PgRsp Is a Novel Redox-Sensing Transcription Regulator Essential for Porphyromonas gingivalis Virulence

Porphyromonas gingivalis is one of the etiological agents of chronic periodontitis. Both heme and oxidative stress impact expression of genes responsible for its survival and virulence. Previously we showed that P. gingivalis ferric uptake regulator homolog affects expression of a gene encoding a putative Crp/Fnr superfamily member, termed P. gingivalis redox-sensing protein (PgRsp). Although PgRsp binds heme and shows the highest similarity to proteins assigned to the CooA family, it could be a member of a novel, separate family of proteins with unknown function. Expression of the pgrsp gene is autoregulated and iron/heme dependent. Genes encoding proteins engaged in the oxidative stress response were upregulated in the pgrsp mutant (TO11) strain compared with the wild-type strain. The TO11 strain showed higher biomass production, biofilm formation, and coaggregation ability with Tannerella forsythia and Prevotella intermedia. We suggest that PgRsp may regulate production of virulence factors, proteases, Hmu heme acquisition system, and FimA protein. Moreover, we observed growth retardation of the TO11 strain under oxidative conditions and decreased survival ability of the mutant cells inside macrophages. We conclude that PgRsp protein may play a role in the oxidative stress response using heme as a ligand for sensing changes in redox status, thus regulating the alternative pathway of the oxidative stress response alongside OxyR.


Introduction
Periodontal diseases, a group of the most abundant human infectious diseases [1][2][3][4], are characterized by reabsorption of tooth-supporting tissues caused by gingiva inflammation induced by virulence factors produced by periodontopathogens and the exaggerated response of the host immune system. There is also growing evidence supporting a link between periodontal diseases and several systemic diseases, such as cardiovascular and respiratory diseases, rheumatoid arthritis, diabetes, osteoporosis, and Alzheimer s disease [5][6][7][8][9][10]. Porphyromonas gingivalis, one of the agents involved in initiation and progression of chronic periodontitis [2][3][4]11], is a Gram-negative, obligatory anaerobic, asaccharolytic bacterium, residing in periodontal pockets of the human oral cavity.
P. gingivalis has developed numerous sophisticated mechanisms for the acquisition of nutrients, enabling its survival as well as its proliferation inside and spread between host cells [12][13][14][15][16]. Among the most important virulence factors are proteases, mainly lysine-and arginine-specific gingipains encoded by kgp, rgpA, and rgpB genes, which play a crucial role in the degradation of host proteins and tissue components [17][18][19]. Together with the heme-uptake Hmu system, gingipains engage in heme acquisition, which is essential for P. gingivalis survival due to the lack of ability to

Oxidative Stress Formation
Overnight cultures of P. gingivalis grown in BM + Hm medium were used to inoculate basal medium supplemented with 0.5 µg/mL menadione and 7.7 µM hemin chloride, with (BM + Cys + Hm) or without (BM-Cys + Hm) cysteine, at starting optical density at 600 nm (OD 600 )~0.2. Bacteria were grown with or without addition of 0.05 mM or 0.25 mM H 2 O 2 added at the culture starting point. Alternatively, H 2 O 2 was added into 4-h bacterial cultures.

Gene Expression Analysis
Gene expression was initially examined in bacteria cultured for 20 h in iron/heme rich (Hm) or depleted (DIP) medium using microarray analysis as reported previously [13,16]. Additionally, reverse transcriptase-quantitative polymerase chain reaction (RT-qPCR) was used as reported previously [16,40]. Total RNA was isolated from 5 × 10 7 to 4 × 10 8 P. gingivalis cells grown for 4, 10, and 24 h in iron/heme rich or iron/heme depleted medium, or one hour after exposure to oxidative stress using the Total RNA Mini Kit (A&A Biotechnology, Gdynia, Poland) and the Clean-Up RNA Concentrator Kit (A&A Biotechnology). Reverse transcription was carried out using total RNA and random hexamers with the SensiFAST cDNA Synthesis Kit (Bioline, London, UK). qPCR was performed using the SensiFAST SYBR No-ROX Kit (Bioline) and LightCycler 96 (Roche, Basel, Switzerland). All primers are listed in Table S1. Samples were analyzed in triplicate in two biological replicates and melting curves were generated to measure the quality of amplified products. Relative changes in gene expression were determined using LightCycler 96 software and P. gingivalis 16S rRNA (PGA7_00000960) as the reference gene.

