The catalytic cycle of nitrous oxide reductase - The enzyme that catalyzes the last step of denitrification.

Accepted Manuscript The catalytic cycle of nitrous oxide reductase — The enzyme that catalyses the last step of denitrification

Cíntia Carreira, Sofia R. Pauleta, Isabel Moura PII: DOI: Reference:

S0162-0134(17)30207-6 doi: 10.1016/j.jinorgbio.2017.09.007 JIB 10326

To appear in:

Journal of Inorganic Biochemistry

Received date: Revised date: Accepted date:

3 April 2017 2 September 2017 8 September 2017

Please cite this article as: Cíntia Carreira, Sofia R. Pauleta, Isabel Moura , The catalytic cycle of nitrous oxide reductase — The enzyme that catalyses the last step of denitrification, Journal of Inorganic Biochemistry (2017), doi: 10.1016/ j.jinorgbio.2017.09.007

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT The catalytic cycle of Nitrous Oxide Reductase - The enzyme that catalyses the last step of denitrification Cíntia Carreira1, Sofia R. Pauleta1*, Isabel Moura1* 1

UCIBIO, REQUIMTE, Departamento de Química, Faculdade de Ciências e

RI

PT

Tecnologia, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal

*Address of the corresponding authors:

SC

Sofia R. Pauleta, [email protected]

NU

Isabel Moura, [email protected]

UCIBIO, REQUIMTE, Departamento de Química, Faculdade de Ciências e Tecnologia,

MA

Universidade Nova de Lisboa, 2829-516 Caparica, Portugal Phone: + 351 212948381

AC

CE

PT E

D

Fax: + 351 212948550

1

ACCEPTED MANUSCRIPT Abstract The reduction of the potent greenhouse gas nitrous oxide requires a catalyst to overcome the large activation energy barrier of this reaction. Its biological decomposition to the inert dinitrogen can be accomplished by denitrifiers through nitrous oxide reductase, the enzyme that catalyses the last step of the denitrification, a pathway of the biogeochemical nitrogen cycle. Nitrous oxide reductase is a multicopper enzyme containing a mixed valence CuA center that can accept electrons from small

PT

electron shuttle proteins, triggering electron flow to the catalytic sulfide-bridged tetranuclear copper “CuZ center”. This enzyme has been isolated with its catalytic

RI

center in two forms, CuZ*(4Cu1S) and CuZ(4Cu2S), proven to be spectroscopic and

SC

structurally different. In the last decades, it has been a challenge to characterize the properties of this complex enzyme, due to the different oxidation states observed for

NU

each of its centers and the heterogeneity of its preparations. The substrate binding site in those two “CuZ center” forms and which is the active form of the enzyme is still a matter of debate. However, in the last years the application of different spectroscopies,

MA

together with theoretical calculations have been useful in answering these questions and in identifying intermediate species of the catalytic cycle.

D

An overview of the spectroscopic, kinetics and structural properties of the two forms of the catalytic "CuZ center" is given here, together with the current knowledge on nitrous

CE

PT E

oxide reduction mechanism by nitrous oxide reductase and its intermediate species.

AC

Keywords: denitrification; greenhouse effect; copper enzyme; nitrous oxide reductase; catalytic cycle; CuZ0

2

ACCEPTED MANUSCRIPT 1. Introduction

1.1 Nitrous oxide reduction by bacteria Anthropogenic emissions of nitrous oxide (N2O) to the atmosphere are one of the main concerns of the 21st century [1, 2]. N2O is a potent greenhouse gas that has a global warming impact 300-fold higher than CO2 and a long half-life in atmosphere, estimated to be 120 years [1, 3]. Its removal from the stratosphere occurs through

PT

photodissociation, generating NOx compounds capable of reacting with ozone, causing its depletion [1, 2]. The exponential increase of N2O in atmosphere and its implication

RI

on the climate change (together with CO2 and CH4) have led to the development of

SC

potential strategies to mitigate emissions, especially those produced by prokaryotes in soils, thereby stabilizing the N2O concentration in the atmosphere [3-6]. As a matter of

NU

fact, about 60 % of N2O emissions arise from the soils through nitrification and denitrification processes, partially due to the increasing use of fertilizers, while its biological reduction can only be performed by nitrous oxide reductase, the last enzyme

MA

of the denitrification pathway and the focus of this review [1, 5, 7]. The abundant nitrogen input and the oxygen limiting tensions stimulate the denitrification pathway [8]. When complete, the bacterial denitrification is a four

D

enzymatic step pathway consisting on dissimilatory reduction of inorganic nitrate (NO3-

PT E

) via nitrite (NO2-) to the nitric oxide (NO) and N2O gases and then to the inert and stable dinitrogen (N2) gas [8] (see Scheme 1). Each reaction is catalyzed by a different metalloenzyme, containing Mo, Fe (heme and non-heme iron) or Cu in its active site, in

CE

a conservative energy process with reduction of metabolites linked with the generation of a proton electrochemical gradient across the inner membrane, which is coupled with

AC

ATP synthesis [4, 8].

As mentioned, the last reaction of the denitrification pathway is catalyzed by nitrous oxide reductase (N2OR), a multicopper enzyme, capable of reducing nitrous oxide to dinitrogen and water in a two-electron two-proton process [7] (Equation 1). N2O + 2 e- + 2 H+ → N2 + H2O

(Eq. 1)

Although the reaction of N2O reduction is highly exergonic (ΔG°’ = - 339.5 kJ mol-1) [7, 9], a catalyst is required to overcome the activation barrier of + 250 kJ mol-1 (in the gas phase) [10]. Besides, N2O is both a weak σ donor and a π acceptor molecule and 3

ACCEPTED MANUSCRIPT therefore its binding to transition metals and breakdown is a challenging reaction [11, 12]. This enzyme is encoded by nosZ, that is included in the nos gene cluster, that comprises several other genes coding for essential proteins required for expression and maturation of N2OR, and are also proposed to sustain a catalytically competent state of N2OR in vivo [13-16]. These genes, highly conserved in the majority of the denitrifying bacteria, are usually arranged as nosRZDFYL and sometimes are preceded or followed by nosX

PT

gene, although the mechanisms involved in its regulation are complex and vary between species [7, 8, 17, 18].

RI

Nevertheless, some denitrifiers lack the nos genes or encode an inactive form of N2OR,

SC

and in the presence of nitrate or nitrite, release N2O to the atmosphere. This incomplete denitrification pathway is still used for energy conservation, as N2O has not been shown

NU

to cause inhibition of any of the enzymes of this pathway and is not involved in its regulation, contrary to nitric oxide, which is a potent cytotoxic and a major regulatory molecule of this pathway [17-19]. Incomplete denitrification has also been observed

MA

under certain environmental conditions, as low pH [20, 21], but this will not be further

1.2 Nitrous Oxide Reductase

D

discussed in this review.

PT E

A periplasmic copper Z-type N2OR (named after nosZ gene) was first isolated from Alcaligenes faecalis in 1972 [22]. However, the requirement of copper in the N2OR respiration was only identified 8 years later, in A. faecalis cultures [23]. Since then, this

CE

now called clade I enzyme has been isolated and characterized from different α-, β- and γ-proteobacteria as periplasmic proteins [24-32]. The ε-proteobacteria Wolinella

AC

succinogenes was reported to express an “atypical” N2OR (classified now as a clade II N2OR), containing an additional c-type heme domain at the C-terminal that has been postulated to function as an electron donor [33, 34]. Although, to date there has not been any report of the isolation of a N2OR from Gram-positive bacteria, it is likely to be a membrane-associated enzyme [35], encoded by an atypical nos gene cluster commonly arranged as nosCZ-ORF-nosDYF-ORF [18, 35, 36], and also classified as a clade II N2OR. Clade II nos clusters consistently exhibit a gene upstream nosZ, which encodes a transmembrane protein, and are frequently associated with c- and b-type cytochromes, as well as iron-sulfur proteins [37]. The different cluster arrangement, as well as the 4

ACCEPTED MANUSCRIPT lacking of nosR and nosX genes in clade II [37], reported as essential to maintain the activity of clade I N2ORs [14, 15, 38], pin pointed to the hypothesis of an alternative electron transfer route via menaquinol [33]. Most of the current knowledge on clade II N2OR is based on genome analysis. This review will focus mainly on clade I N2ORs, describing its biochemical and spectroscopic properties, and in particular the involvement of its unique catalytic center in N2O reduction and its intermediate species.

PT

N2OR is a stable homodimer with a molecular weight of 120 to 160 kDa [8], containing ~ 6 copper atoms per monomer arranged in two distinct copper centers, a binuclear CuA

RI

center, responsible for electron transfer and a sulfide-bridged tetranuclear copper center,

SC

termed “CuZ center”, where the catalysis occurs (Figure 1). CuA center has the ability of accept electrons from small electron shuttle proteins, usually small soluble c-type

NU

cytochromes or cupredoxins [8]. In fact, whole-cells and in vitro studies have been performed to identify the probable physiological electron donors of N2OR in several microorganisms. Studies performed with whole-cells of Rhodobacter capsulatus,

MA

Rhodobacter sphaeroides and Paracoccus denitrificans indicated an oxidation of a cytochrome in the N2O reduction [39, 40]. Moreover, the whole-cells of R. capsulatus

D

in which a gene of cytochrome c2 was deleted, were not able to reduce N2O [41]. These results point out for the role of a c-type cytochrome as the physiological electron donor

PT E

of N2ORs from P. denitrificans, R. capsulatus and R. sphaeroides [29, 39, 41]. In the case of P. pantotrophus N2OR, the enzyme is able to receive electrons from pseudoazurin from the same organism [29, 42], while mammalian cytochrome c can

CE

donate electrons to P. pantotrophus, A. cycloclastes and W. succinogenes N2ORs, as a non-physiological electron donor protein [43-45]. On the other hand, in vitro studies

AC

indicate that cytochrome c552 is the strongest candidate to be the physiological electron donor of Marinobacter hydrocarbonoclasticus N2OR, which was also highlighted in molecular docking studies [46], while in the case of Achromobacter cycloclastes, pseudoazurin was assigned as the probable physiological donor [47, 48]. Although these proteins can donate electrons to N2OR in vitro, being suggested as the probable physiological relevant electron donors, their role in vivo has not yet been demonstrated and the possibility of others proteins with the same role cannot be excluded. “CuZ center”, the catalytic center is distinct from any other copper sites described in the literature, and up-to-now has only been identified in N2OR. 5

ACCEPTED MANUSCRIPT The residues involved in the metal coordination of both copper centers are strictly conserved among N2ORs [9]. The two copper atoms of the CuA center are µ2-bridged by two cysteine residues and also linked to two histidines, a methionine and a tryptophan residue, with each copper atom adopting a distorted tetrahedral geometry [49] (Figure 2). “CuZ center” is a tetranuclear copper center with a central sulfide coordinating the four copper atoms, each of which is coordinated by the side chain of two histidine ligands, with the exception of CuIV, that is only coordinated by one [49,

PT

50] (Figure 3).

CuA center is bound to the C-terminal domain, which is arranged in an antiparallel β-

RI

sandwich fold, a feature common to type I copper proteins [51], while “CuZ center” is

SC

located in the N-terminal domain, which has a seven-bladed β-propeller fold (Figure 1) [49]. In the dimer, these two centers are oriented in a head to tail configuration: in the

NU

same monomer CuA and "CuZ" centers are separate by 40 Å, a long distance for efficient electron transfer, though CuA and "CuZ" centers from each monomer are only 10 Å apart, which confers a functional character to the dimer (Figure 1) [52]. The

MA

function of the calcium ions present in the X-ray structure of all N2ORs was recently elucidated through the studies on the apo-N2OR structure from Shewanella

D

denitrificans. These studies revealed that calcium ions are not required for dimer formation, as calcium binding occurred after the complete assembly of the centers [53].

