1. Introduction
Thiosulfate (S
2O
32-) is a sulfur substrate that is oxidized by the majority of dissimilatory sulfur oxidizers. Its complete oxidation to sulfate is always initiated, and in many cases also completely performed, in the bacterial periplasm and involves the well-studied thiosulfate-oxidizing Sox multienzyme system [
1,
2,
3] (
Figure 1a). Three proteins, SoxYZ, SoxXA and SoxB are required for the initial steps. The
c-type cytochrome SoxXA catalyzes the oxidative formation of a disulfide linkage between the sulfane sulfur of thiosulfate and the persulfurated active site cysteine residue of SoxY [
4]. Then, SoxB catalyzes the hydrolytic release of the sulfone group as sulfate, leaving the original sulfane sulfur of thiosulfate bound to SoxY [
5,
6]. The reaction cycle can be fully completed in the periplasm in organisms containing the hemomolybdoprotein SoxCD, that catalyzes oxidation of SoxY-bound sulfane sulfur to a sulfone, followed again by SoxB-catalyzed hydrolytic release of sulfate [
7].
Many sulfur oxidizers do not contain SoxCD and have a so-called “truncated” Sox system (
Figure 1b) [
2,
8]. For complete oxidation to sulfate, truncated Sox systems can be combined with cytoplasmic sulfur oxidation systems. How the sulfur is transferred into the cytoplasm for further oxidation is still a mystery. The Alphaproteobacterium
Hyphomicrobium denitrificans X
T (DSM 1869
T) is a representative of this group [
9] (
Figure 1b). In this organism, two genes encoding predicted sulfur compound transporters (SoxT1A and SoxT1B) are located in close proximity to the
sox genes and the genes for the cytoplasmic sulfane sulfur-oxidizing heterodisulfide reductase-like (sHdr) system (
Figure 1b). While the
H. denitrificans Sox and sHdr proteins have been shown experimentally to be essential for thiosulfate oxidation [
9,
10,
11], evidence for the proposed sulfur transport has not been provided so far.
Figure 1.
(
a) Model of the complete periplasmic Sox pathway and exemplary
sox gene cluster (A6W98_09510 to A6W98_09585) from the Alphaproteobacterium
Rhodovulum sulfidophilum DSM 1374
T (Rhodobacterales, Rhodobacteraceae) [
12,
13]. SoxS is neither part of the Sox enzyme system nor involved in its regulation [
14]. This periplasmic thiol–disulfide oxidoreductase of the Dsb family prevents SoxYZ inactivation by reducing false mixed disulfides [
15,
16]. (
b) Model of thiosulfate oxidation and a genetic island for sulfur oxidation (Hden_0678 to Hden_0706) in
Hyphomicrobium denitrificans DSM 1869
T (Hyphomicrobiales, Hyphomicrobiaceae) [
9]. The
lip genes encode proteins involved in posttranslational assembly of lipoate on the lipoate-binding LbpA2 protein. The truncated Sox system in the periplasm consists of SoxXY, SoxB and SoxYZ. The sulfane sulfur stemming from thiosulfate and bound to SoxY is transferred to the cytoplasm, possibly via one (or both) of the transporters SoxT1A and Soxt1B, and oxidized to sulfite by the sHdr-LbpA2 system. Sulfite is excreted, probably via TauE, and cannot be effectively oxidized. In panels (
a) and (
b), periplasmic, membrane-bound and cytoplasmic proteins and the encoding genes are shown in green, blue and yellow, respectively. Regulator genes are highlighted in red. The
hyp and
rhd genes encode a predicted cytochrome P450 and a rhodanese-like protein, respectively.
Figure 1.
(
a) Model of the complete periplasmic Sox pathway and exemplary
sox gene cluster (A6W98_09510 to A6W98_09585) from the Alphaproteobacterium
Rhodovulum sulfidophilum DSM 1374
T (Rhodobacterales, Rhodobacteraceae) [
12,
13]. SoxS is neither part of the Sox enzyme system nor involved in its regulation [
14]. This periplasmic thiol–disulfide oxidoreductase of the Dsb family prevents SoxYZ inactivation by reducing false mixed disulfides [
15,
16]. (
b) Model of thiosulfate oxidation and a genetic island for sulfur oxidation (Hden_0678 to Hden_0706) in
Hyphomicrobium denitrificans DSM 1869
T (Hyphomicrobiales, Hyphomicrobiaceae) [
9]. The
lip genes encode proteins involved in posttranslational assembly of lipoate on the lipoate-binding LbpA2 protein. The truncated Sox system in the periplasm consists of SoxXY, SoxB and SoxYZ. The sulfane sulfur stemming from thiosulfate and bound to SoxY is transferred to the cytoplasm, possibly via one (or both) of the transporters SoxT1A and Soxt1B, and oxidized to sulfite by the sHdr-LbpA2 system. Sulfite is excreted, probably via TauE, and cannot be effectively oxidized. In panels (
a) and (
b), periplasmic, membrane-bound and cytoplasmic proteins and the encoding genes are shown in green, blue and yellow, respectively. Regulator genes are highlighted in red. The
hyp and
rhd genes encode a predicted cytochrome P450 and a rhodanese-like protein, respectively.
