Unique Spectroscopic Properties of the H-Cluster in a Putative

Dec 18, 2017 - Sensory type [FeFe] hydrogenases are predicted to play a role in transcriptional regulation by detecting the H2 level of the cellular e...
38 downloads 6 Views 2MB Size
Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES

Article

Unique Spectroscopic Properties of the HCluster in a Putative Sensory [FeFe] Hydrogenase Nipa Chongdar, James A. Birrell, Krzysztof Pawlak, Constanze Sommer, Edward J. Reijerse, Olaf Rüdiger, Wolfgang Lubitz, and Hideaki Ogata J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b11287 • Publication Date (Web): 18 Dec 2017 Downloaded from http://pubs.acs.org on December 18, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of the American Chemical Society is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Unique Spectroscopic Properties of the H-Cluster in a Putative Sensory [FeFe] Hydrogenase Nipa Chongdar,1 James A. Birrell,1 Krzysztof Pawlak,1 Constanze Sommer,1 Edward J. Reijerse,1 Olaf Rüdiger,1 Wolfgang Lubitz,1* Hideaki Ogata1, 2* 1

Max Planck Institute for Chemical Energy Conversion, Stiftstrasse 34-36, D45470 Mülheim an der Ruhr, Germany

2

Institute of Low Temperature Science, Hokkaido University, Kita19 Nishi8, Kita-ku, 060-0819 Sapporo, Japan

ABSTRACT: Sensory type [FeFe] hydrogenases are predicted to play a role in transcriptional regulation by detecting the H2 level of the cellular environment. These hydrogenases contain the hydrogenase domain with distinct modifications in the active site pocket, followed by a Per-Arnt-Sim (PAS) domain. As yet, neither the physiological function nor the biochemical or spectroscopic properties of these enzymes have been explored. Here, we present the characterization of an artificially maturated, putative sensory [FeFe] hydrogenase from Thermotoga maritima (HydS). Artificially maturated HydS showed lower hydrogen conversion activity than prototypical [FeFe] hydrogenases and reduced CO inhibition. Using FTIR spectroelectrochemistry and EPR spectroscopy, three redox states of the active site were identified. The spectroscopic signatures of the most oxidized state closely resemble those of the Hox state from the prototypical [FeFe] hydrogenases, while the FTIR spectra of both singly and doubly reduced states show large differences. The FTIR bands in both the reduced states are strongly red-shifted relative to the Hox state, indicating reduction at the diiron site, but with retention of the bridging CO ligand. The unique functional and spectroscopic features of HydS are discussed with regard to the possible role of altered amino acid residues influencing the electronic properties of the H-cluster.

INTRODUCTION

Hydrogenases are metalloenzymes that catalyze the reversible hydrogen conversion reaction: H2 ⇄ 2H++2e-. These enzymes fall into three classes – [NiFe], [FeFe] and [Fe] hydrogenases according to the metal ion composition at their active site (reviewed in1-6). Among them, [FeFe] hydrogenases are highly active in H2 production (up to 10,000 s-1)7 and H2 oxidation (up to 150,000 s-1),8 and thus these enzymes are attractive candidates for application based research on artificial H2 production. The active center of [FeFe] hydrogenases, known as the H-cluster contains a [4Fe-4S] cluster (referred to as [4Fe-4S]H) that is covalently attached to a unique [2Fe] cluster (called [2Fe]H) (Fig. 1A). In addition to the active center, [FeFe] hydrogenases often have additional [4Fe-4S] clusters (called F-clusters), which provide an electron transfer chain between the H-cluster and the surface of the protein where redox partner proteins can bind (Fig. 1A). Furthermore, some [FeFe] hydrogenases possess accessory FeS cluster containing subunits and form a large protein complex that can interact with several redox partners, either proteins or small molecules. All these factors make [FeFe] hydrogenases structurally and functionally diverse. Various classification schemes of [FeFe] hydrogenases are available in the literature.9-11 In a recent study, all available non-redundant gene sequences of [FeFe] hydrogenases

were analyzed, and the enzymes were classified into three groups: i) prototypical and electron-bifurcating type, ii) ancestral type and iii) sensory type.12 All the [FeFe] hydrogenases that have been extensively studied to date (e.g. DdHydAB from Desulfovibrio desulfuricans, CpI from Clostridium pasteurianum, and CrHydA1 from Chlamydomonas reinhardtii) mostly belong to the prototypical group.8,13-16 Electron bifurcating [FeFe] hydrogenases have also been characterized from Thermotoga maritima, Ruminococcus albus and a few other bacteria.17-21 However, at present, similar studies are lacking for ancestral and sensory type [FeFe] hydrogenases. This work is aimed at the characterization of a sensory type [FeFe] hydrogenase. The sensory type [FeFe] hydrogenases (HydS) are found in anaerobic bacteria, which belong to Firmicutes, Bacteroidates, Spirochaetes or Thermotogae phyla.12 Typically, HydS enzymes contain the H-cluster, three additional [4Fe-4S] clusters - two at the N-terminus, and one at the C-terminal side of the H-cluster (Fig. 1A). The [4Fe-4S] cluster present at the C-terminus is ligated by an unusual Cx2Cx4Cx16C motif. The most interesting feature of HydS is the presence of a PAS (Per-Arnt-Sim) sensory domain at the C-terminus, based on which the sensory function of this particular group of [FeFe] hydrogenases was predicted (Fig. 1A).10 PAS domains are widely used in bacteria for various signal transduction and sensory functions.22 Regulatory [NiFe] hydrogenases also rely on PAS domains for

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 14

Fig. 1. Distinct properties of the sensory [FeFe] hydrogenase. A) Schematic representation of domain organization of sensor type [FeFe] hydrogenase. B) Amino acids forming covalent and non-covalent interactions with the H-cluster are mapped on the crystal structure of the prototypical [FeFe] hydrogenase from CpI (PDB: 4XDC). Residues in the H-cluster interacting pocket that are substituted in TmHydS are highlighted. C) Sequence comparison of H-cluster interacting residues in TmHydS with those of 12 CpI is shown in the upper panel. WebLogo plot (prepared using sequences from reference ) depicting conservation pattern of amino acids surrounding the H-cluster in Group (i) (Prototypical and bifurcating hydrogenase) and Group (iii) (Sensory hydrogenase) is shown in the lower panel. The overall height of each stack of letters in the logo plot indicates sequence conservation at a particular position, and the height of amino acid codes in each stack shows the frequency of occurrence of those amino acids 23 in that position. Cysteine residues coordinating the H-cluster are marked in yellow and the residues forming non-covalent interactions with the H-cluster are marked in black. Residues that are altered in group (iii) compared to group (i) are marked with colored background.

