https://orcid.org/0000-0003-3458-2574
Figures
Abstract
[Fe-S] clusters are ancient and ubiquitous protein co-factors, which contributed to the emergence of life in an anoxic planet. We have recently identified two minimal [Fe-S] biogenesis systems, MIS and SMS, inferred to be ancestral systems dating back to the Last Universal Common Ancestor and which gave rise to the well-studied modern Iron-Sulfur Cluster (ISC), Nitrogen Fixation (NIF), and Sulfur Mobilization (SUF) machineries. The present study focuses on the ancestor SMS from the hyperthermophilic archaeon Methanocaldococcus jannaschii. Biochemical and structural studies showed that SMS is made of a SmsC2B2 heterotetratmer wherein the SmsC subunit hosts both ATP and [Fe-S] cluster binding sites. Binding of ATP and assembly of [Fe-S] were found to be mutually exclusive allowing for a regulatory coupling between binding of both substrates. Mutagenesis and in vitro transfer experiments revealed the key role of SmsC-contained Cys residues in cluster assembly. Strikingly, the SMS system rescued a non-viable Escherichia coli strain lacking endogenous ISC and SUF systems grown under anoxic conditions, in the presence of Na2S, indicating that sulfide is a source of sulfur for SMS. In addition, we predict that most archaea SmsC proteins hold a similar C-terminal [Fe-S] cluster assembly site. Taking into account those unique structural and functional features, we propose a mechanistic model describing how SmsC2B2 assembles and distributes [4Fe-4S] clusters. Altogether this study established SMS as a new bona fide [Fe-S] biogenesis system that operated in anaerobic prokaryotes prior to evolve to SUF after the Great Oxydation Event.
Citation: Dussouchaud M, Martinez-Carranza M, Garcia P-S, Clémancey M, Blondin G, Betton JM, et al. (2025) Ancestral [Fe-S] biogenesis system SMS has a unique mechanism of cluster assembly and sulfur utilization. PLoS Biol 23(6): e3003223. https://doi.org/10.1371/journal.pbio.3003223
Academic Editor: Ann M. Stock, Rutgers University-Robert Wood Johnson Medical School, UNITED STATES OF AMERICA
Received: April 21, 2025; Accepted: May 22, 2025; Published: June 25, 2025
Copyright: © 2025 Dussouchaud et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting information files. A Structure model and map have been deposited in the RCSB Protein Data Bank and the EMDB under the accession code 9H78 and EMD-51913, respectively. Atomic coordinates and structure factors have been deposited in the RCSB Protein Data Bank under the accession codes 9H7Y, 9HBL, and 9H7X.
Funding: This study was supported by the Agence Nationale de la Recherche ANR-22-CE44-0043-First-FeS (to FB), Agence Nationale de la Recherche ANR-11-LABX-0003-01 (to SOC), Agence Nationale de la Recherche ANR-10-LABX-62-IBEID (to FB), Agence Nationale de la Recherche LabEx GRAL (to SOC), Agence Nationale de la Recherche ANR-17-EURE-0003 (to SOC), the Institut Pasteur (to AH, SG, LS, FB), the Centre National Recherche Scientifique (to FB) and the Université Paris Descartes (to MD). This work is also supported by the European Cooperation in Science and Technology FeSImmChemNet (to SD) is acknowledged. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: AMP-PCP, adenosine-5′-[(β,γ)-methyleno]triphosphate; cryo-EM, cryogenic electron microscopy; GOE, Great Oxidation Event; ISC, Iron-Sulfur Cluster; LB, Luria–Bertani broth; LUCA, Last Universal Common Ancestor; MIS, Minimal ISC System; MVA, mevalonate; NIF, Nitrogen Fixation; SMS, Suf Minimal System; SUF, Sulfur Mobilization
Introduction
Iron-sulfur ([Fe-S]) clusters are inorganic entities that contributed to early stages and subsequent evolution of life [1–4]. Life emerged in an anoxic, iron- and sulfur-rich environment. Minerals containing [Fe-S] clusters formed spontaneously under such conditions, such as pyrite (FeS2) or mackinawite (FeS). Very likely, these abiotically produced [Fe-S] clusters contributed to early forms of life and their subsequent evolution by providing nascent living systems with redox power [5,6]. Subsequently, the Great Oxidation Event (GOE) led to increased oxygenation of the atmosphere and limitation of soluble bioavailable iron. In the late 90s, the discovery of ISC, SUF, and NIF machineries led to the widely assumed notion that [Fe-S] biogenesis ought to be catalyzed to mitigate the GOE-caused deleterious conditions [7–12]. However, our recent discovery of Minimal ISC System (MIS) and Suf Minimal System (SMS) and their inference in the Last Universal Common Ancestor (LUCA) suggested the necessity to catalyze biotically [Fe-S] biogenesis even under favorable pre-GOE conditions [13,14]. This discovery raised exciting questions about the functioning of the ancestral systems, SMS and MIS, and their subsequent evolution to give rise to SUF and ISC/NIF, respectively.
In present-day organisms, [Fe-S] clusters are essential cofactors of proteins, controlling multiple essential cellular processes such as DNA replication and repair, protein synthesis, central metabolism, photosynthesis, respiration, and antiviral defenses. The three machineries, ISC, SUF, and NIF have been the focus of multiple studies in both Prokaryotes and Eukaryotes [7–12]. Briefly, these machineries include a cysteine desulfurase, which provides sulfur from l-cysteine, to a scaffold component that assembles the [Fe-S] cluster, which is subsequently delivered to cellular recipient proteins via dedicated [Fe-S] carriers. The MIS system follows these rules as it has a cysteine desulfurase and a scaffold, but no carrier. The simplicity of the SMS raises additional questions, which are discussed below.
SMS is the ancestor of SUF and has been retained mostly in Archaea. With respect to the six components of the SUF system, SMS has only two components, SmsC and SmsB, encoded by the smsCB operon. We previously showed in vitro that SmsCB proteins bind a [Fe-S] cluster, which can be transfered to apo-aconitase, suggesting that SmsCB acts as a scaffold [14]. However, in the absence of a cysteine desulfurase and carrier, it remains unknown how SmsCB makes and distributes [Fe-S] clusters, and what is the source of sulfur. Here, we have carried out a multidisciplinary analysis of the SMS system from Methanocaldococcus jannaschii, a hyperthermophilic methanogenic archaeon. By combining X-ray crystallography and cryogenic electron microscopy (cryo-EM), we determined the structure of the M. jannaschii [4Fe-4S]-bound SmsC2B2 complex. Unexpectedly, the [4Fe-4S] cluster binding site was located in only one of the SmsC subunits at a C-terminal flexible loop region that is disordered in its absence. Mutagenesis and in vitro [Fe-S] cluster transfer experiments revealed the key role of C-terminal located Cys residues in building the cluster. Biochemical analysis revealed a mutual exclusive relationship between ATP binding and [Fe-S] assembly, leading credence to a regulatory interplay within the interaction of SmsC and both of its substrates. Last, in vivo genetic analysis strongly suggested that mineral sulfide acts as a source of sulfur for SMS to assemble [Fe-S] clusters. Our results demonstrated the uniqueness of the SMS machinery and endowed it with a status of new bona fide [Fe-S] biogenesis system. A wide arrays of archaea, including methanogens and gut inhabitants, is relying on SMS to multiply and this study will undoubtedly help in our understanding of these microbial communities.
Results
SmsB and SmsC form an SmsC2B2 hetero-tetramer complex that assembles a [4Fe-4S] cluster
Biochemical characterization of the SmsCB protein was carried out. To purify the SmsCB protein complex, we independently overproduced recombinant His-tagged SmsC and SmsB proteins in Escherichia coli and reassembled the complex in vitro after Ni affinity chromatography and peptide tag removal. A homogenous preparation of the SmsCB complex was then purified by size exclusion chromatography (Fig 1a). Purified SmsC, SmsB, and SmsCB eluted, on a calibrated size exclusion chromatography column, at a volume corresponding to a monomer (SmsC), a homodimer (SmsB2), and a hetero-tetramer (SmsC2B2), respectively (Fig 1a). That the purified SmsC2B2 was under its apo form was shown by a lack of absorption peak in the UV-visible spectrum, corresponding to [Fe-S] cluster, ATP or FAD. It was subsequently transferred to an anoxic chamber and chemical reconstitution of the [Fe-S] bound species (holo-form) was performed using a five-molar excess of iron and sulfur per SmsC2B2. Consistent with our previous report [14], UV-visible spectra of reconstituted complexes displayed an absorption band at 420 nm indicative of the presence of a [4Fe-4S] cluster on the SmsC2B2 complex (Fig 1b). Subsequent quantitation revealed that iron and sulfur content per mol of complex were 2.5 ± 0.1 Fe and 2.4 ± 0.15 S per chemically reconstituted SmsC2B2 complex. These results were consistent with the binding of one [4Fe-4S] cluster per SmsC2B2 complex. Increasing the molar excess of iron and sulfur led to the appearance of more metallic aggregates on the complex, but not binding of another cluster.
