spray conditions used. High resolution spectra were measured using an Exactive EMR mass spectrometer.

134

Figure 6. Possible mechanism of sangivamycin induced exchange of labeled oxygen from solvent into ToyJ (bonds from enzyme to modified-cysteine and amidate ligands are omitted for clarity).

16 R 16 16O R H O R O 16O H R AH C C C H 16O C AH 16O + 16 – 18O B 18OH 16 – O NH H O NH2 2 NH2 NH2 S N S H N S N S N Co Co Co Co – – N – – N SO2 N SO SO2 2 N SO2 S S S S

135

Figure 7. ToyJ is labeled by 18O in a turnover-dependent manner. The full range mass spectrum containing ToyJ in the +8 and +9 charge states and the smaller range used for deconvolution to produce Figure 8 are shown in A) and B), respectively. C) The zoomed in region of the isotopically resolved ToyJ in the +8 charge state after various numbers of turnovers with toyocamycin in 18O-labeled water. High resolution spectra were measured using an Exactive EMR mass spectrometer.

136

Figure 8. Turnover-dependent exchange of 18O from bulk solvent into ToyJ and sangivamycin. Turnover dependent exchange of a single oxygen into enzyme, ToyJ, and product amide, sangivamycin were measured in the presence of various ratios of toyocamycin to ToyJ. In red circles is the percent of the sangivamycin product that has incorporated an 18O-oxygen, and in purple squares is the percent of ToyJ that has incorporated a single 18O- oxygen after increasing numbers of turnovers. The solid curves are generated from theoretical values obtained from KinTek Explorer software according to the depicted kinetic model.

137