their functions.24 Conserved domain analysis suggested that HydS is expressed along with putative regulatory proteins such as serine/threonine phosphatase, histidine kinase etc. and that the operon containing the HydS gene often has a gene to transcribe a prototypical or bifurcating type [FeFe] hydrogenase.12,18 It is presently assumed that transcriptional regulation of the additional hydrogenase gene could be aided by HydS.12,18 Although speculative, some hints regarding the regulatory function of HydS were provided by transcriptional studies on Thermoanaerobacterium saccharolyticum and Ruminococcus albus.18,25 In these species, HydS is found in the same operon as a prototypical [FeFe] hydrogenase, and a separate operon encodes the bifurcating enzyme. Bifurcation allows the recycling of NADH to NAD+ by transferring the electrons to H+ to make H2. However, this is only thermo-

dynamically possible under low partial pressures of H2. At high partial pressures of H2, the prototypical [FeFe] hydrogenase is used to allow thermodynamically favorable proton reduction. Therefore, H2-dependent regulation of the expression of these two hydrogenase genes is essential. Besides having a PAS domain, sensory type [FeFe] hydrogenases also possess alternative amino acids in the Hcluster binding pocket compared to prototypical or bifurcating hydrogenases (Fig. 1B).26 The highly-conserved cysteine residue proximal to the amine group of the azadithiolate bridge (C299 of CpI from Clostridium pasturianum or C169 of CrHydA1), which has been shown to be involved in the catalytic proton transfer pathway,27-29 is replaced by alanine in many HydS enzymes (Fig. 1C). In addition, the two conserved methionine residues (M353

2 ACS Paragon Plus Environment

Page 3 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

and M497, CpI) are also substituted in HydS (Fig. 1C). Mutation of these residues in CrHydA1 and CpI adversely affected the catalytic efficiency of the enzyme.30 Thermotoga maritima (T. maritima) is a hyperthermophilic, anaerobic bacterium that employs [FeFe] hydrogenases to produce H2 by fermenting many different types of carbohydrates.31 The T. maritima genome contains a gene that potentially encodes HydS along with two other genes coding for catalytic [FeFe] hydrogenases. The first gene product is the catalytic subunit of the heterotrimeric electron-bifurcating type (TmHydαβγ) enzyme, and the second gene produces a prototypical type [FeFe] hydrogenase.17,32 The electron-bifurcating enzyme plays the key role for H2 production and is well characterized.33,34 The function of the prototypical type [FeFe] hydrogenase is presently unknown. The sensory type [FeFe] hydrogenase is part of the same operon that transcribes TmHydαβγ.32 Here we provide the first biochemical and spectroscopic studies of a sensory [FeFe] hydrogenase, HydS from Thermotoga maritima (TmHydS). Since the isolation of [FeFe] hydrogenases from the native organism is challenging, we prepared apo-TmHydS, lacking a fully assembled active site by overexpression in E. coli followed by insertion of the [2Fe]H sub-cluster by artificial maturation.35,36

absent, and thus overexpression in E. coli generates apoTmHydS lacking [2Fe]H. However, artificial maturation of apo-hydrogenases can be achieved by incubating them with synthetic [2Fe] pre-cursors in vitro.35 FTIR characterization of the electron-bifurcating [FeFe] hydrogenase (TmHydαβγ) isolated from T. maritima showed that the H-cluster composition of this enzyme is the same as that of other well studied [FeFe] hydrogenases.34 Therefore, the T. maritima MSB8 maturases produce a [2Fe]H subsite with an azadithiolate (ADT) bridge and, since the bacterium possesses only one set of maturases (HydF - WP_004081521.1, HydG - WP_10865292.1, HydE -

RESULTS

Characterization of apo-TmHydS The apo-form of T. maritima HydS lacking the [2Fe]H cluster (apo-TmHydS) was produced following recombinant methods (details of construct preparation, heterologous expression and purification are provided in the supplementary information). After purification, UV-visible spectroscopy of the apo-TmHydS showed a broad absorbance band in the range of 350-500 nm, typical for [4Fe-4S] cluster containing proteins (Fig. 2A). The protein contained 15.0 (± 1.1) moles Fe per mole of protein, indicating the presence of four [4Fe-4S] clusters, in agreement with the expected number and type of FeS clusters. The CW-Xband EPR spectrum of dithionite reduced apo-TmHydS shows the characteristic broad rhombic signal of reduced paramagnetic [4Fe-4S]+ clusters (S=½) (Fig. 2B). Spin quantitation gave two spins per molecule of protein, indicating that under the above experimental conditions an average of two [4Fe-4S] clusters is reduced. Simulation of the EPR spectra of apo-TmHydS suggests that there are two components with similar but slightly different gvalues: 2.050, 1.913, 1.858 and 2.048, 1.942, 1.895. Unlike DdHydAB the EPR spectrum of apo-TmHydS does not show a complex interaction pattern at low temperature (10 K) indicating no large spin coupling between the [4Fe-4S] centers (data not shown).37

Fig. 2. Spectroscopic characterization of apo-TmHydS. A) UV-visible spectroscopy of 9 µM apo-TmHydS in 0.1 M Tris-HCl pH8, 0.15 M NaCl and B) EPR spectroscopy of 150 µM apo-TmHydS at 15 K and 0.1 mW power in 0.1 M Tris-HCl pH8, 0.15 M NaCl, 10 mM sodium dithionite and 10% glycerol. The EPR spectra could be simulated using two components (1:1 ratio) with g-values 2.050, 1.913, 1.858 (component 1) and 2.048, 1.942, 1.895 (component 2). The cumulative simulation is shown with a green line, and the two components are shown below in blue and magenta.

Artificial maturation of TmHydS Maturation of [FeFe] hydrogenases is the process of synthesis and assembly of the [2Fe]H subsite of the active center, which is performed in vivo by three maturases HydG, HydE and HydF.4,38 In E. coli, the maturases are