The obligately heterotrophic
H. denitrificans oxidizes thiosulfate as an additional electron donor during growth on compounds like methanol [
10]. In batch culture, substantial amounts of sulfite are excreted as the product of sHdr-catalyzed sulfur oxidation and accumulate because an enzyme catalyzing efficient sulfite oxidation is not present [
10]. Accumulation of sulfite as an intermediate has also been described for some facultatively autotrophic sulfur oxidizing Alphaproteobacteria, e.g.,
Rhodovulum (previously
Rhodobacter)
sulfidophilum [
17].
(Bi)sulfite (HSO
3-), SO
32- is a highly reactive, strong nucleophile and has many toxic effects. Its strong reducing capacity (
E0` for the sulfate/sulfite couple is -515 mV) contributes to its toxicity and antimicrobial action, which have led to its widespread use as food preservative [
18,
19]. Free sulfite can damage DNA through formation of adducts [
20,
21,
22]. Its toxic effect on mammalian cells has been attributed to the formation of sulfur- and oxygen-based free radicals [
23,
24] which can in turn react with lipids and proteins [
25,
26]. The full Sox pathway or the truncated Sox/sHdr combination may be advantageous, despite the intermediate release of sulfite, for organisms such as
H. denitrificans or
R. sulfidophilum, at low thiosulfate concentrations if removal by other members of the community or chemical oxidation in oxygenated environments keeps sulfite concentrations below inhibitory levels. In any case, the formation of the toxic intermediate sulfite during the oxidation of sulfur compounds, as well as the switching between organic and inorganic electron donors requires fine-tuning to the environmental conditions.
Accordingly, complex regulatory patterns have been reported for facultative sulfur oxidizers, with upregulation usually occurring only in the presence of metabolizable sulfur substrates, whereas the corresponding genes are thought to be always highly expressed in chemolithoautotrophs restricted to the oxidation of sulfur compounds. In
H. denitrificans, and other Alphaproteobacteria that are not restricted to sulfur oxidation, such as
R. sulfidophilum,
Paracoccus pantotrophus or
Pseudaminobacter salicylatoxidans, the ability to oxidize thiosulfate and, depending on the organism, other reduced inorganic and organic sulfur compounds such as sulfide or dimethyl sulfide, is not constitutive but can be induced by the presence of oxidizable sulfur compounds [
10,
17,
27,
28]. While the transcriptional repressor sHdrR is involved in this process in
H. denitrificans [
10]
, genetic and biochemical studies have identified the related SoxR protein as a major regulator in
P. pantotrophus and
P. salicylatoxidans [
27,
28,
29], both of which contain a complete Sox system and are unable to oxidize sulfane sulfur in the cytoplasm.
SoxR is a member of the arsenic repressor (ArsR-SmtB) family of prokaryotic repressors [
30,
31]. Members of the ArsR-SmtB family were originally recognized as metal-responsive transcriptional regulators, but there are also members in this family that have been shown to sense reactive oxygen or sulfur species [
32]. SqrR from
Rhodobacter capsulatus and BigR from
Xylella fastidiosa belong to this group and control the transcription of genes involved in sulfide-dependent photosynthesis and the detoxification of H
2S derived from associated host plants, respectively [
33,
34]. Knowledge about SoxR is comparatively sparse. While binding regions for the transcriptional repressor have been identified in promoter-operator segments within the
sox gene clusters of
P. denitrificans and
P. salicylatoxidans [
27,
28], no information is available on factors that control its DNA-binding capacity. It is therefore completely unclear how SoxR senses the presence of oxidizable sulfur compounds and how it then triggers the transcription of sulfur oxidation genes.
Here, we start to close this knowledge gap by first providing information on the general distribution of complete and truncated Sox systems and their co-occurrence with SoxR. Furthermore, we present genetic information for SoxR function in H. denitrificans, identify target genes and map its binding sites. The DNA-binding properties of the homodimeric repressor and its response to bridging of the sulfur atoms of two conserved cysteine residues by one to three sulfur atoms are characterized via site-directed mutagenesis, mass spectrometry, MalPEG assays, and electrophoretic mobility shift assays (EMSA).
2. Materials and Methods
2.1. Bacterial strains, plasmids, primers, and growth conditions
Table S1 lists the bacterial strains, and plasmids that were used for this study.
Escherichia coli strains were grown on complex lysogeny broth (LB) medium [
35] under aerobic conditions at 37°C unless otherwise indicated.
Escherichia coli. BL21 (DE3) was used for recombinant protein production.
E. coli strains 10-beta and DH5
α were used for molecular cloning.
H. denitrificans strains were cultivated in minimal media containing 24.4 mM methanol kept at pH 7.2 with 100 mM 3-(
N-Morpholino)propanesulfonic acid (MOPS) buffer as described before [
9]. Thiosulfate was added as needed. Antibiotics for
E. coli and
H. denitrificans were used at the following concentrations (in μg ml
-1): ampicillin, 100; kanamycin, 50; streptomycin, 200; chloramphenicol, 25.
