*Article* **The Response Regulator RegA Is a Copper Binding Protein That Covalently Dimerizes When Exposed to Oxygen**

**Nijia Ke and Carl E. Bauer \***

Molecular and Cellular Biochemistry Department, Indiana University, Bloomington, IN 47405, USA; nijke@iu.edu **\*** Correspondence: bauer@iu.edu; Tel.: +1-812-855-6595

**Abstract:** In *Rhodobacter capsulatus*, the histidine kinase RegB is believed to phosphorylate its cognate transcriptional factor RegA only under anaerobic conditions. However, transcriptome evidence indicates that RegA regulates 47 genes involved in energy storage, energy production, signaling and transcription, under aerobic conditions. In this study, we provide evidence that RegA is a copper binding protein and that copper promotes the dimerization of RegA under aerobic conditions. Inductively coupled plasma mass spectrometry (ICP-MS) analysis indicates that RegA binds Cu1+ and Cu2+ in a 1:1 and 2:1 ratio, respectively. Through LC-MS/MS, ESI-MS and non-reducing SDS-PAGE gels, we show that Cu2+ stimulates disulfide bond formation in RegA at Cys156 in the presence of oxygen. Finally, we used DNase I footprint analysis to demonstrate that Cu2+-mediated covalent dimerized RegA is capable of binding to the *ccoN* promoter, which drives the expression of cytochrome *cbb*<sup>3</sup> oxidase subunits. This study provides a new model of aerobic regulation of gene expression by RegA involving the formation of an intermolecular disulfide bond.

**Keywords:** two-component system; RegA; aerobic; copper ion; disulfide bond; DNA binding; gene regulation

#### **1. Introduction**

Just as Ser/Thr/Tyr phosphorylation pathways are among the most common signaling pathways in eukaryotes, two-component systems (TCSs) constitute the most common signal transduction pathway in prokaryotes [1]. Receptor sensor kinases in TCSs sense various stimuli ranging from physical conditions such as light, temperature, redox state, osmolarity and concentration of chemicals such as nutrients and quorum signals [2]. TCSs control numerous cellular activities involved in metabolism, mobility and virulence [1].

In the purple non-sulfur bacterium *Rhodobacter capsulatus*, the RegB/RegA twocomponent system regulates a variety of physiological processes such as photosynthesis, hydrogen fixation, respiration, carbon fixation, nitrogen assimilation, denitrification, hydrogen uptake and aerotaxis [3]. Previous results indicate that RegB senses redox change through two sensing mechanisms. First, RegB has a ubiquinol/ubiquinone binding site located within a membrane-spanning region that monitors the redox state of the ubiquinol pool [4,5]. The binding of oxidized ubiquinone inhibits, while the binding of reduced ubiquinol stimulates, RegB kinase activity [4]. RegB has low affinity to both oxidized ubiquinone and reduced ubiquinol, allowing RegB to readily adjust its activity to the redox state of the ubiquinone pool, not unlike that of a rheostat [4,5]. The second level of RegB control involves direct inhibition of RegB kinase activity by dioxygen via oxidation of a conserved cytosolic cysteine residue (Cys265). The oxidation of Cys265 results in both the formation of a stable sulfenic acid derivative and disulfide bond formation that promotes stable tetramerization of RegB with both oxidized RedB derivatives exhibiting inactive kinase activity [6,7]. Collectively, these results support a long-standing model where RegB only undergoes autophosphorylation and subsequent transfer of the phosphoryl group to its cognate response regulator RegA under anaerobic conditions. Phosphorylated RegA

**Citation:** Ke, N.; Bauer, C.E. The Response Regulator RegA Is a Copper Binding Protein That Covalently Dimerizes When Exposed to Oxygen. *Microorganisms* **2022**, *10*, 934. https://doi.org/10.3390/ microorganisms10050934

Academic Editors: Matthew Sattley and Robert Blankenship

Received: 30 March 2022 Accepted: 28 April 2022 Published: 29 April 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

then binds to DNA as a transcriptional factor, modulating the expression of genes involved in numerous anaerobic energy-generating and energy-utilizing processes [8,9]. While the described model has broad acceptance, RNA-seq data from our laboratory indicate that this model may be too simplistic as RegA is still capable of aerobically regulating ~47 genes, which is a condition in which RegA would not be phosphorylated by RegB [10]. We have therefore investigated whether RegA is capable of regulating gene expression under aerobic unphosphorylated conditions.

