Archaea strains and cultivation media
M. thermolithotrophicus (DSM 2095), M. infernus (DSM 11812) and A. fulgidus (DSM 4304) cells were obtained from the Leibniz Institute DSMZ-German Collection of Microorganisms and Cell Cultures (Braunschweig, Germany). M. thermolithotrophicus and M. infernus were cultivated in the same previously described minimal medium with some modifications23 (see Extended Data for the complete composition of the media).
Anaerobic growth of Archaea
Cell growth was followed spectrophotometrically by measuring the OD600. The purity of the culture was checked by light microscopy. The methanogens were cultivated with 1 × 105 Pa of H2:CO2 at an 80:20 ratio in the gas phase. M. infernus was cultivated at 75 °C in 250 ml glass serum flasks and M. thermolithotrophicus was grown at 65 °C in flasks or fermenters. The serum flasks were not shaken but standing. A. fulgidus was cultivated in anaerobic and sealed 22 ml Hungate tubes, with 0.8 × 105 Pa N2:CO2. DSM 4304 culture (0.5 ml) was grown in 10 ml of classic media (see Supplementary Materials for the complete composition of the media composition) containing a final concentration of 20 mM d/l-lactate. The culture was incubated at 80 °C, standing. All cultures were stored at room temperature in the dark under anaerobic conditions. For the A. fulgidus medium, we found that high molybdate concentrations made it unstable. One of the bottles with a high MoO42− concentration turned yellow (unrelated to O2 contamination) and was omitted, resulting in triplicate instead of quadruplicate cultures (Fig. 1c, right panel).
Adaptation of M. thermolithotrophicus to \({\mathbf{SO}}_{\mathbf{4}}^{\mathbf{2-}}\) and minimal \({\mathbf{SO}}_{\mathbf{4}}^{\mathbf{2-}}\) requirement
M. thermolithotrophicus cells grown on 2 mM Na2S were successively transferred to 10 ml sulfur-free cultivation medium. After two transfers, the carry-over sulfur concentration of the inoculum did not support growth of M. thermolithotrophicus. By supplementing 2 mM Na2SO4, M. thermolithotrophicus growth was resumed. No reducing agent was added to cope with the absence of HS−, which normally establishes a suitable reducing environment. Incubation without shaking is particularly important for reproducibility. Therefore, after inoculation, the cultures were incubated at 65 °C, standing for one night followed by shaking at 180 revolutions per minute (r.p.m.) until they reached their maximum OD600. The gas phase was refreshed after the overnight incubation to maintain the pressure at 1 × 105 Pa of H2:CO2. To measure the minimal \({{\rm{SO}}}_{4}^{2-}\) concentration required to sustain growth, sulfur-limited M. thermolithotrophicus cells (using an inoculum to medium ratio of 1:20) were provided with 2 mM, 1 mM, 0.5 mM, 0.25 mM, 0.1 mM and 0.04 mM Na2SO4. Growth was still observable for cells grown on 0.1 mM but not on 0.04 mM Na2SO4.
\({\mathbf{SO}}_{\mathbf{4}}^{\mathbf{2-}}\) measurements via ion chromatography
Ion chromatography (Methrom ion chromatograph) was used to measure the \({{\rm{SO}}}_{4}^{2-}\) concentrations, analysed via the software IC MagIC Net 3.2. A volume of 8 ml per sample was required, with a maximum concentration of 0.5 mM \({{\rm{SO}}}_{4}^{2-}\). \({{\rm{SO}}}_{4}^{2-}\)-reducing M. thermolithotrophicus cells were therefore grown in 1 l Duran bottles with 100 ml sulfur-free media, which was supplemented with 0.5 mM \({{\rm{SO}}}_{4}^{2-}\) before inoculation. As a negative control, 0.5 mM Na2S-grown M. thermolithotrophicus cells were used, inoculated and collected similarly as the \({{\rm{SO}}}_{4}^{2-}\)-reducing cultures. All samples were taken aerobically and were passed through a 0.45 µM filter (Sartorius). If the cell densities were too high to be filtered, the samples were centrifuged at 13,000 × g for 7 min at 4 °C and the supernatant was taken for ion chromatography measurements. The samples were stored at 4 °C if the measurements were not immediately performed.
