Chemicals and biochemicals were obtained from Sigma-Aldrich, Merck, Roth, VWR, or AppliChem. Materials for molecular biology were purchased from New England BioLabs. Materials and equipment for protein purification were obtained from GE Healthcare, Macherey-Nagel or Millipore. Lead acetate paper was obtained from Macherey-Nagel. Primers were synthesized by Sigma-Aldrich.

Desulfurella acetivorans A63 (DSM 5264) was obtained from the Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ). The cells were grown in medium containing 6.2 mM NH₄Cl, 2.2 mM CaCl₂⋅2H₂O, 1.6 mM MgCl₂⋅6H₂O, 4.4 mM KCl, 1.7 mM KH₂PO₄, 1.3 mM K₂HPO₄, 10 g l^–1 sulfur powder, 1 ml l^–1 SL-10 trace element solution (3 g l^–1 FeCl₂⋅4H₂O, 70 mg l^–1 ZnCl₂, 100 mg l^–1 MnCl₂⋅4H₂O, 4 mg l^–1 CuCl₂⋅2H₂O, 24 mg l^–1 NiCl₂⋅6H₂O, 36 mg l^–1 Na₂MoO₄⋅2H₂O, 30 mg l^–1 H₃BO₃, 224 mg l^–1 CoCl₂⋅6H₂O) and 1 ml l^–1 Wolfe’s vitamin solution (20 mg l^–1 biotin, 20 mg l^–1 folic acid, 100 mg l^–1 pyridoxamine dihydrochloride, 50 mg l^–1 thiamine dihydrochloride, 50 mg l^–1 riboflavin, 50 mg l^–1 nicotinic acid, 50 mg l^–1 DL-Ca-pantothenate, 1 mg l^–1 cyanocobalamin, 50 mg l^–1 4-aminobenzoic acid and 50 mg l^–1 lipoic acid). The medium was prepared without sulfur and vitamin solution, made anaerobic by bubbling with N₂ (100%) and dispensed anaerobically into serum bottles containing sulfur powder. The bottles were sealed with butyl rubber stoppers and aluminum caps and autoclaved for 40 min at 110°C. Before inoculation, the medium was reduced by the addition of Na₂S⋅9H₂O to a final concentration of 0.05% (w/v) and supplemented with vitamins. When cultivated heterotrophically, acetate or propionate was supplemented as sole carbon source to a final concentration of 0.5 and 0.05%, respectively. The gas phase was replaced with N₂/CO₂ (80:20; heterotrophic growth) or H₂/CO₂ (80:20; autotrophic growth) at 1 bar overpressure. The pH was adjusted with bicarbonate to the strain optimum of 6.8–7.0. Cultures were incubated at 55°C while shaking at 130 rpm. Growth was determined using Neubauer counting chambers.

Escherichia coli strains [Top10, Rosetta 2 (DE3)] were grown at 37°C in lysogeny broth medium. Antibiotics were added to the cultures to a final concentration of 100 μg ml^–1 ampicillin and 34 μg ml^–1 chloramphenicol.

Cultures (200 ml) were filtrated aerobically (Whatman GF/D, 47 mm) to remove the elemental sulfur before centrifugation (15,000 × g, 4°C, 20 min). The supernatant was discarded gently and the resulting cell pellet was resuspended with 20 mL of fresh sulfur-free medium. The cell suspension was transferred to a 50 mL centrifuge tube and centrifuged again (21,000 × g, 4°C, 20 min). After having discarded the supernatant, the cells were frozen with liquid nitrogen and stored at −80°C for up to 6 months. To perform enzymatic assays, the frozen cells were thawed on ice and resuspended in 20 mM Tris–HCl (pH 7.8 at 55°C), 5 mM dithiothreitol (DTT), and lysed on ice with the Sonopuls ultrasonic homogenizer (BANDELIN Electronic GmbH; 60% amplitude, 4 min, 1-s pulse, 2-s breaks; total energy input 2,000 kJ). The insoluble cell debris was removed by centrifugation (21,000 × g, 4°C, 20 min).

Acetyl-CoA, propionyl-CoA and succinyl-CoA were synthesized from the corresponding anhydrides and CoA according to Simon and Shemin (1953).

Standard protocols were used for the purification, cloning, transformation and amplification of DNA (Ausubel et al., 1987). The AHF96498 encoding gene was amplified by PCR with Q5 High-Fidelity DNA Polymerase using a forward primer (5′-ATATGAATTCATGGGTCACGGAAAAG-3′) introducing an EcoRI site (bold) and a reverse primer (5′-TAATCTCGAGTTAAAAGGCCAGCTTC-3′) introducing an XhoI site (bold). PCR conditions were as follows: 30 cycles of 30 s denaturation at 98°C, 60 s primer annealing at 64°C, and 30 s elongation at 72°C. The isolated PCR product was treated with the corresponding restrictases and ligated into the expression vector pET23b containing a sequence encoding a C-terminal His₆-tag. The plasmid was transformed into E. coli TOP10 for amplification, followed by purification and sequencing.