(a) SDS-PAGE and size exclusion chromatography profile of purified Sms proteins. Lane (1): V0, lane (2): SmsC2B2, lane (3): SmsC from the SmsC2B2, lane (4): SmsC monomers, lane (5): SmsB homodimers. (b) UV-Vis absorption spectrum of SmsC2B2. SmsC2B2 (38.5 μM) was incubated with 5 equivalents of Fe2+/SmsC2B2, 5 equivalents of Na2S/SmsC2B2, and 3 mM DTT. (c) 6 K Mössbauer spectra (black vertical bars) of SmsC2B2 (350 µM, 3.6 Fe and 3.4 S/SmsC2B2) recorded using a 0.06 T and a 7 T external magnetic field applied parallel to the γ-beam. The simulations assuming a unique iron site are overlaid as thick red solid lines (see text for parameters). (d) UV-Vis absorption spectrum of chemically reconstituted Sms proteins. SmsC (in red) (70 μM) and SmsB (in blue) (62 μM) were incubated with 5 equivalents of Fe2+/SmsC2 or SmsB2, 5 equivalents of Na2S/SmsC2 or SmsB2 and 3 mM DTT. The data underlying this figure can be found in Fig 1 and S1 Data.
Mössbauer spectra of SmsC2B2, recorded at 5.8 K using a 0.06 and a 7 T external magnetic field applied along the γ-rays direction (Fig 1c), were strongly reminiscent of those previously reported for diamagnetic [4Fe-4S] clusters [15]. They could be satisfactorily reproduced assuming a unique iron site, indicating that the four iron ions were equivalent (Fig 1c). Moreover, the nuclear parameters (isomer shift δ = 0.42 ± 0.01 mm s−1, quadrupole splitting ∆EQ = 1.16 ± 0.05 mm s−1, and EFG rhombicity η = 0.7 ± 0.1) were similar to those reported for [4Fe-4S]2+ clusters with at least three coordinated cysteines [16–18]. Indeed, the shoulder observed at 0.7 mm s−1 was better reproduced when considering two different iron sites in a 3:1 ratio (S1 Fig) with nuclear parameters reminiscent of those obtained for the substrate-free aconitase [19]. Interestingly, no [Fe-S] cluster binding was observed on SmsB alone after reconstitution (Fig 1d). In contrast, we could reconstitute an [Fe-S] cluster-bound SmsC (1.7 Fe and 1.6 S/ SmsC monomer (Fig 1d) with spectroscopic properties (UV-visible absorption and Mössbaur spectroscopy) clearly evidencing a [4Fe-4S] diamagnetic cluster (S2 Fig). Altogether, these biochemical and biophysical analyses demonstrated that SmsB and SmsC forms an SmsC2B2 hetero-tetramer complex that assembles an [4Fe-4S] cluster, probably in the SmsC subunit.
X-ray structural analysis of the [Fe-S] cluster-free SmsC2B2 complex
We set up crystallization conditions both under oxic and anoxic conditions and we obtained crystals of SmsC2B2 in both conditions. To obtain a nucleotide-bound structure, the crystals of SmsC2B2 grown in oxic conditions were soaked with 10 mM adenosine-5′-[(β,γ)-imido]triphosphate (AMP-PNP). For the [Fe-S] cluster-bound structure, we chemically reconstituted the protein complex, as described above, and performed crystallization inside the anoxic chamber yielding new crystal forms. Unfortunately, colorless crystals and further iron quantitation showed that the crystallized SmsC2B2 complex had no bound [Fe-S] cluster, but a sulfate anion bound to SmsC. Diffraction data collection and model refinement statistics are summarized in Table 1. Both SO4- and AMP-PNP-bound crystal structures of the SmsC2B2 complex could be superimposed with a r.m.s.d. of 2.5 Å. SmsB and SmsC assembled into a symmetric hetero-tetramer consisting of two subunits of each protein related by a 2-fold symmetry (Fig 2a). The SmsB subunits consist of a right-handed β-helix core domain (residues 33–289) connected by a short linker to a helical hairpin that serves as an anchoring point for SmsC, analogous to the coupling helices of the transmembrane subunits of ABC transporters [20]. The first N-terminal 32 residues of SmsB are disordered and were not included in the final model. However, we observed a weak extra electron density in the different maps, consistent with an α-helix likely belonging to this N-terminal region close to the center of the complex (Fig 2a).
(a) AMP-PNP-loaded SmsC2B2 crystal structure. (b) Close-up on the binding site of the AMP-PNP/Mg2+ showing residues surrounding the binding site of AMP-PNP. (c) Cryo-EM structure of the [Fe-S]-bound SmsC2B2 complex. (d) Close-up on the binding site of the [Fe-S] cluster showing the C218, C239, and C242 coordinating the [Fe-S] cluster. The cluster binding site lies in a solvent-exposed hydrophobic pocket consisting of residues P40, L216, I223, Y235, F234, and P247. (e) The [Fe-S] cluster-bound SmsC2B2 complex exhibits asymmetry. (f) The [Fe-S] bound SmsC displays a folded C-terminal α-helix composed of residues 227–241 and a short terminal loop. (g) Close-up from the apo-SmsC-COOH region till residue 237.
The SmsC subunit, consisting of a RecA-like domain and a helical domain, contains all the structural motifs of the nucleotide-binding (NBD) subunits of the ABC ATPases (i.e., Walker A, Walker B, ABC signature, D-, Q-, and H-loops) [18] (Fig 2b). Most of the interactions between SmsC and the nucleotide are similar to those observed in other NBD structures of ABC transporters [21]. The structure derived from crystals grown in anoxic conditions showed that the nucleotide has been displaced from the ATPase active site by a sulfate anion located into the usual site for the γ-phosphate of ATP. Additionally, SmsC contains a partially folded C-terminal extension comprising a two-cysteine motif (-C-X-X-C-) found in some [Fe-S] cluster-containing proteins of the CIA pathway [22]. The two SmsB subunits tightly interact through an extended dimeric interface, which together with the two SmsB-SmsC contact interfaces stabilizes the hetero-tetrameric assembly [23]. The interface between the SmsC subunits is weaker (the buried surface area is 730 Å2), and likely susceptible to changes upon ATP binding and hydrolysis [24]. Such a conformational change is supported by the structure of the SmsC2 bound to an AMP-PN, which exhibits a “close” conformational state, similar to that observed in many structures of ABC transporters. This “close” conformation is required for the hydrolysis of ATP as it brings essential catalytic residues of the ABC signature motif close to the nucleotide phosphates [20,21] (see below Fig 3, intermediate 3).
SmsC2B2 assembles first a [Fe-S] cluster (ATP not required), on one of the SmsC subunits (a, intermediate 1). Upon interaction, the [Fe-S] cluster is transferred from SmsC2B2 to the targeted apo-protein client. Upon dissociation of the cluster, the C-terminal extension of SmsC unfolds, destabilizing the asymmetric SmsC-SmsC dimer observed in the cryo-EM structure, and allowing the SmsC2B2 complex to open to adopt a conformation like the one observed in the crystal structure (a, intermediate 2). ATP binding triggers conformational changes in the P-loop (b). The SmsC subunits rearrange again to transiently form a tight dimer as seen in the SmsC2 structure favoring ATP hydrolysis (a, intermediate 3), after which ADP and Pi leave SmsC and SmsC is reset to engage into assembling a new [Fe-S] cluster (a, intermediate 1).
Cryo-EM analysis identifies the [4Fe-4S] cluster coordination site in one SmsC subunit
Since attempts to crystallize a [Fe-S] cluster-bound protein complex either in the absence or in the presence of AMP-PNP were unsuccessful, we used cryo-EM to determine its structure. The reconstituted SmsC2B2 complex prepared as described above for crystallization in anoxic conditions, was rapidly vitrified for cryo-EM single particle analysis, and the structure determined reached a global resolution of 2.64 Å. The protein complex prepared in these experimental conditions revealed that the C-terminal residues C218, C239, and C242 of one SmsC subunit coordinate a [4Fe-4S] cluster (Fig 2c). These three cysteine residues bind the cluster in a solvent-exposed hydrophobic pocket consisting of residues P40, L216, I223, Y235, and P247 (Fig 2d). Although the SmsC2B2 complex could in principle contain two such binding sites at the SmsC head-to-tail dimer interfaces, the two SmsC subunits form a strong asymmetric dimer with a single populated [Fe-S] cluster binding site (Fig 2e). All residues coordinating the cluster and forming the hydrophobic pocket are in a single SmsC chain. In contrast with the unoccupied SmsC subunit (apo-SmsC), the [Fe-S] cluster-bound SmsC displays a folded C-terminal α-helix composed of residues 227–241 and a short terminal loop (Fig 2f and 2g). This region comprises the two-cysteine motif (-C239-G240-K241-C242-) coordinating the [Fe-S] cluster, and P247 closing the hydrophobic pocket. Additionally, residues of the C-terminal region extend the dimeric interface between the apo-SmsC and [Fe-S] bound SmsC subunit, in the vicinity of the cluster binding site. This interface consists of a cation–π interaction between the side chains of the K128 in apo-SmsC and the Y243 in [Fe-S] bound SmsC, as well as the side chain-backbone intermolecular hydrogen bonds from the N131 in apo-SmsC and the K236 in [Fe-S] bound SmsC (Fig 2d). As in the crystal structure, the N-terminal of SmsB (residues 1–62) could not be unambiguously resolved in the cryo-EM map of the SmsC2B2 complex.
