Quick and Spontaneous Transformation between [3Fe–4S] and [4Fe–4S] Iron–Sulfur Clusters in the tRNA-Thiolation Enzyme TtuA

Iron–sulfur (Fe–S) clusters are essential cofactors for enzyme activity. These Fe–S clusters are present in structurally diverse forms, including [4Fe–4S] and [3Fe–4S]. Type-identification of the Fe–S cluster is indispensable in understanding the catalytic mechanism of enzymes. However, identifying [4Fe–4S] and [3Fe–4S] clusters in particular is challenging because of their rapid transformation in response to oxidation–reduction events. In this study, we focused on the relationship between the Fe–S cluster type and the catalytic activity of a tRNA-thiolation enzyme (TtuA). We reconstituted [4Fe–4S]-TtuA, prepared [3Fe–4S]-TtuA by oxidizing [4Fe–4S]-TtuA under strictly anaerobic conditions, and then observed changes in the Fe–S clusters in the samples and the enzymatic activity in the time-course experiments. Electron paramagnetic resonance analysis revealed that [3Fe–4S]-TtuA spontaneously transforms into [4Fe–4S]-TtuA in minutes to one hour without an additional free Fe source in the solution. Although the TtuA immediately after oxidation of [4Fe–4S]-TtuA was inactive [3Fe–4S]-TtuA, its activity recovered to a significant level compared to [4Fe–4S]-TtuA after one hour, corresponding to an increase of [4Fe–4S]-TtuA in the solution. Our findings reveal that [3Fe–4S]-TtuA is highly inactive and unstable. Moreover, time-course analysis of structural changes and activity under strictly anaerobic conditions further unraveled the Fe–S cluster type used by the tRNA-thiolation enzyme.


Introduction
Iron-sulfur (Fe-S) clusters, important cofactors of proteins (Fe-S proteins), are responsible for electron transition, protein stabilization, ligand binding, etc. [1]. The Fe-S clusters are essential for various cellular processes such as photosynthesis, respiration, nitrogen fixation, and oxygen sensing [2][3][4][5]. Recently, Fe-S clusters were found to regulate DNA replication, DNA repair [6,7], RNA replication, and RNA regulation [8,9]. These findings suggested that the biological roles of Fe-S clusters are more extensive than speculated previously. Fe-S clusters differ structurally in terms of a varying number of Fe and S atoms, for example, [ [10]. Identification of the type of Fe-S cluster is essential to understand the catalytic mechanism of Fe-S protein. However, accurate structural determination of [4Fe-4S] and [3Fe-4S] is challenging because the structural difference involves only one Fe atom (hereafter termed as the unique Fe), and [4Fe-4S] is sensitive to oxidation and readily decays to [3Fe-4S] [11,12].
In this study, we reconstituted [4Fe-4S]-TtuA and oxidized it to [3Fe-4S]-TtuA under strictly anaerobic conditions. Then we analyzed the structure of the Fe-S cluster using electron paramagnetic resonance (EPR) in time-course experiments and evaluated the activity of these samples. Results from EPR spectroscopy revealed that the [3Fe-4S] cluster in TtuA spontaneously transformed into [4Fe-4S] with the unique Fe in minutes to one hour without an additional free Fe source in the solution. The tRNA-thiolation assay of these samples revealed that while TtuA was inactive immediately after the oxidation of [4Fe-4S]-TtuA to [3Fe-4S]-TtuA, its activity gradually recovered to a level comparable with [4Fe-4S]-TtuA, which was corresponding to an increase in [4Fe-4S]-TtuA. Furthermore, we found that [3Fe-4S]-TtuA could not interact with the sulfur donor at the C-terminus of TtuB, indicating that TtuA requires [4Fe-4S] only for 5-methyl-2-thiouridine (m 5 s 2 U) biosynthesis (tRNA-thiolation) at position 54 of tRNA. Considering the sequence similarity, especially the highly conserved active residues in the TtuA/Ncs6 family, we propose that the unique Fe of [4Fe-4S] is essential for TtuA/Ncs6 family members to catalyze tRNAthiolation. Our findings reveal that the correlation analysis of time-course between the structural changes of cofactor Fe-S clusters and the enzymatic activity of that protein under strictly anaerobic conditions is necessary to reveal the reaction mechanism.  Figure S3).

Results
First, we confirmed whether excess DT affects the structures of Fe-S clusters in TtuA. The linear shape of EPR spectra and the g-value of [4Fe-4S] 1+ -TtuA did not alter until 30 min after reduction, showing that reduction with DT does not cause structural changes in the Fe-S cluster bound to TtuA ( Figure S4).