Growth of M. thermolithotrophicus in a fermenter
M. thermolithotrophicus was grown in three independent fermenters at 60 °C, with 10 mM Na2SO4 as sole sulfur source. For each fermenter, 7 l of anaerobic cultivation medium (see Sulfur-free cultivation medium for Methanococcales) supplemented with 10 mM Na2SO4 was continuously bubbled with H2:CO2 (80:20, 3 l min−1). Under stirring (220 r.p.m.), the medium was inoculated with 360 ml preculture (with an OD600 higher than 3). One hour after inoculation, the culture was stirred at 800 r.p.m. NaOH (1 M) was used as a base to readjust the pH upon acidification, which was controlled using a pH probe. The cells were grown until late exponential phase (OD600 of 6.25–6.8) and then immediately transferred in an anaerobic tent (N2:CO2 atmosphere at a ratio of 90:10). Cells were collected by anaerobic centrifugation for 30 min at 6,000 × g at 4 °C. The highest OD600nm recorded for M. thermolithotrophicus in a SO42−-grown fermenter was 6.8 after 20 h. \({{\rm{SO}}}_{4}^{2-}\) Culture (7 l) with an OD600 of 6.8 yielded 54 g of cells (wet weight). The cell pellet was transferred in a sealed bottle, gassed with 0.3 × 105 Pa N2, flash frozen in liquid N2 and stored at −80 °C.
Synthetic gene constructs
The DNA sequences of the ATP sulfurylase, the APS kinase, the PAP phosphatase and the PAPS reductase α and β subunits from M. thermolithotrophicus were codon optimized for E. coli, synthesized and cloned into pET-28a(+) vectors. For MtATPS, MtAPSK and MtPAPP, the restriction sites NdeI and BamHI were used, with a stop codon (TGA) incorporated before BamHI. For MtPAPSR, a His-tag was placed at the C terminus of the α subunit and a ribosome binding site was inserted between the coding sequences of the α and β subunits. The MtPAPSR construct had the restriction sites NcoI and BamHI, with one stop codon incorporated after the His-tag for the α subunit and one stop codon before BamHI for the β subunit. These steps were performed by GenScript (GenScript). All sequences used are detailed in Supplementary Information under Constructs and gene codon optimization.
Enzyme overexpression and purification
All constructs were overexpressed and purified under aerobic conditions following a similar protocol, except for MtPAPSR which was overexpressed and purified under an anaerobic atmosphere. All enzymes were passed on a HisTrap high-performance column (GE Healthcare), followed, if necessary, by tag cleavage and gel filtration (see Supplementary Materials for the complete protocol).
Protein crystallization
Purified MtATPS, MtAPSK and MtPAPP were kept in 25 mM Tris/HCl pH 7.6, 10% v/v glycerol, 2 mM dithiothreitol and 150 mM NaCl. MtPAPSR was kept in the same buffer without NaCl. Freshly prepared unfrozen samples were immediately used for crystallization. MtATPS, MtAPSK and MtPAPP crystals were obtained under aerobic conditions at 18 °C. MtPAPSR crystals were obtained anaerobically (N2:H2, gas ratio of 97:3) by initial screening at 20 °C. The sitting drop method was performed on 96-well MRC 2-drop crystallization plates in polystyrene (SWISSCI) containing 90 µl of crystallization solution in the reservoir.
Crystallization of MtATPS
MtATPS (0.7 µl) at a concentration of 14 mg ml−1 (MtATPS form 1, Extended Data Table 1) or at a concentration of 27 mg ml−1 (MtATPS form 2) was mixed with 0.7 µl reservoir solution. MtATPS at 27 mg ml−1 was co-crystallized with 2 mM AMPcPP as well as 2 mM Na2SO4. For MtATPS form 1, transparent star-shaped crystals appeared after a few weeks in the following crystallization condition: 35% w/v pentaerythritol ethoxylate (15/4 EO/OH) and 100 mM 2-(N-morpholino)ethanesulfonic acid (MES) pH 6.5. For MtATPS form 2, transparent, long but thin plate-shaped crystals appeared after a few weeks in the following crystallization condition: 20% w/v polyethylene glycol 8000, 100 mM MES pH 6.0 and 200 mM calcium acetate.