The C218, C239, and C242 residues of SmsC were tested for their role in coordinating the [4Fe-4S] cluster. The SmsCC218A, SmsCC239A, SmsCC242A, and SmsCC239-C242AB variants were constructed, purified, and submitted to chemical reconstitution assay (see Table 2). UV-visible spectroscopy of all single-cysteine variants revealed a decreased absorbance at 420 nm, consistent with impaired [Fe-S] cluster coordination (Fig 4a). Despite this reduction, these variants retained high level of iron and sulfur content (Table 2). The double variant SmsC(C239-C242A)2B2 also exhibited a much-diminished absorbance at 420 nm (Fig 4b) while quantitative analysis of metal content indicated that this variant retained only 0.23 ± 0.2 nmol of Fe and 0.3 ± 0.1 nmol of S per nmol of protein complex (Table 2). Together, these results demonstrated that C218, C239, and C242 residues are critical ligands for coordinating a [4Fe-4S] cluster under the tested conditions.
(a) SmsC (84 μM) (black), SmsCC239A (84 μM) (green), SmsCC242A (84 μM) (red), and SmsCC218A (56 μM) (blue) were incubated with 5 equivalents of Fe2+/SmsC, 5 equivalents of Na2S/SmsC, and 3 mM DTT. UV-Vis absorption spectrum of SmsC shows an absorption at 420 nm compared to the variants SmsCC218A, SmsCC239A, and SmsCC242A. (b) SmsC(C239A-C242A)2B2 (50 μM) was incubated with 5 equivalents of Fe2+/SmsC(C239A-C242A)2B2, 5 equivalents of Na2S/SmsC(C239A-C242A)2B2, and 2 mM DTT. UV-Vis absorption spectrum of the SmsC(C239A-C242A)2B2variant shows no absorbance at 420 nm (1 nmol of SmsC(C239A-C242A)2B2 contains 0.23 nmol of iron and 0.3 nmol of sulfur). The data underlying this figure can be found in Fig 4 and S2 Data.
SmsC is an ATPase
SmsC harbors Walker A/B signatures of ABC ATPases. To obtain further functional insights, we performed a thorough enzymological analysis of the influence of ATP on SMS. Whereas SmsC exhibited ATPase activity with a kcat of 0.98 min−1 a variant impaired in the ATP binding site (Walker A motif), SmsCK45R, displayed a weak ATPase activity with a kcat of 0.13 min−1 (Fig 5a). The complex SmsC2B2 displayed an ATPase activity approximately 10-fold higher than SmsC alone (10 min−1), indicating that SmsB stimulated the ATPase activity of SmsC (Fig 5a). Moreover, ligand binding experiments using mantATPyS showed a Kd approximately three times higher for the variant SmsCK45R (8.76 ± 2 nM) than for the wild-type SmsC (3.2 ± 0.2 nM) (Fig 5b). Incubation of SmsC with excess of adenosine-5′-[(β,γ)-methyleno]triphosphate (AMP-PCP) led to SmsC dimer formation (Fig 5c). The SmsCK45R variant was not able to form homodimers upon AMP-PCP binding (Fig 5d), consistent with the notion that binding of ATP is required for dimerization of SmsC.
(a) SmsC (35 µM), SmsCK45R (35 µM), and SmsC2B2 (35 µM) proteins were added to 1 ml of 25 mM Hepes buffer (pH 7.6) containing 100 mM KCl, 5 mM MgSO4, 5 mM phospho-enol pyruvate, 1 mM NADH, 5 UI of PK, and 10 UI of LDH. Then, 1 mM of ATP was added to initiate the reaction at 25 °C. Specific activities were calculated using the molar extinction coefficient of 6.22 mM−1 cm−1 for NADH and the protein concentrations determined from the extinction coefficient. SmsC2B2 displays an ATPase activity 10-fold higher than SmsC. SmsCK45R displays a residual ATPase activity. (b) Equilibirum binding curve. SmsC proteins from 0 to 40 µM and SmsCK45R proteins from 0 to 25 µM were incubated with 400 nM concentration of mantATPyS in HEPES buffer 25 mM, KCl 100 mM, MgSO4 5 mM, pH 7.6. Measurements were taken using a fluorimeter (λexc 355 nm and λem 448 nm for mantATPyS). Dissociation constant was calculated using the equation y = m1 * x/(m2 + x). (c) SmsC proteins first incubated with 10 mM AMP-PCP (blue) or not (red) were purified by SEC. Elution peaks of SmsC, SmsC2, AMP-PCP Sms are indicated on top of the graph. (d) SmsC (blue) and SmsCK45R (red) proteins were incubated with 10 mM of AMP-PCP and purified by SEC. Elution peaks of SmsC, SmsC2, AMP-PCP Sms are indicated on top of the graph. The data underlying this figure can be found in Fig 5 and S3 Data.
ATP and [Fe-S] cluster binding are mutually exclusive
Next, we explored whether [Fe-S] cluster assembly and ATP binding were connected. For this, we asked if we could reconstitute a [Fe-S] cluster onto AMP-PCP-SmsC2B2 bound protein. All our attempts failed, suggesting that the presence of AMP-PCP prevented the binding of the [Fe-S] cluster (Fig 6a). Subsequent quantitation revealed that iron and sulfur content per mol of complex are 0.21 ± 0.3 Fe and 0.29 ± 0.12 S per AMP-PCP-SmsC2B2 complex. Conversely, reconstituted [Fe-S]-SmsC2B2 was incubated with mantATPyS. No increase in fluorescence intensity was observed, indicating that mantATPyS could not bind [Fe-S]-SmsC2B2 (Fig 6b). Last, UV-visible spectroscopy data showed that adding ATP failed to alter [Fe-S]-bound SmsC2B2 stability under oxic conditions, probably because ATP could not bind on [Fe-S]-SmsC2B2. Altogether these results showed that binding of ATP and [Fe-S] clusters are mutually exclusive. Interestingly, no difference in the capacity of SmsC2B2 to transfer a [4Fe-4S] cluster on apo-aconitase in vitro was observed whether ATP was present or not (Fig 6c). The SmsCK45R variant altered in the Walker A box showed impaired ATPase activity, yet retained the capacity to bind a [4Fe-4S] cluster (S3 Fig). Altogether these data revealed an unexpected connection between ATP binding and [Fe-S] cluster coordination, pointing to a coupling in the interaction of SmsC with both of its substrates.
(a) SmsC2B2 (84 μM) (red) and AMP-PCP-SmsC2B2 (84 μM) (blue) were incubated with 5 equivalents of Fe2+/ SmsC2B2, 5 equivalents of Na2S/ SmsC2B2, and 3 mM DTT. UV-Vis absorption spectrum of SmsC2B2 shows an absorption at 420 nm compared to the AMP-PCP-SmsC2B2 bound form. (b) Relative fluorescence measurements. SmsC2B2 (50 μM) (red) and [Fe-S]-SmsC2B2 (50 μM) (blue) were incubated with 21 μM of mantATPyS. Measurements were taken using a fluorimeter (λexc 355 nm and λem 448 nm). (c) Apo-AcnB (0.5 nmol) was incubated with reconstituted SmsC2B2 (1.74 eq.; 2.9 Fe/3.0 S per SmsC2B2) ± 1 mM ATP and 2 mM MgCl₂. Aconitase activity was measured after 1, 2, and 10 min to assess the effect of ATP on [Fe-S] cluster transfer. As a positive control (RecACnB), apo-AcnB was assayed after 30 min with 5 molar excess of Fe2+ and S2− in the presence of 500 µM DTT. The initial velocity (μM isocitrate/min) was measured in duplicate, short bars correspond to mean deviation. The data underlying this figure can be found in Fig 6 and S4 Data.