2.2. Recombinant DNA techniques
Standard techniques for DNA manipulation and cloning were used unless otherwise indicated [
36]. Restriction enzymes, T4 ligase and Q5 polymerase were obtained from New England Biolabs (Ipswich, UK) and used according to the manufacturer’s instructions. Oligonucleotides for cloning were obtained from Eurofins MWG (Ebersberg, Germany). Plasmid DNA from
E. coli was purified using the GenJET Plasmid Miniprep kit (Thermo Scientific, Waltham, USA). Chromosomal DNA from
H. denitrificans strains was prepared using the First-DNA all-tissue Kit (GEN-IAL GmbH, Troisdorf, Germany).
2.3. Construction of plasmid for deletion of soxR in H. denitrificans
For markerless deletion of the
H. denitrificans soxR (Hden_0700) gene by splicing overlap extension (SOE) [
37], PCR fragments were constructed using the primers P1 fwd up hden_0700, P2 rev up hden_0700, P3 fwd down hden 0700 and P4 rev down hden_0700 (
Table S1). The resulting 1.04 kb SOE PCR fragment was cloned into the XbaI and PstI sites of pK18
mobscaB-Tc [
10]. The final construct pK18
mobsacB_Tc_Δs
oxR was electroporated into
H. denitrificans Δ
tsdA and transformants were selected using previously published procedures [
9,
11]. Single crossover recombinants were Cm
r and Tc
r. Double crossover recombinants were Tc
s and survived in the presence of sucrose due to loss of both, the vector-encoded levansucrase (SacB) and the tetracyclin resistance gene.
2.4. Characterization of phenotypes, quantification of sulfur compounds and protein content
Growth experiments with
H. denitrificans were run in 200 ml medium with 24.4 mM methanol and varying concentrations of thiosulfate in 500-ml Erlenmeyer flasks as described in [
10]. Thiosulfate concentrations, protein content and specific thiosulfate oxidation rates were determined by previously described methods [
10,
38]. All growth experiments were repeated three times. Representative experiments with two biological replicates for each strain are shown. All quantifications are based on at least three technical replicates.
2.5. RNA preparation
Total RNA of
H. denitrificans was isolated from cells grown mid-log phase.
H. denitrificans strains Δ
tsdA and Δ
tsdA ΔsoxR were grown in 50 ml methanol-containing medium at 30°C with shaking at 250 rpm in 100-ml Erlenmeyer flasks. Cells from 2 ml were harvested by centrifugation at 16,000 × g for 5 min. The cell pellet was incubated with 500 µl of 10% SDS containing 1 mg ml
-1 lysozyme at room temperature for 5 min. Then 700 µl of TRIzol [
39] was added and the mixture was incubated for another 5 min. This step was followed by addition of 1ml ROTI
®Aqua-P/C/I reagent (Carl Roth GmbH, Karlsruhe, Germany), 10 min incubation and centrifugation at 13,000 × g for 5 min. RNA purification from the supernatant was achieved with the Monarch Total RNA Miniprep Kit (NEB, Frankfurt, Germany). gDNA was removed by treating 10-µl samples with an absorption at 260 nm corresponding to ~1 µg RNA with 1 U of RNase-free DNase I (ThermoFisher, Waltham, MA, USA) in the MgCl
2-containing reaction buffer provided by the manufacturer. RNA concentrations were measured with an Eppendorf NanoDrop Biospectrometer. The absence of gDNA was verified using the primers rpoB-denitf and rpoB-denitr [
40], which bind only to gDNA and not to the corresponding RNA.
2.6. Expression studies based on RT-qPCR
RNA samples of 100 ng were used for RT-qPCR analysis via the Luna Universal One-Step RT-qPCR Kit (NEB) and the CFX Connect
TM real-time detection system (Bio-Rad, Munich, Germany) according to the instructions of the manufacturers. The level of
rpoB mRNA was used as internal standard [
40]. Approximately 200-bp fragments were amplified (see
Table S1 in the supplemental material) with an annealing temperature of 60°C. The RT-qPCR conditions were as follows: 10 min at 55°C (reverse transcription using random nonamer primers), 1 min at 95°C (inactivation of the reverse transcriptase and activation of the polymerase), 40 cycles of 15 s at 95°C, 30 s at 60°C, followed by melting curve analysis, in which the temperature was increased every 10 s by 1°C, from a start at 60°C to 95°C. Analyses of melting curves and calculation of C
t (calculated threshold) values were automatically quantified with The Bio-rad CFX Manager software. C
t values for each point in time were run in triplicate. Relative expression ratios were calculated by the 2
-ΔΔCt method [
41].