3 ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 14

WP_004079975.1), it seems reasonable to assume that the H-cluster composition of TmHydS would be the same as in TmHydαβγ. Therefore, we artificially maturated apoTmHydS with synthetically produced [2Fe]ADT.35 Artificial maturation proceeds by donation of a lone pair of electrons by the thiol side chain of a cysteine ligating the [4Fe-4S]H sub-cluster to one of the two irons of [2Fe]ADT (that becomes the proximal iron Fep). This leads to displacement of one of the CO ligands of Fep from a terminal to a bridging position, and release of one CO (bound to the distal iron (Fed) of [2Fe]ADT) to the solvent. Thus, artificial maturation can be monitored by CO release using a hemoglobin-based assay.35 A typical CO-induced shift in the deoxy-hemoglobin Soret band was observed when apo-TmHydS was incubated with [2Fe]ADT, suggesting that the apo-TmHydS reacted with the synthetic complex (Fig S4). The reaction was complete within 2-3 min of mixing. Fourier Transform Infrared (FTIR) spectroscopy on a mixture of apo-TmHydS and [2Fe]ADT showed sharp peaks appearing at positions distinct from the bands of the free [2Fe]ADT precursor (Fig. S5). These peaks appeared immediately after mixing, and their intensities did not change with longer incubation times (Fig. S5). After removal of the excess [2Fe]ADT, viologen based assays were performed to determine the catalytic activity of holo-TmHydSADT. Catalytic activity of TmHydSADT The holo-TmHydSADT unquestionably shows higher catalytic activity than apo-TmHydS, indicating the success of artificial maturation (Fig. 3A). At 37 °C, holo-TmHydSADT shows a bias for H2 oxidation (turnover frequency of 45.4 ± 5 s-1) over H2 production (turnover frequency of 4.5 ± 0.5 s-1). Both H2 oxidation and H2 production activity of holoTmHydSADT are lower than those of CrHydA1, the most well characterized prototypical [FeFe] hydrogenase, by ≈ 5-fold and ≈ 100-fold, respectively.39 Although the activity of holo-TmHydSADT is low compared to the prototypical [FeFe] hydrogenases (Table S1), this is expected considering the amino acid alterations in the TmHydS H-cluster pocket (Fig. 1b). It was previously observed that variants of these amino acids (C299S, M353L and M497L) diminished the activity of the prototypical [FeFe] hydrogenase CpI.30 Our result is consistent with the findings on Ruminococcus albus where chromatography fractions containing HydS could not be detected using the methyl viologen reduction activity assay.18 It is also interesting to note that regulatory [NiFe] hydrogenases show low activity compared to the catalytic [NiFe] hydrogenases.40 Thus, such low activity of TmHydSADT is consistent with its predicted sensory role. T. maritima is a hyperthermophilic bacterium, which thrives at a temperature of 90 °C and its proteins show optimum activity at high temperature. Therefore, we tested the H2 oxidation and H2 production activity of TmHydSADT at higher temperatures (Fig. 3B). Due to technical limitations, measurement beyond 80 °C was not possible. With increasing temperatures, the specific activ-

ADT

Fig. 3. The catalytic activity of holo-TmHydS for H2 oxidation and H2 production. A) Comparison of the speADT cific activities of apo-TmHydS and holo-TmHydS at 37 °C in 200 mM potassium phosphate buffer pH 8. H2 oxidation (blue bars) was measured by the reduction of 1 mM benzyl viologen (absorbance at 600 nm) in H2 saturated buffer. H2 production (red bars) from 10 mM methyl viologen and 100 mM sodium dithionite was measured by gas chromatography. B) Temperature dependent H2 oxidation and producADT tion activity of holo-TmHydS in the same buffer.

ity of holo-TmHydSADT for both H2 oxidation and H2 production increased. Interestingly, at higher temperature, the catalytic bias shifted somewhat. Although at 80 °C the activity of the enzyme was higher (H2 oxidation and H2 production were ≈ 2 times and ≈ 7 times higher than at 37 °C), it was still low compared to the catalytic [FeFe] hydrogenases.35,41

4 ACS Paragon Plus Environment

Page 5 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

ADT

Fig. 4. FTIR and EPR spectroscopy of artificially maturated TmHydS under various conditions. A) The FTIR spectra of ADT 800 µM holo-TmHydS at 15 °C in the as-isolated form, after oxidation with thionine (≈ 8 mM), reduction with H2 (100% H2, 30 min) or sodium dithionite(Na2S2O4 ≈ 40 mM), and flushing with CO (100% CO, 30 min). All samples were in 50 mM Tris-HCl, 200 mM KCl (pH 8) buffer. B) X-band EPR spectra at 20 K, 0.1 mW microwave power and spectral simulations of holo-TmHydS in the same buffer after thionine oxidation (top) and CO treatment (bottom). In panel A FTIR bands corresponding to an ‘unidentified’ species (low concentration) are marked with asterisks.

FTIR and EPR characterization of holo-TmHydSADT

tions of the FTIR bands in holo-TmHydSADT are quite distinct (Table S2). Therefore, it was not immediately obvious in which oxidation state the H-cluster is. Treatment of holo-TmHydSADT with oxidizing (thionine) and reducing (H2 and sodium dithionite) agents were used to alter the oxidation state of the H-cluster. The FTIR spectrum of the oxidized enzyme (Fig. 4A) is dramatically different from the as-isolated enzyme and closely resembles that of the Hox state spectra observed in C. pasteurianum (2086, 2072, 1971, 1948, 1802 cm-1)44 and C. acetobutylicum (2082, 2070, 1969, 1946, 1801 cm-1)45 [FeFe] hydrogenases (Table S2). The EPR spectrum of this sample also shows the characteristic rhombic signal typical for the Hox state (Table S3) with g-values of 2.113, 2.045 and 2.001 (Fig. 4B). EPR spin-quantification of oxidized holo-TmHydSADT gave approximately one spin per protein molecule (0.9 spin/molecule protein), confirming complete maturation.

Infrared spectroscopy is an effective tool to study the structural and electronic properties of the H-cluster because the CN- and CO ligands associated with the [2Fe]H subsite show stretching frequencies in the range 2150 cm-1 to 1750 cm-1 where the characteristic amide contributions of the protein are absent. In some oxidation states, the Hcluster is paramagnetic and, therefore, can also be studied by EPR spectroscopy. The FTIR spectrum of artificially maturated TmHydSADT in the as-isolated state (under 2% H2) shows peaks at 2055, 2022, 1894, 1872 and 1763 cm-1 (Fig. 4A). These bands are significantly narrower than those of free [2Fe]ADT in solution, indicating successful incorporation of the [2Fe] subsite into the protein (Fig. 4A).35,42,43 Compared to the well-studied prototypical [FeFe] hydrogenases, the posi-

5 ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ADT

Page 6 of 14

ADT

Fig. 5. FTIR spectroelectrochemistry of TmHydS . A) FTIR spectra of TmHydS at selected potentials: –200 mV, Hox -1 -1 ∗ signal (2087, 2079, 1971, 1947, and 1806 cm ) dominates; –440 mV H signal (2055, 2022, 1894, 1871, and 1763 cm ) dominates -1 ∗ ∗ ∗ and at –740 mV H signal (2047, 2013, 1900, 1861, and 1751 cm ) dominates. Peaks corresponding to the H , H and H are shaded in red, blue and green, respectively and the peaks belonging to the unidentified species are marked with asterisks. The experiment was performed with ≈ 1.5 mM protein and at 15 °C in 50 mM Tris/HCl pH 8, 200 mM KCl buffer. B) Reductive titraADT tion curve of holo-TmHydS obtained by plotting intensities of the most prominent CO bands of each redox state against the electrode potential. Solid lines represent the Nernstian curves corresponding to one electron transitions.

broad band at 2090 cm-1 most likely represents two overlapping bands and suggests that the CN- ligands have similar vibrational frequencies in the Hox-CO state, similar to the Hox-CO state of M. elsdenii [FeFe] hydrogenase (Table. S2). The positions of the CO bands resemble those of the Hox-CO state from other [FeFe] hydrogenases (Table S2), and so we assign this as the Hox-CO state. In the Hox-CO state from other [FeFe] hydrogenases the [2Fe]H cluster is EPR active (S = 1/2) and shows a nearly axial EPR spectrum.48 The X-band EPR spectrum of CO-treated holo-TmHydSADT also shows a mixture of two signals (Fig. 4B), originating from the Hox state (g1 = 2.113, g2 = 2. 045, g3 = 2.001) and the Hox-CO state (g1 = 2.047, g2 = 2.018, g3 = 2.007) (Fig. 4B). The formation of the Hox-CO state in TmHydSADT only under 100% CO suggests that the binding affinity of CO to Fed is impaired. In the regulatory [NiFe] hydrogenase, the presence of bulky amino acids in the gas-channel is suggested to restrict the access of gases larger than H2 and thus the enzyme is only weakly inhibited by CO.40,49,50 The g-values of the Hox-CO state in holo-TmHydSADT are shifted compared with other hydrogenases, probably as a consequence of the altered amino acids in its H-cluster binding pocket.