PT E

Thus, calcium binding seems to have an essential role in stabilizing the enzyme structure, since in its absence the domains of the protein stand apart [53].

CE

- Insert Figure 1 here -

AC

From early years, various forms of N2OR have been isolated and named according to the color of their solution as either pink, purple or blue [54]. In-depth investigations over the years determined that those enzyme forms had different redox, spectroscopic and structural properties, due to the oxidation state of CuA center but mainly due to differences in the “CuZ center”. These different forms of "CuZ center" have been attributed to differences in growth conditions and purification strategies [25, 54-57]. Currently, it is generally accepted that N2OR can be isolated with “CuZ center” in two forms, named herein as CuZ(4Cu2S) and CuZ*(4Cu1S). Indeed, the enzyme can be isolated mainly with CuZ(4Cu2S), in the oxidized or reduced state or mainly with CuZ*(4Cu1S) (in only one oxidation state, the resting state), and with CuA center in 6

ACCEPTED MANUSCRIPT either the oxidized or the reduced state. However, in a purified sample of N2OR there is always a mixture of both “CuZ center” forms, ranging so far from 5 % to 80 % of the CuZ(4Cu2S) form [57]. Thus, it has been a challenge to distinguish the spectroscopic features and kinetic parameters of the two forms of "CuZ center". In addition, the catalytic competent form of "CuZ center", as well as the intermediates species and the catalytic mechanism of N2O reduction are still under debate.

PT

1.3 Properties of Copper centers in Nitrous Oxide Reductase CuA center - The electron entry point

RI

CuA center of N2OR is similar to the one identified in cytochrome c oxidase. The

SC

evidence of the presence of a similar CuA center in N2OR was initially pointed out by Kroneck and his colleagues, who used multifrequency EPR spectroscopy to reveal the

NU

properties of N2OR metal centers [58-61]. A direct copper-copper interaction, with similar features to the CuA center of cytochrome c oxidase, was identified, demonstrating that this center is a mixed-valence metal site [1Cu1.5+-1Cu1.5+] [58-60].

MA

The electronic properties of CuA center in N2OR were extensively studied and are now well-known.

The visible spectrum of the oxidized CuA center, in the [1Cu1.5+-1Cu1.5+] redox state,

D

exhibits absorption bands at 480 (ε ≈ 4000 M-1 cm-1), 540 (ε ≈ 4000 M-1 cm-1) and 800

PT E

nm (ε ≈ 3000 M-1 cm-1), while in the reduced state it becomes spectroscopically silent, due to the d10 electronic configuration of the copper atoms [62]. The intense bands at 480 and 540 nm are associated to a S(cys) → Cu charge transfer (CT) in agreement with

CE

the strongest Cu-S(cys) stretching modes observed by resonance Raman [63-65]. The transition at 800 nm is assigned to a mix valence Ψ→Ψ* CT detected in absorption and

AC

magnetic circular dichroism (MCD) spectra and also supported by resonance Raman spectroscopy through an additional enhance vibration at ~130 cm-1 [62, 64, 66]. The X-band EPR spectrum of CuA center in N2OR is characterized by an axial signal, with g║ = 2.18 and g┴ = 2.03 and a 7-line hyperfine pattern in the g║ region, showing intensity ratios of 1:2:3:4:3:2:1, with a A║= 3.83 mT [25, 58]. The hyperfine splitting is due to the mix valence state (S = 1/2), in which an unpaired electron is delocalized over the two copper nuclei (Icu = 3/2) [25, 58, 60]. Nevertheless, the 7-line pattern is not always well resolved since, for the enzyme purified in the presence of oxygen, an overlap of signals from CuZ*(4Cu1S) was detected in this region [7, 67] (vide infra).

7

ACCEPTED MANUSCRIPT Redox titrations followed by EPR or visible absorption spectra have been used to determine the reduction potential of CuA center and a value of + 260 mV vs SHE, at pH 7.5, was estimated for this center in P. stutzeri N2OR [25], in accordance with the reduction potential determined for CuA center of the enzymes isolated from P. pantotrophus [55] and M. hydrocarbonoclasticus [26, 68]. Extended X-ray absorption fine structure (EXAFS) studies on N2OR reinforced the existence of a direct metal-metal bond in the mixed valence CuA center with a distance

PT

of 2.43 Å and provided evidences of a Cu2S2 core with a 2.2 Å Cu-S interactions [69]. These data were validated some years later when the first X-ray structure of a N2OR solved

at

a

2.4

Å

resolution,

for

the

enzyme

isolated

RI

was

from

M.

SC

hydrocarbonoclasticus, confirming the coordination sphere of CuA center and showing identical Cu-Cu (2.47 Å) and Cu-S (2.26 Å - 2.31 Å) distances [49]. The two cysteine

NU

residues (C561 and C565) coordinated, through their Sγ atoms, both coppers atoms (CuA1 and CuA2), being CuA1 also coordinated by Nε2 atom of a histidine (H526) and a Sγ of a methionine (M572) and CuA2 by Nε2 atom of a histidine (H569) and the

MA

carbonyl group of the tryptophan (W563) (Figure 2A). CuA center is bound to the Cterminal cupredoxin folded domain in the loop region between β8 and β9 strands [49] of

D

the β-barrel structure (numbering according to P. denitrificans N2OR structure). Besides the initial M. hydrocarbonoclasticus N2OR structure, higher resolution

PT E

structures of N2ORs also purified in the presence of oxygen were solved for P. denitrificans (1.6 Å) [50] and A. cycloclastes (1.86 Å) [70] N2ORs. In these, CuA center has similar coordination and copper geometries.

CE

More recently, a structure of N2OR purified and crystallized under anoxic conditions was obtained for the enzyme isolated from P. stutzeri [71]. Interestingly, an alternative

AC

conformation of CuA center was reported for most of the monomers in the asymmetric unit of P. stutzeri N2OR, in which the imidazole ring of H583 (equivalent to H526 in P. denitrificans) is rotated ~130°, no longer coordinating CuA1 and forming two hydrogen bonds with two highly conserved residues, S550 and D576 (Figure 2B) [71, 72]. Even in the recently reported structure of the apo-form of S. denitrificans N2OR [53], this histidine was found in the flipped conformation and hydrogen bonded to D576 (numbering according to P. stutzeri N2OR primary sequence), though the causes and function of this different coordination are still poorly understood.

- Insert Figure 2 here 8

ACCEPTED MANUSCRIPT

Furthermore, the conserved D576 residue has been proposed to be the entry point for the electrons donated from the small electron shuttle proteins in P. stutzeri N2OR, based on the docking studies between N2ORs and its redox partners [73]. Indeed, these same studies applied to M. hydrocarbonoclasticus N2OR, P. denitrificans N2OR and A. cycloclastes N2OR with the respective redox partners, also identified this conserved aspartate as being putatively involved in the intermolecular electron transfer pathway

PT

[73].

Moreover, in the case of P. stutzeri N2OR, the mechanism involving the switching of

RI

H583 has been postulated to play a role in the intramolecular electron transfer from

SC

CuA center to the catalytic center, as it is flipped away and non-coordinating CuA1 in the absence of substrate but switches to its coordinating position in the presence of

NU

substrate [71]. This substrate bound structure is also quite intriguing since the enzyme is in the fully oxidized form and no change in the oxidation state of any of the copper

MA

centers was reported upon data acquisition or substrate incubation.

"CuZ center" - The catalytic center

D

The existence of a catalytic center in N2OR, assigned as “CuZ center”, is long known [67, 74]. However, its spectroscopic and structural characteristics took years to discern

PT E

and are still a matter of debate due to its unusual features and difficulties in obtaining an enzyme preparation with a single form of "CuZ center" (either as CuZ(4Cu2S) or as CuZ*(4Cu1S)).

CE

In the early studies, as it was recurrent to obtain enzyme preparations with low copper content of ~8 Cu atoms per functional dimer (currently known to bind 12 Cu atoms

AC

[49]). Therefore, in combination with the available spectroscopic data, it was postulated the existence of a second binuclear copper center, in addition to CuA center, coordinated by a thiolate group [25, 54, 67]. Although, the S bridging character was detected [75], it could only be assigned to an inorganic sulfur, as there are only two conserved cysteine residues in the primary sequence of N2ORs, which are located at the C-terminal domain and coordinate CuA center. The copper atoms of "CuZ center" were proposed to be coordinated by seven conserved histidine residues located at the Nterminal domain. The overall coordination was clarified when the three-dimensional structure of N2OR was solved, in which a central inorganic sulfur atom was identified to

9

ACCEPTED MANUSCRIPT be µ4-bridging the copper atoms (here assigned as CuI-IV) arranged in a distorted tetrahedron [50, 70, 76]. Nevertheless, the electron density coordinating the four copper atoms of "CuZ center" was fitted to an oxygen atom in the first reported N2OR structure [49]. Although the oxygen was assigned as the central atom, early spectroscopic studies had suggested the presence of another copper-sulfide center, with different properties from CuA center [75]. At the same time as the release of the first X-ray structure of M.

PT

hydrocarbonoclasticus N2OR, sulfide quantifications and analysis of resonance Raman spectra undoubtedly showed the presence of labile sulfide in the composition of "CuZ

electron

density,

together

with

sulfur

quantifications

in

M.

SC

center"

RI

center" [63, 77]. Afterwards, new structural data and re-analysis of the first "CuZ

hydrocarbonoclasticus and in P. denitrificans N2ORs showed that the magnitude of the

NU

electron density of the central atom corresponded to a sulfide instead of an oxygen atom [50].

Common to all structures, Cu-S1 bonds have a similar length (~ 2.3 Å), while copper-

MA

copper distances vary, with distances between CuI-CuIV and CuI-CuIII being longer (ranging from 3.3 to 3.6 Å) than CuII-CuIV and CuII-CuIII distances (ranging from 2.4 to 2.8 Å) [50, 70, 78].

D

As mentioned, different properties of the “CuZ center” have been observed mostly due

PT E

to the heterogeneity of "CuZ center" in the enzyme preparation, CuZ(4Cu2S) and CuZ*(4Cu1S). However, the understanding that the oxidation states of each of these forms can be studied by selectively reducing the other form, simultaneously with the

CE

reduction of CuA center (using sodium ascorbate, dithionite or reduced viologens, as reducing agents), and that the amount of each “CuZ center” form in an enzyme

AC

preparation can be estimated by EPR spin quantification, has led to unprecedented advances in the spectroscopic and structural characterization of "CuZ center" [26, 79]. These properties, as well as the identification of which is the catalytically relevant form of this center and the catalytic intermediate species in the catalytic cycle will be discussed in the next sections.