The amplified expression vector was used to transform E. coli Rosetta 2 (DE3). The cells were grown at 37°C in lysogeny broth medium with ampicillin and chloramphenicol. Expression was induced at an optical density (OD₆₀₀ nm) of 0.5–0.8 with 1 mM isopropyl-ß-D-thiogalactopyranoside (IPTG), and the temperature was lowered to 20°C. The cells were harvested after overnight growth and stored at −20°C until use.

Frozen cells were suspended in a triple volume of 20 mM Tris–HCl (pH 7.8), containing 20 mM NaCl and 0.1 mg ml^–1 DNase I. The cell suspensions were lysed by a threefold passage through a chilled French pressure cell (103 MPa). To remove thermally unstable E. coli proteins, the supernatant (cell extract) has been subjected to a heat precipitation step (80°C; 15 min). The resulting cell lysate was centrifuged (100,000 × g; 4°C; 60 min), filtered and used for protein purification.

The heterologously produced His-tagged protein encoded by the AHF96498 gene was purified by affinity chromatography using a gravity flow separation column (Econo-column, 1.0 cm × 30 cm glass chromatography column, Bio-Rad) loaded with 2-mL Protino Ni-NTA agarose matrix (Macherey-Nagel). The column was equilibrated with 20 mM Tris–HCl (pH 7.8) containing 20 mM NaCl. Cell extract was applied to the column and incubated for 15 min at 4°C. To elute unwanted proteins, the column was washed first with the same equilibration buffer, then two times with the same buffer containing 10 mM imidazole. The recombinant enzyme was eluted with the same buffer containing 500 mM imidazole. The enzyme was concentrated using a 10 K Vivaspin Turbo 4 and stored at −20°C with glycerol (50%, v/v). The identity of the purified recombinant protein was confirmed at the IZKF Core Unit Proteomics Münster with tryptic in-gel digestion and mass spectrometric analysis using Synapt G2 Si coupled to M-Class (Waters Corp, Eschborn, Germany).

CoA-transferase activity of the purified AHF96498 gene product and in D. acetivorans cell extracts was measured using ultra high performance liquid chromatography (UHPLC). All the activities were measured at 55°C and the reaction mixtures (40 μl) contained 100 mM Tris–HCl (pH 7.8), 5 mM MgCl₂, 5 mM DTT, and purified enzyme or cell extract. Activities were tested with acetyl-CoA, succinyl-CoA or propionyl-CoA toward acetate, succinate or propionate. In addition, the purified CoA-transferase was tested also with acetyl-CoA toward butyrate, 4-hydroxybutyrate, formate, methylsuccinate, DL-malate, and glutarate. The reaction was stopped after 1 min by the addition of 20 μl of 1 M HCl/10% acetonitrile (sample:stop solution, 1:1 [v/v]). The specific activities were calculated by considering the peaks of consumed and formed CoA-esters. The concentration of the latter was calculated by multiplying the starting substrate concentration by the relative abundance [%] of the formed CoA-ester integrated peak area.

Succinyl-CoA synthetase activity was measured using UHPLC as the CoA-dependent formation of succinyl-CoA from succinate and CoA. The assay mixture contained 100 mM Tris–HCl (pH 7.8), 5 mM DTT, 5 mM MgCl₂, 20 mM succinate, 5 mM ATP, 1 mM CoA, and cell extract.

Acetate kinase activity was measured spectrophotometrically following acetate-dependent oxidation of NADH at 365 nm in an assay containing (in 0.3 ml) 100 mM Tris–HCl (pH 7.8), 5 mM DTT, 5 mM MgCl₂, 20 mM acetate, 5 mM ATP, 0.5 mM NADH, 5 mM PEP, 5 μ pyruvate kinase (rabbit muscle, Sigma P9136), 25 μ lactate dehydrogenase (rabbit muscle, Sigma L2500), and cell extract. The activity of AMP-forming acetyl-CoA synthetase was tested by including 5 μ myokinase and 0.5 mM CoA in the same reaction mixture.

Phosphotransacetylase activity was measured spectrophotometrically in an assay coupled with endogenous citrate synthase and malate dehydrogenase following NAD reduction at 365 nm as the CoA- and acetyl-phosphate dependent acetyl-CoA formation in the reaction mixture containing 100 mM Tris–HCl (7.8), 5 mM DTT, 5 mM NAD, 10 mM malate, 10 mM acetyl-phosphate, 1 mM CoA and cell extract.