Enzymatic Activity of Oxidized TtuA was Recovered by Reconstitution of [4Fe-4S]-TtuA
We analyzed the enzymatic activity of oxidized [4Fe-4S]-TtuA under strictly controlled anaerobic conditions. Because [3Fe-4S] transforms to [4Fe-4S] cluster within one hour (Figure 1d), limiting the reaction time is important. We optimized the reaction conditions, including TtuA and TtuB concentrations and temperature, to monitor the enzymatic activity in a relatively short reaction time.
We measured the synthesis of m 5 s 2 U using oxidized [4Fe-4S]-TtuA in the presence and absence of K 3 [Fe(CN) 6 ]. While m 5 s 2 U was detectable using [4Fe-4S]-TtuA, nearly no m 5 s 2 U was synthesized at 5 min after oxidation, indicating that [3Fe-4S]-TtuA had no enzymatic activity. In contrast, the amount of m 5 s 2 U synthesis increased with time (1, 2, and 24 h after oxidation), irrespective of the fact that excess K 3 [Fe(CN) 6 ] was removed from the solution (Figures 2 and S6). When TtuA activity was normalized to 100% before oxidation, it was restored to 30%-50% in 1 h, >70% in 2 h, and 70%-80% in 24 h (Table 2). Combining these observations with EPR results, according to which [3Fe-4S] was transformed to [4Fe-4S] cluster with time ( Figure 1d, Table 1), it is apparent that TtuA activity increased as the amount of [4Fe-4S] cluster bound to TtuA increased. Furthermore, TtuA activity did not recover to 100% with or without excess K 3 [Fe(CN) 6 ]. This finding indicates that some [3Fe-4S] clusters in TtuA degraded into free Fe and S, resulting in the reconstitution of [4Fe-4S]-TtuA and the formation of apo-TtuA. m s U was synthesized at 5 min after oxidation, indicating that [3Fe-4S]-TtuA had n zymatic activity. In contrast, the amount of m 5 s 2 U synthesis increased with time (1, 2 24 h after oxidation), irrespective of the fact that excess K3[Fe(CN)6] was removed the solution (Figures 2 and S6). When TtuA activity was normalized to 100% befor dation, it was restored to 30%-50% in 1 h, >70% in 2 h, and 70%-80% in 24 h (Tab Combining these observations with EPR results, according to which [3Fe-4S] was formed to [4Fe-4S] cluster with time ( Figure 1d, Table 1), it is apparent that TtuA ac increased as the amount of [4Fe-4S] cluster bound to TtuA increased. Furthermore, activity did not recover to 100% with or without excess K3[Fe(CN)6]. This finding ind that some [3Fe-4S] clusters in TtuA degraded into free Fe and S, resulting in the rec tution of [4Fe-4S]-TtuA and the formation of apo-TtuA.   6 ]. The tRNA-thiolation activity of TtuA before oxidation was normalized to 100%. All data are presented with standard deviation values (N = 3, red asterisk (*): one time was measured from a different lot and scaling by the activity of [Before oxidation]). The quantification result of this assay is shown in Table 2.   [24]. Th spectra are normalized for comparison and the raw data are shown in Figure S8.

Formation of a Functional Fe-S Cluster Bound to TtuA
Enzymes possess a wide variety of Fe-S clusters. Identification of the Fe-S c type is a primary step in understanding the catalytic mechanism of an enzyme.  (Figures 2 and S6). matic activity did not completely recover because the reaction mixture contained TtuA, which was generated by the degradation of [3Fe-4S] cluster (Figure 4). Consi together, we conclude that only [4Fe-4S]-TtuA has enzymatic activity, and the uniq  Figure 1). Because apo-TtuA is an inactive form [25,26] (Figures 2 and S6). Enzymatic activity did not completely recover because the reaction mixture contained apo-TtuA, which was generated by the degradation of [3Fe-4S] cluster ( Figure 4). Considered together, we conclude that only [4Fe-4S]-TtuA has enzymatic activity, and the unique Fe is essential as a binding site for the sulfur donor TtuB-COSH for transferring sulfur to tRNA.
We compared the positions of key residues S55, D59, C130, C133, K137, D161, and C222 in tRNA-thiolation enzymes by superposing crystal structures of TthTtuA with those of MmaNcs6, EcoMnmA, and models of EcoTtcA and MmaThiI (Fe-S cluster type) which was predicted using AlphaFold [39]. The superimpositions showed that all key residues, excluding D161 and C222 of TtuA, are located at similar positions ( Figure S10). Conservation of key residues and structural similarity of the active site reveals that [4Fe-4S] cluster binding of TtuA is likely to be shared by members of the TtuA/Ncs6 family, MnmA, and MmaThiI. Our results are consistent with the previous studies, which have reported that TtcA and MnmA are [4Fe-4S]-binding enzymes [40].

