*3.2. Cu2+ Promotes the Dimerization of RegA Aerobically*

*3.2. Cu2+ Promotes the Dimerization of RegA Aerobically* Since Cu2+ binds reduced RegA in a 0.5:1 ratio, we next investigated whether Cu2+ was capable of stimulating dimerization of RegA. For this analysis, monomeric RegA was loaded with Cu2+ via buffer exchange aerobically and immediately run through a nonreducing SDS-PAGE gel with no DTT added to the samples. As shown in Figure 2, the Since Cu2+ binds reduced RegA in a 0.5:1 ratio, we next investigated whether Cu2+ was capable of stimulating dimerization of RegA. For this analysis, monomeric RegA was loaded with Cu2+ via buffer exchange aerobically and immediately run through a non-reducing SDS-PAGE gel with no DTT added to the samples. As shown in Figure 2, the addition of Cu2+ indeed promotes the dimerization of RegA. As observed from the non-reducing SDS-PAGE gel in Figure 3A, Cu2+ mediated dimerization of RegA is also

addition of Cu2+ indeed promotes the dimerization of RegA. As observed from the non-

A.

B.

quite rapid, with the majority of RegA dimerized after only one minute of Cu2+ treatment in the presence of air. In contrast, exposure of RegA to air in the absence of Cu2+ did not lead to any significant dimerization even after 4 h of exposure, confirming that Cu2+ in the presence of oxygen serves a critical role in mediating the dimerization of RegA (Figure 3B). rapid, with the majority of RegA dimerized after only one minute of Cu2+ treatment in the presence of air. In contrast, exposure of RegA to air in the absence of Cu2+ did not lead to any significant dimerization even after 4 h of exposure, confirming that Cu2+ in the presence of oxygen serves a critical role in mediating the dimerization of RegA (Figure 3B). Dimer 75 50 37

rapid, with the majority of RegA dimerized after only one minute of Cu2+ treatment in the presence of air. In contrast, exposure of RegA to air in the absence of Cu2+ did not lead to any significant dimerization even after 4 h of exposure, confirming that Cu2+ in the presence of oxygen serves a critical role in mediating the dimerization of RegA (Figure 3B).

*Microorganisms* **2022**, *10*, x FOR PEER REVIEW 5 of 13

*Microorganisms* **2022**, *10*, x FOR PEER REVIEW 5 of 13

**Figure 2.** Cu2+ indispensably promotes the dimerization of RegA. (**A**) Non-reducing SDS-PAGE gel of apo-RegA and RegA treated with divalent metals. + denotes addition of RegA protein whereas – denotes where metal is not added. (**B**) Histogram of RegA dimer:monomer ratios determined by **Figure 2.** Cu2+ indispensably promotes the dimerization of RegA. (**A**) Non-reducing SDS-PAGE gel of apo-RegA and RegA treated with divalent metals. + denotes addition of RegA protein whereas –denotes where metal is not added. (**B**) Histogram of RegA dimer:monomer ratios determined by image J. and Mn2+ following the same protocol. The non-reducing SDS-PAGE gel results in Figure 2 show that none of the other six divalent metals tested stimulated RegA dimerization, indicating that only Cu2+ is capable of promoting dimerization of RegA.

0min 1min 10min 30min 1h 2h 4h **Marker** Time 0min 1min 10min 30min 1h 2h 4h **Marker** Dimer/monomer0.6 0.8 1 0.6 0.8 1 Dimer/monomer C. D. **Figure 3.** Copper catalyzes the dimerization of RegA likely with the help of oxygen. (**A**) Non-reducing SDS-PAGE gel of RegA treated with Cu2+ in the air after different time periods. (**B**) Non-reducing SDS-PAGE gel of RegA in the absence of Cu2+ in the air after different time periods. (**C**) Dimer:monomer ratio of RegA treated with Cu2+ in the air after different time periods. (**D**) Dimer:monomer **Figure 3.** Copper catalyzes the dimerization of RegA likely with the help of oxygen. (**A**) Non-reducing SDS-PAGE gel of RegA treated with Cu2+ in the air after different time periods. (**B**) Non-reducing SDS-PAGE gel of RegA in the absence of Cu2+ in the air after different time periods. (**C**) Dimer:monomer ratio of RegA treated with Cu2+ in the air after different time periods. (**D**) Dimer:monomer ratio of RegA in the absence of Cu2+ in the air after different time periods. The dimer:monomer ratio was calculated based on the intensity of bands quantified through Image J v1.51.