When reduced with H2, the FTIR spectrum of holoTmHydSADT remains unchanged with respect to the ‘as isolated’ sample (Fig. 4A). Even treatment with the stronger reducing agent sodium dithionite did not show any change in the FTIR spectrum, indicating that in the ‘as-isolated’ state TmHydSADT is in the reduced form. The EPR spectrum of the sodium dithionite reduced holoTmHydSADT is very similar to that observed for apoTmHydS and shows signals typical for reduced [4Fe-4S] clusters (Fig. S6). All [FeFe] hydrogenases studied to date bind CO in the Hox state to form the Hox-CO state.14,46 Often the CO release from the [2Fe]ADT complex during artificial maturation is incomplete, due to rebinding of the released CO to the open coordination site of Fed and thus the Hox-CO signal is detected in FTIR spectra of the artificially maturated as-isolated protein.35,41,47 However, after artificial maturation and during all oxidative and reductive treatments no traces of a Hox-CO type FTIR signature could be detected in TmHydSADT, suggesting that CO binding to TmHydS is very weak. Flushing a thionine oxidized sample with CO for 30 min. gave a mixture of the Hox state and a new state with FTIR bands at positions 2090 (broad), 2016, 1973, 1964, and 1805 cm-1 (Fig. 4A). The

6 ACS Paragon Plus Environment

Page 7 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

[FeFe] hydrogenases are highly sensitive to molecular oxygen. O2 initially attacks the distal Fe (Fed) of [2Fe]H and causes degradation of the active center.16,51 On the other hand, the regulatory [NiFe] hydrogenase is O2 tolerant.40 To test how TmHydSADT behaves in the presence of O2, we exposed the protein (as isolated sample) to air in an Attenuated Total Reflectance (ATR) cell and followed the reaction by FTIR spectroscopy (Fig. S7). Initially, upon exposing the protein to air, the intensity of the vibrational bands corresponding to the reduced state decreased, and the FTIR bands corresponding to the oxidized state appeared (Fig. S7A-C). Further exposure of TmHydSADT to air led to a decrease in intensity of the oxidized species bands as well and finally, only the bands corresponding to the minor ‘unidentified’ species (bands marked with asterisks in Fig. 4A) remained (Fig. S7A). It should be noted here that no Hox-CO like signal was detected for TmHydSADT during the course of O2 inactivation. Under similar experimental conditions, the rate of O2 inactivation determined for the ‘as isolated’ state of TmHydSADT was similar to that of CrHydA1 in the HsredH+ state (Fig. S7B and S7C). The air–exposed TmHydSADT sample, which showed vibrational bands of the ‘unidentified’ species, did not show any H2 oxidation activity, indicating that this species is inactive. The above observation shows that O2 can reach and damage the active site of TmHydS like other [FeFe] hydrogenases. TmHydS is, therefore, an O2 sensitive enzyme (Fig. S7E).

of the Hox-state decreased by approximately 40% indicating some degradation of the H-cluster at high potentials. However, unlike in CrHydA1, formation of the Hox-CO state was not observed (Fig. S8).54 The titration curve obtained for both reductive and oxidative titrations could be fitted using the Nernst equation corresponding to two successive one electron transfers, implying that the state observed between –400 and –500 mV is a one electron reduced state of Hox, and the state observed at potentials lower than –500 mV is reduced by two electrons (Fig. 5B and Fig. S9). The first and second one-electron reductions occurred at –300 ± 10 mV and –570 ± 10 mV, respectively. As the FTIR signatures of both the one and two electron reduced states of TmHydSADT have not been observed before in [FeFe] hydrogenase, we refer to these states as ∗ ∗ H and H . Spectroscopic characterization of TmHydSPDT To obtain further insight into the reduction pathway of TmHydS, it was maturated with a [2Fe]H analogue in which the azadithiolate bridge (ADT) is replaced by PDT (propanedithiolate).42,43 In CrHydA1 and DdHydAB this variant only showed reduction of the [4Fe-4S]H subcluster.54,55 This was explained by the inability to protonate the PDT ligand thus preventing proton coupled rearrangement towards a reduced [2Fe]H subsite. 54,55 The PDT activated variant serves as a good model to study the oxidized (H ) and singly reduced (unprotonated) state (H ).54-56 In this redox transition, only the [4Fe-4S]H subcluster is reduced causing only small redshifts (≈ 10 cm-1) in the FTIR bands. The results of FTIR and EPR spectroscopy, and FTIR specroelectrochemistry on TmHydSPDT are summarized in Fig. S11 and Fig. 6. Under oxidizing conditions (thionine treatment or at less negative potential, ≈ -200 mV vs SHE), the FTIR spectrum of TmHydSPDT resembles closely that of the Hox spectrum of TmHydSADT (Fig. S11A and Fig. 6A). The EPR spectrum of thionine treated TmHydSPDT also showed similar g-values to the Hox state of TmHydSADT (Fig S11B). Upon reduction of TmHydSPDT either by treatment with sodium dithionite or electrochemically, a mixture of states appeared (Fig. S11A and Fig. 6A), which showed distinct FTIR bands compared to the oxidized state. 15N labeling of the CN- ligands identified that the bands at 2078, 2070, 2048 and 2012 cm-1 arise from CN- vibrations (Fig. S12). A closer analysis reveals that five of the FTIR bands present in the reduced sample (2078, 2070, 1966, 1935, 1790 cm-1), are red-shifted slightly (≈ 10 cm-1), and the other four (2048, 2012, 1883, 1862 cm-1) are strongly red-shifted (≈ 80 cm-1), compared with the Hox state. While, the first species is reminiscent of the well-known (unprotonated) Hred state, the FTIR bands of the second reduced species TmHydSPDT closely resemble ∗ that of the H state of TmHydSADT. EPR analysis of sodium dithionite reduced TmHydSPDT also indicated reduction of the H-cluster to an EPR silent state (Fig. S11B). Interestingly, in FTIR spectroelectrochemistry of TmHydSPDT, when the potential was further decreased