A mineral sulfur source mediates cluster formation by SmsCB in vivo
No cysteine desulfurase encoding genes are present in the SMS encoding operon or anywhere else in the genome of M. jannaschii, raising the question of the source of sulfur for feeding the SMS system. Early work has proposed that H2S could act as a sulfur source for [Fe-S] proteins in Methanococcus maripaludis [25]. To investigate whether H2S could be a bona fide sulfur source for M. jannaschii SMS system as well in vivo, we tested the capacity of the M. jannaschii smsCB operon to complement an E. coli FBE605 (Table 3) recipient strain lacking both functional ISC and SUF machineries. FBE605 strain has ∆iscUA∆sufABCDSE deletions, which renders it non-viable as it cannot maturate the isoprenoid synthesizing [Fe-S] containing IspG and IspH proteins. Previously, FBE605 strain was made viable by bringing an ectopically version of eucaryotic genes encoding an arabinose-controlled, mevalonate (MVA)-dependent [Fe-S] independent isoprenoid synthesizing pathway [26,27]. Thus, the resulting FBE605 strain can grow if both arabinose and MVA are present in the medium or if a [Fe-S] producing system is provided. We observed that the M. jannaschii smsCB operon was indeed able to rescue viability of the ∆iscUA∆sufABCDSE strain, but only if Na2S was added to the medium and the strains incubated under anoxic conditions (Fig 7a and 7b). Importantly, the M. jannaschii smsCK45RsmsB operon, defective in ATPase activity, was unable to complement the E. coli ∆iscUA∆sufABCDSE strain. It is worth mentioning that substituting Na2S for l-cysteine did not enable complementation, either under aerobic or anaerobic conditions (S4 Fig). Altogether, these results demonstrated that SMS acts as a [Fe-S] biogenesis system in vivo and strongly support the view that it uses sulfide as a source of sulfur.
Spot test assay for growth indifferent culture dilutions of E. coli. ΔiscUAΔsuf MEV carrying the empty pBAD vector (lane 1), the pBAD vector carrying the Escherichia coli sufABCSDE operon (lane 2), the pBAD vector carrying the Methanocaldococcus jannaschii smsCB operon (lane 3), and the pBAD vector carrying the M. jannaschii smsCK45RB operon (lane 4). Medium was LB supplemented with 0.2% arabinose (a) or 0.2% arabinose and 5 mM Na2S (b), in oxic or anoxic conditions.
Diversity and [Fe-S] cluster binding capacities of SmsC C-terminal region
Based upon the structural analysis of SmsC described above, we defined a C-terminal region (referred to as CTR below) ranging from residue D212 to the COOH terminus, including the beta-beta-alpha region. A multiple sequence alignment of archaeal SmsC sequences showed that the beta-beta-alpha region is relatively conserved from D212 to G240 residues (Fig 8a) whereas the extreme C-terminal region is highly variable (Fig 8a). About 75% of archaeal SmsC sequences (292 out of 415) were found to harbor the C-X-X-C motif (see above). Interestingly, Cys residues were found to be enriched in the CTR (Fig 8b) but vary in number (from 0 to 5), position, and spacing. Alphafold predictions of the structures of the Sms CTRs harboring different arrangements of Cys residues showed they all harbor several Cys residues close to each other in the structural models, being compatible with a [Fe-S] binding cluster capacity (Fig 9). This analysis revealed that, despite a large variability in sequence, the capacity to bind a [Fe-S] cluster is likely to be a conserved feature of SmsC proteins.
(a) Multiple sequence alignment of SmsC sequences showing that the beta-beta-alpha region is well conserved from D212 to G240 residues. Number of cysteine residues per site in the multiple alignement of 1,387 SmsC. (b) Subsample of SmsC CTR showing the diversity in terms of cysteine arrangements.
Predictions harbor several Cys residues close to each other in the structural models, being compatible with a [Fe-S] binding cluster capacity. Alphafold prediction of the SmsCB proteins of Clostridium tetani (Ctetani), Clostridium novyi (Cnovyi), Methanopyrus kandleri (Mkandleri), Archaeoglobus fulgidus (Afulgidus), Methanococcus voltae (Mvoltae), Dehalobacter sp (Dsp), and Syntrophus aciditrophicus (Saciditrophic). Cysteine residues are shown in red.
Conclusions
Previously we identified two minimal [Fe-S] biogenesis systems, MIS and SMS, which we proposed to be ancestors of modern ISC, NIF, and SUF. The MIS system had already received some attention, although under a different appellation [28,29]. In the present study, we focused on SMS from the archaeon M. jannaschi. SMS is a two-component system, proposed to be the ancestor of SUF, which comprises six components. SufSE acts as heterodimeric cysteine desulfurase. SufC, an ABC ATPase, which, with SufB and D paralogs, forms a heteromeric scaffold assembling [Fe-S] cluster. SufA, acts as a carrier delivering [Fe-S] clusters to cellular client targets. M. jannaschi SmsC and SmsB share 32% and 14% sequence identity with E. coli SufC and SufB, respectively, and SmsC, like SufC, harbors Walker A/B signatures of ABC ATPases. Thus, the question is to know how in the absence of a cysteine desulfurase (SufSE), of one of the scaffold component (SufD) and of a carrier (SufA), does SmsCB make and distribute [Fe-S] clusters.
Globally, the three-dimensional structure of the SmsC2B2 complex from M. jannaschii is similar to that of the SufBC2D complex from E. coli [30]. However, a major and unexpected finding based upon our structural, spectroscopic, and mutagenesis analyses is the demonstration that the [Fe-S] cluster binding site on SmsC2B2 complex locates in SmsC and in addition in only one SmsC subunit. This departs from the SufBC2D system wherein the [Fe-S] cluster is thought to bind at the SufBD interface [30,31], although no conclusive structural study has been produced to date supporting this claim. Despite several attempts, we obtained no evidence for [Fe-S] binding onto SmsB. Moreover, preventing SmsC to bind [Fe-S] cluster by mutagenesis was sufficient to hamper [Fe-S] binding by the whole SmsC2B2 complex. We found that the three Cys residues, C218, C239, and C242 coordinate the [4Fe-4S] cluster. Given the solvent-exposed location of the assembly site, a solvent molecule might contribute to additional liganding. The C239 and C242 residues are part of the C-terminal unstructured and flexible CTR, which is enriched in C-X-X-C motifs, in all SmsC sequences examined. Interestingly, CTR sequences show a wide diversity, yet we predict that CTR are very likely to enable cognate SmsC to bind [Fe-S] clusters. This is another striking difference from SufC, which has no such flexible CTR. Another specificity of SmsC2B2 lies in the fact that it assembles a single [4Fe-4S] cluster per complex. This departs from SufBC2D, NifU, or IscU, which assemble a [2Fe-2S] or a [3Fe-3S] cluster [30–35]. The reason why orthodox scaffolds bind [2Fe-2S] clusters lies in their capacity to cooperate with A-type carriers [33]. These carriers receive two [2Fe-2S] clusters from the scaffold and convert them to [4Fe-4S] clusters which they eventually transfer to apo-targets. E. coli A-type carriers include IscA, SufA, ErpA, and NfuA proteins. Analysis of the genome of M. jannaschii failed to identify such A-type carriers. Hence, it is possible that SmsC2B2 directly transfers the [4Fe-4S] cluster to its target, independently of a carrier. Alternatively, SmsC2B2 might team up with Mrp, a putative scaffold/carrier whose role remains ill-defined and for which a cognate structural gene was identified in M. jannaschii genome [14]. Another issue concerns the maturation of [2Fe-2S] clusters proteins if such species arise at all in M. jannaschii. This is a fully open question as only [4Fe-4S] proteins have been characterized in this organism or other related archaea and it is impossible to predict from genome analysis whether a given protein harbors a [2Fe-2S] or [4Fe-4S] cluster.
Our biochemical and structural data indicated that [Fe-S] cluster binding to one SmsC subunit induced a strong asymmetry within its dimeric interface by folding the C-terminal helical region. Thus, we propose that SmsC2B2 assembles first a [Fe-S] cluster, on one of the SmsC subunits (Fig 3a, intermediate 1). Upon interaction the [Fe-S] cluster is transferred from SmsC2B2 to the targeted apo-protein client. Upon dissociation of the cluster, the C-terminal extension of SmsC unfolds, destabilizing the asymmetric SmsC-SmsC dimer observed in the cryo-EM structure, and allowing the SmsC2B2 complex to open up to adopt a conformation like the one observed in the crystal structure (Fig 3a, intermediate 2). ATP binding to this conformer triggers conformational changes in the P-loop (Fig 3b). The SmsC subunits rearrange again to transiently form a tight dimer as seen in the SmsC2 structure favoring ATP hydrolysis (Fig 3a, intermediate 3), after which ADP and Pi leave SmsC and SmsC is reset to engage into assembling a new [Fe-S] cluster (Fig 3a, intermediate 1).