(min)

**Figure 3.** Copper catalyzes the dimerization of RegA likely with the help of oxygen. (**A**) Non-reducing SDS-PAGE gel of RegA treated with Cu2+ in the air after different time periods. (**B**) Non-reducing SDS-PAGE gel of RegA in the absence of Cu2+ in the air after different time periods. (**C**) Dimer:monomer ratio of RegA treated with Cu2+ in the air after different time periods. (**D**) Dimer:monomer

0 0.2 0.4

0 50 100 150 200 250

50 37

25 20KDa

0 0.2 0.4

To determine whether Cu2+ uniquely plays a role in promoting dimerization of RegA, we also treated RegA with six other divalent metals including Mg2+, Ca2+, Co2+, Ni2+, Zn2+ and Mn2+ following the same protocol. The non-reducing SDS-PAGE gel results in Figure 2 show that none of the other six divalent metals tested stimulated RegA dimerization, indicating that only Cu2+ is capable of promoting dimerization of RegA. *Microorganisms* **2022**, *10*, 0 6 of 13 *Microorganisms* **2022**, *10*, x FOR PEER REVIEW 6 of 13 ratio of RegA in the absence of Cu2+ in the air after different time periods. The dimer:monomer ratio was calculated based on the intensity of bands quantified through Image J v1.51.

#### *3.3. RegA Dimerization Requires Both Cu2+ and Molecular Oxygen 3.3. RegA Dimerization Requires Both Cu2+ and Molecular Oxygen 3.3. RegA Dimerization Requires Both Cu2+ and Molecular Oxygen*

We next addressed whether molecular oxygen is a requirement for Cu2+ dimerization of RegA. For this analysis, we tested whether Cu2+ can promote the dimerization of RegA under anaerobic conditions by treating RegA monomers with Cu2+ in an anaerobic chamber with degassed buffers for 10 min. The reactions were quenched with EDTA (or passed through filters) to remove Cu2+ prior to analysis of the monomer/dimer state of RegA using electrospray ionization mass spectrometry (ESI-MS). As shown in Figure 4, the RegA anaerobically treated with Cu2+ and a monomer control without Cu2+ treatment both showed identical ESI-MS profiles exhibiting a monomer molecular weight of 20,207 daltons. This indicates that both Cu2+ and atmospheric levels of molecular oxygen are necessary to promote the dimerization of RegA. We next addressed whether molecular oxygen is a requirement for Cu2+ dimerization of RegA. For this analysis, we tested whether Cu2+ can promote the dimerization of RegA under anaerobic conditions by treating RegA monomers with Cu2+ in an anaerobic chamber with degassed buffers for 10 min. The reactions were quenched with EDTA (or passed through filters) to remove Cu2+ prior to analysis of the monomer/dimer state of RegA using electrospray ionization mass spectrometry (ESI-MS). As shown in Figure 4, the RegA anaerobically treated with Cu2+ and a monomer control without Cu2+ treatment both showed identical ESI-MS profiles exhibiting a monomer molecular weight of 20,207 daltons. This indicates that both Cu2+ and atmospheric levels of molecular oxygen are necessary to promote the dimerization of RegA. We next addressed whether molecular oxygen is a requirement for Cu2+ dimerization of RegA. For this analysis, we tested whether Cu2+ can promote the dimerization of RegA under anaerobic conditions by treating RegA monomers with Cu2+ in an anaerobic chamber with degassed buffers for 10 min. The reactions were quenched with EDTA (or passed through filters) to remove Cu2+ prior to analysis ofthe monomer/dimer state of RegAusing electrospray ionization mass spectrometry (ESI-MS). As shown in Figure 4, the RegA anaerobically treated with Cu2+ and a monomer control without Cu2+ treatment both showed identical ESI-MS profiles exhibiting a monomer molecular weight of 20,207 daltons. This indicates that both Cu2+ and atmospheric levels of molecular oxygen are necessary to promote the dimerization of RegA.

**Figure 4.** ESI-MS spectrum of RegA in the presence or absence of Cu2+ anaerobically. (**A**) RegA molecular weight determination by ESI-MS (20,207 daltons) in the presence of Cu2+ without oxygen. (**B**) RegA molecular weight determination by ESI-MS (20,207 daltons) in the absence of Cu2+ without oxygen. **Figure 4.** ESI-MS spectrum of RegA in the presence or absence of Cu2+ anaerobically. (**A**) RegA molecular weight determination by ESI-MS (20,207 daltons) in the presence of Cu2+ without oxygen. (**B**) RegA molecular weight determination by ESI-MS (20,207 daltons) in the absence of Cu2+ without oxygen. **Figure 4.** ESI-MS spectrum of RegA in the presence or absence of Cu2+ anaerobically. (**A**) RegA molecular weight determination by ESI-MS (20,207 daltons) in the presence of Cu2+ without oxygen. (**B**) RegA molecular weight determination by ESI-MS (20,207 daltons) in the absence of Cu2+ without oxygen.