One clue for a possible alternative mechanism for controlling the activity of RegA in the absence of phosphorylation by RegB may be the transcriptional co-localization of *regB* and *regA* with *senC*. In species that contain the group 1 RegB/RegA system, *regA* and/or *regB* is invariably cotranscribed with *senC* [11,12]. SenC is a known copper-binding protein with proposed copper chaperone activity [13–15]. In *R. capsulatus*, SenC has a role in the assembly of copper containing *cbb*<sup>3</sup> cytochrome oxidase [13–15]. However, SenC also has sequence similarity to a family of oxidoreductases that have thiol-disulfide bond forming activity [16]. Furthermore, group 1 RegA orthologs also contain an invariant conserved Cys (Cys156 in *R. capsulatus*) located immediately upstream of a DNA binding helix-turn-helix domain [12,17]. Thus, we explored whether the activity of RegA could be modulated by copper and/or the oxidation of Cys156. Our results indicate that RegA is capable of binding Cu2+ and that exposure of RegA-Cu2+ to O<sup>2</sup> results in disulfide bond formation at Cys156, which stimulates dimerization and DNA binding.

#### **2. Materials and Methods**

#### *2.1. Overexpression of RegA*

The construction of a RegA overexpression vector follows the same method of construction of a RegA\* overexpression vector as described in Du et al. (1998) [18] with the exception that the 95th residue of the RegA\* codon was mutated from Alanine to Serine, which is the wild-type RegA codon. The resulting plasmid pET29CBD::*regA*, which also contains a chitin-binding domain fused to the C-terminal of RegA, was then transformed into the overexpression strain BL21(DE3) and grown overnight by shaking at 250 rpm at 37 ◦C. The overnight culture was then subcultured (1:50 ratio) into 1 L of Terrific Broth (TB) containing 50 µg/mL kanamycin. When the OD600nm reached 0.5 to 0.8, the expression of RegA was induced by the addition of 0.4 mL 1M IPTG per liter culture and then shaken at 250 rpm at 37 ◦C for 4 h. Cells were then harvested by centrifuging at 10,000× *g* for 15 min. Cell pellets were stored at −80 ◦C prior to RegA purification.

### *2.2. Purification of RegA from E. coli*

Frozen cell pellets from 2 L of RegA culture were resuspended in 50 mL lysis/wash buffer (20 mM Tris-HCl pH 8.0, 500 mM NaCl) and digested with 400 µL 2 mg/mL DNaseI via homogenization on ice and rotation at 4 ◦C for an hour. The cell lysate was then disrupted by three passages through an emulsifier and centrifuged at 17,210× *g* for 30 min. RegA containing an intein chitin-binding domain was purified by passing through a 10 mL chitin affinity column after equilibrating the column with 10 column volumes of lysis buffer, and then washed with another 10 column volumes of lysis/wash buffer. The intein chitin-binding domain was cleaved off RegA by purging the chitin affinity column with 3 column volumes of cleavage buffer (20 mM Tris-HCl pH 8.0, 500 mM NaCl, 30 mM DTT) and incubated at room temperature overnight. Tagless RegA was then eluted with 3 column volumes of elution buffer (20 mM Tris-HCl (pH 8.0), 500 mM NaCl). Finally, reduced RegA monomer was treated with gel filtration chromatography using a Superose 12 column equilibrated with 20 mM Tris-HCl pH 8.0, 250 mM NaCl, 5 mM TCEP, 10 mM EDTA and 10% glycerol. Finally, RegA was concentrated to 2 to 3 mg/mL and stored at −80 ◦C.

#### *2.3. Loading RegA with Metal through Buffer Exchange*

An aliquot of purified RegA was concentrated to 100 µL in a 500 µL centrifugal filter (Amicon, Burlington, MA, USA) with the 10 kDa cutoff by centrifuging at 17,000× *g* for 8 min. RegA was then buffer exchanged into 20 mM Tris-HCl pH 8.0 250 mM NaCl by repeating the steps of adding 400 µL fresh buffer and centrifuging at 17,000× *g* five times with final centrifugation run until the sample volume reached 100 µL. The concentration of RegA was then measured using the Bradford method. The RegA sample was then anaerobically loaded with metals by incubating for 5 min with 400 µL 20 mM Tris-HCl pH 8.0 250 mM NaCl containing 3× metal to 1× RegA's concentration. CuCl and CuSO<sup>4</sup> solutions were prepared inside the anaerobic chamber using degassed ddH2O for copper salt preparation. The protein buffer (20 mM Tris-HCl pH 8.0, 500 mM NaCl) used for buffer exchange was also degassed and purged with nitrogen before being brought into the anaerobic chamber. Excessive metal not bound to RegA was then removed from the buffer by repeating Amicon filter centrifugation eight times (maximum speed until ~100 µL concentrated sample was left) with each step involving the addition of 400 µL fresh buffer.