1.4 Structural and spectroscopic features of CuZ(4Cu2S) N2OR can be isolated with “CuZ center” mainly as CuZ(4Cu2S), in an oxidized [2Cu2+2Cu1+] or reduced [1Cu2+-3Cu1+] state, together with either an oxidized or reduced CuA center. This form of the “CuZ center” can be easily reduced or oxidized and the 10

ACCEPTED MANUSCRIPT reduction potential of the pair [2Cu2+-2Cu1+] / [1Cu2+-3Cu1+] was estimated to be + 60 mV vs SHE at pH 7.5, in P. pantotrophus N2OR [55]. The crystallographic structures of P. stutzeri N2OR, isolated with CuZ(4Cu2S) in the oxidized state were solved at 1.7 and 2.1 Å resolution and reported for the first time in 2011 (Figure 3A) [71]. The histidine residues H326, H382, H433, H130 and H178, coordinate the copper atoms by Nε2 and the other two, H494 and H129, coordinate the copper atoms via Nδ1 (numbered according to P. stutzeri N2OR primary sequence) [71].

PT

These residues are located either in the blades (H129, H130, H178, H382 and H433) or on the top of the β-propeller domain (H326 and H494) [49, 50, 71]. Although, the core

RI

structure of "CuZ center" is conserved, in CuZ(4Cu2S) a second sulfur atom was

SC

modeled into the electron density between CuI and CuIV in the position where a waterderived ligand had been identified in the other existing structures (vide infra) [71].

NU

Typical metal-sulfur bonds were observed with CuI-S2 and CuIV-S2 being 2.5 Å and 2.3 Å, respectively [71].

MA

- Insert Figure 3 here –

Spectroscopically, the oxidized [2Cu2+-2Cu1+] state is characterized by a strong

D

absorption band at 545 nm (ε ≈ 5000 M-1 cm-1) and a shoulder at 635 nm [55, 75, 80],

PT E

corresponding to five transitions at room temperature, assigned to S→Cu CT (Figure 4, Table 1) [80]. This CT character was addressed by resonance Raman spectroscopy upon excitation at 568 or 676 nm, whose profiles exhibited enhanced vibrational modes at

CE

350 and 405 cm-1, that are S isotope-sensitive. In particular, excitation at 568 nm, showed a prevalent Cu-S stretching mode at 350 cm-1, while similar intensities between

AC

both vibrational modes were observed upon laser excitation at 676 nm [63, 80]. The resonance Raman profiles combined with Density Functional Theory (DFT) calculations associated the nature of the transitions to the presence of a sulfide (µ2S2-) edge ligand that has a pKa ≤ 3 [80]. Moreover, upon reduction of CuA center with sodium ascorbate, neither MCD or EPR spectra presented signals of the oxidized CuZ(4Cu2S) confirming the diamagnetic behavior (S = 0) of this oxidation state [55]. The [1Cu2+-3Cu1+] state is characterized by a strong absorption band in the visible spectrum at 670 nm (ε ≈ 3000 M-1 cm-1) (Figure 4, Table 1) [55]. Different transitions were observed by simultaneous analyses of low temperature absorption and MCD spectra, which were assigned to S→Cu CT transitions, inter-valence (IT) transitions, 11

ACCEPTED MANUSCRIPT low Cu d→d transitions and His→Cu CT [67, 80, 81]. Additionally, the resonance Raman spectrum, acquired by excitation at 676 or 697 nm, presented enhanced Cu-S stretching modes, particularly at 203, 378, 450 and 492 cm-1, being the 378 cm-1 vibration assigned to the µ4S, since it is common to the CuZ*(4Cu1S) profile (vide infra) [80]. On the other hand, vibrations at 203 cm-1 were assigned to the Cu-S stretching modes of the edge sulfur ligand, while 450 and 492 cm-1 vibrations are D2Oisotope sensitive and assigned as S-H bending modes, suggesting that this oxidation

PT

state has a thiolate (µ2SH-) edge ligand between CuI and CuIV [80] (Table 1). The resonance Raman profiles obtained at pH 7.8 and pH 10 showed unperturbed behavior

SC

RI

of S-H bending modes, suggesting that the thiolate has a pKa ≥ 11 [80].

NU

- Insert Table 1 here -

The reduced CuZ(4Cu2S), [1Cu2+-3Cu1+], is characterized by an X-band EPR axial spectrum due to the delocalized spin density (STotal = 1/2) over the copper dx2-y2 orbitals

MA

[80]. The simulation of the derivative spectrum estimated a g║ = 2.152, a g┴ =2.042 and three identical hyperfine constants, of A║ = 5.6 mT, were required to explain the five

D

hyperfine lines observed in the EPR spectrum [80]. Moreover, the requirement of the three hyperfine constants is attributed to the spin density being distributed over CuI (17

PT E

%), CuII (11 %) and CuIV (10 %), in a 2:1:1 ratio, since these atoms are localized in the same plane as the sulfur ligands [80]. Besides the oxidation states described above, no other reduced states have been reported

CE

for CuZ(4Cu2S). Indeed, even upon prolonged incubation with reduced viologens, the fully reduced [4Cu1+] state cannot be attained by CuZ(4Cu2S) [79], which can be

AC

attributed to the presence of the sulfur atom in the CuI-CuIV edge.

1.5 Structural and spectroscopic features of CuZ*(4Cu1S) A N2OR sample with “CuZ center” mainly as CuZ*(4Cu1S) is usually obtained when its purification is performed from long term frozen crude extracts [49, 50, 70, 76], with either CuA center oxidized or reduced. Such an enzyme sample can also be obtained in two other conditions, isolated in the absence of oxygen from a P. denitrificans double mutant in nirXnosX [15, 38] or from cells grown at low pH (Carreira et al., unpublished results).

12

ACCEPTED MANUSCRIPT Up-to-date, N2OR structures containing “CuZ center” mainly as CuZ*(4Cu1S) have been

reported

for

the

enzyme

isolated

from

three

microorganisms

M.

hydrocarbonoclasticus, P. denitrificans and A. cycloclastes (Figure 3B) [49, 50, 70, 76]. The coordination sphere of the tetranuclear CuZ*(4Cu1S) is very similar to CuZ(4Cu2S), except that in the edge of CuI-CuIV there is a water-derived ligand. The comparison between the three known structures of CuZ*(4Cu1S) highlights that the difference lies in the CuI-CuIV edge: an oxygen atom bound to CuIV, was assigned to a

PT

water molecule or a hydroxide, in the cases of M. hydrocarbonoclasticus and P. denitrificans N2ORs, while two oxygen atoms bridged to a different copper atom (CuI

RI

and CuIV) were modeled in the electron density of A. cycloclastes N2OR [50, 70, 76].

SC

The spectroscopic properties of the resting CuZ*(4Cu1S) have been investigated in detail [26, 63, 81, 82]. The MCD spectra of the CuZ*(4Cu1S) form, acquired at variable

NU

temperature and field, were simulated for a STotal = 1/2, indicating a paramagnetic behavior for the resting state [83]. The oxidation state of the resting CuZ*(4Cu1S) was determined by Cu K-edge X-ray absorption spectroscopy to be [1Cu2+-3Cu1+], attending

MA

to the possible oxidation states of the copper atoms for a STotal = 1/2 [83]. The resting CuZ*(4Cu1S) [1Cu2+-3Cu1+] is characterized by a strong absorption band at 640 nm (ε ≈ 4000 M-1 cm-1) (Figure 4, Table 1) and intense transitions in the visible

D

region in both CD and MCD spectra, assigned to S→Cu CT, His→Cu CT, Cu→Cu IT

PT E

and also high energy d→d transitions [26, 67, 77, 81, 82]. The resonance Raman spectrum at pH ~7.3, excited at 624.4 nm (or 647 nm, which showed similar vibrational frequencies) presented three vibrational modes at 366, 386 and 415 cm-1, assigned to

enriched with

CE

Cu-S stretching vibrations, as they exhibited 16

O/18O, no

18

S shift [63, 82]. When the sample is

O shifts were observed, upon excitation at 600 nm [81].

Nevertheless, at pH 10.5 a 9 cm-1

AC

34

16

O/18O shift of the 415 cm-1 peak was detected,

suggesting a Cu-O vibrational mode at high pH [81]. Selective reduction of CuA center with sodium ascorbate enabled the determination of the EPR features of the resting CuZ*(4Cu1S). Q- and X-band EPR showed axial signals and hyperfine splitting in the g║ region, however the X-band hyperfine pattern was not well resolved due to the broadening of the signals (Figure 4), which could be improved in the second derivative spectrum [80, 83]. Thus, accurate g-values, g║ = 2.16 and g┴ = 2.04, were determined based on the Q-band EPR spectra of M. hydrocarbonoclasticus N2OR, which presented similar values to those previous reported for P. pantotrophus N2OR (Table 1) [83, 84]. Two hyperfine coupling constants were estimated (A║ = 6.1 13

ACCEPTED MANUSCRIPT mT and A║ = 2.4 mT) due to the unpaired spin density delocalized over two coppers in a ~5:2 ratio [83]. The crystallographic structure of M. hydrocarbonoclasticus N2OR has been used in DFT calculations to investigate the electronic configuration of the ground-state of CuZ*(4Cu1S) [83, 85]. Taking into account that the spin distribution is affected by the ligand in the CuI-CuIV edge, different models were considered. Initially a model with a water molecule in the CuI-CuIV edge was proposed, in which the spin density would be

PT

distributed over CuI and CuII atoms [83]. Other models, involving the following ligands, have been considered: a H2O bound to CuI, a OH- in the CuI-CuIV edge, a OH- in the

RI

CuI-CuIV edge bonded to a lysine residue, or a OH- ligand H-bonding to a protonated

SC

lysine residue [81]. The DFT model of CuZ*(4Cu1S) that better explains the spectroscopic properties has a OH- ligand occupying the CuI-CuIV edge, being closer to CuI (2.00 Å) than to CuIV (2.09 Å). Two other residues, K397 and E435, located in the vicinity of the center, were

NU

included in this model, being hydrogen bonded to each other (but not to the metal center) (Figure 3B) [81, 86].

MA

At high pH, there is a shift of 9 cm-1 in the resonance Raman spectra, which is now known to be due to the deprotonation of the K397 residue [81, 86]. The deprotonation of this nearby residue, induces shorter bonds of OH- with CuI and CuIV atoms, increasing the vibrational energy.

D

Thus, the edge of CuI-CuIV in CuZ*(4Cu1S) is proposed to be occupied with a

PT E

hydroxide that is mainly bound to CuI, as the spin density mainly resides on this atom (26 %), with CuIV (13 %) having a smaller contribution [80]. The dominant spin density attributed to CuI has been explained by the higher coordination number of this atom,

CE

relative to the other copper atoms, since it is coordinated by 4 ligands [81].

AC


Contrary to CuZ(4Cu2S), the resting CuZ*(4Cu1S), in the [1Cu2+-3Cu1+] oxidation state, is not easily reduced or oxidized. However, it can be fully reduced to the [4Cu1+] oxidation state after prolonged incubation (3 – 5 h) with a large excess of a reduced viologen [46, 79, 87]. The reduction potential of the pair [1Cu2+-3Cu1+] / [4Cu1+] has not yet been determined, since reducing the resting CuZ*(4Cu1S) in a potentiometric redox titration is very difficult, even though CuZ*(4Cu1S) in the [4Cu1+] state can be easily oxidized to the resting state upon addition of potassium ferricyanide [68]. In the [4Cu1+] oxidation state, CuZ*(4Cu1S) does not contribute to the absorption spectrum in 14

ACCEPTED MANUSCRIPT the visible region, due to the d10 configuration of its copper atoms. Likewise, both MCD and EPR features are absent, since the species are diamagnetic with no unpaired electrons (S = 0).