We next addressed whether the dimer formed by Cu2+ treated RegA involves a covalent disulfide bond or instead is a non-covalent interaction. For this analysis, Cu2+ + O<sup>2</sup> mediated RegA dimer was isolated away from monomer RegA by gel filtration chromatography and then treated with either DTT or EDTA alone or both. As is indicated in Figure 5, the addition of excess EDTA, which would remove Cu2+ ,did not affect dimerization. This is contrasted by the conversion of RegA dimers to its monomeric form by the addition DTT, which would indicate that Cu2+ likely stimulates disulfide bond formation. For confirmation, we also undertook liquid chromatography–tandem mass spectrometry (LC-MS/MS) of trypsin-digested Cu2+-formed RegA dimer, which showed the presence of a peptide fragment VYELCDR containing an intermolecular disulfide bond at Cys156 (Figure 5B). Interestingly, RegA only has one cysteine residue at the 156th position, with this cysteine residue conserved among all the α-proteobacteria that contain RegA homologs. Finally, analysis of the presence of Cu in the isolated RegA dimers with ICP-MS shows no significant amounts of Cu present in covalently linked dimerized RegA. We next addressed whether the dimer formed by Cu2+ treated RegA involves a covalent disulfide bond or instead is a non-covalent interaction. For this analysis, Cu2+ + O<sup>2</sup> mediated RegA dimer was isolated away from monomer RegA by gel filtration chromatography and then treated with either DTT or EDTA alone or both. As is indicated in Figure 5, the addition of excess EDTA, which would remove Cu2+, did not affect dimerization. This is contrasted by the conversion of RegA dimers to its monomeric form by the addition DTT, which would indicate that Cu2+ likely stimulates disulfide bond formation. For confirmation, we also undertook liquid chromatography–tandem mass spectrometry (LC-MS/MS) of trypsin-digested Cu2+-formed RegA dimer, which showed the presence of a peptide fragment VYELCDR containing an intermolecular disulfide bond at Cys156 (Figure 5B). Interestingly, RegA only has one cysteine residue at the 156th position, with this cysteine residue conserved among all the α-proteobacteria that contain RegA homologs. Finally, analysis of the presence of Cu in the isolated RegA dimers with ICP-MS shows no significant amounts of Cu present in covalently linked dimerized RegA. We next addressed whether the dimer formed by Cu2+ treated RegA involves a covalent disulfide bond or instead is a non-covalent interaction. For this analysis, Cu2+ + O<sup>2</sup> mediated RegA dimer was isolated away from monomer RegA by gel filtration chromatography and then treated with either DTT or EDTA alone or both. As is indicated in Figure 5, the addition of excess EDTA, which would remove Cu2+, did not affect dimerization. This is contrasted by the conversion of RegA dimers to its monomeric form by the addition DTT, which would indicate that Cu2+ likely stimulates disulfide bond formation. For confirmation, we also undertook liquid chromatography–tandem mass spectrometry (LC-MS/MS) of trypsin-digested Cu2+-formed RegA dimer, which showed the presence of a peptide fragment VYELCDR containing an intermolecular disulfide bond at Cys156 (Figure 5B). Interestingly, RegA only has one cysteine residue at the 156th position, with this cysteine residue conserved among all the α-proteobacteria that contain RegA homologs. Finally, analysis of the presence of Cu in the isolated RegA dimers with ICP-MS shows no significant amounts of Cu present in covalently linked dimerized RegA.

**Figure 5.** Cu2+ can promote RegA to dimerize with the formation of a disulfide bond in vitro. (**A**) Non-reducing SDS-PAGE gel of Cu2+ treated RegA dimer treated with EDTA or DTT. (**B**) Disulfide bond formation in Cu2+treated RegA dimer observed in LC-MS/MS. **Figure 5.** Cu2+ can promote RegA to dimerize with the formation of a disulfide bond in vitro. (**A**) Non-reducing SDS-PAGE gel of Cu2+ treated RegA dimer treated with EDTA or DTT. (**B**) Disulfide bond formation in Cu2+treated RegA dimer observed in LC-MS/MS.