Crystallization of MtAPSK
MtAPSK (0.7 µl) at a concentration of 17.6 mg ml−1 was mixed with 0.7 µl reservoir solution and co-crystallized with 2 mM MgCl2. Transparent, plate-shaped crystals appeared after a few weeks in the following crystallization condition: 20% w/v polyethylene glycol 3350 and 100 mM tri-sodium citrate pH 5.5. MtAPSK was also crystallized with 2 mM MgCl2 and 2 mM APS but the obtained structures of those crystals were of lower resolution and without any substrate or product present in the active site.
Crystallization of MtPAPSR
MtPAPSR (0.7 µl) at a concentration of 20 mg ml−1 was mixed with 0.7 µl reservoir solution and co-crystallized with FAD (0.5 mM final concentration). The crystal used for phasing was a brown flat square and appeared after a few days in the following crystallization condition: 40% v/v 2-methyl-2,4-pentanediol and 100 mM Tris/HCl pH 8.0.
The crystal used to refine at high resolution was brown with an elongated plate shape. It appeared after a few days in the following crystallization condition: 35% v/v 2-methyl-2,4-pentanediol, 100 mM Tris pH 7.0 and 200 mM NaCl. Before transfer to liquid N2, the crystal was soaked in 10 mM disodium 3’-phosphoadenosine 5’-phosphate for 7 min.
Crystallization of MtPAPP
MtPAPP (0.7 µl) at a concentration of 20 mg ml−1 was mixed with 0.7 µl reservoir solution and co-crystallized with Tb-Xo4 (10 mM final concentration), MnCl2 (2 mM final concentration) and 2 mM PAP. The Tb-Xo4 is a nucleating/phasing agent50, which should increase the crystallization performance; however, in this case, the same crystalline form was obtained in the absence of the compound and diffracted to similar resolution. Transparent, bipyramid crystals appeared after a few weeks in the following crystallization condition: 1.6 M tri-sodium citrate.
X-ray crystallography and structural analysis
MtPAPSR crystal handling was done inside the Coy tent under anaerobic atmosphere (N2:H2, 97:3); the other crystals were handled under aerobic conditions. The crystals were directly plunged in liquid nitrogen or were soaked for 5–30 s in their crystallization solution supplemented with a cryoprotectant before being frozen in liquid nitrogen. For MtATPS form 2, 30% glycerol was used as cryoprotectant. For MtAPSK, 25% ethylene glycol was used as cryoprotectant.
Crystals were tested and collected at 100 K at different synchrotrons (Extended Data Table 1). Data were processed with autoPROC51 except for MtPAPP, which gave better statistics with indexation by the X-ray Detector Software (XDS) and the scaling step performed with SCALA52. All data collection statistics are provided in Extended Data Table 1. MtATPS forms 1 and 2, MtAPSK and MtPAPP were solved by using PHENIX with the following templates: 1V47 (ATPS from T. thermophilus) for MtATPS form 1, MtATPS form 1 for MtATPS form 2 and 5CB6 (APS kinase from Synechocystis sp.) for MtAPSK. For MtPAPP, the template was created de novo using AlphaFold 2 (ref. 32).
For MtPAPSR, an X-ray fluorescence spectrum on the Fe K-edge was measured to optimize the data collection at the appropriate wavelength. Datasets were collected at 1.73646 Å for the single-wavelength anomalous dispersion experiment. Native datasets were collected at a wavelength of 0.97625 Å on another crystal. Data were processed and scaled with autoPROC51. Phasing, density modification and automatic building were performed with CRANK-2 (ref. 53).
All models were manually rebuilt with COOT and further refined with PHENIX54,55. During the refinement, non-crystallographic symmetry and translational-libration screw were applied. For all structures except for ATPS form 1, hydrogens were added in riding position in the last refinement cycle. Hydrogens were removed in the final deposited models.