1.6 Substrate binding site The combined studies on the structural, kinetics and spectroscopic properties of N2OR pin pointed the CuI-CuIV water-derived edge as the probable substrate binding site in

PT

CuZ*(4Cu1S) [49, 82, 85, 88]. The unveiling of crystallographic structures of N2OR was one of the major findings for this assumption, showing that Cu IV is coordinated by

RI

the side chain of only one histidine, while the other copper atoms are coordinated by

SC

two histidines, together with the presence of a water-derived ligand in the CuI-CuIV edge, regarded as a labile molecule that would be replaced by the substrate nitrous oxide

NU

[66, 76].

EPR studies on the resting CuZ*(4Cu1S) have identified a spin density delocalized over CuI and CuIV that favors the N2O interaction, and supports this hypothesis [83].

MA

Additionally, DFT calculations on "CuZ center" models performed by Solomon and collaborators, suggested a µ-1,3-N2O coordination as the most favorable binding mode of nitrous oxide at the CuI-CuIV edge, in the lowest energy configuration of

D

CuZ*(4Cu1S) in the [4Cu1+] oxidation state [87]. In this configuration the terminal N-

PT E

atom of the N2O molecule is coordinating CuI and the O atom is bonded to CuIV, forming a ~139° angle [66, 85-87] (Figure 5). The proposed substrate binding site was also strengthened by the X-ray structure of an

CE

inhibited form of A. cycloclastes N2OR, solved at 1.7 Å resolution [70]. The exposure of N2OR to the inhibitor iodide lead to the formation of a blue adduct characterized by a

AC

strong absorption band at 650 nm [70]. In the structure of the adducted-N2OR, the iodide was modeled in the electron density maps between CuI and CuIV [70]. The effect of iodide in the specific activity of N2OR is unknown, but it can rapidly and strongly bind to a synthetic µ4-sulfide tetranuclear copper complex, a model of CuZ*(4Cu1S), in the [4Cu1+] oxidation state, originating an unreactive species [89].


Others exogenous ligands, including fluoride, cyanide, azide, nitric oxide, carbon monoxide and hydrogen peroxide, presented different reactivities and spectroscopic 15

ACCEPTED MANUSCRIPT effects on the interaction with the enzyme [24, 54, 76, 90, 91]. Indeed, some of those inhibitors bind irreversibly to the enzyme or chelate the copper atoms, as observed by Haltia and co-authors in the structures of N2OR crystals treated with cyanide and hydrogen peroxide, in which a lower occupancy of the copper atoms of the “CuZ center” was reported [76]. The binding mode of the other exogenous ligands is not known, and it remains to be elucidated whether their binding induces any conformational change in the copper centers or in the enzyme. These studies will

PT

provide additional insights into the complex catalytic mechanism of this enzyme and clarify the substrate interaction with the enzyme.

RI

There is, to our knowledge, only one N2OR structure crystallized in the presence of

SC

nitrous oxide, which was solved for P. stutzeri N2OR (at 2.24 Å resolution) after pressurization of anaerobically prepared purple crystals with nitrous oxide [71]. In this

NU

preparation, CuA center is in the [1Cu1.5+-1Cu1.5+] oxidation state and "CuZ center" as CuZ(4Cu2S) is in the [2Cu2+-2Cu1+] oxidation state [71], a form of the enzyme that is not catalytically competent. In this structure, nitrous oxide is not coordinated to any

MA

metal center, being located above the CuII-µ4S-CuIV face in a linear mode, between both CuA and CuZ(4Cu2S) centers, and not even close to the edge position of CuI-CuIV that

D

is in this case is occupied by a sulfur atom (Figure 5) [71, 78]. Einsle and collaborators proposed a hydrophobic channel for substrate access to this location and reorientation of

PT E

the N2O molecule is envisaged through the residues F621 and M627 (Figure 5) [71, 72]. The only significant difference between the two P. stutzeri N2OR structures, in the presence and absence of substrate, is, as mentioned, at the CuA center, with H583

CE

imidazole ring being rotated and coordinating CuA1 atom in the presence of nitrous oxide [71], position that was observed in the others structures of N2OR with "CuZ

AC

center" as CuZ*(4Cu1S). Thus, there is no reported change in the oxidation state of either copper center upon the binding of substrate (vide infra).

1.7 The catalytic cycle Specific activity and activation mechanism of N2OR The N2OR specific activity is expressed in µmol N2O reduced. min-1. mg-1 N2OR and can be determined by spectrophotometric or gas chromatography assays [92, 93]. In the spectrophotometric assays the rate of N2O reduction is indirectly determined by following the oxidation of reduced viologen dyes at 600 nm. These assays can either be initiated upon addition of the enzyme or water saturated-N2O to the reaction mixture, 16

ACCEPTED MANUSCRIPT which contains the reduced viologens and nitrous oxide or the enzyme, respectively [46, 92]. In alternative, the activity can be directly measured based on the substrate (N 2O) consumption and/or product (N2) formation by gas chromatography [25, 93]. The rate of nitrous oxide reduction by P. stutzeri crude cell extracts, determined using the spectrophotometric assay, ranged from 48 to 72 µmol N2O reduced. min-1. mg-1 N2OR [25, 94]. Nevertheless, the activity of P. stutzeri N2OR purified in the presence and absence of oxygen, measured in vitro using the spectrophotometric assay, is on

PT

average 1.8 and 4.3 µmol N2O reduced. min-1. mg-1 N2OR, respectively [25]. Intriguingly, none of the activities determined for the isolated enzyme, independently of

RI

the proportion between the two "CuZ center" forms, are in line with the N2O reduction

SC

rates obtained for the crude extract, suggesting that those values correspond to an unready state of the enzyme [57]. These specific activity values are in agreement with

NU

the ones measured for N2ORs purified from P. pantotrophus [55], Alcaligenes xylosoxidans [27] and Pseudomonas aeruginosa [43, 95], using methyl or benzyl viologen as electron donors (ranging from 1 to 10 µmol N2O reduced. min-1. mg-1

MA

N2OR).

As one can imagine, these spectrophotometric assays are far from being physiologic, as

D

the reduction potential of the electron donors used (methyl viologen, - 450 mV vs SHE, pH 7.0 and benzyl viologen, - 374 mV vs SHE, pH 7.0 [96]) are below of what is

PT E

expected to be found in the periplasm of Gram-negative bacteria. Nevertheless, physiologically relevant electron donors have been tested and identified that are able to donate electrons to N2OR in in vitro activity assays [46, 48]. However, even in the

CE

presence of reduced electron donor proteins, hypothesized to induce conformational changes in N2OR [44], the specific activities cannot explain the high N2O-reduction

AC

rates attained when using either crude cell extracts or whole-cells [14, 25, 41, 44, 46]. Nonetheless, a large excess of electron donor protein over the enzyme (~26000-fold) was recently reported to substantially increase the activity of A. cycloclastes N2OR [48], but no molecular explanation was provided nor similar observations have been reported for any other enzyme. To be fair, in vivo other mechanisms might be involved or the existence of other relevant electron donor protein cannot be ruled out. On the other hand, a high specific activity, of 160 µmol N2O reduced. min-1. mg-1 N2OR, was reported for the enzyme isolated from W. succinogenes (clade II N2OR) [34], an indication that it does not require activation. This could be attributed to the fact 17

ACCEPTED MANUSCRIPT that this enzyme does not require a soluble electron donor protein, being able to receive electrons directly from membrane proteins via menaquinol [33], with the additional domain playing a role in the electron transfer pathway via CuA center to "CuZ center". In fact, clade I N2ORs may require additional accessory proteins to attain such activity, possibly NosR, since this protein was shown to be essential for the whole cells to reduce N2O [14] or may require the formation of a supra-complex as recently reported in P. aeruginosa [97]. Thus, what is clear is the requirement of a chemical activation step of

PT

the purified clade I N2OR in order to reach its maximum catalytic activity [87, 98]. Enzyme activation has been proposed to be attained through two processes: prolonged

RI

incubation with reduced viologens or with an alkaline solution, though the activity

SC

values attained with this later process are not as high as the ones with the former, vide infra [25, 87, 92].

NU

The pH has a distinct effect on the enzyme activation and N2O reduction [99]. In fact, alkaline activation followed by the activity assays at neutral pH seems to be divergent between N2ORs isolated from different species: ~ 1.5 fold increase, after 14 h at pH 9.0,

MA

for P. pantotrophus N2OR with “CuZ center” as CuZ*(4Cu1S)/CuZ(4Cu2S) [29], ~ 1.31.4 fold increase, after 3 h at pH 10.0, for A. xylosoxidans N2OR with “CuZ center” as CuZ(4Cu2S) [27], ~ 14 fold increase, after 16 h at pH 9.8, for P. stutzeri N2OR with

D

“CuZ center” as CuZ(4Cu2S) [25], or ~ 50-75 fold increase after incubation at pH 10.0

PT E

for P. aeruginosa N2OR with “CuZ center” as CuZ(4Cu2S) [95]. Thus, when the enzyme is incubated at alkaline pHs the specific activity increases more than when it is activated at neutral or more acidic pH, in particular for CuZ(4Cu2s), in spite of the

CE

heterogeneous increment in activity between N2ORs isolated from different organisms [99]. The exact molecular mechanism for the alkaline activation is unknown, but could

AC

be related with the deprotonation of a residue nearby "CuZ center" (vide infra) [99]. Incubation with a reduced viologen increases the activity by two orders of magnitude [87, 98]. Nevertheless, the activation of N2OR is not only dependent on the incubation with a compound with a very low reduction potential but also on the type of reducing agent and on the incubation time with the reduced viologen: dithionite (with a reduction potential of - 471 mV vs SHE at pH 7.0 [100]), by itself it does not activate N2OR (with "CuZ center" in any of the two forms) [87, 98], while N2OR with "CuZ center" as CuZ*(4Cu1S) after 3 h incubation in the presence of excess of reduced methyl viologen attains maximum specific activity.

18

ACCEPTED MANUSCRIPT This activation process is related to the reduction of CuZ*(4Cu1S) (k ~ 1.2x10-3 s-1 at pH 7, in the presence of reduced methyl viologen), in which CuZ*(4Cu1S) in the [1Cu2+-3Cu1+] oxidation state is reduced to [4Cu1+] [87]. However, it is important to point out that this is a very slow reduction rate to be part of the catalytic cycle [79, 85]. The reduction process is marked by a decrease in the absorbance at 640 nm and can be followed by the intensity decay of the EPR signal that is correlated with the reduction of the "CuZ center" as CuZ*(4Cu1S), in the [1Cu2+-3Cu1+] to the diamagnetic [4Cu1+]

PT

state [87]. The decrease of the EPR signal, is concomitant with the increase in activity, suggesting that [4Cu1+] state is the in vitro catalytic relevant oxidation state of

RI

CuZ*(4Cu1S) [87].

SC

A pKa of 9.0 ± 0.2 for the reduction process (from [1Cu2+-3Cu1+] to [4Cu1+]) was determined by following the rate of the EPR signal decay at different pHs [81]. It is

NU

important to mention that the features of the visible spectrum do not disappear completely, due to the presence of a small amount of N2OR with "CuZ center" as CuZ(4Cu2S), since it is not possible to have a N2OR sample with a homogeneous form

MA

of "CuZ center", and CuZ(4Cu2S) is not fully reduced by reduced methyl viologen (nor

D

sodium dithionite) to the [4Cu1+] oxidation state [79].