2.7. Cloning, site-directed mutagenesis, overproduction, and purification of recombinant SoxR proteins
The
soxR gene was amplified from
H. denitrificans genomic DNA with primers adding a sequence for an N-terminal Strep-tag and cloned between the NdeI and HindIII sites of pET-22b(+), resulting in pET22b-SoxR-Strep. Cysteine to serine exchanges were implemented with the Q5 Site-Directed Mutagenesis Kit (NEB) according to the manufacturer’s instructions and using the primers listed in
Table S1. Recombinant SoxR proteins were overproduced in
E. coli BL21(DE3) containing plasmids pET22b-SoxR-Strep, pET-22b-SoxR C
50S, pET-22b-SoxR C
116S, and pET-22b-SoxR C
50S C
116S. The cells were grown in 1-l Erlenmeyer flasks at 37°C in 400 ml LB medium containing ampicillin up to an OD600 of 0.5-0.6. Expression of
soxR was induced by adding 0.5 mM IPTG. IPTG-induced
E. coli cells were grown over night at 20 °C. Cells were harvested at 14,000 × g for 30 min. Three ml lysis buffer (100 mM Tris-HCl buffer pH 7.0, 5 mM EDTA containing a spatula tip of deoxyribonuclease I and protease inhibitor) were added per g wet weight for homogenization. Cell lysis was achieved by sonification and followed by centrifugation (16,100 × g, 30 min, and 4°C) and ultracentrifugation (145,000 × g, 1 h, 4°C). The supernatant was applied to a Strep-tactin affinity chromatography column equilibrated with buffer W (100 mM Tris-HCl, pH 8.0, 150 mM NaCl). The column was washed with six volumes of buffer W and eluted with buffer E (100 mM Tris-HCl, pH 8.0, 150 mM NaCl, 2.5 mM D-desthiobiotin). The protein was assessed for its purity by 12.5 % SDS-PAGE. Pure SoxR proteins were stored on ice in buffer W. Buffer exchange was achieved with Amicon
® Ultra-3K centrifugal filters.
2.8. Electrophoretic mobility shift assays (EMSA)
Gel electrophoretic mobility shift assays are used to detect interactions between proteins and nucleic acids. In the assay, solutions of protein and nucleic acid are combined and the resulting mixtures are subjected to polyacrylamide under native conditions. After electrophoresis, the distribution of nucleic acid species is determined. In general, protein-nucleic acid complexes migrate more slowly than the corresponding free nucleic acid [
39]. The binding reaction mixture (15 μl final volume), contained purified SoxR wildtype or variant protein in various concentrations (up to 700 nM), 2 μl 50 % glycerol and 1.5 μl 10 × binding buffer (100 mM Tris-HCl, 500 mM KCl, 10 mM DTT, 5 % glycerol, pH 8.0). Reaction mixtures were pre-incubated for 20 min at room temperature followed by a further 30 min incubation at 30°C after adding the DNA probe to a final concentration of 17 nM. The DNA probes consisted of a 362-bp fragment covering the entire intergenic region between the
shdrR (Hden_0682) and the
soxT1A (Hden_0681) genes, a 180-bp fragment representing the central part of the first product (created with primers EMSA-Fr2-Fr and EMSA_Fr3-Rev), a 177-bp fragment situated between the
shdrR and the
lipS1 (Hden_0683) gene, a 173-bp fragment situated between the
lipX (Hden_0687) and
dsrE3C (Hden_0688) genes, a 176-bp fragment located between the
tusA (Hden_0698) and
hyp (Hden_0697) genes, and a 151-bp fragment situated between the
soxA (Hden_0703) and
soxY (Hden_0704) genes. All primers used are listed in
Table S1. The reaction mixtures were loaded onto 6 % native polyacrylamide gels after these had been pre-run at 100 V for 1 h at 4 °C with 0.25 × TBE buffer (25 mM Tris-borate, 0.5 mM EDTA). The loaded gels were electrophoresed in 0.25 × TBE with 0.5 % glycerol at 180 V for 1h at 4 °C). Gels were subsequently stained for 20 min with SYBR green I. The bands corresponding to SoxR-bound and free DNAs were visualized with a ChemiDoc Imaging System (BioRad).
2.9. Gel permeation chromatography
The size exclusion chromatography column Superdex™ 75 Increase 10/300 GL (Cytiva, Freiburg, Germany) was calibrated using Blue dextran (2,000 kDa), conalbumin (75 kDa), bovine serum albumin (67 kDa), ovalbumin (43 kDa), lactoglobulin (35 kDa), carbonic anhydrase (29 kDa) chymotrypsin (23 kDa) and ribonuclease (13.7 kDa). The calibration curve was plotted using the gel-phase distribution coefficient (kav) versus the logarithm of molecular weight. kav = (Ve-V0/Vc-V0) where Ve = elution volume, V0 = column void volume (7.94 ml based on Blue dextran elution volume), Vc geometric column volume (24 ml). The column was run in 50 mM Tris-HCl, pH 8.0, 150 mM NaCl at a flow rate of 0.8 ml min-1 using an Äkta FPLC system.
2.10. Preparation of polysulfides
A polysulfide stock solution was prepared according to Ikeda et al. [
42] by mixing 1.2 g NaHS × H
20 and 0.16 g sulfur powder with 3 ml oxygen-free water in a closed 10 ml serum bottle under a nitrogen atmosphere for 1 h at room temperature. Then, the volume was filled up to 10 ml with oxygen-free water. Based on an average length for the resulting polysulfides of four sulfur atoms, their concentration is 0.5 M in the final solution that can be kept at room temperature for many months. If necessary, the polysulfide solution was diluted with oxygen-free water and immediately used for persulfuration reactions.