The present study uncovered unexpected functional coupling between the binding of ATP and the assembly of [Fe-S] cluster in SmsC. AMP-PCP-bound SmsC2B2 was unable to accommodate the [Fe-S] cluster and mantATPyS could not bind [Fe-S]-SmsC2B2. This finding provided evidence that the binding of ATP and the [Fe-S] cluster to SmsC2B2 are mutually exclusive. Interestingly, mutual exclusion between [Fe-S] and the ATP binding sites might be due to their close vicinity in the structure. The physiological value of alternated bindings of ATP and Fe-S clusters might be to serve as a homeostatic control. [Fe-S] proteins being key in energetic producing pathways, one might surmise that when enough [Fe-S] proteins are functioning, ATP level is high and SMS activity ought to be reduced. On the contrary, a reduced level of ATP, reflecting reduced activities of [Fe-S] proteins, will cease to antagonize [Fe-S] assembly on SmsCB, enhanced [Fe-S] biogenesis will ensue and ATP level will be replenished.
A major difference of SMS with respect to the four well-studied [Fe-S] cluster biogenesis machineries, including the other ancestor minimal MIS system, is that there is no identified cysteine desulfurase component in M. jannaschii genome. This raises the question of the source of sulfur. Here, we showed that adding Na2S in culture media allows SMS to carry out [Fe-S] biogenesis in an E. coli strain lacking its own biogenesis systems, ISC and SUF, indicating that SMS can use sulfide for building [Fe-S] clusters. How E. coli integrates exogenously provided sulfide into its sulfur metabolism and in particular how sulfide is taken up remain unclear. Previous work by the Imlay lab [36] argued that sulfide freely equilibrates across membranes. Also, a recent transciptomic analysis [37] highlighted the role of sulfide sensing ygaV transcriptional regulator in E. coli grown under oxygen-limiting and/or sulfide-enriched conditions. Expression of a large number of genes was found modified in the presence of sulfide. Thus, an exciting possibility is that some of these genes might contribute to the SmsCB-mediated complementation reported here and this is being investigated in our lab. This result is consistent with the early proposal that SMS-containing archaea growing in H2S-rich environments might build [Fe-S] clusters using exogenous sulfide as a sulfur source [25,38]. Interestingly, we observed complementation only under anaerobiosis, whereas the SUF system is able to mature proteins under both aerobiosis and anaerobiosis. This further supports the hypothesis that SMS operated in anaerobic prokaryotes and that SUF evolved from it as adaptation to an increasingly oxygenic atmosphere after the GOE.
Materials and methods
Bacterial strains and plasmid construction
The E. coli strains, plasmids, and oligonucleotides used in this study are listed in Table 3. Bacterial strains were routinely grown in aeration at 37 °C in Luria–Bertani broth (LB). When required, ampicillin was added at 100 μg ml−1. The nucleotide sequences were codon-optimized for expression in E. coli and genes have been ordered from Twist Bioscience (TWB) (Table 4).
For purification experiments, the smsC and smsB genes of M. jannaschii were cloned in the pET6His-Tev vector [39,40]. These constructions enabled the production of N-terminal 6His-tagged SmsC and N-terminal 6His-tagged SmsB proteins, respectively. For complementation experiments, the smsCB genes of M. jannaschii were cloned in the pBAD [41]. Finally, all the plasmids were sequenced using Eurofins oligonucleotides T7 and T7term for pET vectors or pBAD-FP and pTrcHis-RP oligonucleotides for pBAD vectors.
Site-directed mutagenesis by PCR
Site-directed mutagenesis was performed by PCR with the Stratagene pfuUltra enzyme, the plasmid to be mutated as atemplate, and two complementary oligonucleotides containing the mutation. The amplification products were digested with 10 units of DpnI enzyme before being used for transformation into NEB5α.
In vivo complementation assay
The strain MG1655 ∆iscUA∆sufABCDSE (FBE605) is a derivative of E. coli MG1655 in which iscUA and sufABCDSE genes containing regions are deleted and which contains eucaryotic genes encoding MVA-dependent isoprenoid synthesis pathway [26,27,42]. This strain is viable on LB supplemented with 0.2% arabinose and 0.5 mM MVA (mevalonolactone, Sigma-Aldrich). FBE605 was sequentially transformed with pBAD plasmids derivatives (Table 3). Transformants were selected aerobically at 37 °C on LB agar in the presence of MVA and arabinose. Cells were pelleted, washed, resuspended in LB, and used as a normalization of OD600 = 1. Serial dilutions in 1 ml final volume were performed until 10−6, and 2 μl drops were plated on LB plates supplemented or not with 5 mM Na2S. For anaerobic growth, plates were placed in a jar with AnaeroGen bags from ThermoScientific.
Expression and purification of SmsB and SmsC from M. jannaschii
SmsB and SmsC proteins were produced in strain MG1655(DE3) (FBE682) growing under oxic conditions in LB medium. Expression was carried out in MG1655(DE3) at 30 °C until OD600 reached 0.8–0.9, and induced at 28 °C with 0.1 mM IPTG during 3 h. Pellets were stored at −20 °C. Harvested cells were resuspended in washing buffer (25 mM Tris-HCl pH 7.8, 100 mM NaCl, 25 mM imidazole). Cells were lysed by Cell Disruption System at 25,000 Psi and the lysate was recovered with MgCl2 5 mM and benzonase at 6,240 UI. Then, the lysate was cleared by centrifugation (12.000g, 20 min, 4 °C). Supernatants were loaded onto a 5 ml HisTrap HP (Cytiva). The column was washed using buffer (25 mM Tris-HCl pH 7.8, 100 mM NaCl, 25 mM imidazole) and proteins were eluted with washing buffer containing 300 mM imidazole. Fractions containing SmsC and/or SmsB proteins were pooled and dialyzed overnight in 50 mM Tris-HCl pH 7.8, 100 mM NaCl, 1 mM dithiothreitol (DTT) buffer. Treatment with recombinant TEV protease was also added to remove the N-terminal His-tag. The dialysate was recovered and loaded onto a His-binding (Ni-NTA-Agarose) pre-equilibrated with 50 mM Tris-HCl pH 7.8 and NaCl 10 mM. The eluate was concentrated using PES column 10,000 MWCO. This was injected on a calibrated Hiload 600 Superdex 200 column (50 mM Tris-HCl pH 7.8, 100 mM NaCl, 1 mM DTT). SmsB was mainly homodimeric in solution (71,200 kDa) and SmsC monomeric (28,600 kDa). Then, SmsB2 et SmsC proteins were mixed with molar amounts corresponding to SmsC2B2 and incubated overnight, before injecting again on a Hiload 600 Superdex 200 column (50 mM Tris-HCl pH 7.8, 100 mM NaCl, 1 mM DTT). From this size exclusion chromatography, resulted different peaks, corresponding to the reformed SmsC2B2 complex (128,400 Da) and SmsC (28,600 Da) that could be in excess or have been detached from the complex itself. Desired fractions were combined, concentrated, aliquoted, and checked for purity by SDS–PAGE. Each time, analysis of the size exclusion chromatography fractions by SDS-PAGE revealed the protein purity. Protein concentration was determined via UV-visible spectroscopic analysis with Jasco V-730 spectrophotometer, by using ε280 = 0.714 M−1 cm−1 and ε280 = 0.335 M−1 cm−1 for SmsC and SmsB, respectively. The purification yield was approximately 13 mg/L for SmsC2 and 8 mg/mL for SmsB. The UV-visible absorbance spectrum of protein purified under oxic conditions presented no characteristic absorption bands indicative of a metal-bound protein.
[Fe-S] cluster reconstitution of SmsCB proteins
All steps were performed under anoxic conditions inside a Jacomex glovebox (<1 ppm oxygen). SmsCB proteins were pretreated under anoxic conditions with 3 mM DTT or TCEP in Tris-HCl 50 mM pH 7.8, NaCl 10 mM. Then, 5 molar excess of ammonium iron(II) sulfate hexahydrate and sodium sulfide were added per nmol of SmsC2B2. Unbound iron and sulfide were removed by passage on a NAP-25 column (Cytiva). UV-visible spectra (250–750 nm) were recorded on an Avantes Havalight DHS spectrophotometer connected to the glove box by optical fibers to monitor cluster formation. Colorimetric assays were used to measure the iron [43] and sulfur [44] content after reconstitution.
For [Fe-S] cluster reconstitution on SmsC variants, the same procedure was performed on SmsC proteins obtained from recombinant His-tagged SmsC (produced from pVP428, pVP430, pVP487 plasmids) and SmsCB (produced from pVP367 plasmid) proteins in E. coli.
A similar procedure was performed for the 57FeS reconstitution of the SmsC (12.6 mg) and SmsC2B2 (9 mg), for Mössbauer spectroscopic studies. To reach 467 µM SmsC2 and 350 µM SmsC2B2, we used concentration Amicon Ultra device. 4 mM of DTT was used in the pretreatment and 4 equivalents of 95% 57Fe-enriched Mohr’s salt ((NH4)2Fe(SO4)2·6H2O) and Na2S were used per SmsC2B2.