Homotypic Biofilm Formation and Coaggregation Assay
P. gingivalis was grown in 96-well plates as described previously [41]. Briefly, fresh BM + Hm or BM + DIP medium was inoculated with the overnight cultures at starting OD 600~0 .2 and bacteria were grown for 24 h. Non-attached bacterial cells were washed out three times with 20 mM phosphate buffer, pH 7.4, containing 140 mM NaCl (PBS). Then, 1% crystal violet (Roth, Frederikssund, Denmark) was used to visualize the biofilm formed. Coaggregation assay was performed as described previously [16]. Briefly, overnight liquid cultures of P. gingivalis, P. intermedia, T. forsythia, and S. gordonii were centrifuged (8000× g, 20 min, 4 • C) and washed with PBS. Subsequently, optical density at 600 nm (OD 600 ) of the cultures was adjusted to 1.0 using PBS. The cultures of P. gingivalis and respective bacteria were mixed at a 1:1 ratio and incubated anaerobically for 6 h. The coaggregation rate was monitored by measurement of the decrease in OD 600 . Autoaggregation of the monocultures was used as a control.

Proteolytic Activity Assay
To determine total proteolytic activity of P. gingivalis cultures, azocasein (Sigma-Aldrich, St. Louis, MO, USA) as the proteolytic substrate was used as described previously [16]. Briefly, 50 µL of 2.5% azocasein solution and 30 µL of 0.5% NaHCO 3 solution (pH 8.3) were added to 20 µL of P. gingivalis culture samples, following by incubation at 37 • C for 30 min. The reaction was stopped by precipitation of undigested proteins with 400 µL of 5% trichloroacetic acid and then centrifuged (10,000× g, 5 min). Subsequently, 240 µL of 0.5 M NaOH was added to 400 µL of the supernatant and absorbance at 440 nm was measured. Proteolytic activity exhibited by the wild-type strain under given conditions was considered as 100%.

Cell Culture and Infection Assay
THP-1 (human acute monocytic leukemia) (ATCC TIB-202) cell line was obtained from the American Type Culture Collection (ATCC). THP-1cells were grown in 12-well plates in RPMI-1640 medium (BioWest, Nuaillé, France) with the addition of 2 mM L-glutamine (BioWest), 10% heat-inactivated fetal bovine serum (BioWest), 100 U/mL penicillin and 100 µg/mL streptomycin as described previously [13]. THP-1 cells were differentiated toward macrophages by supplementing the culture medium without antibiotics with 0.01 µg/mL phorbol 12-myristate 13-acetate (PMA) (Sigma-Aldrich) for 48 h prior to infection. P. gingivalis cells were grown to the early stationary phase in BM + Hm medium, centrifuged (4000× g, 20 min, 4 • C) and washed twice with PBS. Macrophages were infected with P. gingivalis at multiplicity of infection (MOI) of 100 [16]. After 4 h, the cells were washed twice with PBS and one out of the three portions was lysed with water to count live bacterial cells present inside macrophages and attached to them. The two remaining portions were further incubated with the fresh culture medium, supplemented with 300 µg/mL gentamicin and 200 µg/mL metronidazole. After 1 h, one portion of the cells was washed three times with PBS and lysed with water to count live bacterial cells residing inside macrophages. The last portion of the cells was collected 8 h after infection, washed three times with PBS, and lysed with water to count live bacterial cells present inside macrophages. All samples were serially diluted and plated onto Schaedler blood agar plates to determine colony-forming units (CFU).