The catalytically competent forms

PT E

The hypothesis that other oxidation states of N2OR "CuZ center" are able to react with N2O was recently investigated [79], and not all are catalytically competent. For instance, CuZ(4Cu2S) in the [2Cu2+-2Cu1+] state and resting CuZ*(4Cu1S) in the

CE

[1Cu2+-3Cu1+] state, do not react with N2O even when CuA center is reduced (no spectroscopic changes were observed in the presence of N2O) [79]. On the other hand,

AC

N2OR with "CuZ center" as CuZ(4Cu2S) in the [1Cu2+-3Cu1+] state and with CuA center reduced, slowly reacts with N2O (k = 0.6 h-1), as supported by MCD and resonance Raman studies [79]. In this reaction, both CuA and CuZ(4Cu2S) centers were re-oxidized to the [1Cu1.5+-1Cu1.5+] and [2Cu2+-2Cu1+], respectively, in a two-electron process [79]. However, such low turnover numbers suggests that CuZ(4Cu2S) in the [1Cu2+-3Cu1+] state is not relevant for the catalytic cycle and thus is unlikely to be the active form in vivo [57, 79], possibly having a role as a protective form of the enzyme when either the substrate or electrons are not available to complete the catalytic cycle. The fully reduced state of CuZ*(4Cu1S), [4Cu1+], is obtained after prolonged exposure of N2OR to an excess of reduced (methyl or benzyl) viologen. N2OR with "CuZ center" 19

ACCEPTED MANUSCRIPT is this fully reduced state reacts with the substrate even after removal of reduced methyl viologen, demonstrated by the partial re-appearance of the visible and EPR spectral features and supported by GC-MS detection of the product (dinitrogen), when a

15

N

labeled N2O was used [98]. The spectroscopic evidences are the appearance of the CuA center features in the visible spectrum, at 480, 540 and 800 nm, and of a band at 640 nm, corresponding to the re-oxidation of the CuZ*(4Cu1S) to [1Cu2+-3Cu1+] oxidation state, which is associated with the re-emergence of a hyperfine pattern in the EPR

PT

spectrum, involving two electrons [98].

The dependence of reduced viologen and N2O concentration in the reaction, as well as

RI

the pH dependence has been investigated by steady-state kinetics [43, 46, 99]. High

SC

turnover numbers of 321 s-1 and 162.9 s-1 were obtained for M. hydrocarbonoclasticus N2OR [46] and A. cycloclastes N2OR [43], respectively, showing that [4Cu1+] is

NU

kinetically competent in reducing N2O. The specific activity of A. cycloclastes N2OR is pH dependent, with two pKa values at 6.8 and 9.8 [99]. The specific activity of M. hydrocarbonoclasticus N2OR is also pH dependent, with a maximum at pH 8 and a bell-

MA

shaped profile, with two pKa values of 6.19 ± 0.05 and 9.15 ± 0.05 [46, 86]. In the presence of its physiologic relevant electron donor, cytochrome c552, the maximum

D

activity is attained at lower pHs, with a pKa of 8.3 [46]. The pH dependence of the specific activity of P. pantotrophus N2OR in the presence of mammalian cytochrome c

PT E

exhibited, similarly to M. hydrocarbonoclasticus N2OR, higher activities at lower pHs, with a maximum observed at pH ~ 5.6 [29]. In single turnover experiments, an intermediate species was identified and named CuZ0

CE

[68]. This intermediate species was observed after addition of equimolar amount of N2O to fully reduced N2OR. The rate of formation of this species has a lower limit of 200 s-1,

AC

which is compatible with the turnover number of M. hydrocarbonoclasticus N2OR determined in the steady-state kinetics [79]. CuZ0 is characterized by the emergence of a strong absorption band at 680 nm (ε ≈ 2000 M-1 cm-1, identified after subtraction of CuA and CuZ(4Cu2S) contributions), which is concomitant with the oxidation of CuA center, as bands at 480, 540 and 800 nm were detected (Table 1) [86]. In fact, this intermediate species can only be observed during the two initial minutes after addition of N2O, as it decays at a rate of kdecay ~ 5x10-3 s-1 [68]. This decay is accompanied by a decrease of the intensity at 680 nm, which is concomitant with the increasing intensity at 640 nm band, characteristic of the resting CuZ*(4Cu1S) [68]. Furthermore, this intermediate is a competent species in the rapid 20

ACCEPTED MANUSCRIPT turnover as it decays at the same rate as the reduction rate of N2O in the steady-state kinetics assays [68]. Recently, the nature and spectroscopic characterization of this intermediate species was addressed [86]. MCD and absorption spectra allowed the identification of six transitions assigned as µ4S2-→Cu CT transitions, His→Cu CT transitions and also d→d transitions [86]. Two of the three transitions assigned as µ4S2-→Cu CT transitions, showed different relative intensities compared to those observed for the resting CuZ*(4Cu1S)

PT

[82, 86]. Furthermore, in the CuZ0 spectra the transition corresponding to the spin delocalization onto CuI is lower relative to the transition corresponding to the spin

RI

delocalization onto CuIV, as opposed to what was observed in the resting CuZ*(4Cu1S),

SC

suggesting a more homogeneous copper spin distribution between Cu I and CuIV in CuZ0 [82, 86].

NU

The X-band EPR spectrum (S = 1/2) exhibits similar g-values to those determined for the resting CuZ*(4Cu1S) (Table 1) [68]. The axial signal with g║ = 2.177 > g┴ = 2.05 > 2.0 indicates that the spin resides in the Cu dx2-y2 orbitals [68, 86]. A 6-lines hyperfine

MA

pattern was identified in the g║ region and simulated with two equal hyperfine constants (A║= 4.2 mT), indicative, contrary to what was observed in the resting CuZ*(4Cu1S), of

D

an unpaired spin equally distributed over the two copper atoms, confirming the analysis of the absorption and MCD spectra presented before [86]. In the resonance Raman

PT E

spectra, upon excitation at 676 nm, an intense vibration and a weak shoulder were identified at 426 and 413 cm-1, respectively, and assigned to Cu-S stretching vibrations [86]. Moreover, the vibration at 413 cm-1 showed a 3 cm-1 shift in solvent O18 isotope-

CE

sensitive and also a 36 % increase in intensity, predicting a ligand with an oxygen atom in the CuI-CuIV edge [86].

AC

Two computational models were hypothesized for CuZ0 in M. hydrocarbonoclasticus N2OR, one with a H2O edge ligand and another with a OH- coordinated terminally to CuIV and hydrogen bonded to K397 [86]. The spectral features of CuZ0 are only reproduced by the model with a OH- coordinated terminally to CuIV (1.93 Å) and stabilized by a H-bond to the protonated lysine (K397), which in turn strongly interacts with the negatively charged glutamate residue (E435) [86].

Catalytic Cycle Different spectroscopies, kinetic assays and DFT calculations were used to unravel and characterize intermediate species in the catalytic cycle of N2OR. Based on our current 21

ACCEPTED MANUSCRIPT knowledge, the activation and catalytic mechanism of CuZ*(4Cu1S) N2OR is presented in Figure 6. This mechanism has been proposed for "CuZ center" as CuZ*(4Cu1S) since the CuZ(4Cu2S) form of this center, even in the reduced state, cannot be considered relevant for catalysis (based on its specific activity), and up-to-now no activation mechanism has been proposed, which can be envisaged to require a sulfur displacement to allow the substrate to bind at the "CuZ center".

PT

In steady-state kinetics assays fast turnover numbers (kcat ~300 s-1) were observed [46], implying that the species involved in the catalytic site have to be rapidly reduced, thus,

RI

the resting CuZ*(4Cu1S) cannot be part of this mechanism due to its slow reduction (k

SC

= 1.2 × 10-3 s-1, pH 7) [87]. During the slow activation process of N2OR with reduced methyl viologen, CuA center is reduced to the [1Cu1+-1Cu1+] oxidation state and the

NU

resting CuZ*(4Cu1S) is reduced to [4Cu1+] state (Intermediate 1).

MA


Upon addition of equimolar amount of N2O to the fully reduced [4Cu1+] state, the

D

terminal N atom of the N2O molecule binds to CuI, in a linear configuration [86]. The N-O bond breakdown is initiated with an elongation of the N-O bond, which induces a

PT E

rearrangement of its structure, where the oxygen binds to CuIV and forms a H-bond with the protonated K397 (Intermediate 2). Thus, the substrate molecule establishes a µ-1,3N2O coordination to the CuI-CuIV edge forming a 139º angle [87, 101].

CE

The release of N2 requires two electrons transferred from the fully reduced CuZ*(4Cu1S) [4Cu1+], donated via CuIV and one proton transfer, in a process involving

AC

all the copper atoms from the “CuZ center” [86]. First, one electron is required to cleave the N-O bond and then, a proton is transferred from K397 to the oxygen, resulting in a hydroxide ligand coordinated to CuIV, which triggers the transfer of the second electron to cleave the CuI-N bond and subsequent release of N2 molecule [86]. Once N2 is release, the K397 is re-protonated with a proton from the solvent and a coupled e-/H+ transfer from the CuA center forms the CuZ0 species (Intermediate 3). In this metastable intermediate (with a + 6.4 kcal mol-1 higher free energy than the resting state CuZ*(4Cu1S)), the K397 is positioned to establish one H-bond with the hydroxide ligand coordinated to CuIV and another H-bond to the E435, stabilizing the CuZ0 species [86]. 22

ACCEPTED MANUSCRIPT In order to complete the catalytic cycle, CuZ0 intermediate is reduced by rapid intramolecular electron transfer (kIET > 0.1 s-1) via CuA center [86]. This reduction of the intermediate species was very recently addressed through the addition of sodium ascorbate. In fact, the intramolecular reduction of CuZ0 occurs at least 104 faster than intramolecular reduction of resting CuZ*(4Cu1S). This suggests that CuZ0 can be rapidly reduced via electron transfer from CuA center [86]. Two possible intramolecular electron transfer routes have been proposed for M.

PT

hydrocarbonoclasticus N2OR, based on molecular docking simulations (Figure 5) [73, 102]. In one pathway the electron would be transferred from W563, coordinating CuA

RI

center, to the F564 entering the catalytic center via the oxygen positioned in the Cu I-

SC

CuIV edge. The other pathway would involve electron transfer from the CuA ligand H569 to M570, and subsequently to H128, which coordinates CuII of “CuZ center”.

NU

One the other hand, in the absence of an external electron donor, the H-bond of CuZ0 to K397 is disrupted and the hydroxide binds also to CuI, and CuZ0 decays (kdecay ~ 5x10-3 s-1, pH 7.6) to the resting CuZ*(4Cu1S) form [68]. This inactivation is favored at higher

MA

pH, considering that K397 would be deprotonated [81, 86].

D

1.8 Concluding remarks

N2OR has a relevant environmental role in the N-cycle, as being the last enzyme of the

PT E

denitrification pathway and being directly involved in nitrous oxide detoxification. Its catalytic center has a completely distinct nature as a sulfide-bridged tetranuclear copper

last 35 years.