#### *2.4. ICP-MS*

The concentrations of protein were measured before analysis. At least 20 µL protein of 20 µM or 500 µL buffer was required to conduct inductive coupled plasma mass spectrometry (ICP-MS). Protein samples of RegA were diluted 100-fold with 2.5% HNO<sup>3</sup> (Sigma) into a final volume of 3 mL, while using Pure Plus Internal Standard Mix (100 ppb, PerkinElmer, Waltham, MA, USA) as an internal standard. Using hydrogen (55Mn, <sup>56</sup>Fe, <sup>59</sup>Co, <sup>60</sup>Ni detection) or helium (24Mg, <sup>63</sup>Cu, <sup>66</sup>Zn detection) as collision gases to remove possible interferences with the internal reference <sup>45</sup>Sc or <sup>72</sup>Ge, samples were analyzed with an Agilent 8800 QQQ ICP-MS device. Based on standard curves of all analyzed metals generated with Pure Plus Multi-Element Calibration Standard 3 (0.5–100 ppb, PerkinElmer), metal concentrations of protein or buffer samples were calculated.

#### *2.5. ESI-MS*

RegA were incubated with a gradient concentration of CuSO<sup>4</sup> in one-fold, three-fold, five-fold and ten-fold of RegA concentration, respectively. Measures of 30 µL of these Cu2+ treated RegA samples, apo-RegA and buffer were separately injected and sealed in glass vials and brought out of the anaerobic chamber. The electrospray ionization mass spectrometry (ESI-MS) spectra of the samples were generated using an LC (C<sup>4</sup> reverse phase)-MS (Synapt G2S HDMS) instrument and analyzed with Mass Lynx v4.1.

#### *2.6. LC-MS/MS*

Gel bands were diced into 1 mm cubes and incubated for 45 min at 57 ◦C. To maintain potential disulfide bonds, the samples were neither reduced nor alkylated. A solution containing 1 µg trypsin, in 25 mM ammonium bicarbonate was added and the samples were digested for 12 h at 37 ◦C. The resulting peptides were desalted using a ZipTip (Millipore, Billerica, MA, USA). The samples were dried down and injected into an Eksigent HPLC coupled to an LTQ Orbitrap XL (Thermo Fischer Scientific, Waltham, MA, USA). The peptides were separated using a 75 micron, 15 cm column packed in-house with C18 resin (Michrom Bioresources, Auburn, CA, USA) at a flow rate of 300 nl/min. A one-hour gradient was run from Buffer A (2% acetonitrile, 0.1% formic acid) to 60% Buffer B (100% acetonitrile, 0.1% formic acid). The Orbitrap was configured to acquire a survey scan over the mass range 300–2000 at a resolution of 30,000. This was followed by collision-induced dissociation mass spectrometry (CID MS/MS) on the top-five most-intense precursor ions above a threshold of 1000 counts.

The resulting MS/MS peaklists were searched against the Uniprot *Rhodobacter capsulatus* database using Protein Prospector (v5.10.14). Acetylation of the protein amino terminus, oxidation of methionine and pyroglutamine formation of peptide amino terminal glutamine and disulfide bond formation were set as variable modifications. A mass

tolerance of 20 ppm was used for precursor ions and 0.6 Da was used for fragment ions. Peptide expectation values were set to <0.05. a 299 bp DNA fragment containing the promoter region of *ccoN* was PCR amplified using a 6-FAM labeled forward primer and a HEX labeled reverse primer. A 50 nM DNA probe

terminus, oxidation of methionine and pyroglutamine formation of peptide amino terminal glutamine and disulfide bond formation were set as variable modifications. A mass tolerance of 20 ppm was used for precursor ions and 0.6 Da was used for fragment ions.

DNase I footprint analysis was performed as described by Willett et al. [19]. Briefly,

*Microorganisms* **2022**, *10*, x FOR PEER REVIEW 4 of 13

Peptide expectation values were set to <0.05.