Spectroelectrochemical FTIR of TmHydSADT Spectroelectrochemical FTIR has been widely used to study the equilibrium redox properties of several [NiFe] and [FeFe] hydrogenases.14,52,53 Here, we used this method to characterize the potential dependence of the redox states in holo-TmHydS. The protein sample was mixed with a redox mediator cocktail, loaded into a transmission FTIR cell containing a three – electrode system for poising the potential (see experimental section) and FTIR spectra were measured at each potential during reductive and oxidative titrations. The reductive titration of TmHydSADT commenced at open circuit potential (OCP) around –200 mV vs. SHE. The FTIR spectrum at this potential is similar to the spectrum of the thionine oxidized Hox sample (Fig. 5A). As the potential was lowered the intensities of this set of bands decreased and a new set of bands appeared that were the same as in the reduced samples (Fig. 5A). These bands reached their highest intensities around – 440 mV. When the potential was lowered even further, new peaks, most of which are slightly red shifted with respect to the –440 mV spectrum, started to appear and attain their highest intensity around –700 mV (Fig. 5A). No more shifts in the FTIR bands were observed upon decreasing the potentials further to –740 mV. By gradually increasing the potential the protein could be reversibly re-oxidised (Fig. S8). After re-oxidation, the FTIR band intensities of the species obtained at around –200 mV had decreased by only about 5% compared to the first spectrum of the reductive titration (Fig. S8). When the potential was increased further to +230 mV, the signal intensity

7 ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 14

PDT

Fig. 6. FTIR spectroelectrochemistry of TmHydS . A) Selected FTIR spectra at -205, -545 and -785 mV are shown. Peaks ∗ ∗ shaded in red, orange, blue and green belongs to Hox, Hred, H and H states, respectively. The experiment was performed with ≈ 1.1 mM protein and at 15 °C in 50 mM Tris/HCl pH 8, 200 mM KCl buffer. B) Changes in FTIR absorbance with changing -1 -1 -1 -1 electrode potential at peak position 1944 cm , 1935 cm , 1862 cm and 1853 cm are shown. Solid lines represent the n=1 Nernst fits corresponding to the model shown above. Mid-point potentials extracted for the transitions from the fits ∗ ∗ ∗ ∗ are: H /H = -440 ± 10 mV, H / H = -460 ± 10 mV, H / H = -670 ± 10 mV and H / H = -690 ± 10 mV.

( 10.58 The midpoint potential extracted for the ∗ ∗ H /H transition in TmHydSADT is –570±10 mV (Fig. 5A and Fig. 7A). This potential is comparable to the midpoint potential of H H /H H (–540 mV) observed in DdHydAB,14,41 indicating that the two enzymes have similarly negative [4Fe-4S]H subsite redox potentials. The extremely negative apparent redox potential of the [4Fe4S]H sub-cluster in DdHydAB was recently attributed to redox anti-cooperativity with the proximal F-cluster.55 Spectroscopic analysis of TmHydS suggests that reduction of the [2Fe]H subsite occurs without protonation of the bridging amine group of the ADT bridge and thus electronic rearrangement of the H-cluster is uncoupled from protonation. This observation indicates that the protein environment plays a substantial role in proton-coupled events occurring at the H-cluster. Replacement of residues important for the proton transfer channel, one of which is the cysteine in the vicinity of the amine group of the ADT bridge, most likely impairs protonation of the Hcluster in TmHydS (Fig. S15 and S16). According to our current understanding, it is unclear how TmHydS achieves the necessary protonation/deprotonation steps required for H2 conversion. Exchange of the cysteine in the proton channel for serine abolished activity in CrHydA1.30 Therefore, alternative proton transfer pathways must be operational in TmHydS. Analysis of a homology model of TmHydS reveals the presence of a tyrosine residue (Y223 replaced by F417 in CpI) and a serine residue (S267 replaced by M497 in CpI) in the neighborhood of the ADT bridge (Fig. S17). We speculate that these two residues could act as immediate proton acceptors and aid in deprotonation of the ADT bridge in TmHydS during H2 oxidation. During H2 production, on the other hand, protonation of the H-cluster is strongly disfavored even in the most reduced states. This may explain why there is such a large catalytic bias toward H2 oxidation in TmHydS. One of the most striking features of the IR spectra in the reduced states of TmHydSADT is the conservation of the bridging CO ligand. This appears to be a consequence of the high redox potential of [2Fe]H that allows its reduction even without protonation. Alterations in amino acids surrounding the H-cluster pocket are likely to be responsible for such a high redox potential. One such change in amino acid identity exists nearby the bridging CO ligand. The methionine (M353, CpI) in the prototypical [FeFe] hydrogenases is replaced by glycine in TmHydS (G177) (Fig. 1B). Interaction of methionine with the H-cluster may increase electron density on the bridging CO forcing it to adopt a terminal configuration when the [2Fe]H subcluster is reduced. This, in turn, may facilitate protonation of the ADT ligand. Although mutation of this methionine by leucine in CrHydA1 leads to only minor changes

Page 10 of 14

Fig. 7. Schematic representation of redox-states obADT served for the H-cluster of TmHydS (A) and PDT TmHydS (B). The midpoint potentials of the redox transition at pH 8 and 15 °C are given. The cube represents the – [4Fe-4S]H cluster. The terminal CO and CN ligands are not shown for clarity. The detailed reasoning for the assignment of these oxidation states is given in the text.

in the FTIR and EPR spectra,30 replacing it with glycine may lead to more dramatic effects due to the differences in physico-chemical properties of the two amino acid residues. Studying a G177M variant of TmHydS in future experiments would provide important insight into the role of this bridging-terminal transition of the CO ligand during H2 conversion. Low but detectable H2 conversion activity of TmHydS proves beyond doubt that this protein can indeed detect hydrogen and hence may act as a hydrogen sensor. The Km value of TmHydSADT for H2 appears to be quite high (> 400 µM) (Fig. S18), which is similar to other [FeFe] hydrogenases.63 However, a relatively high poten∗ tial for the H /H transition (–300mV) may allow reduction of the [2Fe]H, even in the presence of trace amounts of H2. Here it is useful to compare the midpoint potential of the active site of TmHydSADT (defined as ∗ ∗ ∗ E1/2= /   /2 ; E1/2 (pH 8) =-435 mV) / with the proton reduction thermodynamic potential (E2H+/H2 (pH 8) = -472 mV).64 A more positive value for the active site indicates that the active site would thermodynamically oxidize H2 more easily than evolve it. The physiological role of TmHydS may be in the H2dependent transcriptional regulation of both Tmhydαβγ (the gene of electron bifurcating [FeFe] hydrogenase present in the same operon as TmHydS) or in the transcriptional regulation of the prototypical hydrogenase in a separate operon. The reduced TmHydS formed in presence of H2 may downregulate expression of Tmhydαβγ, while in the absence of H2 the oxidized TmHydS is