CE

center and thus it has been a major subject of research on the bioinorganic field in the

Over the years, different kinetic and spectroscopic properties of this multicopper

AC

reductase were observed, pin pointing the existence of two distinct forms of the “CuZ center”, CuZ*(4Cu1S) and CuZ(4Cu2S). Moreover, the recent X-ray structural data revealed that those two forms are structurally different, being distinguished by a second sulfur atom at the CuI-CuIV edge in the CuZ(4Cu2S) form. The relevance of the CuZ(4Cu2S) form in the catalytic cycle of the enzyme in vivo, remains to be elucidated, since this form is unlikely to be involved in the rapid N2O turnover either in its oxidized or reduced state, due to its low specific activity that does not explains the high reduction rate observed by whole cells. The involvement of this form of "CuZ center" has been postulated by the fact that this form of the catalytic center has been observed in N2OR when the enzyme is isolated in the absence of oxygen from cells highly active in the 23

ACCEPTED MANUSCRIPT reduction of nitrous oxide. However, an activation mechanism has not yet been either observed or proposed, which could be envisaged as including a sulfur displacement from the CuI-CuIV edge allowing substrate binding directly to the metal center. Furthermore, this form of "CuZ center" could be regarded as having a protective role, as the binding of the sulfur atom might occur in the absence of substrate or electron supply, as a protection from nitric oxide that is transiently formed in actively denitrifying cells.

PT

Progress in understanding the catalytic cycle of nitrous oxide reductase with “CuZ center” as CuZ*(4Cu1S) has been made in the past decade with the identification of

RI

catalytic competent forms and their relevance in the rapid reduction of N2O. Indeed,

SC

spectroscopic studies coupled with DFT calculations were essential to clarify the nature of intermediate species of the catalytic cycle, so far two have been identified and

NU

characterized: fully reduced [4Cu1+] state and CuZ0.

The next challenges will be to answer the raising questions as how a [4Cu1+] "CuZ center" is formed in vivo and which is the mechanism and the structural changes

MA

involved. This process still needs to be addressed and might require others proteins encoded in the nos gene cluster, such as NosR, which has been shown to be essential to

PT E

Acknowledgments

D

maintain in vivo the ability of cells to reduce nitrous oxide.

We thank Fundação para a Ciência e Tecnologia for the financial support through the project PTDC/BBB-BQB/0129/2014 (IM) and the scholarship SFRH/BD/87898/2012

CE

(CC). This work was supported by the Unidade de Ciências Biomoleculares AplicadasUCIBIO, which is financed by national funds from FCT/MEC (UID/Multi/04378/2013)

AC

and co-financed by the ERDF under the PT2020 Partnership Agreement (POCI-010145-FEDER-007728). SRP is an IF fellow supported by FCT.

Author contribution CC wrote the manuscript with contributions from SRP, who also planned the manuscript, and IM critically revised the manuscript.

24

ACCEPTED MANUSCRIPT References [1] in: S. Solomon, D. Qin, M. Manning, Z. Chen, M.

Marquis, K. B. Averyt, M.

Tignor, H.L. Miler (Eds.), IPCC fourth assessment report: climate change, Cambridge (2007. [2] A.R. Ravishankara, J.S. Daniel, R.W. Portmann, Science 326 (2009) 123-125. [3] D.J. Wuebbles, Science 326 (2009) 56-57. [4] D. Richardson, H. Felgate, N. Watmough, A. Thomson, E. Baggs, Trends

PT

Biotechnol. 27 (2009) 388-397.

Trans. R. Soc. Lond. B Biol. Sci. 367 (2012) 1157-1168.

RI

[5] A.J. Thomson, G. Giannopoulos, J. Pretty, E.M. Baggs, D.J. Richardson, Philos.

SC

[6] P. Smith, D. Martino, Z. Cai, D. Gwary, H. Janzen, P. Kumar, B. McCarl, S. Ogle, F. O'Mara, C. Rice, B. Scholes, O. Sirotenko, M. Howden, T. McAllister, G. Pan, V.

NU

Romanenkov, U. Schneider, S. Towprayoon, M. Wattenbach, J. Smith, Philos. Trans. R. Soc. Lond. B Biol. Sci. 363 (2008) 789-813.

[7] W.G. Zumft, P.M. Kroneck, Adv. Microb. Physiol. 52 (2007) 107-227.

MA

[8] W.G. Zumft, Microbiol. Molec. Biol. Rev. 61 (1997) 533-616. [9] S.R. Pauleta, S. Dell’Acqua, I. Moura, Coord. Chem. Rev. 257 (2013) 332-349.

D

[10] K. Jones, The chemistry of nitrogen, in: J. C. Bailar, H. J. Emelöus, R. Nyholm, A.F. Trottman-Dickenson (Eds.), Comprehensive Inorganic Chemistry, Pergamon

PT E

Press, Oxford, 1975, pp. 147–388.

[11] W.B. Tolman, Angew. Chem. Int. Ed. Engl. 49 (2010) 1018-1024. [12] D.F.-T. Tuan, R. Hoffmann, Inorg. Chem. 24 (1985) 871-876.

CE

[13] W.G. Zumft, J. Mol. Microbiol. Biotechnol. 10 (2005) 154-166. [14] P. Wunsch, W.G. Zumft, J. Bacteriol. 187 (2005) 1992-2001.

AC

[15] P. Wunsch, H. Körner, F. Neese, R.J. van Spanning, P.M. Kroneck, W.G. Zumft, FEBS Lett. 579 (2005) 4605-4609. [16] P. Wunsch, M. Herb, H. Wieland, U.M. Schiek, W.G. Zumft, J. Bacteriol. 185 (2003) 887-896. [17] S. Spiro, Philos. Trans. R. Soc. Lond. B Biol. Sci. 367 (2012) 1213-1225. [18] M.J. Torres, J. Simon, G. Rowley, E.J. Bedmar, D.J. Richardson, A.J. Gates, M.J. Delgado, Adv. Microb. Physiol. 68 (2016) 353-432. [19] H. Arai, M. Mizutani, Y. Igarashi, Microbiology 149 (2003) 29-36. [20] B. Liu, P.T. Mørkved, A. Frostegård, L.R. Bakken, FEMS Microbiol. Ecol. 72 (2010) 407-417. 25

ACCEPTED MANUSCRIPT [21] R.N. Van Den Heuvel, S.E. Bakker, M.S.M. Jetten, M.M. Hefting, Geobiology 9 (2011) 294-300. [22] T. Matsubara, H. Iwasaki, J. Biochem. 71 (1972) 747-750. [23] H. Iwasaki, T. Saigo, T. Matsubara, Plant Cell Physiol. 21 (1980) 1573-1584. [24] S.W. Snyder, T.C. Hollocher, J. Biol. Chem. 262 (1987) 6515-6525. [25] C.L. Coyle, W.G. Zumft, P.M. Kroneck, H. Körner, W. Jakob, Eur. J. Biochem. 153 (1985) 459-467.

PT

[26] M. Prudêncio, A.S. Pereira, P. Tavares, S. Besson, I. Cabrito, K. Brown, B. Samyn, B. Devreese, J. Van Beeumen, F. Rusnak, G. Fauque, J.J.G. Moura, M. Tegoni, C.

RI

Cambillau, I. Moura, Biochemistry 39 (2000) 3899-3907.

SC

[27] S. Ferretti, J.G. Grossmann, S.S. Hasnain, R.R. Eady, B.E. Smith, Eur. J. Biochem. 259 (1999) 651-659.

NU

[28] K. Sato, A. Okubo, S. Yamazaki, J. Biochem. 124 (1998) 51-54. [29] B.C. Berks, D. Baratta, J. Richardson, S.J. Ferguson, Eur. J. Biochem. 212 (1993) 467-476.

MA

[30] C.L. Hulse, B.A. Averill, Biochem. Biophys. Res. Commun. 166 (1990) 729-735. [31] K. Yamaguchi, A. Kawamura, H. Ogawa, S. Suzuki, J. Biochem. 134 (2003) 853-

D

858.

[32] A.G. McEwan, A.J. Greenfield, H.G. Wetzstein, J.B. Jackson, S.J. Ferguson, J.

PT E

Bacteriol. 164 (1985) 823-830.

[33] J. Simon, E. Oliver, P.M.H. Kroneck, W.G. Zumft, FEBS Lett. 569 (2004) 7-12. [34] S. Teraguchi, T.C. Hollocher, J. Biol. Chem. 264 (1989) 1972-1979.

CE

[35] K. Heylen, J. Keltjens, Front. Microbiol. 3 (2012) 1-27.

46-52.

AC

[36] X. Liu, C. Gao, A. Zhang, P. Jin, L. Wang, L. Feng, FEMS Microbiol. Lett. (2008)

[37] R. Sanford, D.D. Wagner, Q. Wu, J.C. Chee-Sanford, S.H. Thomas, C. CruzGarcía, G. Rodríguez, A. Massol-Deyáe, K.K. Krishnanif, K.M. Ritalahtig, S. Nisseng, K.T. Konstantinidisb, E.E. Löfflerg, Proc. Natl. Acad. Sci. U. S. A. 109 (2012) 1970919714. [38] N.F. Saunders, J.J. Hornberg, W.N. Reijnders, H.V. Westerhoff, S. de Vries, R.J. van Spanning, J. Bacteriol. 182 (2000) 5211-5217. [39] M. Itoh, K. Matsuura, T. Satoh, FEBS Lett. 251 (1989) 104-108. [40] F.C. Boogerd, H.W. van Verseveld, A.H. Stouthamer, FEBS Lett 113 (1980) 279284. 26

ACCEPTED MANUSCRIPT [41] D.J. Richardson, L.C. Bell, A.G. McEwan, J.B. Jackson, S.J. Ferguson, Eur. J. Biochem. 199 (1991) 677-683. [42] J.W.B. Moir, S.J. Ferguson, Microbiol. 140 (1994) 389-397. [43] K. Fujita, J.M. Chan, J. Bollinger, M.L. Alvarez, D.M. Dooley, J. Inorg. Biochem. 101 (2007) 1836-1844. [44] T. Rasmussen, T. Brittain, B.C. Berks, N.J. Watmough, A.J. Thomson, Dalton Trans. (2005) 3501-3506.

PT

[45] C.-S. Zhang, T.C. Hollocher, Biochim. Biophys. Acta 1142 (1993) 253-261. [46] S. Dell'acqua, S.R. Pauleta, E. Monzani, A.S. Pereira, L. Casella, J.J.G. Moura, I.

RI

Moura, Biochemistry 47 (2008) 10852-10862.

SC

[47] M.Y. Liu, M.C. Liu, W.J. Payne, J. Legall, J. Bacteriol. 166 (1986) 604-608. [48] K. Fujita, M. Hirasawa-Fujita, D.E. Brown, Y. Obara, F. Ijima, T. Kohzuma, D.M.

NU

Dooley, J. Inorg. Biochem. 115 (2012) 163-173.

[49] K. Brown, M. Tegoni, M. Prudêncio, A.S. Pereira, S. Besson, J.J.G. Moura, I. Moura, C. Cambillau, Nat. Struct. Biol. 7 (2000) 191-195.

MA

[50] K. Brown, K. Djinovic-Carugo, T. Haltia, I. Cabrito, M. Saraste, J.J.G. Moura, I. Moura, M. Tegoni, C. Cambillau, J. Biol. Chem. 275 (2000) 41133-41136.

D

[51] M.E. Murphy, P.F. Lindley, E.T. Adman, Protein Sci. 6 (1997) 761-770. [52] C.C. Page, C.C. Moser, X. Chen, P.L. Dutton, Nature 402 (1999) 47-52.

PT E

[53] L.K. Schneider, O. Einsle, Biochemistry 55 (2016) 1433-1440. [54] J. Riester, W.G. Zumft, P.M. Kroneck, Eur. J. Biochem. 178 (1989) 751-762.

815.