For inactivation experiments, the purified CoA-transferase (20 μg) was treated with NaBH₄. The enzyme was preincubated with 1 mM acetyl-CoA for 10 min at 55°C, while an assay without acetyl-CoA was used as a control. NaBH₄ in 1 M NaOH was added to both reactions at a final concentration of 10 mM, followed by the addition of HCl to a final concentration of 10 mM (Friedmann et al., 2006). The reactions were further incubated for 10 min. Finally, 2 μg of the enzyme was used for the measurement of acetyl-CoA:propionate CoA-transferase activity.

CoA and CoA-esters were detected with Agilent 1290 Infinity II UHPLC using a reversed-phase C18 column (Agilent InfinityLab Poroshell 120 EC-C18 1.9 μm 2.1 mm × 50 mm column). The following acetonitrile gradient in 10 mM potassium phosphate buffer (pH 7) with a flow rate of 0.55 ml min^–1, was used: from 2 to 8% at 0–2.66 min; from 8 to 30% at 2.66–3.33 min; from 30 to 2% at 3.33–3.68 min; 2% at 3.68–5 min. Retention times were: succinyl-CoA, 0.7 min; CoA, 0.9 min; acetyl-CoA, 1.7 min; propionyl-CoA, 2.4 min; butyryl-CoA, 3.4 min. Reaction products and standard compounds were detected by UV absorbance at 260 nm with a 1290 Infinity II diode array detector (Agilent) and the amount of product was calculated from the relative peak area. The identification of the CoA esters was based on co-chromatography with standards and analysis of the UV spectra of the products.

Query sequences for the database searches were obtained from the NCBI database. For the alignment, we took the sequences of experimentally characterized CoA-transferases used by Hackmann (2021) (96 sequences) in his analysis of CoA-transferases and supplemented them with sequences from Pseudomonas aeruginosa (Sasikaran et al., 2014), Paraburkholderia xenovorans (Kronen et al., 2015), from Desulfurellaceae species (17 sequences), from a closed genome of a sulfate-reducing bacterium (1 sequence) and from sulfate- and sulfur-reducing bacteria where the CoA-transferase variant of the oxidative TCA cycle was shown experimentally (7 sequences) (Thauer et al., 1989; Segura et al., 2008; Mollaei et al., 2021). The genomes of sulfate reducers selected were at the “Finished” status. Only for the family Desulfurellaceae, all the genomes available from the NCBI database have been considered. The BLASTP searches (Altschul et al., 1990) were performed via the NCBI BLAST server¹ and via the Integrated Microbial Genomes and Microbiomes system.² The phylogenetic tree was constructed by using the maximum likelihood method and Jones-Taylor-Thornton (JTT) matrix-based model (Jones et al., 1992) in MEGA11 (Tamura et al., 2021). Altogether, 123 amino acid sequences were used in this analysis. In addition, the same data set was used for the construction of a tree with the maximum-parsimony statistical method and the Subtree-Pruning-Regrafting (SPR) search method.

DNA sequence determination of purified plasmids was performed by Eurofins (Ebersberg, Germany). Protein concentration was measured according to the Bradford method (Bradford, 1976) using BSA as a standard. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE; 12.5%) was performed as previously described (Laemmli, 1970). Proteins were visualized using Coomassie blue staining (Zehr et al., 1989). K_m and V_max values were calculated by GraphPad Prism 5 software.

Desulfurella acetivorans was described as a thermophilic sulfur-reducing bacterium oxidizing acetate through the TCA cycle, while no significant growth was shown with propionate (Bonch-Osmolovskaya et al., 1990; Schmitz et al., 1990). Later, the capability of this species to grow autotrophically with molecular hydrogen was also demonstrated (Pradella et al., 1998). In our hands, this bacterium could grow autotrophically as well as heterotrophically on acetate or propionate. Acetate-grown cells had only very low activity of succinyl-CoA synthetase, whereas this activity in propionate- and autotrophically grown cells was much higher (Table 1). In contrast, acetate-grown cells showed high activity of succinyl-CoA:acetate CoA-transferase, suggesting that acetate activation in D. acetivorans proceeds through the CoA-transfer from succinyl-CoA. The activity of this enzyme in autotrophically- and propionate-grown cells was 4 and 34 times lower than in acetate-grown cells, respectively (Table 1). The participation of a CoA-transferase in acetate activation is further supported by the low activity of enzymes involved in alternative acetate activation pathways. Indeed, the acetate kinase activity was 4 times lower than the succinyl-CoA:acetate CoA-transferase activity (Table 1). Since the addition of myokinase to the reaction mixture to detect AMP formation did not result in any detectable difference in the spectrophotometric assays, the activity of the AMP-forming acetyl-CoA synthetase was absent in acetate-grown D. acetivorans cells (or was below the detection limit of 0.02 μmoles min⁻¹ mg^–1 protein). Our comparison of proteomes of acetate- and propionate-grown cells revealed that one of the four CoA-transferases present in D. acetivorans genome (AHF96498) was 50-fold down-regulated (Pettinato, König and Berg, manuscript in preparation). Therefore, we decided to characterize this putative succinyl-CoA:acetate CoA-transferase biochemically.