10 ACS Paragon Plus Environment

Page 11 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

formed, which may upregulate its expression. In recent work on the H2 producing bacterium Thermoanaerobacterium saccharolyticum it was proposed that the HydS protein may regulate the expression of alcohol dehydrogenase.65 Thus, it is also possible that TmHydS participates in other regulatory pathways that are governed by partial pressure of H2 in the cell. How formation of the reduced/ oxidized [2Fe]H subsite affects transcriptional regulation of other genes is beyond the scope of this work. Nevertheless, it can be speculated that the FeS cluster present at the C-terminus of the H-cluster in TmHydS is involved in the detection of the redox changes of the H-cluster and transmits these signals to the PAS domain. PAS domains are proposed to detect chemical and physical stimuli and propagate signal transduction via tertiary or quaternary structural changes.66 Similarly, the PAS domain present in TmHydS is likely to receive a stimulus through the FeS cluster when the [2Fe]H subsite is reduced, which could induce some structural changes in the protein so that it can bind a serine/threonine phosphatase and thereby regulate transcription. Structural studies of the oxidized and reduced states of TmHydS could shed light on this proposition. As it is believed that HydS functions by responding to the partial pressure of H2,18 in vitro studies to test the interaction of TmHydS with potential partners such as serine/threonine phosphatase under different concentration of H2 would provide important clues into its mechanism.

typical class, and may ultimately provide clues for designing novel hydrogen conversion catalysts for industrial purposes. EXPERIMENTAL SECTION EPR spectroscopy: For X-band (9.63 GHz) EPR spectroscopy, 200 μl of sample was transferred to X-band quartz EPR tubes and frozen in liquid nitrogen. The spectra were recorded on a Bruker ELEXSYS E500 CW EPR spectrometer. Cryogenic temperatures were maintained with liquid He using an Oxford ESR900 helium flow cryostat. The measurement parameters were: modulation frequency, 100 kHz, modulation amplitude, 7.46 G, time constant, 81.92 ms, conversion time, 81.92 ms. The temperature and microwave power were varied and are given in the figure legends. All the spectra were processed using home written programs in the MATLABTM environment. EPR simulations were also carried out in MATLABTM using the ‘esfit’ fitting function from the Easyspin package.67 Spin quantitation was achieved by comparison with a 1 mM CuSO4, 10 mM EDTA standard. Activity assays: Hydrogen production was measured by gas chromatography on a 6890 Series GC System (Agilent Technologies) using a molecular sieve 5 Å PLOT column. 10 µg of holo-TmHydS was added to 10 mM methyl viologen reduced with 100 mM sodium dithionite in 200 mM potassium phosphate buffer pH 8 in 2.5 mL plastic tubes with rubber stoppers. The reaction mixture was purged with argon for 5 min and incubated at the desired reaction temperature for 5 min before 0.3 mL of the headspace gas was extracted for analysis. Desired reaction temperatures were achieved by incubating the reaction vials in a temperature-controlled water bath. Hydrogen content was quantified by comparison with a 100% H2 standard. All values are the average of three measurements after subtracting the value of a blank measurement. Hydrogen oxidation was measured by following the reduction of 1 mM benzyl viologen in H2- saturated 200 mM phosphate buffer pH 8 with 1.5 µg holo-TmHydS. The measurements were performed in 1.5 ml plastic cuvettes using an Ocean Optics DH-mini UV-Vis-NIR light source and a USB2000 + XR1-ES detector, operated by the Spectra Suite software. The desired reaction temperatures were achieved using a temperature controlled cuvette holder (CUV-QPOD-2E-ABSKIT, Ocean Optics). The specific activity of the protein was measured by the initial rate of change of absorbance at 600 nm due to the reduction of benzyl viologen. All values are the average of three measurements after subtracting the value of the blank measurement. The other details are described in the figure legends.

CONCLUSION The sensory type [FeFe]-hydrogenase of T. maritima was successfully prepared using recombinant methods followed by artificial maturation. The artificially reconstituted holo-TmHydSADT showed low H2-conversion activity compared to the prototypical or electron bifurcating [FeFe]-hydrogenases, supporting its sensory role. We observed that holo-TmHydSADT showed a low affinity towards CO, which suggests that CO binding at the open coordination site of the H-cluster in TmHydS is impaired. From FTIR spectroelectrochemistry, we could identify ∗ ∗ three different redox states:H , H , and H of ADT ∗ TmHydS . Our data suggest that in the H , and ∗ H states, even though the [2Fe]H is in Fe(I)Fe(I) configuration, the bridging CO is retained. This observation implies that reduction of the [2Fe]H in TmHydSADT proceeds without protonation of the bridging ADT ligand. This hypothesis was supported by FTIR spectroscopic results on TmHydSPDT, where reduction at the [2Fe]H subsite was also observed, unlike in other well characterized [FeFe] hydrogenases. The high midpoint potential of ∗ the H /H transition in TmHydSADT is possibly related to changes in the H-cluster surrounding. The high midpoint potential of the active site means that even trace amounts of H2 can reduce the [2Fe]H subsite of TmHydS. Thus, a large number of adaptations to the protein surrounding in TmHydS perturb the H-cluster properties specializing this [FeFe] hydrogenase for its signalling role. These effects also advance our understanding of the role of the protein surrounding in hydrogenases of the proto-

FTIR spectroscopy and spectroelectrochemistry: For FTIR spectroscopy, 10 μl of sample was placed between two CaF2 windows (Korth Kristalle, Altenholz), separated by a 50 μm Teflon spacer. These windows were then accommodated in a FTIR cell and fixed with rubber rings.

11 ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

FTIR spectra of the samples were recorded using a Bruker IFS 66v/S FTIR spectrometer equipped with a liquid nitrogen cooled Bruker mercury cadmium telluride (MCT) detector. Spectra were collected at 15 °C in the double sided, forward-backward mode with 1000 scans, and a resolution of 2 cm-1, an aperture setting of 2 mm and scan velocity of 20 kHz. Spectroelectrochemical FTIR was carried out in the same spectrometer set-up, but using a home built electrochemical IR cell, constructed according to an original design by Moss et. al.68 Protein samples (≈ 1 - 1.5 mM, 54 µl) mixed with 0.5 mM of each redox mediator (potassium indigo trisulphonate, anthraquinone-1,5-disulfonic acid, anthraquinone-2-sulfonate, benzyl viologen, methyl viologen and 1,1´,2,2´-Tetramethyl-[4,4´-bipyridine]-1,1´-diium iodide) and loaded between two CaF2 windows on an electrochemically reduced gold mesh working electrode (approximately 50 µm thick) in electrical contact with a platinum counter electrode. An Ag/AgCl (1 M KCl) electrode was used as a reference and was calibrated before and after measurement with (hydroxymethyl)ferrocine (E = 436 mV). The potential was monitored using an Autolab PGSTAT101 potentiostat controlled by Nova software. The sample was equilibrated at a particular potential until the current flowing through the cell reached equilibrium, which took ≈ 30–75 min. This was followed by measurement of the FTIR spectra at that potential. All the FTIR spectra were processed using home written routines in MATLABTM environment. The oxygen sensitivity experiment was carried out in an anaerobic chamber (100% N2), at 20 °C by placing 5 µl of the samples on an ATR cell (Harrick BioATR cell II). The protein samples were exposed to oxygen by injecting 1 ml air into the headspace of the ATR cell. The changes in FTIR spectra of the samples after air exposure were recorded using the Bruker Tensor 27 system also equipped with a liquid nitrogen cooled MCT detector.