2.11. Redox treatments, persulfuration reactions, MalPEG gel-shift assays and mass spectrometry
5 µg protein was treated with DTT (1 mM and 5 mM for samples analyzed by mass spectrometry and EMSA, respectively) for reduction, 5 mM CuC12 for oxidation, 0.5 mM polysulfide for persulfuration and 1 mM MalPEG (methoxy-polyethylene glycol maleimide, MW 10000 g mol-1) for PEGylation or 5 mM iodoacetamide for carbamidomethylation in a final volume of 15 µl containing 100 mM Tris-HCl, pH 8.0, 150 mM NaCl. When polysulfide, MalPEG and DTT were applied consecutively, concentrations were 0.5 mM, 5 mM and 1 mM, respectively. When polysulfide and DTT were applied consecutively concentrations were 0.5 mM and 10 mM, respectively. Protein samples used in EMSA experiments were reacted with the reagents for 20 min at 25°C. Samples analyzed by SDS-PAGE were incubated with each reagent for 15 min at 30°C. Reactions were either stopped by addition of 5 µl 4 x non-reducing Roti®-Load2 (Carl Roth GmbH) and subjected to 15% SDS–PAGE without boiling the sample or analyzed by mass spectrometry. For MS, samples of 20 µl were desalted by ZiptipC4 Pipette tips (Merck Millipore, Darmstadt, Germany) and measured by MALDI-TOF at the Core Facility Protein synthesis & BioAnalytics, Pharmaceutical Institute, University of Bonn.
2.12. Distribution of Sox systems and SoxR: dataset generation and anaylsis
Archaeal and bacterial genomes were downloaded from Genome Taxonomy Database (GTDB, release R207). In GTDB, all genomes are sorted according to validly published taxonomies, they are pre-validated and have high quality (completeness minus 5*contamination must be higher than 50%). One representative of each of the current 65,703 species clusters was analyzed. It should be noted that GTDB is built on recently standardized bacterial and archaeal taxonomies derived by normalization of the evolutionary distance between taxonomic levels [
43,
44]. Among the bacteria, 148 phyla are currently distinguished. For the Archaea, GTDB lists 16 phyla. Open reading frames were determined using Prodigal [
45] and subsequently annotated for SoxR, other Sox proteins and clustering of the respective genes via HMS-S-S using default conditions [
46]. Chromatiaceae and Ectothiorhodospiraceae were treated as exceptions as they do not contain contiguous
sox clusters but the thiosulfate-oxidizing capabilities and the functionality of the Sox proteins have been experimentally established for relevant species [
47,
48].
4. Discussion
In this study, we collected a wealth of new information on the transcriptional repressor SoxR. We show that among the investigated more than 70,000 prokaryotic genomes bona-fide soxR (i.e., genetically linked to the genes for the SoxYZ sulfur-binding protein and/or catalytic Sox components) occurs exclusively among the bacterial phylum Proteobacteria, where it is more frequently found in gene clusters for complete than for truncated Sox systems. Based on the available data, it is difficult to draw general conclusions from this observation. However, it appears that a number of bacteria that operate the truncated Sox system, such as the green and purple sulfur bacteria or members of the Aquificota are dedicated sulfur-oxidizing chemolithoautrophs without much need for sophisticated transcriptional regulation of the sulfur oxidation machinery.
We show that in the model Alphaproteobacterium
H. denitrificans SoxR is not only involved in the transcriptional regulation of true
sox genes but that it also affects the transcription of a number of other genes. In particular, the
shdr genes, which encode the cytoplasmic sulfur-oxidizing multi-enzyme system required for sulfane sulfur oxidation that cannot be achieved by the truncated hyphomicrobial Sox system, are co-controlled by SoxR. How it interacts with a second, related repressor, sHdrR, that affects the transcription of the same genes [
10] is an important research question for the future.
The expression levels of
sox as well as of
shdr and associated genes are increased by thiosulfate in wildtype cells and elevated in the
soxR-deficient
H. denitrificans mutant irrespective of the presence of thiosulfate (
Figure 4). DNA binding in vitro and probably also transcriptional repression in living cells involve thiol modifications. This can be concluded from the observation that the DNA-binding activity of recombinant SoxR is strongly reduced upon incubation with polysulfide which leads to persulfuration of the regulator as proven by reaction with MalPEG (
Figure 9) and mass spectrometry (
Table 1,
Supplementary Figure S1). In polysulfide-treated SoxR, the two conserved cysteine residues can neither be modified by MalPEG nor by iodoacetamide. In addition, polysulfide treatment increases the mass of wildtype SoxR by 32, 64 or 96 Da. These findings can be fully explained by the formation of an intramolecular tri-, tetra- or pentasulfide bond formed upon interaction with reactive sulfane sulfur species. Thus, SoxR clearly is not a simple redox sensor switching between dithiol and disulfide state but has been identified as a transcriptional regulator sensing reactive sulfane sulfur species (
Figure 10), similar to the related SqrR protein from
R. capsulatus [
34,
67]. Notably, the substitution of the two crucial conserved cysteine residues leads to a different outcome for SoxR as compared to SqrR: The lack of Cys
50 causes complete loss of DNA binding in SoxR, whereas lack of Cys
116 creates a variant that tightly binds to its target DNA and is insensitive to persulfuration. In SqrR, both equivalent Cys-Ser variants are DNA-binding competent and do not respond to persulfuration as a signal [
34]. Clearly, this difference inspires future research that should also include detailed inspection of the conformational changes triggered by formation of a sulfur bridge and resulting in detachment of SoxR from its target DNA (
Figure 10).