Plasmid Construction, Overexpression, and Purification of Recombinant PgRsp Protein
For overexpression of PgRsp protein in fusion with N-terminally attached His-tag and maltose-binding protein (MBP), pMAL_c5x plasmid (New England Biolabs), modified as described previously (pMAL c5x_His plasmid), was used [40]. The pgrsp gene was amplified by PCR using primers listed in Table S1 and cloned into the pMAL c5x_His plasmid using restriction-free cloning [42]. Final expression pMAL c5x_PgRsp plasmids were verified by DNA sequencing (Microsynth).
To obtain modified PgRsp protein with single amino acid substitutions, appropriate modifications of pMAL c5x_PgRsp plasmid were performed using the QuikChange II Site-Directed Mutagenesis Kit (Agilent Technologies, Santa Clara, CA, USA) and primers listed in Table S1. The resulting plasmids were sequenced and used to transform E. coli cells.
PgRsp protein variants in fusion with 6His-MBP protein were overexpressed in E. coli Rosetta 2 (DE3) strain grown in Terrific Broth (TB), supplemented with 35 µg/mL chloramphenicol and 100 µg/mL ampicillin. After transformation, bacteria were grown at 37 • C with shaking at 220 rpm until cultures reached OD 600 = 1.0. Subsequently, cultures were cooled down to 16 • C and protein overexpression was induced by addition of IPTG (Roth) to a final concentration of 0.5 mM. Proteins were overexpressed for 16-20 h. The cells were collected by centrifugation (4000× g, 20 min, 4 • C) and the pellets were stored at −20 • C until used. The bacteria were re-suspended in 25 mM HEPES, pH 7.8, containing 300 mM NaCl, and lysed by sonication (UP100H Hielscher Ultrasonics, at 4 • C). Then, the lysate was centrifuged (40,000× g, 15 min, 4 • C) and the soluble fraction was loaded onto an amylose resin (New England Biolabs) column. After washing, 6His-MBP-PgRsp protein was eluted with 25 mM HEPES, pH 7.8, containing 300 mM NaCl and 25 mM maltose. The protein was then transferred to 25 mM HEPES, pH 7.8, containing 80 mM NaCl, 5% glycerol and 2 mM CaCl 2 using Amicon Ultra-15, 10.000 NMWL (Millipore, Burlington, MA, USA). Subsequently, the protein was subjected to digestion with Factor Xa (New England Biolabs) according to the manufacturer's protocol (New England Biolabs), but with addition of 0.01% SDS. Untagged PgRsp protein was obtained by binding of released 6His-MBP to TALON resin (Clontech, Mountain View, CA, USA). Purified PgRsp protein was stored in 25 mM HEPES, pH 7.8, containing 300 mM NaCl and 10% glycerol at −20 • C until used. Protein purification process was monitored by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) in 12% polyacrylamide gels and staining with Coomassie Brilliant Blue G-250 (CBB G-250). Concentration of PgRsp protein was determined by measuring absorbance at 280 nm using the empirical molar absorption coefficient (ε 280 = 15.33 mM −1 cm −1 ) determined in this study as reported by others [43].

Protein-Heme Complex Formation
Hemin chloride (ICN Biomedicals, Costa Mesa, CA, USA) solution was freshly prepared and protein-heme complexes were formed as described previously [20]. UV-visible spectra were monitored using a double beam Jasco V-750 spectrophotometer (10 mm path length). Heme binding to the PgRsp protein (5 or 10 µM in PBS) was analyzed under air (oxidized) or reduced conditions, the latter formed by addition of 10 mM sodium dithionite (Sigma-Aldrich) and a layer of mineral oil (Sigma-Aldrich).

Electrophoretic Mobility Shift Assay
For the electrophoretic mobility shift assay (EMSA), the LightShift Chemiluminescent EMSA Kit (Thermo Scientific) was used according to the manufacturer s protocol. Briefly, biotin-labeled DNA fragments of selected promoter regions were amplified by PCR using primers listed in Table S1. Then, 1 ng of DNA was added to 1× binding buffer with the addition of 2.5% glycerol, 5 mM MgCl 2 , 50 ng/µL poly (dI-dC) and varying concentrations of heme, purified PgRsp protein, non-biotinylated DNA and sodium dithionite. Samples were incubated for 20 min at room temperature, and subsequently subjected to electrophoresis for 1 h at 150 V in 0.25× TBE buffer (25 mM Tris, 25 mM boric acid, 0.5 mM EDTA, pH 8.6) on pre-run for 30 min 6% polyacrylamide gels. After electrophoresis, samples were transferred onto a nylon membrane (Bionovo, Legnica, Poland) using wet transfer in the presence of 1× TBE buffer (300 mA, 30 min, 4 • C) and subjected to crosslinking for 10 min with UV radiation at a wavelength of 274 nm. The protocol was continued according to the manufacturer s (Thermo Scientific) guidelines and labeled nucleic acids were visualized using the appropriate reagents and Chemidoc MP Imaging System (Bio-Rad, Hercules, CA, USA).