Page 12 of 14

Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Funding Sources The work was supported by the Max Planck Society and in part by JSPS KAKENHI grant number 16K21748 (H.O.). ACKNOWLEDGMENT We thank Ingeborg Heise for synthesis of (Et4N)2[Fe2(ADT)(CO)4(CN)2], and (Et4N)2[Fe2(PDT) (CO)4(CN)2], Alaa Alsheikh Oughli for synthesis of 1,1´,2,2´-tetramethyl-[4,4´-bipyridine]-1,1´-diium iodide, Miriam Frenzer for gas chromatographic measurements and Yvonne Brandenburger for technical assistance. REFERENCES (1)Lubitz, W.; Ogata, H.; Rudiger, O.; Reijerse, E. Chem. Rev. 2014, 114, 4081-4148. (2)Ogata, H.; Lubitz, W.; Higuchi, Y. J. Biochem. 2016, 160, 251258. (3) Shima, S.; Thauer, R. K. Chem. Rec. 2007, 7, 37-46. (4) Peters, J. W.; Schut, G. J.; Boyd, E. S.; Mulder, D. W.; Shepard, E. M.; Broderick, J. B.; King, P. W.; Adams, M. W. Biochim. Biophys. Acta 2015, 1853, 1350-1369. (5)Winkler, M.; Esselborn, J.; Happe, T. Biochim. Biophys. Acta 2013, 1827, 974-985. (6) Shafaat, H. S.; Rudiger, O.; Ogata, H.; Lubitz, W. Biochim. Biophys. Acta 2013, 1827, 986-1002. (7)Glick, B. R.; Martin, W. G.; Martin, S. M. Can. J. Microbiol. 1980, 26, 1214-1223. (8) Hatchikian, E. C.; Forget, N.; Fernandez, V. M.; Williams, R.; Cammack, R. Eur. J. Biochem. 1992, 209, 357-365. (9)Vignais, P. M.; Billoud, B. Chem. Rev. 2007, 107, 4206-4272. (10) Calusinska, M.; Happe, T.; Joris, B.; Wilmotte, A. Microbiology 2010, 156, 1575-1588. (11) Poudel, S.; Tokmina-Lukaszewska, M.; Colman, D. R.; Refai, M.; Schut, G. J.; King, P. W.; Maness, P. C.; Adams, M. W. W.; Peters, J. W.; Bothner, B.; Boyd, E. S. Biochim. Biophys. Acta 2016, 1860, 1910-1921. (12) Greening, C.; Biswas, A.; Carere, C. R.; Jackson, C. J.; Taylor, M. C.; Stott, M. B.; Cook, G. M.; Morales, S. E. ISME J. 2016, 10, 761-777. (13) Nicolet, Y.; Piras, C.; Legrand, P.; Hatchikian, C. E.; Fontecilla-Camps, J. C. Structure 1999, 7, 13-23. (14) Roseboom, W.; De Lacey, A. L.; Fernandez, V. M.; Hatchikian, E. C.; Albracht, S. P. J. Biol. Inorg. Chem. 2006, 11, 102-118. (15) Kamp, C.; Silakov, A.; Winkler, M.; Reijerse, E. J.; Lubitz, W.; Happe, T. Biochim. Biophys. Acta 2008, 1777, 410-416. (16) Swanson, K. D.; Ratzloff, M. W.; Mulder, D. W.; Artz, J. H.; Ghose, S.; Hoffman, A.; White, S.; Zadvornyy, O. A.; Broderick, J. B.; Bothner, B.; King, P. W.; Peters, J. W. J. Am. Chem. Soc. 2015, 137, 1809-1816. (17) Schut, G. J.; Adams, M. W. J Bacteriol 2009, 191, 4451-4457. (18) Zheng, Y.; Kahnt, J.; Kwon, I. H.; Mackie, R. I.; Thauer, R. K. J. Bacteriol. 2014, 196, 3840-3852.

ASSOCIATED CONTENT This material is available free of charge via the Internet at http://pubs.acs.org. Details regarding construction of the clone, overexpression, purification and artificial maturation are provided in the supplementary information. FTIR data of O2 denaturation of TmHydSADT, comparison of TmHydSADT with other [FeFe] hydrogenases, FTIR and EPR spectroscopic data of TmHydSPDT are also included. AUTHOR INFORMATION Corresponding Author * [email protected] * [email protected] Notes The authors declare no competing financial interest.

12 ACS Paragon Plus Environment

Page 13 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

(19) Soboh, B.; Linder, D.; Hedderich, R. Microbiology 2004, 150, 2451-2463. (20) Huang, H.; Wang, S.; Moll, J.; Thauer, R. K. J. Bacteriol. 2012, 194, 3689-3699. (21) Schuchmann, K.; Muller, V. J. Biol. Chem. 2012, 287, 3116531171. (22)Henry, J. T.; Crosson, S. Ann. Rev. Microbiol., Vol 65 2011, 65, 261-286. (23) Crooks, G. E.; Hon, G.; Chandonia, J. M.; Brenner, S. E. Genome Res. 2004, 14, 1188-1190. (24) Buhrke, T.; Lenz, O.; Porthun, A.; Friedrich, B. Mol. Microbiol. 2004, 51, 1677-1689. (25)Shaw, A. J.; Hogsett, D. A.; Lynd, L. R. J. Bacteriol. 2009, 191, 6457-6464. (26) Ballor, N. R.; Paulsen, I.; Leadbetter, J. R. Microb. Ecol. 2012, 63, 282-294. (27)Cornish, A. J.; Gartner, K.; Yang, H.; Peters, J. W.; Hegg, E. L. J. Biol. Chem. 2011, 286, 38341-38347. (28) Mulder, D. W.; Guo, Y.; Ratzloff, M. W.; King, P. W. J. Am. Chem. Soc. 2017, 139, 83-86. (29) Mulder, D. W.; Ratzloff, M. W.; Bruschi, M.; Greco, C.; Koonce, E.; Peters, J. W.; King, P. W. J. Am. Chem. Soc. 2014, 136, 15394-15402. (30) Knorzer, P.; Silakov, A.; Foster, C. E.; Armstrong, F. A.; Lubitz, W.; Happe, T. J. Biol. Chem. 2012, 287, 1489-1499. (31) Juszczak, A.; Aono, S.; Adams, M. W. J. Biol. Chem. 1991, 266, 13834-13841. (32) Pan, G.; Menon, A. L.; Adams, M. W. J. Biol. Inorg. Chem. 2003, 8, 469-474. (33) Verhagen, M. F.; O'Rourke, T.; Adams, M. W. Biochim. Biophys. Acta 1999, 1412, 212-229. (34)Greene, B. L., Schut, G.J., Adams, M.W.W, Dyer, R.B ACS Catalysis 2017, 7, 2145−2150. (35) Esselborn, J.; Lambertz, C.; Adamska-Venkatesh, A.; Simmons, T.; Berggren, G.; Noth, J.; Siebel, J.; Hemschemeier, A.; Artero, V.; Reijerse, E.; Fontecave, M.; Lubitz, W.; Happe, T. Nat. Chem. Biol. 2013, 9, 607-609. (36) Birrell, J. A., Rudiger, O., Reijerse, E.J., Lubitz, W. Joule 2017, 1, 61-76. (37) Voordouw, G.; Hagen, W. R.; Kruse-Wolters, K. M.; van Berkel-Arts, A.; Veeger, C. Eur. J. Biochem. 1987, 162, 31-36. (38) Shepard, E. M.; Mus, F.; Betz, J. N.; Byer, A. S.; Duffus, B. R.; Peters, J. W.; Broderick, J. B. Biochemistry 2014, 53, 4090-4104. (39) Kuchenreuther, J. M.; Grady-Smith, C. S.; Bingham, A. S.; George, S. J.; Cramer, S. P.; Swartz, J. R. PLoS One 2010, 5, e15491. (40)Bernhard, M.; Buhrke, T.; Bleijlevens, B.; De Lacey, A. L.; Fernandez, V. M.; Albracht, S. P.; Friedrich, B. J. Biol. Chem. 2001, 276, 15592-15597. (41) Birrell, J. A.; Wrede, K.; Pawlak, K.; Rodriguez-Macia, P.; Rudiger, O.; Reijerse, E. J.; Lubitz, W. Isr. J. Chem.2016, 56, 852863. (42) Berggren, G.; Adamska, A.; Lambertz, C.; Simmons, T. R.; Esselborn, J.; Atta, M.; Gambarelli, S.; Mouesca, J. M.; Reijerse, E.; Lubitz, W.; Happe, T.; Artero, V.; Fontecave, M. Nature 2013, 499, 66-69. (43)Siebel, J. F.; Adamska-Venkatesh, A.; Weber, K.; Rumpel, S.; Reijerse, E.; Lubitz, W. Biochemistry 2015, 54, 1474-1483. (44)Chen, Z.; Lemon, B. J.; Huang, S.; Swartz, D. J.; Peters, J. W.; Bagley, K. A. Biochemistry 2002, 41, 2036-2043. (45) Morra, S.; Maurelli, S.; Chiesa, M.; Mulder, D. W.; Ratzloff, M. W.; Giamello, E.; King, P. W.; Gilardi, G.; Valetti, F. Biochim. Biophys. Acta 2016, 1857, 98-106. (46)Lemon, B. J.; Peters, J. W. Biochemistry 1999, 38, 1296912973. (47) Caserta, G.; Adamska-Venkatesh, A.; Pecqueur, L.; Atta, M.; Artero, V.; Roy, S.; Reijerse, E.; Lubitz, W.; Fontecave, M. Biochim. Biophys. Acta 2016, 1857, 1734-1740.