The physiological processes involving the various sulfane sulfur-responsive regulators characterized so far [
33,
34,
68,
69,
70,
71] differ fundamentally from that controlled by SoxR. The former mainly regulate stress responses, sense intracellular and extracellular reactive sulfur species and ensure upregulation of H
2S oxidation genes for the purpose of detoxification, i.e., they control the removal of excess sulfide and sulfane sulfur thus contributing to cell survival in the presence of external reactive sulfur species. In contrast, SoxR regulates dissimilatory sulfur metabolism and enables the use of reduced sulfur compounds such as thiosulfate as electron donors for lithotrophic or mixotrophic growth.
As pointed out earlier, thiosulfate oxidation is initiated in the periplasm and it is highly unlikely that thiosulfate itself serves as the signalling molecule. Instead, SoxR responds to the presence of low concentrations of sulfane sulfur, which was provided as polysulfide in our in vitro assays. A working hypothesis for how this signal reaches its destination inspired by the arrangement of the respective genes in
H. denitrificans (
Figure 1b) is presented in
Figure 10. It is conceivable that the sulfur bound to the sulfur carrier protein SoxYZ in the periplasm in the course of thiosulfate oxidation reaches the cytoplasm via a YedE-like SoxT transporter [
72]. The periplasmic thiol-disulfide oxidoreductase SoxS [
16] could be involved in the transfer of the sulfane sulfur to the transporter. Once in the cytoplasm, the sulfur transferase TusA [
73] is a possible acceptor protein for the sulfur, which could be passed on from there to SoxR, possibly involving the cytochrome P450 encoded by gene Hden_0697.
In conclusion, our study shows that SoxR allows H. denitrificans to adapt to changes in thiosulfate availability via thiol persulfidation chemistry and the formation of an intramolecular sulfur bridge, which may involve transporters and sulfurtransferases encoded in the same genetic island. Clearly, much remains to be learned about this regulator not only in terms of signal transduction but also in terms of crosstalk with its counterpart sHdrR.
Figure 2.
Occurrence of complete and trunctated Sox systems among five bacterial phyla. Simultaneous presence of SoxR is indicated in black.
Figure 2.
Occurrence of complete and trunctated Sox systems among five bacterial phyla. Simultaneous presence of SoxR is indicated in black.
Figure 3.
Growth and thiosulfate consumption of H. denitrificans ΔtsdA (black circles and lines), ΔtsdA ΔshdrR (black triangles and lines) and ΔtsdA ΔsoxR (red boxes and lines). cultures were grown on methanol-containing medium (24.4 mM methanol) without (open symbols) or with 2 mM thiosulfate (filled symbols). Precultures contained either no thiosulfate (not induced, broken lines) or 2 mM thiosulfate (induced, solid lines). In the lower panels, thiosulfate concentrations for the different cultures are given. Symbol assignment is the same as in the upper panels. Specific thiosulfate (TS) oxidation rates are depicted in the same color code. Error bars indicating SD are too small to be visible for determination of biomass.
Figure 3.
Growth and thiosulfate consumption of H. denitrificans ΔtsdA (black circles and lines), ΔtsdA ΔshdrR (black triangles and lines) and ΔtsdA ΔsoxR (red boxes and lines). cultures were grown on methanol-containing medium (24.4 mM methanol) without (open symbols) or with 2 mM thiosulfate (filled symbols). Precultures contained either no thiosulfate (not induced, broken lines) or 2 mM thiosulfate (induced, solid lines). In the lower panels, thiosulfate concentrations for the different cultures are given. Symbol assignment is the same as in the upper panels. Specific thiosulfate (TS) oxidation rates are depicted in the same color code. Error bars indicating SD are too small to be visible for determination of biomass.
Figure 4.
(
a) Relative mRNA levels of twelve genes located in the
shdr-sox genetic island (depicted in panel (
b)) from
H. denitrificans for the Δ
tsdA reference strain in the absence (gray columns) and presence of thiosulfate (white columns), as assessed by RT-qPCR. Results for
H. denitrificans Δ
tsdA Δ
soxR are shown by black columns. Results were adjusted using
H. denitrifcans rpoB, which encodes the β-subunit of RNA polymerase, as an endogenous reference according to [
40]. (
b) DNA regions tested in EMSA assays for SoxR binding are indicated as black rectangles below the hyphomicrobial
shdr-sox genetic island. Fragment sizes: 362 bp for the
soxT1A-shdrR intergenic region, 177 bp and 173 bp for the regions upstream of
lipS1 and
dsrE3C, respectively. The fragments downstream of
tusA and between
soxA and
soxY had sizes of 176 bp and 151 bp, respectively. (
c) EMSA analysis of Strep-tagged SoxR with upstream promoter sequence probes of sulfur oxidation related genes as specified in (
b). 17 nM DNA probes were incubated with different amounts of SoxR (300 and 700 nM).