Phylogenetic Analyses
To ascribe the protein encoded by the PGA7_00004090 gene to a family of the Crp/Fnr superfamily, phylogenetic analysis was performed using comparison of homologous sequences defined by BLAST search (NCBI server). The sequences were aligned by Clustal Omega and the phylogenetic tree was constructed using the EvolView server [44][45][46]. The prediction of PgRsp protein structure was performed using the Phyre2 server [47] and the Swiss PDB Viewer program [48] was used to construct three-dimensional protein models.

Statistical Analysis
Unpaired Student s t-test or one-way ANOVA (analysis of variance) with post hoc Tukey honest significant difference comparison was used to analyze the data (GraphPad Prism version 5). Results are shown as mean ± standard deviation. All experiments were performed at least three times in two replicates. Statistically significant results were indicated with asterisks (* p < 0.05; ** p < 0.01; *** p < 0.001).

PgRsp-a Heme-Binding Member of a Novel Family of Crp/Fnr Superfamily Transcription Regulators
To assign the potential function of the PgRsp protein, first bioinformatics approaches were employed. Searching for PgRsp-encoding sequences in P. gingivalis strains revealed that amino acid sequences deposited in databases for several P. gingivalis strains (e.g., W83, AJW4, ATCC 33277, 381, HG66, A7A1-28) predicted a 207-amino-acid version of the PgRsp protein, whereas in the A7436 strain a 283-amino-acid protein was predicted [49]. Our theoretical approaches and preliminary analysis of the 5 UTR region of the mRNA encoding PgRsp using 5 RACE (data not shown) led us to the conclusion that in the A7436 strain the pgrsp gene encodes a 207-amino-acid long protein. Therefore, in this study for further theoretical and experimental approaches we used this protein version.
Next, the amino acid sequence of the PgRsp protein was compared with homologous proteins from different families of the Crp/Fnr superfamily. Construction of the phylogenetic tree showed that the protein encoded by the pgrsp gene was not ascribed to any of the best-characterized families. The highest amino acid sequence similarity (~5-20%) was found when compared with proteins belonging to the CooA and NnrR families (Figures 1 and 2). Alignment of the sequences of the closest homologs to the PgRsp protein showed 19.5% and 15.1% homology to proteins from Rhodobacter sphaeroides and Brucella canis, respectively, assigned as members of the NnrR family, as well as 20.5% and 20% homology to the best characterized CooA family proteins from Rhodospirillum rubrum (RrCooA) and Azotobacter vinelandii, respectively.
Some proteins belonging to the Crp/Fnr superfamily, including CooA and Dnr proteins, may bind heme [31,33,50]. Preliminary observations based on the visual inspection of the color exhibited by the purified PgRsp protein samples confirmed, at least in part, such a possibility. To ascertain whether PgRsp is a heme-binding protein, the purified protein was titrated with heme under air (oxidized) or reduced conditions. The λ max in the Soret region for the PgRsp-heme complex under oxidized or reduced conditions was 412 or 425 nm, respectively (Figure 3a,b). The maxima for the Q bands were clearly seen at 531 and 559 nm only for the PgRsp-heme complex under reduced conditions ( Figure 3b). In contrast, under oxidized conditions weak maxima at 533 and 620 nm were observed (Figure 3a,b). Similar spectra were reported for RrCooA protein [51]. To further study the PgRsp-heme complex, the difference spectra were generated (Figure 3c,d). Although maxima of Q bands under oxidized conditions were present at 537 and 563 nm and an additional maximum at 648 nm (Figure 3c), reduction of the heme significantly increased Q band maxima at 529 and 560 nm ( Figure 3d). However, no changes in UV-visible spectra were observed when Cys76, Met78, or His85 was singly replaced by Ala residue (data not shown). In addition, all the purified PgRsp protein variants were visible as red colored samples, suggesting the presence of heme bound complexes (data not shown). Homologous protein from P. gingivalis (PgRsp) is marked with a red frame. The proteins are shown by the bacterial species and NCBI accession number. Amino acid sequences were identified using BLAST search and compared using Clustal Omega and Simple Phylogens online tools. The phylogenetic tree was created using the online tool EvolView. Further, the PgRsp amino acid sequence was compared with sequences of representative homologous proteins from the CooA family, as they showed the highest amino acid similarity (Figures 1 and 2). The heme-binding motifs