#### *3.4. Covalently Dimerized RegA Is Capable of Binding DNA 3.4. Covalently Dimerized RegA Is Capable of Binding DNA*

Phosphorylation of response regulators by cognate sensor kinases is known to stimulate dimerization, which facilitates DNA binding. Given that unphosphorylated RegA monomers have low DNA binding activity [20], we next addressed whether RegA dimers formed by disulfide bond formation are also capable of binding to DNA. For this analysis, we performed DNase I footprint assays with Cu2+ stimulated RegA disulfide dimer to the *ccoN* promoter region that controls cytochrome *cbb<sup>3</sup>* oxidase expression. This promoter was chosen for analysis based on previous transcriptomic analyses, which showed that RegA regulates the expression of the *ccoN* promoter under aerobic conditions when RegB Phosphorylation of response regulators by cognate sensor kinases is known to stimulate dimerization, which facilitates DNA binding. Given that unphosphorylated RegA monomers have low DNA binding activity [20], we next addressed whether RegA dimers formed by disulfide bond formation are also capable of binding to DNA. For this analysis, we performed DNase I footprint assays with Cu2+ stimulated RegA disulfide dimer to the *ccoN* promoter region that controls cytochrome *cbb*<sup>3</sup> oxidase expression. This promoter was chosen for analysis based on previous transcriptomic analyses, which showed that RegA regulates the expression of the *ccoN* promoter under aerobic conditions when RegB kinase activity is suppressed [3,10].

kinase activity is suppressed [3,10]. Footprint results shown in Figure 6 demonstrate that covalently dimerized unphosphorylated RegA can bind to the *ccoN* promoter region, as exhibited by excellent DNase I peak suppression with 2 µM of dimerized RegA. Two binding sites highlighted in red are readily observed on both the forward and the reverse strands (Figure 6B,D, respectively). The two forward strand protection sites observed with Cu2+ treated RegA dimers (−80 bp to −76 bp and −74 bp to −49 bp upstream of the transcription initiation site, respectively) are at the same location as previously reported with a constitutively dimerized and active RegA variant called RegA\* [3]. On the reverse strand, RegA dimers also bind at two binding sites (−74 bp to −69 bp and −66 bp to −52 bp upstream of the transcription initiation site), which are just slightly narrower than the binding site observed with RegA\* dimers (−74 bp to −49 bp) [3]. Footprint analysis with varying amounts of dimerized RegA shows that Cu2+ stimulated RegA dimer binds to the ccoN promoter with an apparent Kd of 1 μM ~ 4-fold lower than that observed with RegA\* (0.25 μM) [3]. These results demonstrate that covalently dimerized unphosphorylated RegA indeed binds to the *ccoN* promoter at just a slightly lower affinity than phosphorylated noncovalent dimers do. Footprint results shown in Figure 6 demonstrate that covalently dimerized unphosphorylated RegA can bind to the *ccoN* promoter region, as exhibited by excellent DNase I peak suppression with 2 µM of dimerized RegA. Two binding sites highlighted in red are readily observed on both the forward and the reverse strands (Figure 6B,D, respectively). The two forward strand protection sites observed with Cu2+ treated RegA dimers (−80 bp to −76 bp and −74 bp to −49 bp upstream of the transcription initiation site, respectively) are at the same location as previously reported with a constitutively dimerized and active RegA variant called RegA\* [3]. On the reverse strand, RegA dimers also bind at two binding sites (−74 bp to −69 bp and −66 bp to −52 bp upstream of the transcription initiation site), which are just slightly narrower than the binding site observed with RegA\* dimers (−74 bp to −49 bp) [3]. Footprint analysis with varying amounts of dimerized RegA shows that Cu2+ stimulated RegA dimer binds to the ccoN promoter with an apparent Kd of 1 µM ~ 4-fold lower than that observed with RegA\* (0.25 µM) [3]. These results demonstrate that covalently dimerized unphosphorylated RegA indeed binds to the *ccoN* promoter at just a slightly lower affinity than phosphorylated noncovalent dimers do.