All models were validated using MolProbity56. Data collection and refinement statistics, as well as PDB identification codes for the deposited models and structure factors are listed in Extended Data Table 1. Figures were generated with PyMOL (Schrödinger). The metal in MtATPS was modelled as zinc using CheckMyMetal57.
High-resolution clear native PAGE (hrCN PAGE)
To visualize the expression levels of MtFsr when cells were grown on different sulfur sources, hrCN PAGE was performed. M. thermolithotrophicus cultures (2 × 10 ml) were supplemented with either 2 mM Na2S, 2 mM Na2SO3, 2 mM Na2S and 2 mM Na2SO4, or 2 mM Na2SO4 as sulfur substrates and grown for one night at 65 °C, standing. Cells were collected by anaerobic centrifugation at 6,000 × g for 20 min at room temperature and the cell pellets were resuspended in 2 ml lysis buffer (50 mM tricine pH 8.0 and 2 mM sodium dithionite). The cells were sonicated 4× at 70% intensity for 10 s, followed by a 30 s break (MS 73 probe, SONOPULS Bandelin). The hrCN PAGE was run anaerobically and the protocol is detailed in Extended Data under hrCN PAGE preparation. One gel with an 8–15% acrylamide gradient was run (shown in Fig. 5b) and another one with a 5–15% acrylamide gradient (see Source Data Fig. 5).
Coupled enzyme activity of MtATPS/MtAPSK
The activity of both enzymes was determined by the production of ADP which was coupled to NADH oxidation via pyruvate kinase and lactate dehydrogenase58. The assays were performed in a final volume of 100 µl 96-well deep-well plates and spectrophotometrically monitored (Omega multimode microplate reader) at 360 nm at 35 °C. KH2PO4 (100 mM) at pH 7.0, supplemented with 1.5 mM MgCl2 and 100 mM KCl, was used as a buffer. For NADH, a molar extinction coefficient of 4,546.7 cm−1 M−1 was experimentally determined for the above-named conditions. To the buffer, 1 mM NADH, 2.5 mM Na2SO4, 1 mM phosphoenolpyruvate (PEP), 2 mM ATP, 2 U inorganic pyrophosphatase (Saccharomyces cerevisiae, 10108987001, Sigma-Aldrich), 1.1 U ml−1 lactate dehydrogenase, 0.8 U ml−1 pyruvate kinase (rabbit muscle, P0294, Sigma-Aldrich) and 0.5 mg ml−1MtAPSK (all final concentrations) were added. The reaction was started by the addition of 0.5 mg ml−1MtATPS. Addition of 0.02 mM Na2MoO4 did not affect activity (0.116 ± 0.027 µmol of oxidized NADH min−1 mg−1), but the addition of 2 mM Na2MoO4 resulted in a decrease (0.068 ± 0.019 µmol of oxidized NADH min−1 mg−1). All assays were performed in triplicates.
MtPAPP enzyme assay
The activity of the MtPAPP was determined by the production of orthophosphate, which was quantified using the malachite green phosphate assay kit (Sigma-Aldrich) by the formation of a green complex. The assays were performed in 96-well deep-well plates and the absorbance at 620 nm was spectrophotometrically followed (Omega multimode microplate reader). Tris/HCl (25 mM) at pH 7.64 was used as a buffer. Buffer, 40 µM PAP or 90 µM of AMP/ADP/ATP/APS or PPi, 1 mg ml−1 bovine serum albumin, 50 µM MnCl2 and/or 50 µM MgCl2 (final concentration) were mixed in a 1.5 ml Eppendorf tube on ice. Previously frozen MtPAPP (0.5 µg ml−1 final concentration) was added and the mixture (final volume of 40 µl) was immediately incubated for 5 min at 40 °C. Next, 14 µl of the reaction mix was diluted in 66 µl of filtered Milli-Q H2O and immediately flash frozen in liquid N2 to quench the reaction. Then, 20 µl of malachite green reagent was added to the samples, the mixture was incubated at room temperature for 30 min and the formation of the green complex was measured at 620 nm. All assays were performed in triplicates. The measurements presented in Fig. 3a come from two different experiments (left and right subpanels). Both experiments were performed at two different days with the same enzyme preparation.