The incubation lasted 3 h. Unbound iron and sulfide were removed by passage on a NAP-25 column (Cytiva). A Mössbauer cup in Delrin was immediately filled and frozen in cooled 2-methyl-butane inside the glove box and kept in liquid nitrogen until the measurement. The recorded UV-visible spectrum is shown in S1 Fig for the complex SmsC2B2 and was similar to that displayed in Fig 1D for SmsC.
ATPase activity assays
The ATPase activity of purified proteins was measured by an enzyme-coupled spectrophotometric assay [45]. The coupling between ATP hydrolysis and NADH oxidation was achieved by excess of pyruvate kinase (PK) and lactate dehydrogenase (LDH) and by monitoring the decrease of NADH at 340 nm using a Jasco V-730 spectrophotometer. SmsC samples (n = 3) were added to 1 ml of 25 mM Hepes buffer (pH 7.6) containing 100 mM KCl, 5 mM MgSO4, 5 mM phospho-enol pyruvate, 1 mM NADH, 5 UI of PK, and 10 UI of LDH. Then, 1 mM of ATP was added to initiate the reaction at 25 °C. Specific activities were calculated using the molar extinction coefficient of 6.22 mM−1 cm−1 for NADH and the SmsC protein concentration determined from its extinction coefficient as previously described. For Kd measurement, a concentration range of SmsC proteins from 0 to 40 µM and SmsCK45R proteins from 0 to 25 µM were incubated with a fixed 400 nM concentration of mantATPyS (2′/3′-O-(N-Methyl-anthraniloyl)-adenosine-5′-(γ-thio)-triphosphate, Triethylammonium salt from Jena Bioscience λmax 255/355 nm, ɛ 23.3/5.8 L mmol−1 cm−1 (Tris-HCl pH 7.5), λexc 355 nm, λem 448 nm) [46] in HEPES buffer 25 mM, KCl 100 mM, MgSO4 5 mM, pH 7.6. Then, λexc 355 nm and λem 448 nm were used for mantATPyS fluorescence measurements.
Then dissociation constant was calculated using the following equation [y = m1 * x/(m2 + x)], with y being the observed fluorescence signal, x the concentration of free ligand, m1 the maximum binding signal, and m2 the dissociation constant (Kd) [46]. The same λexc 355 nm and λem 448 nm parameters were used for relative fluorescence intensity measurements, with the exception of a scan from 400 to 600 nm to obtain the spectrum of relative fluorescence intensity. For this experiment, a fixed concentration of 50 µM of proteins and 21 µM of mantATPyS were used.
Mössbauer spectroscopy
Mössbauer spectra were recorded at 5.8 K on a strong-field Mössbauer spectrometer equipped with an Oxford Instruments Spectromag 4000 cryostat containing an 8 T split-pair superconducting magnet. The spectrometer was operated in a constant acceleration mode in transmission geometry. Velocity and absorption values were obtained after the classical folding procedure of the crude recorded data (channel number and counts of photons per channel). The error is the square root value of the counts expressed in percentage versus the counts associated with the baseline. The isomer shift is referenced against that of a metallic iron foil at room temperature. The spectra were analyzed with a home-made program and simultaneously simulated [47,48]. Similar simulated spectra can be obtained using the WMOSS Mössbauer Spectral Analysis Software (www.wmoss.org, 2012–2013, Web Research, Edina).
In vitro [Fe-S] transfer to aconitase
To assess aconitase activation [49], apo-AcnB from E. coli (0.5 nmol) was pretreated with DTT. Then, apo-AcnB was desalted using a Microbiospin column (Biorad) and incubated under anoxic conditions in a glove box at 18 °C in aconitase buffer (50 mM Tris-HCl pH 7.6) with 1.74 equivalents of the reconstituted SmsC2B2 (2.9 Fe and 3.0 S/SmsC2B2) or 1.70 equivalents reconstituted SmsC2(K45R)B2 (3.0 Fe and 3.1 S/SmsC2(K45R)B2) in order to provide 5 Fe and 5 S atoms per AcnB monomer. After 30 min of incubation, aconitase activity was assessed as described [50]. Briefly, AcnB, SmsC2B2, and SmsC2(K45R)B2 mixtures were added to 0.6 mM MnCl2, 25 mM citrate, 0.5 U isocitric dehydrogenase, 0.25 mM NADP+, 50 mM Tris-HCl pH 7.6, in a 100 μl final volume and NADPH formation was monitored at 340 nm by UV-visible absorption spectroscopy. To assess the effect of ATP during [Fe-S] transfer, pre-reduced apo-AcnB (0.5 nmol) was incubated with 1.74 equivalents of reconstituted SmsC2B2 (2.9 Fe and 3.0 S/ SmsC2B2) in the presence or absence of 1 mM ATP and 2 mM MgCl2. After 1, 2, or 10 min incubation the aconitase activity was measured. The AcnB activity corresponds to the initial velocity of isocitrate production (µM/min). The positive control corresponds to the activity of the chemically reconstituted AcnB (RecAcnB) prepared by incubating apo-AcnB with 5 molar excess of ferrous iron and sulfur for 30 min in the presence of 500 µM DTT (Activity of 76 ± 2 µM/min). The experiment was performed in duplicate.
X-ray crystallography: Crystallization, data collection, and structure determination
Screenings of crystallization conditions were performed in sitting-drop 96-well Greiner plates at the Crystallography Core Facility of the Institut Pasteur [51]. Crystallization hits were optimized in 24-well plates using the hanging drop method. Colorless monoclinic crystals (space group P21) of the SmsB2C2 complex grew in wells containing 0.1 M ammonium sulfate, 0.3 M sodium formate, 0.1 M sodium acetate, 3% w/v PGA (Na+ form, LM) and 3% w/v PEG20000 at pH 5.0 in the reservoir under anoxic conditions. Orthorhombic crystals (space group C2221) of SmsB2C2 were obtained in 5% w/v PEG3350, 10% w/v Tacsimate under oxic conditions and soaked in the reservoir solution supplemented with AMP-PNP for 10 min before freezing in liquid nitrogen for diffraction data collection. Crystals of SmsC2 in complex with AMP-PN grew in wells containing 0.1 M NaCl, 0.1 M Bicine pH 9, 30% w/v PEG MME 2K.
X-ray diffraction data were collected at beamlines PROXIMA 1 and PROXIMA 2A (Synchrotron SOLEIL, St. Aubin, France) and processed with autoPROC [52]. The crystal structures of the SmsB2C2 and SmsC were solved by the molecular replacement method with Phaser [53], using trimmed AlphaFold3 [54] models as search probe. The final models were obtained through interactive cycles of manual model building with Coot [55] and reciprocal space refinement with Buster [56].
Atomic coordinates and structure factors have been deposited in the RCSB Protein Data Bank under the accession codes 9H7Y, 9HBL, and 9H7X.
Cryo-EM sample preparation
SmsC2B2 samples prepared and sealed under anoxic conditions were quickly vitrified for cryo-EM experiments on previously glow-discharged Quantifoil R0.6/1 Cu-mesh 300 grids, using a Vitrobot Mk. IV (Thermo Fisher Scientific), blotting for 4 seconds at 100% humidity and 22 °C before plunge freezing.
Cryo-EM data acquisition, image processing, and model building.
Data acquisition was carried out in a Titan Krios electron microscope equipped with a Falcon 4i direct electron detector with a Selectris X energy filter (Thermo Fisher Scientific). Two datasets were acquired: 10,883 movies with no stage tilt and 3,565 movies with 30° stage tilt. Patch motion correction and CTF estimation were carried out in CryoSPARC v4.5.1 [57], as well as the downstream image processing. Data acquisition parameters and model-building statistics are summarized in Table 5, and cryo-EM data processing workflow is depicted in S5 Fig.
Our crystal structure of SmsC2B2 was used as an initial model for model building of the [Fe-S] cluster-loaded SmsC2B2 cryo-EM structure. Model building and refinement were carried out in Coot v0.9.8 [55] and Phenix v1.20.1-4487-000 [58]. Structure model and map have been deposited in the RCSB Protein Data Bank and the EMDB under the accession code 9H78 and EMD-51913, respectively.
Supporting information
S1 Fig. Mössbauer analysis on SmsC2B2 bound to [Fe-S].
(a) UV-visible spectrum of SmsC2B2 reconstituted for Mössbauer analysis. (b) 6 K Mössbauer spectra (black vertical bars) of SmsC2B2 (350 µM, 3.6 Fe and 3.4 S/ SmsC2B2) recorded using a 0.06 T (A) and a 7 T (B) external magnetic field applied parallel to the γ-beam. The simulations assuming two iron sites in a 3:1 ratio are overlaid as thick red solid lines and the major and minor contributions are displayed above as blue and mauve thin solid lines, respectively. The nuclear parameters are: major component (75%): δ1 = 0.42 ± 0.01 mm s−1, ∆EQ,1 = 1.23 ± 0.05 mm s−1 and η1 = 0.8 ± 0.1; minor component (25%): δ2 = 0.40 ± 0.01 mm s−1, ∆EQ,2 = 0.83 ± 0.05 mm s−1 and η2 = 0.2 ± 0.1. The data underlying this figure can be found in S1 Fig and S5 Data.
https://doi.org/10.1371/journal.pbio.3003223.s001
(TIFF)
S2 Fig. Mössbauer analysis on SmsC bound to [Fe-S].