The Change of Fe-S Clusters under Strict Regulation of Oxidation-Reduction
Fe-S clusters are present in diverse forms in the Fe-S proteins. Fe-S clusters are sensitive to oxidation and thus decay readily. Their instability causes incorrect structural determination of Fe-S clusters even with the smallest amount of oxygen contaminant during the experiments. Our findings demonstrated that the Fe-S cluster bound to TtuA is sensitive to oxidation-reduction levels. The [3Fe-4S] clusters quickly and spontaneously transformed into [4Fe-4S] clusters, even under strictly anaerobic conditions. Unexpectedly, in TtuA, the [3Fe-4S] cluster is more unstable than the [4Fe-4S] cluster, which is more oxygen-sensitive and has a unique Fe. Hence, misidentification of the Fe-S clusters may happen without correlation analysis between the structure of Fe-S clusters and the enzymatic activity. Therefore, it is necessary to strictly maintain anaerobic conditions and perform time-course tracking to comprehensively understand the reaction mechanism.
Many proteins that employ the Fe-S cluster are overlooked as apo-type proteins because Fe-S clusters are easily degraded due to their oxygen sensitivity. Although bioinformatics approaches have recently been used to predict the structure and binding site of Fe-S clusters, experiments are required to confirm these results [41]. The current study demonstrated that [3Fe-4S]-TtuA is inactive and spontaneously transforms into an active form of [4Fe-4S]-TtuA. A time-course analysis of the structure of Fe-S clusters under controlled anaerobic conditions minimizes the risk of incorrectly predicting the reaction mechanism and leads to an accurate understanding of the catalytic mechanism of enzymes with Fe-S clusters.

Expression of TtuA
TtuA, from Thermus thermophilus (HB27 strain), was expressed as a C-terminal His6tagged protein in E. coli (B834 DE3 strain) using the pET26 vector expression system (Novagen) [25]. We cultured recombinant E. coli in 3 L of lysogeny broth (Miller) containing 25 µg/mL kanamycin at 37 • C and 150 rpm until the absorption at 600 nm (OD600) reached 0.6. Then we induced overexpression of TtuA with 1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) after cold shock and cultured the cells at 25 • C and 150 rpm for 16 h. The cells were collected by centrifugation at 5000× g for 30 min and stored at −30 • C.

Purification of TtuA
Lysis of recombinant E. coli and purification of TtuA were performed under strictly anaerobic conditions (Vinyl Anaerobic Chamber, COY) with 5% hydrogen and 95% nitrogen as described [25]. The collected cells were lysed by sonication on ice for 45 min in the purification buffer (50 mM HEPES-KOH [pH 7.6], 200 mM ammonium sulfate, 50 mM ammonium acetate, 5 mM magnesium chloride, 10% (v/v) glycerol, and 7 mM 2-mercaptoethanol) containing 0.1% Triton X-100. Then the cells were heat treated at 70 • C for 20 min, and the precipitates were removed by centrifugation at 7000× g for 1 h with a 0.22-µm filter (Millipore). The supernatant was loaded onto a Ni-affinity chromatography column (1 mL His-Trap HP; GE Healthcare) equilibrated with the purification buffer. Non-specifically bound proteins were removed using the wash buffer (purification buffer containing 50 mM imidazole). The target proteins were eluted with a gradient of 50-500 mM imidazole in the purification buffer and further purified on an SEC column (HiLoad 16/60 Superdex 200, GE Healthcare) equilibrated with the purification buffer. Purity was checked using 15% SDS-PAGE and luminescence images were analyzed with the Amersham Imager 680 (GE Healthcare). For reconstituting [4Fe-4S]-TtuA, we incubated TtuA with 5 mM dithiothreitol (DTT) for 10 min at room temperature (RT). Then, 9-fold molar excess of ferric chloride (FeCl 3 ) was added, and the mixture was incubated for 10 min at RT. Subsequently, 9-fold molar excess Na 2 S was added, and the mixture was incubated for 3 h at RT. The iron sulfide precipitate was removed by centrifugation at 7000× g for 10 min with a 0.22-µm filter. Reconstituted TtuA was concentrated using the Amicon Ultra Centrifugal Filter (30-kDa cutoff; Millipore). Excess FeCl 3 and Na 2 S were removed using the Sephadex PD-10 desalting column (GE Healthcare) equilibrated with the purification buffer.
For reconstituting [3Fe-4S]-TtuA, 6-fold molar excess K 3 [Fe(CN) 6 ] was added to [4Fe-4S]-TtuA and incubated at RT for 10 min. The precipitates were removed using centrifugation at 7000× g for 30 s and a 0.22-µm filter. K 3 [Fe(CN) 6 ] and free Fe were removed with a Sephadex PD-10 desalting column equilibrated with the purification buffer. All treatments were performed under anaerobic conditions (oxygen concentration was less than 1 ppm). Oxidized TtuA was immediately used for EPR or activity assay. To