Coupled MtPAPSR assay
Since PAPS is unstable at high temperatures, we first tried to determine the activity of MtPAPSR in the direction of PAPS production, as previously described for dissimilatory APS reductases for APS production38. PAPS oxidation was determined in 50 mM Tris/HCl buffer (pH 7.5) containing 5 mM Na2SO3, 2 mM PAP or 2 mM AMP (final concentrations) and 3.27 µg ml−1MtPAPSR. The reaction was started with a final concentration of 0.5 mM K3Fe(CN)6. The decrease in absorbance at 420 nm was measured and corrected for the background reaction without enzyme. No activity was detected. Therefore, we used the physiological reaction to monitor MtPAPSR activity. To perform the coupled MtPAPSR assay, the enzymes needed to be purified at the same time and immediately used for the assay (see Supplementary Materials for the detailed purification protocol for the enzymes used in this assay).
MtPAPSR activity assays were carried out in an anaerobic atmosphere (100% N2) at 45 °C. The assays were performed in 200 µl final volume in 96-well deep-well plates and spectrophotometrically monitored on a SPECTROstar Nano microplate reader. HEPES (50 mM, pH 7.0) supplemented with 50 mM KCl, 1.5 mM MnCl2 and 1.5 mM MgCl2 was used as a buffer. Reduced methyl viologen (MVred, 0.5 mM) served as an electron donor for MtPAPSR. The molar extinction coefficient (ε600nm = 8,133.3 cm−1 M−1) was experimentally determined using the above-named conditions and by reducing methyl viologen with 2 mM sodium dithionite. For the assay, methyl viologen was reduced with carbon monoxide by the CO-dehydrogenase from Clostridium autoethanogenum according to a previously published protocol59. CO was exchanged for N2 and the MVred was immediately used for the assay. To the buffer and MVred, 5 mM ATP, 1 mM sodium dithionite, 0.2 U pyrophosphatase (E. coli, MFCD00131379, Sigma-Aldrich), 0.127 mg ml−1MtATPS, 0.12 mg ml−1MtAPSK, 0.1 mg ml−1MtPAPP and 0.0645 mg ml−1MtPAPSR were added. The reaction was started by the addition of 5 mM Na2SO4 and followed by oxidation of MVred at 600 nm. All assays were performed in triplicates.
Sulfite reductase activity in cell extracts
To determine the sulfite reductase activity from M. thermolithotrophicus, cultures were grown on either 2 mM Na2S, 2 mM Na2SO3 or Na2SO4 in 10 ml of the above-mentioned medium in serum flasks. Cells (9 ml) were collected in late exponential phase (OD600: 3.45 for 2 mM Na2S, 3.91 for 2 mM Na2SO3, 3.37 for Na2SO4) by centrifugation at 6,000 × g for 10 min at 4 °C. The supernatant was discarded and the cell pellets were frozen in liquid N2. The pellets were then resuspended in 1 ml 0.5 M KH2PO4 pH 7.0. The cells were lysed by sonication (2× 10 s at 50% intensity, probe MS73, SONOPULS Bandelin), followed by centrifugation at 4 °C at 15,600 × g. The supernatant was passed through a 0.2 µm filter and the protein concentration was determined by the Bradford method (6.63 mg ml−1 for 2 mM Na2S, 6.14 mg ml−1 for 2 mM Na2SO3 and 6.31 mg ml−1 for Na2SO4). The activity assays were performed under an anaerobic atmosphere (100% N2) at 50 °C in 96-well deep-well plates and spectrophotometrically monitored (SPECTROstar Nano microplate reader). The assay mixture contained 0.5 M KH2PO4 pH 7.0, 118 µM MVred (final concentration, previously reduced with the equimolar amount of sodium dithionite) and 30 µM Na2SO3 (final concentration). Under these conditions, a molar extinction coefficient of ε600nm = 9,840 cm−1 M−1 was experimentally determined. The reaction was started by the addition of 0.05 µg of cell extract, followed by oxidation of MVred at 600 nm. All assays were performed in triplicates.
Phylogenetic trees
For a detailed description of the phylogenetic analysis, see Supplementary Materials60.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.