(a) UV-visible spectrum of SmsC reconstituted for Mössbauer analysis. (b) 6 K Mössbauer spectra (black vertical bars) of SmsC (935 µM, 1.7 Fe and 1.6 S/ SmsC) recorded using a 0.06 T (A) and a 7 T (B) external magnetic field applied parallel to the γ-beam. The blue solid line was calculated assuming a diamagnetic Fe site that accounts for 90 ± 5% of the total iron content. The nuclear parameters are: isomer shift = 0.45 ± 0.01 mm s−1, quadrupole splitting ΔEQ = 1.11 ± 0.05 mm s−1 and EFG rhombicity = 0.6 ± 0.1. They are strongly reminiscent of those determined for cysteine-coordinated [4Fe-4S]2+ clusters. The remaining area (≈10% of the iron content) may correspond to high-spin FeII impurities. The data underlying this figure can be found in S2 Fig and S6 Data.
https://doi.org/10.1371/journal.pbio.3003223.s002
(TIFF)
S3 Fig. SmsCK45R binds [Fe-S] cluster.
UV-Vis absorption spectrum of SmsCK45R. SmsCK45R (42 μM) was incubated with 5 equivalents of Fe2+/SmsC(K45R)2, 5 equivalents of Na2S/SmsC(K45R)2 and 3 mM DTT. The data underlying this figure can be found in S3 Fig and S7 Data.
https://doi.org/10.1371/journal.pbio.3003223.s003
(TIFF)
S4 Fig. L-Cysteine is not a source of sulfur for the SmsCB complex.
Spot test assay for growth indifferent culture dilutions of Escherichia coli. ΔiscUAΔsuf MEV carrying the empty pBAD vector (lane 1), the pBAD vector carrying the E. coli sufABCSDE operon (lane 2), and the pBAD vector carrying the Methanocaldococcus jannaschii smsCB operon (lane 3). Medium was LB supplemented with 0.2% arabinose and different concentrations of l-Cysteine (0.1, 0.5, and 1 mM) as indicated, in oxic or anoxic conditions.
https://doi.org/10.1371/journal.pbio.3003223.s004
(TIFF)
S1 File. Raw Image.
SDS-PAGE as analyzed prior to annotation (see legend Fig 1a for details).
https://doi.org/10.1371/journal.pbio.3003223.s006
(PDF)
S1 Data. Raw Data corresponding to Fig. 1a-1d. Spectroscopy analysis of SmsC2B2 complex.
https://doi.org/10.1371/journal.pbio.3003223.s007
(ZIP)
S2 Data. Raw Data corresponding to Fig. 4ab. Residues C219, C239, and C242 of SmsC act as ligands of the [Fe-S] cluster.
https://doi.org/10.1371/journal.pbio.3003223.s008
(ZIP)
S3 Data. Raw Data corresponding to Fig. 5a-5d. Role of ATP binding/hydrolysis in SmsCB.
https://doi.org/10.1371/journal.pbio.3003223.s009
(ZIP)
S4 Data. Raw Data corresponding to Fig. 6a-6c.
ATP and [Fe-S] cluster binding are mutually exclusive.
https://doi.org/10.1371/journal.pbio.3003223.s010
(ZIP)
S5 Data. Raw Data corresponding to S1 Fig ab. Mössbauer analysis on SmsC2B2 bound to [Fe-S].
https://doi.org/10.1371/journal.pbio.3003223.s011
(XLSX)
S6 Data. Raw Data corresponding to S2 Fig ab. Mössbauer analysis on SmsC bound to [Fe-S].
https://doi.org/10.1371/journal.pbio.3003223.s012
(XLSX)
S7 Data. Raw Data corresponding to S3 Fig. SmsCK45R binds [Fe-S] cluster.
https://doi.org/10.1371/journal.pbio.3003223.s013
(CSV)
Acknowledgments
We thank members of the SAMe and of the EMBC units for discussion. We also thank Marc Fontecave for lending us the anaerobic crystallography plateform at the Collège de France, and Ludovic Pecqueur for his help with part of the project.
References
- 1. Fontecave M. Iron–sulfur clusters: ever-expanding roles. Nat Chem Biol. 2006;2:171–4.
- 2. Wächtershäuser G. Groundworks for an evolutionary biochemistry: the iron-sulphur world. Prog Biophys Mol Biol. 1992;58(2):85–201. pmid:1509092
- 3. Imlay JA. Iron-sulphur clusters and the problem with oxygen. Mol Microbiol. 2006;59(4):1073–82. pmid:16430685
- 4. Boyd ES, Thomas KM, Dai Y, Boyd JM, Outten FW. Interplay between oxygen and Fe-S cluster biogenesis: insights from the Suf pathway. Biochemistry. 2014;53(37):5834–47. pmid:25153801
- 5. Spietz RL, Payne D, Szilagyi R, Boyd ES. Reductive biomining of pyrite by methanogens. Trends Microbiol. 2022;30(11):1072–83. pmid:35624031
- 6. Larson J, Tokmina-Lukaszewska M, Payne D, Spietz RL, Fausset H, Alam MG, et al. Impact of mineral and non-mineral sources of iron and sulfur on the metalloproteome of Methanosarcina barkeri. Appl Environ Microbiol. 2024;90(8):e0051624. pmid:39023267
- 7. Roche B, Aussel L, Ezraty B, Mandin P, Py B, Barras F. Reprint of: Iron/sulfur proteins biogenesis in prokaryotes: formation, regulation and diversity. Biochim Biophys Acta. 2013;1827(8–9):923–37. pmid:23660107
- 8. Braymer JJ, Freibert SA, Rakwalska-Bange M, Lill R. Mechanistic concepts of iron-sulfur protein biogenesis in Biology. Biochim Biophys Acta Mol Cell Res. 2021;1868(1):118863. pmid:33007329
- 9. Tsaousis AD. On the origin of iron/sulfur cluster biosynthesis in eukaryotes. Front Microbiol. 2019;10:2478. pmid:31781051
- 10. Py B, Barras F. Building Fe-S proteins: bacterial strategies. Nat Rev Microbiol. 2010;8(6):436–46. pmid:20467446
- 11. Esquilin-Lebron K, Dubrac S, Barras F, Boyd JM. Bacterial approaches for assembling iron-sulfur proteins. mBio. 2021;12(6):e0242521. pmid:34781750
- 12. Allen RM, Chatterjee R, Madden MS, Ludden PW, Shah VK. Biosynthesis of the iron-molybdenum cofactor of nitrogenase. Crit Rev Biotechnol. 1994;14(3):225–49. pmid:7954845
- 13. Dussouchaud M, Barras F, Ollagnier de Choudens S. Fe-S biogenesis by SMS and SUF pathways: a focus on the assembly step. Biochim Biophys Acta Mol Cell Res. 2024;1871(7):119772. pmid:38838856
- 14. Garcia PS, D’Angelo F, Ollagnier de Choudens S, Dussouchaud M, Bouveret E, Gribaldo S, et al. An early origin of iron-sulfur cluster biosynthesis machineries before Earth oxygenation. Nat Ecol Evol. 2022;6(10):1564–72. pmid:36109654
- 15. Boll M, Fuchs G, Meier C, Trautwein A, Lowe DJ. EPR and Mössbauer studies of benzoyl-CoA reductase. J Biol Chem. 2000;275:31857–68.
- 16. Ollagnier-de Choudens S, Sanakis Y, Hewitson KS, Roach P, Münck E, Fontecave M. Reductive cleavage of S-adenosylmethionine by biotin synthase from Escherichia coli. J Biol Chem. 2002;277(16):13449–54. pmid:11834738
- 17. Lee K-H, Saleh L, Anton BP, Madinger CL, Benner JS, Iwig DF, et al. Characterization of RimO, a new member of the methylthiotransferase subclass of the radical SAM superfamily. Biochemistry. 2009;48(42):10162–74. pmid:19736993
- 18. Dilg AW, Capozzi F, Mentler M, Iakovleva O, Luchinat C, Bertini I, et al. Comparison and characterization of the [Fe4S4]2+/3+ centre in the wild-type and C77S mutated HiPIPs from Chromatium vinosum monitored by Mössbauer, 57Fe ENDOR and EPR spectroscopies. J Biol Inorg Chem. 2001;6(3):232–46. pmid:11315559
- 19. Kent TA, Dreyer JL, Kennedy MC, Huynh BH, Emptage MH, Beinert H. Mössbauer studies of beef aconitase: evidence for facile interconversions of iron-sulfur clusters. Proc Natl Acad Sci U S A. 1982;79:1096–100.