Expression and Purification of TtuB-COSH
TtuB from T. thermophilus (HB27 strain) was expressed as a C-terminal intein-tagged protein in E. coli (B834 DE3 strain) with the pTYB1 vector (New England BioLabs) as described [25]. We cultured recombinant E. coli in 3 L of lysogeny broth containing 100 µg/mL ampicillin at 37 • C and 150 rpm until OD600 reached 0.6. Then we induced overexpression of recombinant TtuB with 1 mM IPTG after cold shock and cultured the cells at 25 • C and 150 rpm for 16 h. The cells were collected by centrifugation at 5000× g for 30 min and stored at −30 • C.
We sonicated the collected cells for 30 min in the purification buffer (20 mM Tris-HCl [pH 8.5 at 25 • C] and 500 mM NaCl) with 0.1% Triton X-100, 0.5 mg/mL lysozyme (Sigma), and 0.1 mg/mL DNase I (Sigma). The precipitates were removed by centrifugation at 40,000× g for 30 min with a 0.22-µm filter. The supernatant was loaded onto a chitin resin (New England BioLabs) equilibrated with the purification buffer. Non-specifically bound proteins were removed using 20 column volumes of the purification buffer. To cleave the intein tag from the TtuB-intein fusion protein, 1 column volume of the purification buffer containing 50 mM ammonium sulfide ((NH 4 ) 2 S) was added to the chitin resin and incubated at RT for 20 h.
TtuB was eluted in a cold room (7 • C) with the purification buffer and then concentrated using the Amicon Ultra Centrifugal Filter (3-kDa cutoff). Excess (NH 4 ) 2 S was removed with the Sephadex PD-10 desalting column equilibrated with the storage buffer (14 mM Tris-HCl [pH 8.5] at 25 • C, 350 mM NaCl, and 30% (v/v) glycerol). The fractions con-taining TtuB were collected, concentrated using the Amicon Ultra Centrifugal Filter , and stored at −80 • C. We confirmed the purity of TtuB by 15% (v/v) SDS-PAGE.
To confirm the presence of sulfur at the C-terminus of TtuB, we removed salts from TtuB solution on ice with the C4 ZipTip (Millipore) and performed matrix-assisted laser desorption/ionization-time-of-flight mass spectrometry (MALDI-TOF MS) using Ultra-flexIII (Bruker) with sinapinic acid as the matrix [25]. The concentration of TtuB was determined by the Bradford method [42], using Protein Assay Dye Reagent Concentrate (Bio-rad) because TtuB from T. thermophilus does not contain tyrosine and tryptophan.

EPR Spectroscopy of the Fe-S Cluster Bound to TtuA
To analyze the structure of the Fe-S cluster bound to TtuA in time-course, we prepared EPR samples under strictly anaerobic conditions. Immediately after [3Fe-4S]-TtuA was prepared, we divided fresh 0.  To analyze the interaction between [3Fe-4S]-TtuA and TtuB-COSH, we prepared EPR samples using a slightly modified version of our protocol [24]. Fresh [3Fe-4S]-TtuA was mixed with TtuB to a final concentration of 0.5 mM and incubated at 25 • C for 5 min under strictly anaerobic conditions. The sample mixture was incubated with a 5-fold molar excess DT at 25 • C for 10 min, aliquoted into quartz EPR tubes, and then frozen. The EPR samples were stored in liquid nitrogen and transported to the Center for Experimental Science and Analysis, Saga University, or Institute for Molecular Science, Okazaki. Continuous wave (CW) X-band EPR spectra were measured in the CW mode at~9.59 GHz using the ELEXSYS E580 spectrometer (Bruker) equipped with the ESR 910 continuous helium flow cryostat (Oxford Instruments) and E500 spectrometer (Bruker) equipped with the ESR 900 continuous helium flow cryostat (Oxford Instruments). The experimental parameters were 12 K for both [4Fe-4S] 1+ and free Fe 3+ , and 40 K for [3Fe-4S] 1+ ; 1 mW microwave power; 100 kHz field modulation; and 10 G modulation amplitude.