present in CooA proteins include Cys or His and Pro or Ala residues ( Figure 2). Although amino acid sequence comparison showed the possibility of a similar motif for heme binding in the PgRsp, the structure modeling based on RrCooA protein did not include it. Therefore, a theoretical three-dimensional model of PgRsp was constructed, using known structures of proteins from the Crp/Fnr superfamily as templates ( Figure 4) and for final modeling, Crp homolog from Eubacterium rectale, showing the highest amino acid sequence coverage and homology ( Figure 2) was used (PDB ID: 3DV8). The modeled structure showed fold typical for Crp proteins, comprising N-and C-terminal (DNA-binding domain) domains, both connected by an alpha helix ( Figure 4). Besides Cys and His, CooA homologs use N-terminal Pro or Ala as the second heme-binding ligand in the non-active state [52,53]. Neither PgRsp nor its closest homologs have such an amino acid residue at the N terminus ( Figure 2).
Our results suggest that PgRsp protein may bind heme in a different manner and thus exhibit a different mechanism of gene regulation at the molecular level compared to the known heme-binding proteins comprising the Crp/Fnr superfamily. Further, the PgRsp amino acid sequence was compared with sequences of representative homologous proteins from the CooA family, as they showed the highest amino acid similarity (Figures 1 and 2). The heme-binding motifs present in CooA proteins include Cys or His and Pro or Ala residues ( Figure 2). Although amino acid sequence comparison showed the possibility of a similar motif for heme binding in the PgRsp, the structure modeling based on RrCooA protein did not include it. Therefore, a theoretical three-dimensional model of PgRsp was constructed, using known structures of proteins from the Crp/Fnr superfamily as templates ( Figure 4) and for final modeling, Crp homolog from Eubacterium rectale, showing the highest amino acid sequence coverage and homology ( Figure 2) was used (PDB ID: 3DV8). The modeled structure showed fold typical for Crp proteins, comprising Nand C-terminal (DNA-binding domain) domains, both connected by an alpha helix ( Figure 4). Besides Cys and His, CooA homologs use N-terminal Pro or Ala as the second heme-binding ligand in the non-active state [52,53]. Neither PgRsp nor its closest homologs have such an amino acid residue at the N terminus ( Figure 2).
Our results suggest that PgRsp protein may bind heme in a different manner and thus exhibit a different mechanism of gene regulation at the molecular level compared to the known heme-binding proteins comprising the Crp/Fnr superfamily. Microorganisms 2019, 7, x FOR PEER REVIEW 9 of 23

Pgrsp Gene Expression is Iron/Heme-and PgFur-Dependent
The pgrsp gene expression analysis carried out in the wild-type A7436 strain in relation to iron/heme availability and bacterial growth phase demonstrated that it was significantly increased in bacteria cultured for a prolonged time in the iron/heme-depleted medium (1.76-fold change determined by microarray analysis and data shown in Figure 5a). Analysis of the influence of growth phase on pgrsp gene expression showed the highest levels of mRNA encoding PgRsp produced in bacteria cultured under heme/iron excess during the middle exponential growth phase (10 h) in comparison to the early growth phase (4 h), whereas the lowest mRNA levels were produced at the early stationary phase (24 h) (Figure 5b). No significant influence of growth phases was observed on the pgrsp gene expression when bacteria were grown under low-iron/heme conditions.
Recently, we reported that P. gingivalis PgFur is essential for P. gingivalis virulence and regulates, among many other functions, expression of the pgrsp gene [13,16]. Therefore, we suggest that PgRsp may cooperate with PgFur protein in regulation of bacterial virulence. This suggestion was supported by microarray analysis demonstrating decreased pgrsp gene expression in the pgfur mutant strain, especially when bacteria were grown in iron/heme-depleted conditions (−2.78-fold change). Theoretical three-dimensional model of PgRsp protein structure. PgRsp shows fold typical for Crp/Fnr homologs: it contains a C-terminal DNA-binding domain connected with an N-terminal, sensor domain, and alpha-helix connecting the two domains, which is responsible for protein dimerization. The structure was modeled using the Phyre2 web-based modeling server based on the three-dimensional protein structure (PDB ID: 3DV8) of the Eubacterium rectale Crp/Fnr superfamily member as a template. Green-PgRsp theoretical model; red-experimentally solved structure of E. rectale protein.