5' - GATGCGACCGTTCGTCGCACCGT - 3'

**Figure 6.** DNaseI footprinting of Cu2+-treated RegA dimer to the *ccoN* promoter. (**A**) DNase I digestion of the *ccoN* promoter top strand without RegA. (**B**) DNase I digestion of the *ccoN* promoter top strand with 2 µM Cu2+ treated unphosphorylated RegA dimer. (**C**) DNase I digestion of the *ccoN* promoter bottom strand without RegA. (**D**) DNase I digestion of the *ccoN* promoter bottom strand with 2 µM Cu2+ treated unphosphorylated RegA dimer. Areas of RegA protection are highlighted **Figure 6.** DNaseI footprinting of Cu2+-treated RegA dimer to the *ccoN* promoter. (**A**) DNase I digestion of the *ccoN* promoter top strand without RegA. (**B**) DNase I digestion of the *ccoN* promoter top strand with 2 µM Cu2+ treated unphosphorylated RegA dimer. (**C**) DNase I digestion of the *ccoN* promoter bottom strand without RegA. (**D**) DNase I digestion of the *ccoN* promoter bottom strand with 2 µM Cu2+ treated unphosphorylated RegA dimer. Areas of RegA protection are highlighted with a black bar.

#### **4. Discussion**

with a black bar.

**4. Discussion**

The response regulator paradigm is that phosphorylation of a response regulator by its cognate histidine kinase promoted dimerization and subsequent DNA binding activity. This study shows that RegA from *R. capsulatus* binds copper and that bound Cu2+ in the presence of O<sup>2</sup> leads to intermolecular disulfide bond formation. Covalently dimerized RegA is capable of binding DNA at sub μM concentrations, which likely provides RegA the ability to aerobically regulate gene expression in the absence of phosphorylation. The response regulator paradigm is that phosphorylation of a response regulator by its cognate histidine kinase promoted dimerization and subsequent DNA binding activity. This study shows that RegA from *R. capsulatus* binds copper and that bound Cu2+ in the presence of O<sup>2</sup> leads to intermolecular disulfide bond formation. Covalently dimerized RegA is capable of binding DNA at sub µM concentrations, which likely provides RegA the ability to aerobically regulate gene expression in the absence of phosphorylation.

The formation of disulfide bonds has been adopted by several transcription factors to control their DNA binding activities in response to redox changes. Some well-characterized examples are OxyR, which, in response to hydrogen peroxide or S-nitrosothiols, induces the expression of over ten genes encoding proteins [21] against oxidative and nitrosative stress; and *oxyS*, a gene that encodes a nontranslated RNA involved in DNA repair. In both cases, exposure to oxygen stimulates disulfide bond oxidation that affects DNA binding activity [22]. In *R. capsulatus*, CrtJ is an aerobic repressor of photosystem genes that has regulatory cysteins that are oxidized in the presence of molecular oxygen The formation of disulfide bonds has been adopted by several transcription factors to control their DNA binding activities in response to redox changes. Some well-characterized examples are OxyR, which, in response to hydrogen peroxide or S-nitrosothiols, induces the expression of over ten genes encoding proteins [21] against oxidative and nitrosative stress; and *oxyS*, a gene that encodes a nontranslated RNA involved in DNA repair. In both cases, exposure to oxygen stimulates disulfide bond oxidation that affects DNA binding activity [22]. In *R. capsulatus*, CrtJ is an aerobic repressor of photosystem genes that has regulatory cysteins that are oxidized in the presence of molecular oxygen [23–25]. Besides

the use of disulfide bond formation, other transcription factors have utilized changes in the redox state of iron to sense the presence or absence of oxygen. For example, FNR, a global regulator of over a hundred genes in *E. coli*, contains a [4Fe-4S] iron–sulfur cluster that disassembles under aerobic conditions, which disrupts dimerization and subsequent binding to DNA [21,26,27]. In the case of RegA, not only does copper itself become oxidized by oxygen from Cu+1 to Cu+2, but, in addition, the presence of both Cu2+ and oxygen are needed for disulfide bond formation. In some respects, this may be regarded as a hybrid of the above-described mechanisms of redox sensing. example, FNR, a global regulator of over a hundred genes in *E. coli*, contains a [4Fe-4S] iron–sulfur cluster that disassembles under aerobic conditions, which disrupts dimerization and subsequent binding to DNA [21,26,27]. In the case of RegA, not only does copper itself become oxidized by oxygen from Cu+1 to Cu+2 , but, in addition, the presence of both Cu2+ and oxygen are needed for disulfide bond formation. In some respects, this may be regarded as a hybrid of the above-described mechanisms of redox sensing. *4.1. A Model Describing How Copper and Oxygen Promote RegA Disulfide Bond Formation*

[23–25]. Besides the use of disulfide bond formation, other transcription factors have utilized changes in the redox state of iron to sense the presence or absence of oxygen. For

*Microorganisms* **2022**, *10*, x FOR PEER REVIEW 9 of 13

#### *4.1. A Model Describing How Copper and Oxygen Promote RegA Disulfide Bond Formation* Taking all of our results into consideration, we can propose a model of how Cu2+ and