The purified recombinant AHF96498 gene product catalyzed CoA-transferase reaction. We examined the ability of this enzyme to utilize different CoA-donors (succinyl-CoA, acetyl-CoA and propionyl-CoA) and CoA-acceptors (acetate, succinate, propionate, butyrate, formate, 4-hydroxybutyrate, methylsuccinate, glutarate, DL-malate). The enzyme was active with all three tested CoA-esters and showed the highest activity with propionate and acetate, lower activity with 4-hydroxybutyrate and butyrate and (almost) no activity with formate, methylsuccinate, malate and glutarate (Table 2). The activity with succinyl-CoA and acetate did not follow a Michaelis-Menten kinetic (Figure 2) but increased linearly with the rise of succinate/succinyl-CoA concentrations (40 μmoles min⁻¹ mg^–1 protein at 2 mM succinyl-CoA and 20 mM acetate, Table 2). The high affinities of the enzyme for acetate/acetyl-CoA (K_m of 0.4/0.2 mM, respectively) further strengthened our suggestion that its main physiological function is acetate activation and succinate formation in the course of the oxidative TCA cycle. However, it could also use propionyl-CoA as a CoA-donor. Moreover, the highest activity of the enzyme was detected with propionate and acetyl-CoA or acetate and propionyl-CoA as substrates (Table 2). This activity could also be measured in D. acetivorans cell extracts and it showed the same trend in autotrophically vs. acetate-/propionate-grown cells as with succinyl-CoA and acetate (Tables 1, 3). Nevertheless, acetyl-CoA:propionate (or propionyl-CoA:acetate) CoA-transferase activity does not have any physiological sense in autotrophic or acetate-grown cells due to the unavailability of propionate/propionyl-CoA and can be explained by the promiscuity of the AHF96498 gene product. Importantly, the K_m values of the heterologously produced enzyme to the substrates were comparable to the apparent K_m values in cell extracts, further confirming the major role of this enzyme in the D. acetivorans TCA cycle (compare Tables 2, 3). To understand the reaction mechanism of the protein, an inactivation test with NaBH₄ has been performed. The reduction of a glutamyl-CoA intermediate by borohydride would lead to the consequent loss of activity by inactivation of the catalytic glutamate residue, which is typical for class I CoA-transferases (Solomon and Jencks, 1968; Tung and Wood, 1975; Moore and Jencks, 1982). The CoA-free enzyme did not show any reduction of activity, whereas the enzyme pre-incubated with acetyl-CoA and subsequently treated with borohydride retained only 4.5% of the original activity, thus identifying AHF96498 as a class I CoA-transferase.

To better understand the evolution of D. acetivorans CoA-transferases, we performed the phylogenetic analysis of CoA-transferases belonging to different families. The resulting phylogenetic tree (Figure 3) was similar to the tree published by Hackmann (2021), where a new six-family classification of CoA-transferases was proposed, elaborating the previous three-family classification of Heider (2001). The characterized succinyl-CoA:acetate CoA-transferase encoded by the AHF96498 gene belonged to the OXTC1/family I CoA-transferases; close homologs of this protein were found in all sequenced Desulfurellaceae, suggesting that they all use the CoA-transferase variant of the TCA cycle. Interestingly, the CoA-transferases of species in which this variant of the TCA cycle was discovered (i.e., Desulfuromonas acetoxidans and Desulfobacter hydrogenophilus) belonged to another family (Cat1/family I; Figure 3). Notably, most of the Cat1 proteins show a preference for short acyl-CoAs (C₂–C₄), with 13 of them being already characterized as succinyl-CoA:acetate CoA-transferases (Hackmann, 2021). The phylogenetic analysis also revealed that two other CoA-transferases from D. acetivorans, AHF97294, and AHF97575, belonged to the Cat1/family I and the Frc/family III CoA-transferases, respectively, whereas the CoA-transferase AHF96963 belonged to the Gct/family I cluster (Figure 3).