Figure 4.
(
a) Relative mRNA levels of twelve genes located in the
shdr-sox genetic island (depicted in panel (
b)) from
H. denitrificans for the Δ
tsdA reference strain in the absence (gray columns) and presence of thiosulfate (white columns), as assessed by RT-qPCR. Results for
H. denitrificans Δ
tsdA Δ
soxR are shown by black columns. Results were adjusted using
H. denitrifcans rpoB, which encodes the β-subunit of RNA polymerase, as an endogenous reference according to [
40]. (
b) DNA regions tested in EMSA assays for SoxR binding are indicated as black rectangles below the hyphomicrobial
shdr-sox genetic island. Fragment sizes: 362 bp for the
soxT1A-shdrR intergenic region, 177 bp and 173 bp for the regions upstream of
lipS1 and
dsrE3C, respectively. The fragments downstream of
tusA and between
soxA and
soxY had sizes of 176 bp and 151 bp, respectively. (
c) EMSA analysis of Strep-tagged SoxR with upstream promoter sequence probes of sulfur oxidation related genes as specified in (
b). 17 nM DNA probes were incubated with different amounts of SoxR (300 and 700 nM).
Figure 5.
Amino sequence alignment of SoxR homologs. Accession numbers/locus tags and references in the order of appearance: (Hden_0700 [
9], Hden_0682, [
9,
10], WP_010893290 [
33], HLYU_VIBCH [
64], WP_019171658 [
28], ADE85198 [
34], b2667 [
65].
Figure 5.
Amino sequence alignment of SoxR homologs. Accession numbers/locus tags and references in the order of appearance: (Hden_0700 [
9], Hden_0682, [
9,
10], WP_010893290 [
33], HLYU_VIBCH [
64], WP_019171658 [
28], ADE85198 [
34], b2667 [
65].
Figure 6.
Conformation of SoxR and its variants as analyzed by (panels a and b) non-reducing SDS-PAGE analysis and (c) gel permeation chromatography. For the experiments shown in panels a and b 5 µg of SoxR or its variants were incubated in 15 µl 100 mM Tris-HCl, pH 8.0, 150 mM NaCl with either 1 mM DTT or 5 mM Cucl2 for 20 min at room temperature, mixed with 5 µl non-reducing Roti®-Load2 (Carl Roth GmbH) and run on 15% SDS polyacrylamide gels. The wildtype SoxR protein is shown in panels a and b to allow direct comparison with protein variants on different gels. In (c) the elution profiles upon gel filtration on Superdex 75 Increase 10/300 are depicted for SoxR, solid line; SoxR Cys50Ser, dotted line; SoxR Cys116Ser, dashed line; SoxR Cys50Ser Cys116Ser, dashed-dotted line. SoxR and SoxR Cys116Ser dimers elute at a kav of 0.2 corresponding to a molecular mass of 36.7 kDa, whereas SoxR Cys50Ser and SoxR Cys50Ser Cys116Ser elute earlier (kav = 0.174, 41.9 kDa) indicating a more open conformation. The resolution of the column does not allow clear separation of the different tetrameric conformations (kav 0.086 to 0.093, corresponding to 65.9 to 63.6 kDa).
Figure 6.
Conformation of SoxR and its variants as analyzed by (panels a and b) non-reducing SDS-PAGE analysis and (c) gel permeation chromatography. For the experiments shown in panels a and b 5 µg of SoxR or its variants were incubated in 15 µl 100 mM Tris-HCl, pH 8.0, 150 mM NaCl with either 1 mM DTT or 5 mM Cucl2 for 20 min at room temperature, mixed with 5 µl non-reducing Roti®-Load2 (Carl Roth GmbH) and run on 15% SDS polyacrylamide gels. The wildtype SoxR protein is shown in panels a and b to allow direct comparison with protein variants on different gels. In (c) the elution profiles upon gel filtration on Superdex 75 Increase 10/300 are depicted for SoxR, solid line; SoxR Cys50Ser, dotted line; SoxR Cys116Ser, dashed line; SoxR Cys50Ser Cys116Ser, dashed-dotted line. SoxR and SoxR Cys116Ser dimers elute at a kav of 0.2 corresponding to a molecular mass of 36.7 kDa, whereas SoxR Cys50Ser and SoxR Cys50Ser Cys116Ser elute earlier (kav = 0.174, 41.9 kDa) indicating a more open conformation. The resolution of the column does not allow clear separation of the different tetrameric conformations (kav 0.086 to 0.093, corresponding to 65.9 to 63.6 kDa).
Figure 7.