CE

[55] T. Rasmussen, B.C. Berks, J.N. Butt, A.J. Thomson, Biochem. J. 364 (2002) 807-

[56] S. Dell'Acqua, S.R. Pauleta, J.J.G. Moura, I. Moura, Philos. Trans. R. Soc. Lond. B

AC

Biol. Sci. 367 (2012) 1204-1212. [57] S.R. Pauleta, C. Carreira, I. Moura, Insights into Nitrous Oxide Reductase, in: I. Moura, J.J.G. Moura, L. Maia, S.R. Pauleta (Eds.), Metalloenzymes in Denitrification: Applications and Environmental Impacts, RSC, 2017, pp. 141-169. [58] P.M. Kroneck, W.A. Antholine, J. Riester, W.G. Zumft, FEBS Lett. 242 (1988) 7074. [59] P.M. Kroneck, W.E. Antholine, D.H. Kastrau, G. Buse, G.C. Steffens, W.G. Zumft, FEBS Lett. 268 (1990) 274-276. [60] W.E. Antholine, D.H. Kastrau, G.C. Steffens, G. Buse, W.G. Zumft, P.M. Kroneck, Eur. J. Biochem. 209 (1992) 875-881. 27

ACCEPTED MANUSCRIPT [61] P.M. Kroneck, W.A. Antholine, J. Riester, W.G. Zumft, FEBS Lett. 248 (1989) 212-213. [62] J.A. Farrar, F. Neese, P. Lappalainen, P.M.H. Kroneck, M. Saraste, W.G. Zumft, A.J. Thomson, J. Am. Chem. Soc. 118 (1996) 11501-11514. [63] M.L. Alvarez, J. Ai, W. Zumft, J. Sanders-Loehr, D.M. Dooley, J. Am. Chem. Soc. 123 (2001) 576-587. [64] D.R. Gamelin, D.W. Randall, M.T. Hay, R.P. Houser, T.C. Mulder, G.W. Canters,

PT

S.d. Vries, W.B. Tolman, Y. Lu, E.I. Solomon, J. Am. Chem. Soc. 120 (1998) 52465263.

RI

[65] C.R. Andrew , J. Han , S.d. Vries , J.v.d. Oost , B.A. Averill , T.M. Loehr , J.

SC

Sanders-Loehr, J. Am. Chem. Soc. 116 (1994) 10805-10806.

[66] E.I. Solomon, D.E. Heppner, E.M. Johnston, J.W. Ginsbach, J. Cirera, M. Qayyum,

NU

M.T. Kieber-Emmons, C.H. Kjaergaard, R.G. Hadt, L. Tian, Chem. Rev. 114 (2014) 3659-3853.

[67] J.A. Farrar, A.J. Thomson, M.R. Cheesman, D.M. Dooley, W.G. Zumft, FEBS

MA

Lett. 294 (1991) 11-15.

[68] S. Dell'Acqua, S.R. Pauleta, P.M. Paes de Sousa, E. Monzani, L. Casella, J.J.

D

Moura, I. Moura, J. Biol. Inorg. Chem. 15 (2010) 967-976. [69] J.M. Charnock, A. Dreusch, H. KoÈrner, F. Neese, J. Nelson, A. Kannt, H. Michel,

PT E

C.D. Garner, P.M. Kroneck, W.G. Zumft, Eur. J. Biochem. 267 (2000) 1368-1381. [70] K. Paraskevopoulos, S.V. Antonyuk, R.G. Sawers, R.R. Eady, S.S. Hasnain, J. Mol. Biol. 362 (2006) 55-65.

CE

[71] A. Pomowski, W.G. Zumft, P.M.H. Kroneck, O. Einsle, Nature 477 (2011) 234237.

AC

[72] L.K. Schneider, A. Wüst, A. Pomowski, L. Zhang, O. Einsle, Met. Ions Life Sci. 14 (2014) 177-210.

[73] S. Dell'acqua, I. Moura, J.J.G. Moura, S.R. Pauleta, J. Biol. Inorg. Chem. 16 (2011) 1241-1254. [74] W.G. Zumft, T. Matsubara, FEBS Lett. 148 (1982) 107-112. [75] J.A. Farrar, W.G. Zumft, A.J. Thomson, Proc. Natl. Acad. Sci. U. S. A. 95 (1998) 9891-9896. [76] T. Haltia, K. Brown, M. Tegoni, C. Cambillau, M. Saraste, K. Mattila, K. DjinovicCarugo, Biochem. J. 369 (2003) 77-88.

28

ACCEPTED MANUSCRIPT [77] T. Rasmussen, B.C. Berks, J. Sanders-Loehr, D.M. Dooley, W.G. Zumft, A.J. Thomson, Biochemistry 39 (2000) 12753-12756. [78] A. Wüst , L. Schneider, A. Pomowski, W.G. Zumft, P.M.H. Kroneck, O. Einsle, Biol. Chem. 393 (2012) 1067-1077. [79] E.M. Johnston, S. Dell'Acqua, S. Ramos, S.R. Pauleta, I. Moura, E.I. Solomon, J. Am. Chem. Soc. 136 (2014) 614-617. [80] E.M. Johnston, S. Dell'Acqua, S.R. Pauleta, I. Moura, E.I. Solomon, Chem. Sci. 6

PT

(2015) 5670-5679.

[81] S. Ghosh, S.I. Gorelsky, S.D. George, J.M. Chan, I. Cabrito, D.M. Dooley, J.J.G.

RI

Moura, I. Moura, E.I. Solomon, J. Am. Chem. Soc. 129 (2007) 3955-3965.

SC

[82] P. Chen, I. Cabrito, J.J.G. Moura, I. Moura, E.I. Solomon, J. Am. Chem. Soc. 134 (2002) 10497-10507.

NU

[83] P. Chen, S. DeBeer George, I. Cabrito, W.E. Antholine, J.J. Moura, I. Moura, B. Hedman, K.O. Hodgson, E.I. Solomon, J. Am. Chem. Soc. 124 (2002) 744-745. [84] V.S. Oganesyan, T. Rasmussen, S. Fairhurst, A.J. Thomson, Dalton Trans. (2004)

MA

996-1002.

[85] P. Chen, S.I. Gorelsky, S. Ghosh, E.I. Solomon, Angew. Chem. Int. Ed. Engl. 43

D

(2004) 4132-4140.

[86] E.M. Johnston, C. Carreira, S. Dell'Acqua, S. Ghosh, S.R. Pauleta, I. Moura, E.I.

PT E

Solomon, J. Am. Chem. Soc. 139 (2017) 4462-4476. [87] S. Ghosh, S.I. Gorelsky, P. Chen, I. Cabrito, J.J.G. Moura, I. Moura, E.I. Solomon, J. Am. Chem. Soc. 125 (2003) 15708-15709.

CE

[88] I. Moura, S.R. Pauleta, J.J. Moura, J. Biol. Inorg. Chem. 13 (2008) 1185-1195. [89] B.J. Johnson, S.V. Lindeman, N.P. Mankad, Inorg. Chem. 53 (2014) 10611-10619.

AC

[90] D.M. Dooley, M.L. Alvarez, A.C. Rosenzweig, R.S. Hollis, W.G. Zumft, Inorg. Chem. Acta 242 (1996) 239-244. [91] P.M.H. Kroneck, J. Riester, W.G. Zumft, W.E. Antholine, Biol. Metals 3 (1990) 103-109. [92] J.K. Kristjansson, T.C. Hollocher, J. Biol. Chem. 255 (1980) 704-707. [93] K. Frunzke, W.G. Zumft, J. Chromatogr. A 299 (1984) 477-483. [94] H. Körner, K. Frunzke, K. Döhler, W.G. Zumft, Arch. Microbiol. 148 (1987) 2024. [95] C.K. SooHoo, T.C. Hollocher, J. Biol. Chem. 266 (1991) 2203-2209. [96] P. Wardman, Free Radic. Res. Commun. 14 (1991) 57-67. 29

ACCEPTED MANUSCRIPT [97] J.M. Borrero-de Acuña, M. Rohde, J. Wissing, L. Jänsch, M. Schobert, G. Molinari, K.N. Timmis, M. Jahn, D. Jahn, J. Bacteriol. 198 (2016) 1401-1413. [98] J.M. Chan, J. Bollinger, C.L. Grewell, D.M. Dooley, J. Am. Chem. Soc. 126 (2004) 3030-3031. [99] K. Fujita, D.M. Dooley, Inorg. Chem. 46 (2007) 613-615. [100] S.G. Mayhew, Eur. J. Biochem. 85 (1978) 535-547. [101] M.Z. Ertem, C.J. Cramer, F. Himo, P.E. Siegbahn, J. Biol. Inorg. Chem. 17

PT

(2012) 687-698.

[102] R.M. Almeida, S. Dell' Acqua, I. Moura, S.R. Pauleta, J.J.G. Moura, Electron

RI

Transfer and Molecular Recognition in Denitrification and Nitrate Dissimilatory

SC

Pathways, in: I. Moura, J.J.G. Moura, L. Maia, S.R. Pauleta (Eds.), Metalloenzymes in

AC

CE

PT E

D

MA

NU

Denitrification: Applications and Environmental Impacts, RSC, 2017, pp. 252-286.

30

ACCEPTED MANUSCRIPT Scheme 1

NO3- [+5]

Nitrate reductase

NO2- [+3]

Nitrite reductase

NO [+2]

NO reductase

N2O [+1]

N2O reductase

N2 [0]

Scheme 1 - Scheme of the denitrification pathway. The enzymes involved in catalysis

PT

of each reaction are identified above the arrow. The oxidation state of the nitrogen

AC

CE

PT E

D

MA

NU

SC

RI

atoms is indicated in brackets.

31

ACCEPTED MANUSCRIPT

Table 1 - Summary of "CuZ center" properties. The different oxidation states of the four copper atoms in "CuZ center" are presented for each of the "CuZ center" forms, together with their spectroscopic and kinetic properties. Form

of

"CuZ

Center" Oxidized CuZ(4Cu2S) Reduced CuZ(4Cu2S)

Resting CuZ*(4Cu1S)

Oxidation state [2Cu2+:2Cu1+:2S]2+

CuI - CuIV

Spin

Turnover

Visible

edge ligand

State

Number

Absorptionb

bridging S2-

[1Cu2+:3Cu1+:2S]1+

bridging SH-

CuZ*(4Cu1S)

S = 1/2

[1Cu2+:3Cu1+:S:OH]2+ bridging OH-

I R

(≈ 5000 M-1cm-1)

SC

PT

[1Cu2+:3Cu1+:S:OH]2+ bridging OH-

E C

C A

[4Cu1+:S]2+

Empty/H2O

U N

0.6 h-1

A M

S = 1/2

D E

CuZ0 a Fully reduced

none

T P

670 nm

Intermediate species

S=0

545 mn

none

S = 1/2

> 200 s-1

S=0

> 200 s-1

(≈ 3000 – 4400 M-1cm1

)

640 nm (≈ 4000 M-1cm-1)

680 nm (≈ 2000 M-1cm-1)

No bands

EPR

Ref. [25, 55, 75,

silent

80]

g‖=2.150, g┴=2.035 A‖=5.6 mTc

55, 56, 75, 79, 80, 83]

g‖=2.160, g┴=2.040 A‖=6.1

[24, 25, 31,

[26, 55, 67,

mT/A‖=2.4 75, 77, 83,

mTd g‖=2.177, g┴=2.05 A‖=4.2 mTe

silent

85]

[68, 86]

[25,

79-81,

87]

Notes: aCuZ0 is an intermediate of CuZ*(4Cu1S) catalytic cycle. bMolar extinction coefficient is given by concentration of N2OR monomer. c

Considering 3 identical hyperfine coupling constants. dWith a 5/2 ratio. eConsidering 2 identical hyperfine coupling constants.