Pgrsp Gene Expression is Iron/Heme-and PgFur-Dependent
The pgrsp gene expression analysis carried out in the wild-type A7436 strain in relation to iron/heme availability and bacterial growth phase demonstrated that it was significantly increased in bacteria cultured for a prolonged time in the iron/heme-depleted medium (1.76-fold change determined by microarray analysis and data shown in Figure 5a). Analysis of the influence of growth phase on pgrsp gene expression showed the highest levels of mRNA encoding PgRsp produced in bacteria cultured under heme/iron excess during the middle exponential growth phase (10 h) in comparison to the early growth phase (4 h), whereas the lowest mRNA levels were produced at the early stationary phase (24 h) (Figure 5b). No significant influence of growth phases was observed on the pgrsp gene expression when bacteria were grown under low-iron/heme conditions.
Recently, we reported that P. gingivalis PgFur is essential for P. gingivalis virulence and regulates, among many other functions, expression of the pgrsp gene [13,16]. Therefore, we suggest that PgRsp may cooperate with PgFur protein in regulation of bacterial virulence. This suggestion was supported by microarray analysis demonstrating decreased pgrsp gene expression in the pgfur mutant strain, especially when bacteria were grown in iron/heme-depleted conditions (−2.78-fold change).

PgRsp Plays a Role in Regulation of P. gingivalis Virulence
To characterize the importance of the PgRsp protein in the production of P. gingivalis virulence factors, a pgrsp deletion mutant strain (TO11) and complemented mutant strain (TO11 + pgrsp) were constructed, and expression of the virulence factors chosen in regard to their importance in the colonization of the oral cavity and participation in synergistic mechanisms providing nutrients for P. gingivalis and other periodontopathogens was analyzed. Examination of bacterial growth rates showed slight differences between TO11 and A7436 strains grown in iron/heme-rich medium. The mutant TO11 strain grew similarly to the wild-type A7436 strain until it reached the early stationary phase, exhibiting higher biomass production during prolonged growth time (Figure 6a). Complementation of the TO11 strain (Figure 6b) resulted in partial correction of the mutant's phenotype (Figure 6a).
In order to initiate infection, P. gingivalis must interact with other periodontopathogens and host cells. Important virulence factors which play a role in this process are hemagglutinins, gingipains, and fimbriae [17,19,21,54]. To examine the potential influence of PgRsp on the production of P. gingivalis surface proteins, coaggregation ability of P. gingivalis and selected oral bacteria, P. intermedia, T. forsythia, and S. gordonii, was analyzed. We observed faster coaggregation of the mutant TO11 strain with P. intermedia and T. forsythia in comparison to the wild-type A7436 strain ( Figure  6c). No differences were observed in interaction of TO11 and A7436 strains with S. gordonii ( Figure  6c). Additionally, the homotypic biofilm formation ability of both strains was monitored. In comparison to the A7436 strain, the TO11 strain formed higher biofilm structures (Figure 6d), which could be correlated with higher bacterial biomass production in the case of the TO11 strain ( Figure  6a).