(48)Bennett, B.; Lemon, B. J.; Peters, J. W. Biochemistry 2000, 39, 7455-7460. (49) Duche, O.; Elsen, S.; Cournac, L.; Colbeau, A. FEBS J. 2005, 272, 3899-3908. (50) Ash, P. A.; Liu, J.; Coutard, N.; Heidary, N.; Horch, M.; Gudim, I.; Simler, T.; Zebger, I.; Lenz, O.; Vincent, K. A. J. Phys. Chem. B 2015, 119, 13807-13815. (51) Stripp, S. T.; Goldet, G.; Brandmayr, C.; Sanganas, O.; Vincent, K. A.; Haumann, M.; Armstrong, F. A.; Happe, T. Proc. Natl. Acad. Sci. U S A 2009, 106, 17331-17336. (52)Silakov, A.; Kamp, C.; Reijerse, E.; Happe, T.; Lubitz, W. Biochemistry 2009, 48, 7780-7786. (53) De Lacey, A. L.; Pardo, A.; Fernandez, V. M.; Dementin, S.; Adryanczyk-Perrier, G.; Hatchikian, E. C.; Rousset, M. J. Biol. Inorg. Chem. 2004, 9, 636-642. (54) Adamska-Venkatesh, A.; Krawietz, D.; Siebel, J.; Weber, K.; Happe, T.; Reijerse, E.; Lubitz, W. J. Am. Chem. Soc. 2014, 136, 11339-11346. (55)Rodriguez-Macia, P.; Pawlak, K.; Rudiger, O.; Reijerse, E. J.; Lubitz, W.; Birrell, J. A. J. Am. Chem. Soc. 2017, 139, 15122-15134. (56) Artz, J. H.; Mulder, D. W.; Ratzloff, M. W.; Lubner, C. E.; Zadvornyy, O. A.; LeVan, A. X.; Williams, S. G.; Adams, M. W. W.; Jones, A. K.; King, P. W.; Peters, J. W. J. Am. Chem. Soc. 2017, 139, 9544-9550. (57)Lubitz, W.; Reijerse, E.; van Gastel, M. Chem. Rev. 2007, 107, 4331-4365. (58) Sommer, C.; Adamska-Venkatesh, A.; Pawlak, K.; Birrell, J. A.; Rudiger, O.; Reijerse, E. J.; Lubitz, W. J. Am. Chem. Soc. 2017, 139, 1440-1443. (59) Senger, M.; Laun, K.; Wittkamp, F.; Duan, J.; Haumann, M.; Happe, T.; Winkler, M.; Apfel, U. P.; Stripp, S. T. Angew. Chem. Int. Ed. Engl. 2017, 56 (60) Mebs, S.; Senger, M.; Duan, J.; Wittkamp, F.; Apfel, U. P.; Happe, T.; Winkler, M.; Stripp, S. T.; Haumann, M. J. Am. Chem. Soc. 2017, 139, 12157-12160. (61) Liu, Z. P.; Hu, P. J. Am. Chem. Soc. 2002, 124, 5175-5182. (62) Yu, L.; Greco, C.; Bruschi, M.; Ryde, U.; De Gioia, L.; Reiher, M. Inorg. Chem. 2011, 50, 3888-3900. (63) Fourmond, V.; Baffert, C.; Sybirna, K.; Dementin, S.; AbouHamdan, A.; Meynial-Salles, I.; Soucaille, P.; Bottin, H.; Leger, C. Chem. Commun. 2013, 49, 6840-6842. (64)Fourmond, V.; Baffert, C.; Sybirna, K.; Lautier, T.; Abou Hamdan, A.; Dementin, S.; Soucaille, P.; Meynial-Salles, I.; Bottin, H.; Leger, C. J. Am. Chem. Soc. 2013, 135, 3926-3938. (65) Eminoglu, A.; Murphy, S. J.; Maloney, M.; Lanahan, A.; Giannone, R. J.; Hettich, R. L.; Tripathi, S. A.; Belduz, A. O.; Lynd, L. R.; Olson, D. G. Biotechnol. Biofuels 2017, 10, 282. (66) Moglich, A.; Ayers, R. A.; Moffat, K. Structure 2009, 17, 12821294. (67) Stoll, S.; Schweiger, A. J. Magn. Reson. 2006, 178, 42-55. (68) Moss, D.; Nabedryk, E.; Breton, J.; Mantele, W. Eur. J. Biochem. 1990, 187, 565-572.

13 ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 14

TOC figure

ACS Paragon Plus Environment

14