Taking all of our results into consideration, we can propose a model of how Cu2+ and oxygen likely promote covalent dimerization of RegA as follows (Figure 7): (i) anaerobically, RegA monomer binds Cu1+, possibly by accepting Cu1+ transfer from an intracellular copper chaperone such as from CopZ [28] or other copper chaperones. (ii) When the environment changes from anaerobic to aerobic conditions, oxygen oxidizes RegA bound Cu1+ into Cu2+. Cu2+ with the help of O<sup>2</sup> then promotes subsequent disulfide bond formation. (iii) Disulfide bond dimerized RegA then binds to a subset of promoters. oxygen likely promote covalent dimerization of RegA as follows (Figure 7): (i) anaerobically, RegA monomer binds Cu1+, possibly by accepting Cu1+ transfer from an intracellular copper chaperone such as from CopZ [28] or other copper chaperones. (ii) When the environment changes from anaerobic to aerobic conditions, oxygen oxidizes RegA bound Cu1+ into Cu2+. Cu2+ with the help of O2 then promotes subsequent disulfide bond formation. (iii) Disulfide bond dimerized RegA then binds to a subset of promoters.

**Figure 7.** Proposed model of the mechanism of how Cu2+ promotes the dimerization of RegA in **Figure 7.** Proposed model of the mechanism of how Cu2+ promotes the dimerization of RegA in vivo.

vivo*.* Recently, Alphabet's DeepMind released the artificial intelligence program AlphaFold that predicts a protein's tertiary structures based on a protein's primary sequence [29]. Analysis of a putative RegA structure derived from AlphaFold shows two well-defined domains (Figure 8). The largest right-side domain in Figure 8 is very similar to response regulator receiver domains that have previously been solved by X-ray crystallography [1]. This phosphorylatable receiver domain is linked by a long, flexible linker region containing five prolines to a DNA binding domain shown on the left. The two smallest left-most helices exhibit a high degree of structural similarity to the RegA helix-turn-helix DNA binding helices as previously solved by NMR [17]. The larger helix after the flexible linker region just before the two small DNA binding helices contains the universally conserved Cys156 as shown in yellow. Given the flexibility of this domain from the receiver domain, this Cys should be available for oxidation into an intermolecular disulfide by molecular oxygen. We suspect that the role of Cu2+ may be a structural role where it binds Recently, Alphabet's DeepMind released the artificial intelligence program AlphaFold that predicts a protein's tertiary structures based on a protein's primary sequence [29]. Analysis of a putative RegA structure derived from AlphaFold shows two well-defined domains (Figure 8). The largest right-side domain in Figure 8 is very similar to response regulator receiver domains that have previously been solved by X-ray crystallography [1]. This phosphorylatable receiver domain is linked by a long, flexible linker region containing five prolines to a DNA binding domain shown on the left. The two smallest left-most helices exhibit a high degree of structural similarity to the RegA helix-turn-helix DNA binding helices as previously solved by NMR [17]. The larger helix after the flexible linker region just before the two small DNA binding helices contains the universally conserved Cys156 as shown in yellow. Given the flexibility of this domain from the receiver domain, this Cys should be available for oxidation into an intermolecular disulfide by molecular oxygen. We suspect that the role of Cu2+ may be a structural role where it binds and holds two RegA monomers in such a position that this helix from two RegA's are positioned in close proximity, thereby allowing oxygen to effectively promote disulfide bond formation.

and holds two RegA monomers in such a position that this helix from two RegA's are positioned in close proximity, thereby allowing oxygen to effectively promote disulfide

bond formation.

**Figure 8.** RegA tertiary structure as predicted from the DeepMind released artificial intelligence program AlphaFold. The DNA binding helix-turn-helix domain is on the left and the receiver domain is on the right. Cys156 that undergoes disulfide bond formation is highlighted in yellow. **Figure 8.** RegA tertiary structure as predicted from the DeepMind released artificial intelligence program AlphaFold. The DNA binding helix-turn-helix domain is on the left and the receiver domain is on the right. Cys156 that undergoes disulfide bond formation is highlighted in yellow.