- 20. Locher KP. Mechanistic diversity in ATP-binding cassette (ABC) transporters. Nat Struct Mol Biol. 2016;23(6):487–93. pmid:27273632
- 21. Zaitseva J, Oswald C, Jumpertz T, Jenewein S, Wiedenmann A, Holland IB, et al. A structural analysis of asymmetry required for catalytic activity of an ABC-ATPase domain dimer. EMBO J. 2006;25(14):3432–43. pmid:16858415
- 22. Stehling O, Jeoung J-H, Freibert SA, Paul VD, Bänfer S, Niggemeyer B, et al. Function and crystal structure of the dimeric P-loop ATPase CFD1 coordinating an exposed [4Fe-4S] cluster for transfer to apoproteins. Proc Natl Acad Sci U S A. 2018;115(39):E9085–94. pmid:30201724
- 23. Krissinel E, Henrick K. Inference of macromolecular assemblies from crystalline state. J Mol Biol. 2007;372(3):774–97. pmid:17681537
- 24. Davidson AL, Dassa E, Orelle C, Chen J. Structure, function, and evolution of bacterial ATP-binding cassette systems. Microbiol Mol Biol Rev. 2008;72(2):317–64, table of contents. pmid:18535149
- 25. Liu Y, Sieprawska-Lupa M, Whitman WB, White RH. Cysteine is not the sulfur source for iron-sulfur cluster and methionine biosynthesis in the methanogenic archaeon Methanococcus maripaludis. J Biol Chem. 2010;285(42):31923–9. pmid:20709756
- 26. Campos N, Rodríguez-Concepción M, Sauret-Güeto S, Gallego F, Lois LM, Boronat A. Escherichia coli engineered to synthesize isopentenyl diphosphate and dimethylallyl diphosphate from mevalonate: a novel system for the genetic analysis of the 2-C-methyl-d-erythritol 4-phosphate pathway for isoprenoid biosynthesis. Biochem J. 2001;353:59–67.
- 27. Loiseau L, Gerez C, Bekker M, Ollagnier-de Choudens S, Py B, Sanakis Y, et al. ErpA, an iron sulfur (Fe S) protein of the A-type essential for respiratory metabolism in Escherichia coli. Proc Natl Acad Sci U S A. 2007;104(34):13626–31. pmid:17698959
- 28. Tokumoto U, Kitamura S, Fukuyama K, Takahashi Y. Interchangeability and distinct properties of bacterial Fe-S cluster assembly systems: functional replacement of the isc and suf operons in Escherichia coli with the nifSU-like operon from Helicobacter pylori. J Biochem. 2004;136(2):199–209. pmid:15496591
- 29. Deere TM, Prakash D, Lessner FH, Duin EC, Lessner DJ. Methanosarcina acetivorans contains a functional ISC system for iron-sulfur cluster biogenesis. BMC Microbiol. 2020;20(1):323. pmid:33096982
- 30. Hirabayashi K, Yuda E, Tanaka N, Katayama S, Iwasaki K, Matsumoto T, et al. Functional dynamics revealed by the structure of the SufBCD complex, a novel ATP-binding Cassette (ABC) protein that serves as a scaffold for iron-sulfur cluster biogenesis. J Biol Chem. 2015;290(50):29717–31. pmid:26472926
- 31. Yuda E, Tanaka N, Fujishiro T, Yokoyama N, Hirabayashi K, Fukuyama K, et al. Mapping the key residues of SufB and SufD essential for biosynthesis of iron-sulfur clusters. Sci Rep. 2017;7(1):9387. pmid:28839209
- 32. Veronesi G, Pérard J, Clémancey M, Gerez C, Duverger Y, Kieffer I, et al. Multimodal spectroscopic analysis of the Fe-S clusters of the as-isolated Escherichia coli SufBC2D complex. Inorg Chem. 2024;63(19):8730–8. pmid:38687645
- 33. Lill R, Freibert S-A. Mechanisms of mitochondrial iron-sulfur protein biogenesis. Annu Rev Biochem. 2020;89:471–99. pmid:31935115
- 34. Chandramouli K, Johnson MK. HscA and HscB stimulate [2Fe-2S] cluster transfer from IscU to apoferredoxin in an ATP-dependent reaction. Biochemistry. 2006;45(37):11087–95. pmid:16964969
- 35. Schulz V, Steinhilper R, Oltmanns J, Freibert SA, Krapoth N, Linne U. Nat Commun. 2024;15:3269.
- 36. Korshunov S, Imlay JA. Quantification of hydrogen sulfide and cysteine excreted by bacterial cells. Bio Protoc. 2018;8(10):e2847. pmid:29955619
- 37. Hori K, Balasubramanian R, Masuda S. Sulfide-responsive transcription control in Escherichia coli. Microorganisms. 2025;13(2):344. pmid:40005711
- 38. Whitman WB, Ankwanda E, Wolfe RS. Nutrition and carbon metabolism of Methanococcus voltae. J Bacteriol. 1982;149(3):852–63. pmid:6801012
- 39. Wahl A, My L, Dumoulin R, Sturgis JN, Bouveret E. Antagonistic regulation of dgkA and plsB genes of phospholipid synthesis by multiple stress responses in Escherichia coli. Mol Microbiol. 2011;80(5):1260–75. pmid:21463370
- 40. Studier FW, Moffatt BA. Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J Mol Biol. 1986;189(1):113–30. pmid:3537305
- 41. Guzman LM, Belin D, Carson MJ, Beckwith J. Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. J Bacteriol. 1995;177(14):4121–30. pmid:7608087
- 42. Vinella D, Brochier-Armanet C, Loiseau L, Talla E, Barras F. Iron-sulfur (Fe/S) protein biogenesis: phylogenomic and genetic studies of A-type carriers. PLoS Genet. 2009;5(5):e1000497. pmid:19478995
- 43. Fish WW. Rapid colorimetric micromethod for the quantification of complexed iron in biological samples. Methods Enzymol. 1988;158:357–64.
- 44. Beinert H. Semi-micro methods for analysis of labile sulfide and of labile sulfide plus sulfane sulfur in unusually stable iron–sulfur proteins. Anal Biochem. 1983;131:373–8.
- 45. Sehgal P, Olesen C, Møller JV. ATPase activity measurements by an enzyme-coupled spectrophotometric assay. Methods Mol Biol. 2016;1377:105–9.
- 46. Yakamavich JA, Baker TA, Sauer RT. Asymmetric nucleotide transactions of the HslUV protease. J Mol Biol. 2008;380(5):946–57. pmid:18582897
- 47. Carboni M, Clémancey M, Molton F, Pécaut J, Lebrun C, Dubois L. Biologically relevant heterodinuclear iron–manganese complexes. Inorg Chem. 2012;51:10447–60.
- 48.
Charavay C, Segard S, Edon F, Clémancey M, Blondin G. SimuMoss Software. Univ. Grenoble Alpes, CEA, CNRS; 2012.
- 49. Jordan PA, Tang Y, Bradbury AJ, Thomson AJ, Guest JR. Biochemical and spectroscopic characterization of Escherichia coli aconitases (AcnA and AcnB). Biochem J. 1999;344 Pt 3(Pt 3):739–46. pmid:10585860
- 50. Gardner PR, Fridovich I. Inactivation-reactivation of aconitase in Escherichia coli: a sensitive measure of superoxide radical. J Biol Chem. 1992;267(13):8757–63. pmid:1315737
- 51. Weber P, Pissis C, Navaza R, Mechaly AE, Saul F, Alzari PM, et al. High-throughput crystallization pipeline at the crystallography core facility of the institut pasteur. Molecules. 2019;24.
- 52. Vonrhein C, Flensburg C, Keller P, Sharff A, Smart O, Paciorek W, et al. Data processing and analysis with the autoPROC toolbox. Acta Crystallogr D Biol Crystallogr. 2011;67(Pt 4):293–302. pmid:21460447
- 53. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser crystallographic software. J Appl Crystallogr. 2007;40:658–74.
- 54. Abramson J, Adler J, Dunger J, Evans R, Green T, Pritzel A, et al. Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature. 2024;630(8016):493–500. pmid:38718835
- 55. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004;60(Pt 12 Pt 1):2126–32. pmid:15572765
- 56.
Bricogne G, Blanc E, Brandl M, Flensburg C, Keller P, Paciorek W. BUSTER version 2.11.1. Cambridge, UK: Global Phasing Ltd; 2011.
- 57. Punjani A, Rubinstein JL, Fleet DJ, Brubaker MA. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat Methods. 2017;14(3):290–6. pmid:28165473
- 58. Liebschner D, Afonine PV, Baker ML, Bunkóczi G, Chen VB, Croll TI, et al. Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Crystallogr D Struct Biol. 2019;75(Pt 10):861–77. pmid:31588918