#### *2.7. DNase I FootPrinting Assay* was incubated at room temperature for 15 min with a gradient concentration of purified

*2.7. DNase I FootPrinting Assay*

DNase I footprint analysis was performed as described by Willett et al. [19]. Briefly, a 299 bp DNA fragment containing the promoter region of *ccoN* was PCR amplified using a 6-FAM labeled forward primer and a HEX labeled reverse primer. A 50 nM DNA probe was incubated at room temperature for 15 min with a gradient concentration of purified proteins in 20 mM Tris-HCl pH 8.0, 2 mM MgCl2, 0.5 mM CaCl2, 0.1 mg/mL BSA of a total volume of 20 µL and then digested with 5 µL of DNaseI for another 15 min at room temperature. The reactions were quenched by the addition of 25 µL 0.5 M EDTA pH 8.0. The digested fragments were recovered using a Min Elute PCR purification kit (Qiagen, Hilden, Germany) and then eluted with 15 µL elution buffer. The samples were detected with a 3730 DNA Analyzer (Applied Biosystem, Waltham, MA, USA) using 500 LIZTM Size Standard and eventually analyzed with the Peak Scanner Software v2.0. proteins in 20 mM Tris-HCl pH 8.0, 2 mM MgCl2, 0.5 mM CaCl2, 0.1 mg/mL BSA of a total volume of 20 μL and then digested with 5 μL of DNaseI for another 15 min at room temperature. The reactions were quenched by the addition of 25 μL 0.5 M EDTA pH 8.0. The digested fragments were recovered using a Min Elute PCR purification kit (Qiagen, Hilden, Germany) and then eluted with 15 μL elution buffer. The samples were detected with a 3730 DNA Analyzer (Applied Biosystem, Waltham, MA, USA) using 500 LIZTM Size Standard and eventually analyzed with the Peak Scanner Software v2.0. **3. Results**

#### **3. Results** *3.1. RegA Binds Copper Anaerobically*

#### *3.1. RegA Binds Copper Anaerobically* We first addressed whether RegA can bind copper in the Cu1+ or Cu2+ states. Since

We first addressed whether RegA can bind copper in the Cu1+ or Cu2+ states. Since Cu1+ is only stable in the absence of oxygen (oxygen oxidizes Cu1+ to Cu2+), we assayed the ability of RegA to bind these two redox states of Cu under anaerobic conditions. For this analysis, we anaerobically exposed RegA to threefold molar excess of either Cu1+ or Cu2+ followed by removal of unbound Cu using centrifugal filtration mediated buffer exchange in an anaerobic chamber. After unbound Cu1+ and Cu2+ were removed, we measured the concentration of copper that remained bound to RegA using inductively coupled plasma mass spectrometry (ICP-MS). As shown in Figure 1, RegA anaerobically binds approximately stoichiometric amount of Cu1+, which is indicated by a copper ion over RegA monomer ratio of 1.02 <sup>±</sup> 0.21. In contrast, the ratio of Cu2+ over RegA is very close to 0.5 (0.45 ± 0.03), suggesting that two RegA's bind to one molecule of Cu2+ anaerobically. Taken together, these results indicate that RegA is indeed capable of binding copper. Cu1+ is only stable in the absence of oxygen (oxygen oxidizes Cu1+ to Cu2+), we assayed the ability of RegA to bind these two redox states of Cu under anaerobic conditions. For this analysis, we anaerobically exposed RegA to threefold molar excess of either Cu1+ or Cu2+ followed by removal of unbound Cu using centrifugal filtration mediated buffer exchange in an anaerobic chamber. After unbound Cu1+ and Cu2+ were removed, we measured the concentration of copper that remained bound to RegA using inductively coupled plasma mass spectrometry (ICP-MS). As shown in Figure 1, RegA anaerobically binds approximately stoichiometric amount of Cu1+, which is indicated by a copper ion over RegA monomer ratio of 1.02 ± 0.21. In contrast, the ratio of Cu2+ over RegA is very close to 0.5 (0.45 ± 0.03), suggesting that two RegA's bind to one molecule of Cu2+ anaerobically. Taken together, these results indicate that RegA is indeed capable of binding copper.

**Figure 1.** Ratio of concentration of copper ion over RegA anaerobically. **Figure 1.** Ratio of concentration of copper ion over RegA anaerobically.