Due to the capability of copper ions to undergo reversible oxidation of Cu1+ to Cu2+ , copper acts as an essential cofactor for several Cu metalloenzymes such as cytochrome *cbb<sup>3</sup>* oxidase encoded by the *ccoN* operon that RegA aerobically and anaerobically regulates and superoxide dismutase [30]. However, copper unbound to protein is also quite toxic. For example, Cu1+ tends to mediate the generation of radical oxygen species (ROS) through Fenton reaction [31,32], while Cu2+ can nucleate protein aggregation, leading to amyloid formation in mammals [30]. Hence, the transportation and distribution of copper is precisely regulated [33]. Apart from copper importers and exporters, copper ions are often delivered to copper-required proteins through copper chaperones [34] in the form of Cu1+. It is therefore plausible that one or several copper chaperones may directly interact and transfer copper to RegA. SenC is unlikely to itself deliver copper to RegA as SenC's copper binding domain is located in the periplasm [14,15]. Besides SenC [13–15], the identified copper chaperones in *Rhodobacter* include a Cu1+ chaperone Cox11 [35] in *Rhodobacter sphaeroides*, a PCuAC-like periplasmic chaperone PccA [15], and a recently identified intracellular Cu1+ chaperone CopZ [28], in *Rhodobacter capsulatus*. Since RegA is an intracellular protein, and the copper-binding site of SenC is in the periplasm [14], we suspect that cytoplasmic CopZ most likely transfers Cu1+ to RegA in *R. capsulatus*. Due to the capability of copper ions to undergo reversible oxidation of Cu1+ to Cu2+ , copper acts as an essential cofactor for several Cu metalloenzymes such as cytochrome *cbb*<sup>3</sup> oxidase encoded by the *ccoN* operon that RegA aerobically and anaerobically regulates and superoxide dismutase [30]. However, copper unbound to protein is also quite toxic. For example, Cu1+ tends to mediate the generation of radical oxygen species (ROS) through Fenton reaction [31,32], while Cu2+ can nucleate protein aggregation, leading to amyloid formation in mammals [30]. Hence, the transportation and distribution of copper is precisely regulated [33]. Apart from copper importers and exporters, copper ions are often delivered to copper-required proteins through copper chaperones [34] in the form of Cu1+. It is therefore plausible that one or several copper chaperones may directly interact and transfer copper to RegA. SenC is unlikely to itself deliver copper to RegA as SenC's copper binding domain is located in the periplasm [14,15]. Besides SenC [13–15], the identified copper chaperones in *Rhodobacter* include a Cu1+ chaperone Cox11 [35] in *Rhodobacter sphaeroides*, a PCuAC-like periplasmic chaperone PccA [15], and a recently identified intracellular Cu1+ chaperone CopZ [28], in *Rhodobacter capsulatus*. Since RegA is an intracellular protein, and the copper-binding site of SenC is in the periplasm [14], we suspect that cytoplasmic CopZ most likely transfers Cu1+ to RegA in *R. capsulatus*.

#### *4.2. Both Aerobic and Anaerobic Conditions Control RegA Activity 4.2. Both Aerobic and Anaerobic Conditions Control RegA Activity*

Most previous studies have focused on the role of RegA, and its cognate sensor kinase RegB, in controlling anaerobic gene expression [8,9]. These studies demonstrated that the phosphorylation activity of RegB is controlled by both redox changes in the ubiquinone pool as well as by oxidation of a conserved cystine (Figure 9) [4–7]. RegB has six membrane-spanning helices, three of which are known to be involved in binding both oxidized and reduced ubiquinone [4,5,36]. When bound to oxidized ubiquinone, the kinase activity is low, whereas when bound to reduced ubiquinol, the kinase activity is high [4] (Figure 9). Additionally, there is a redox-active Cys in the four-helix bundle that links the membrane-spanning domain to the kinase domain [6,7]. In the presence of oxygen, this Cys undergoes oxidation either to a sulfonic acid derivative or to a disulfide, either of which inhibits kinase activity (Figure 9) [6,7]. Most previous studies have focused on the role of RegA, and its cognate sensor kinase RegB, in controlling anaerobic gene expression [8,9]. These studies demonstrated that the phosphorylation activity of RegB is controlled by both redox changes in the ubiquinone pool as well as by oxidation of a conserved cystine (Figure 9) [4–7]. RegB has six membranespanning helices, three of which are known to be involved in binding both oxidized and reduced ubiquinone [4,5,36]. When bound to oxidized ubiquinone, the kinase activity is low, whereas when bound to reduced ubiquinol, the kinase activity is high [4] (Figure 9). Additionally, there is a redox-active Cys in the four-helix bundle that links the membranespanning domain to the kinase domain [6,7]. In the presence of oxygen, this Cys undergoes oxidation either to a sulfonic acid derivative or to a disulfide, either of which inhibits kinase activity (Figure 9) [6,7].