32

ACCEPTED MANUSCRIPT Figure Legends Figure 1 - Structure of M. hydrocarbonoclasticus N2OR. The surface is colored according to the subunit. One monomer is represented in blue and the other in violet, with the C-terminal light colored and the N-terminal dark colored. The copper atoms are represented as dark blue spheres and the sulfur atoms are represented as yellow spheres. The calcium ions in the dimer are evidenced as grey small spheres. The distance

PT

between CuA center and “CuZ centers” is represented. Figure was prepared with DS

RI

visualizer 4.5 using PDB ID 1QNI.

SC

Figure 2 - Structures of CuA center of N2OR. Representation of the CuA center of P. denitrificans N2OR (A) and P. stutzeri N2OR (B). In Panel B, H583 does not coordinate

NU

CuA1 atom. The copper atoms are represented as blue dark spheres and numbered 1-2. Figure was prepared with DS visualizer 4.5 using PDB ID 1FWX (A) and 3SBP (B).

MA

Figure 3 - Structures of the “CuZ center” of N2OR. (A) Representation of the “CuZ center” as CuZ(4Cu2S) from P. stutzeri N2OR with first and second coordination sphere

D

residues. This form contains two sulfur atoms named S1 and S2. (B) Representation of the CuZ*(4Cu1S) from P. denitrificans N2OR with the first and second coordination

PT E

sphere residues. This form contains one sulfur atom named S1 and one oxygen atom (from either a hydroxide or a water molecule) at the CuI-CuIV edge. The copper atoms are represented as dark blue spheres and numbered I-IV. Figure was prepared with DS

CE

visualizer 4.5 using PDB ID 3SBP (A) and 1FWX (B).

AC

Figure 4 – Visible absorption and EPR spectroscopic characterization of N2OR. Visible (A) and X-band EPR (B) spectra of different forms of N2OR isolated from M. hydrocarbonoclasticus. The spectra of N2OR with mainly CuZ(4Cu2S) in the oxidized [2Cu2+-2Cu1+] state (I) and dithionite reduced, [1Cu2+-3Cu1+] state (II), and the spectra of N2OR with mainly CuZ*(4Cu1S) in the resting [1Cu2+-3Cu1+] (III) and dithionite reduced [1Cu2+-3Cu1+] (IV) state, are shown in Panels A and B. The parameters of the EPR spectra were acquired at 30 K, with a 5 G modulation amplitude and a 9.66 GHz microwave frequency.

33

ACCEPTED MANUSCRIPT Figure 5 – Substrate binding mode in N2OR. (A) The N2O was modeled in the CuI-CuIV edge in a µ-1,3-bridging mode. The N2O binding site was proposed for A. cycloclastes N2OR structure, based on DFT calculations and spectroscopic features of the “CuZ center” as CuZ*(4Cu1S), in the [4Cu1+] oxidation state. (B) The N2O molecule is located between CuA and “CuZ” centers. P. stutzeri crystals of N2OR, with oxidized CuA center and with the “CuZ center” as CuZ(4Cu2S), in the [2Cu2+-2Cu1+] oxidation state, were pressurized with N2O. The surface at the dimer interface, as well as the

PT

relevant residues involved in the substrate binding mode are represented and colored according to the subunit by pink and blue. The N2O atoms are colored according to the

RI

element. Figures were prepared using DS visualizer 4.5 using PDB ID 2IWF (A) and

SC

3SBR (B).

NU

Figure 6 - Activation mechanism and catalytic cycle of N2O reduction by N2OR with “CuZ center” as CuZ*(4Cu1S). The main catalytic cycle is represented by solid arrows and the activation and inactivation pathways are represented by dash arrows. Residues

AC

CE

PT E

D

MA

are numbered according to M. hydrocarbonoclasticus N2OR mature primary sequence.

34

ACCEPTED MANUSCRIPT Figures Figure 1

PT

40Å

10Å

Ca2+

AC

CE

PT E

D

MA

NU

SC

RI

CuZ

35

ACCEPTED MANUSCRIPT Figure 2

B

A

D576

C565

C622

H526

W563

W620

CuA2

S550

M629

M572

C618

AC

CE

PT E

D

MA

NU

SC

C561

H583

RI

CuA2

CuA1

H626

CuA1

PT

H569

36


A

B H326

H270 H382

CuI

K397 H433

S1

CuIV

CuIII

H130 CuII H178

E435

H494

CuII

CuIII

H80

H128 H79

AC

CE

PT E

D

MA

NU

H129

S1

H437

SC

E492

H376

PT

S2

CuIV

CuI OH-

RI

K454

H325

37


A

B 6000

5000 II

3000

PT

4000

III

2000 IV

SC

II

1000 0

550

650

750

850

300

320

340

360

380

Magnetic Field (mT)

CE

PT E

D

MA

Wavelength (nm)

950 280

NU

450

IV

RI

I

III

AC

Ɛ (M-1 cm -1 per monomer)

I

38


A

B

M629

H543

M589

H583 CuA1

CuA1 C578

C618

CuA2

PT

C582

CuA2

W580

RI

K412

H452

H178

NU

CuII N2O

S1

H391

K454

S2

S1

CuI CuIII H433

H130

H382

CE

PT E

H340

D

H94

MA

CuIII

H494

CuIV

H326

CuI

H285

H129 CuII

E492

N2O

M627 CuIV

F621

AC

H142

SC

E450

F581

H93

W620

H626

H586

M587

C622

39


2+

N

N

CuI S

CuIII

CuIV

O K397H+

H2O

CuII

Cu III

Cu CuOI I

CuI 2+ HO S

Cu I

N2

3+

N

S Cu IVCuIII

N

2+

Cu Cu I I

CuIII CuIV H CuZ0 S S OO S S Cu Cu CuII Cu Cu II Cu Cu Intermediate 3 III II III Cu IV CuIIIIII IV Cu K397H+ N CuO3+ IV K397H+ Cu II e-Fully reduced Cu I 2 Cu II Cu II CuHII O Intermediate 1H2O 2 SE435 + E435 E CuIII e- H Cu IV Slow +inactivation in the H+ H N2 N2 absence of reductants 2+ Cu II Cu pk = 7.1 0.8 a I H2O kIET > 0.1 s-1 , e-/H+ CuIII

Slow activation by reduced methyl viologen (k ~ 1.2x10-3 s-1, pH 7.3) pk a = 9.0 0.2

HO Cu IV 3+

H2 O Cu IV

Cu I S

CuIII

Cu II

HO Cu IV

Cu I S Cu II

SC

O

kIET < 1x10-5 s-1

e-

N

2+

N

H2O

S

NU

S

2+

N

pH 7.6

-3

-1

kdecay ~ 5x10 s , pH 7.6 Cu III CuAox Cu IV 3+ 3+ 3+ CuoxI Cu I CytCu HO Cu I Cu I c-552/ Asc HOII HO CuAred S S S S Cu CuCyt Cu IIICu c-552/ Asc red III IV IV Cu III III Resting CuCu IV K397H+ Cu Cu II Cu II CuZ*(4Cu1S) IICu II 3+ eCu I E435 H2 O S + + CuIII e- H e- H Cu IV Cu II

3+

HO

Cu III

CE

Cu IV

CuI

Cyt c-552/ Asc ox

N2

Cu IV

Cu I S

3+

Cu III

Cu II

AC

H2O

CuAox

e-/H+

MA

H2O

Cyt c-552/ Asc red

D

N2O

N

PT E

CuIV

H+

CuAred

RI

S

S

N2O

S Cu III CuN IV 2+ N Cu CuIII O

Intermediate 2

S Cu III OCu IV Cu III Cu IV Cu II Cu II

E435

2+

CuI

H2O

2+

NCu I Cu I

N

CuII

N2O

N

N

PT

O

40

ACCEPTED MANUSCRIPT Graphical Abstract

Nitrous oxide reductase is a multicopper enzyme containing a CuA and a tetranuclear copper sulfide-bridged “CuZ center", that catalyzes N2O reduction. "CuZ center" exists in two forms, CuZ(4Cu2S) and CuZ*(4Cu1S), with the later proven to be the relevant

10Å

40Å

SC

CuA center

RI

PT

species in the reaction mechanism that involves CuZ0 as an intermediate species.

AC

CE

PT E

D

MA

NU

CuZ center

41

ACCEPTED MANUSCRIPT Highlights 

Nitrous oxide reductase catalyzes the last step of the denitrification pathway



The catalytic center is a tetranuclear copper sulfide center, named "CuZ center"



This enzyme can be isolated with its catalytic center in two forms, CuZ and CuZ* CuZ*(4CuS) is the catalytically relevant form of "CuZ center"



CuZ0, a catalytically active intermediate species in the [1Cu2+-3Cu1+] oxidation

PT



AC

CE

PT E

D

MA

NU

SC

RI

state

42

Periplasmic and membrane-bound respiratory nitrate reductases in Thiosphaera pantotropha. The periplasmic enzyme catalyzes the first step in aerobic denitrification.

Modeling nitrous oxide production and reduction in soil through explicit representation of denitrification enzyme kinetics.

The structure of lipoyl synthase, a remarkable enzyme that performs the last step of an extraordinary biosynthetic pathway.

Evaluating the Role of Microbial Internal Storage Turnover on Nitrous Oxide Accumulation During Denitrification.

Determination of the active form of the tetranuclear copper sulfur cluster in nitrous oxide reductase.

The nitrate-sensing NasST system regulates nitrous oxide reductase and periplasmic nitrate reductase in Bradyrhizobium japonicum.

Increased nitrous oxide accumulation by bioelectrochemical denitrification under autotrophic conditions: kinetics and expression of denitrification pathway genes.

Interaction of dichloromethane (methylene chloride) with the nitrous oxide reductase from Wolinella succinogenes.

Probing the electrostatics of active site microenvironments along the catalytic cycle for Escherichia coli dihydrofolate reductase.

Dual isotope and isotopomer signatures of nitrous oxide from fungal denitrification--a pure culture study.

Simulating the catalytic effect of a designed mononuclear zinc metalloenzyme that catalyzes the hydrolysis of phosphate triesters.

An apparently allosteric effect involving N2O with the nitrous oxide reductase from Wolinella succinogenes.

Nitrous oxide production and consumption by denitrification in a grassland: Effects of grazing and hydrology.

Structure and properties of the catalytic site of nitric oxide reductase at ambient temperature.

Burkholderia glumae ToxA Is a Dual-Specificity Methyltransferase That Catalyzes the Last Two Steps of Toxoflavin Biosynthesis.

Structural snapshots illustrate the catalytic cycle of β-galactocerebrosidase, the defective enzyme in Krabbe disease.

Nitrous oxide and the middle ear.

The effects of nitrous oxide on the developing hamster embryos.

The neurotoxicity of nitrous oxide: the facts and "putative" mechanisms.

Nitrous oxide.

Denitrification and oxygen respiration in biofilms studied with a microsensor for nitrous oxide and oxygen.

Reverse Electron Transfer Completes the Catalytic Cycle in a 2,3,5-Trifluorotyrosine-Substituted Ribonucleotide Reductase.

A model of the copper centres of nitrous oxide reductase (Pseudomonas stutzeri). Evidence from optical, EPR and MCD spectroscopy.

Inhibition effect of graphene oxide on the catalytic activity of acetylcholinesterase enzyme.