(a) EMSA of the 151-bp soxA-soxY intergenic fragment with increasing amounts of untreated (upper panel) SoxR or SoxR pre-incubated with polysulfide in a molar ratio SoxR/polysulfide of 1:1 (lower panel), (b) EMSA of the 151-bp soxA-soxY intergenic fragment with SoxR pre-incubated with increasing amounts of polysulfide (upper panel) or thiosulfate (lower panel).
Figure 7.
(a) EMSA of the 151-bp soxA-soxY intergenic fragment with increasing amounts of untreated (upper panel) SoxR or SoxR pre-incubated with polysulfide in a molar ratio SoxR/polysulfide of 1:1 (lower panel), (b) EMSA of the 151-bp soxA-soxY intergenic fragment with SoxR pre-incubated with increasing amounts of polysulfide (upper panel) or thiosulfate (lower panel).
Figure 8.
(a) EMSA of the 151-bp soxA-soxY intergenic fragment (17 nM) with 700 nM SoxR wildtype and variant proteins as isolated, reduced with DTT, oxidized with CuCl2, treated with polysulfide and sequentially treated with polysulfide and DTT. (b) EMSA of the 180 bp central part of the soxT1A-shdrR intergenic fragment (17 nM) with 300 nM SoxR wildtype and variant proteins as isolated, oxidized with CuCl2 or treated with polysulfide.
Figure 8.
(a) EMSA of the 151-bp soxA-soxY intergenic fragment (17 nM) with 700 nM SoxR wildtype and variant proteins as isolated, reduced with DTT, oxidized with CuCl2, treated with polysulfide and sequentially treated with polysulfide and DTT. (b) EMSA of the 180 bp central part of the soxT1A-shdrR intergenic fragment (17 nM) with 300 nM SoxR wildtype and variant proteins as isolated, oxidized with CuCl2 or treated with polysulfide.
Figure 9.
Analysis of H. denitrificans SoxR cysteines with MalPEG gel-shift assays in non-reducing SDS-PAGE. Results are shown for the wildtype, single (Cys50Ser and Cys116Ser) and double (Cys50Ser Cys116Ser) mutants after MalPEG treatment of the as-isolated, reduced and oxidized states as well as after pre-incubation with polysulfide. Polysulfide and MalPEG reacted samples were furthermore reduced with DTT.
Figure 9.
Analysis of H. denitrificans SoxR cysteines with MalPEG gel-shift assays in non-reducing SDS-PAGE. Results are shown for the wildtype, single (Cys50Ser and Cys116Ser) and double (Cys50Ser Cys116Ser) mutants after MalPEG treatment of the as-isolated, reduced and oxidized states as well as after pre-incubation with polysulfide. Polysulfide and MalPEG reacted samples were furthermore reduced with DTT.
Figure 10.
Proposed signal transduction pathway and mode of action of the homodimeric SoxR repressor protein. Established sulfur-binding proteins are printed in black.
Figure 10.
Proposed signal transduction pathway and mode of action of the homodimeric SoxR repressor protein. Established sulfur-binding proteins are printed in black.
Table 1.
Mass spectrometry of SoxR and variants after treatment with modifying agents. CAM, carbamidomethylation; S, sulfur. Calculated masses for Strep-Tagged SoxR and SoxR Cys50Ser, SoxR Cys116Ser and SoxR Cys50Ser Cys116 without the initiator methionine are 15212.54 Da, 15197.54 Da, 15197.54 Da and 15182.54 Da, respectively.
Table 1.
Mass spectrometry of SoxR and variants after treatment with modifying agents. CAM, carbamidomethylation; S, sulfur. Calculated masses for Strep-Tagged SoxR and SoxR Cys50Ser, SoxR Cys116Ser and SoxR Cys50Ser Cys116 without the initiator methionine are 15212.54 Da, 15197.54 Da, 15197.54 Da and 15182.54 Da, respectively.
Treatment |
SoxR Mass [Da] (addition: [Da]) |
SoxR C50S Mass [Da] (addition: [Da]) |
SoxR C116S Mass [Da] (addition: [Da]) |
SoxR C50S C116S Mass [Da] (addition: [Da]) |
Native |
15212.8 |
15197.3 |
15198.2 |
15182.3 |
DTT reduced |
15212.5 |
15199.3 |
15198.8 |
nd |
CuCl2 oxidized |
15210.5 |
15196.7 |
15196.9 |
15182.0 |
Iodoacetamide |
15328.0 (2 CAM: 2 × 57.07) |
15255.2 (1 CAM: 57.07) |
15255.22 (1 CAM: 57.07) |
nd |
Polysulfide |
15212.5 15244.7 (1 S: 32) 15276.4 (2 S: 64) 15308.0 (3 S: 96) |
15198.9 15230.0 (1 S: 32) |
15198.0 15230.3 (1 S: 32) 15261.1 (2 S: 64) |
15182.0 |
Polysulfide + Iodoacetamide |
15212.9 15244.3 (1 S: 32) 15275.2 (2 S: 64) 15306.0 (3 S: 96) |
15198.0 15286.0 (1 S + 1 CAM: 90) |
15197.6 15228.5 (1 S: 32) 15285.4 (1 S + 1 CAM: 90) |
15180.5 |