**Figure 9.** Model of RegB/RegA two-component system in sensing redox changes and regulating gene expressions anaerobically (**left**) and hypothetic model of how RegA regulates gene expressions **Figure 9.** Model of RegB/RegA two-component system in sensing redox changes and regulating gene expressions anaerobically (**left**) and hypothetic model of how RegA regulates gene expressions aerobically (**right**).

Evidence that RegA itself may undergo redox control was recently revealed when it was observed that RegA controls the expression of ~47 genes under aerobic growth conditions when RegB kinase activity is inhibited [10]. Interestingly, all 47 genes that are aerobically regulated by unphosphorylated RegA also belong to the group of 707 genes that are anaerobically regulated by phosphorylated RegA [10]. We suspect that there must be a subtle difference between aerobic disulfide dimerized unphosphorylated RegA and anaerobic phosphorylated RegA that differentiates which promoters that these two variations of RegA function with. For example, the binding of Cu2+ and O<sup>2</sup> dimerized RegA to the *ccoN* promoter appears to be slightly lower than that observed with the constitutively active variant RegA\*. Even such a subtle difference in binding affinity may have significant effects on the promoters that RegA controls. There could also be differences by which these two differently dimerized variants interact with RNA polymerase. Further additional comparative studies will have to be undertaken to reveal how promoter activity is Evidence that RegA itself may undergo redox control was recently revealed when it was observed that RegA controls the expression of ~47 genes under aerobic growth conditions when RegB kinase activity is inhibited [10]. Interestingly, all 47 genes that are aerobically regulated by unphosphorylated RegA also belong to the group of 707 genes that are anaerobically regulated by phosphorylated RegA [10]. We suspect that there must be a subtle difference between aerobic disulfide dimerized unphosphorylated RegA and anaerobic phosphorylated RegA that differentiates which promoters that these two variations of RegA function with. For example, the binding of Cu2+ and O<sup>2</sup> dimerized RegA to the *ccoN* promoter appears to be slightly lower than that observed with the constitutively active variant RegA\*. Even such a subtle difference in binding affinity may have significant effects on the promoters that RegA controls. There could also be differences by which these two differently dimerized variants interact with RNA polymerase. Further additional comparative studies will have to be undertaken to reveal how promoter activity is differentially regulated by these variants.

differentially regulated by these variants. **Author Contributions:** Conceptualization, N.K., C.E.B.; methodology, N.K., C.E.B.; formal analysis, N.K., C.E.B.; investigation, N.K., C.E.B. resources, C.E.B.; writing—original draft preparation, N.K.; writing—review and editing, C.E.B.; supervision, C.E.B.; project administration, C.E.B.; funding acquisition, C.E.B. All authors have read and agreed to the published version of the manuscript. **Author Contributions:** Conceptualization, N.K. and C.E.B.; methodology, N.K. and C.E.B.; formal analysis, N.K. and C.E.B.; investigation, N.K. and C.E.B. resources, C.E.B.; writing—original draft preparation, N.K.; writing—review and editing, C.E.B.; supervision, C.E.B.; project administration, C.E.B.; funding acquisition, C.E.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by The National Institutes of Health grant GM040941 awarded to C.E.B. **Funding:** This research was funded by The National Institutes of Health grant GM040941 awarded to C.E.B.

**Data Availability Statement:** Not applicable. **Data Availability Statement:** Not applicable.

sensor kinase RegB. *J. Biol. Chem.* **2006**, *281*, 6768–6775. https://doi.org/10.1074/jbc.m509687200.

aerobically (**right**).

**Acknowledgments:** We thank Jiefei Wang and Jordan, Matthew Robert from Giedroc D. P.'s lab for ICP-MS analysis, Giovannei Gonzalez-Gutierrez for Deep mind structural analysis of RegA, and Jonathan Trinidad for MS-MS analysis. A special thanks to Jiangchuan Shen for giving advice on buffer exchange and how to work with redox active metals. **Acknowledgments:** We thank Jiefei Wang and Jordan, Matthew Robert from Giedroc D. P.'s lab for ICP-MS analysis, Giovannei Gonzalez-Gutierrez for Deep mind structural analysis of RegA, and Jonathan Trinidad for MS-MS analysis. A special thanks to Jiangchuan Shen for giving advice on buffer exchange and how to work with redox active metals.

**Conflicts of Interest:** The authors declare no conflict of interest. **Conflicts of Interest:** The authors declare no conflict of interest.

#### **References References**