PgRsp Plays a Role in Regulation of P. gingivalis Virulence
To characterize the importance of the PgRsp protein in the production of P. gingivalis virulence factors, a pgrsp deletion mutant strain (TO11) and complemented mutant strain (TO11 + pgrsp) were constructed, and expression of the virulence factors chosen in regard to their importance in the colonization of the oral cavity and participation in synergistic mechanisms providing nutrients for P. gingivalis and other periodontopathogens was analyzed. Examination of bacterial growth rates showed slight differences between TO11 and A7436 strains grown in iron/heme-rich medium. The mutant TO11 strain grew similarly to the wild-type A7436 strain until it reached the early stationary phase, exhibiting higher biomass production during prolonged growth time (Figure 6a). Complementation of the TO11 strain (Figure 6b) resulted in partial correction of the mutant's phenotype (Figure 6a).
In order to initiate infection, P. gingivalis must interact with other periodontopathogens and host cells. Important virulence factors which play a role in this process are hemagglutinins, gingipains, and fimbriae [17,19,21,54]. To examine the potential influence of PgRsp on the production of P. gingivalis surface proteins, coaggregation ability of P. gingivalis and selected oral bacteria, P. intermedia, T. forsythia, and S. gordonii, was analyzed. We observed faster coaggregation of the mutant TO11 strain with P. intermedia and T. forsythia in comparison to the wild-type A7436 strain (Figure 6c). No differences were observed in interaction of TO11 and A7436 strains with S. gordonii (Figure 6c). Additionally, the homotypic biofilm formation ability of both strains was monitored. In comparison to the A7436 strain, the TO11 strain formed higher biofilm structures (Figure 6d), which could be correlated with higher bacterial biomass production in the case of the TO11 strain (Figure 6a).
To determine whether fimbriae could be involved in this process, expression of the fimA gene, encoding the main component of P. gingivalis long fimbriae [21] was examined. Surprisingly, fimA gene expression was lower in the TO11 strain in comparison to the A7436 strain (Table 1). It is likely that lower amounts of fimbriae could enable exposure or induce higher expression of other proteins involved in the interaction between P. gingivalis and other bacterial cells, finally increasing the coaggregation capacity of the mutant cells.   Gingipains and Hmu heme-acquisition system proteins are considered the most important virulence factors of P. gingivalis [12,17,[55][56][57]. Therefore, expression of hmuY, rgpA, rgpB and kgp genes in the TO11 strain in comparison to the A7436 strain was examined. Although no significant changes in the hmuY gene expression were observed (Table 1), higher expression of the kgp gene and lower expression of rgpB and rgpA genes were found, especially in bacteria grown under low-iron/heme conditions. During prolonged culture in iron/heme-rich medium, expression of rgpB and rgpA genes was slightly higher in the pgrsp mutant strain in comparison to the wild-type strain (Table 1). Moreover, the expression of the PGA7_00031810 gene encoding a putative protease was examined and it showed decreased expression in the TO11 strain (Table 1). To correlate gene expression with the activity of gene products, total proteolytic activity was determined. The TO11 strain showed higher proteolytic activity in iron/heme-rich medium and lower in iron/heme-depleted medium (Figure 6e).
Altogether, these results suggest that PgRsp might regulate expression of important virulence factors of P. gingivalis. Based on our results one may assume that PgRsp is involved in the control of expression of genes engaged in interaction between P. gingivalis cells and other periodontopathogens.

PgRsp is Involved in the Response to Oxidative Stress
Sequence similarity of PgRsp to CooA proteins and the observed phenotype of the mutant pgrsp strain led us to conclude that PgRsp may be involved in sensing extracellular stresses. Therefore, the expression of selected genes, namely oxyR, sod, tpx, bcp, pgdps, ahpC, rbr, pgfur, and ustA, was analyzed. We observed increased expression of all but pgdps and pgfur genes in the TO11 strain in comparison to the A7436 strain during prolonged growth, especially under iron/heme-depleted conditions (Table 1). To further examine the potential role played by PgRsp particularly in the oxidative stress response, we analyzed the influence of different redox conditions on bacterial growth. Bacteria were cultured in rich culture media but in the absence of cysteine (culture medium component providing reduced conditions and protecting P. gingivalis from atmospheric oxygen) and with different H 2 O 2 concentrations, added at the initiation of bacterial cultures. We observed growth retardation of the TO11 strain compared to the A7436 strain during bacterial growth in rich culture media lacking cysteine (conditions inducing mild aerobic stress) and in rich culture media lacking cysteine and with the addition of 0.05 mM H 2 O 2 (conditions inducing high oxidative stress). Higher concentration of H 2 O 2 (0.25 mM) was lethal for the TO11 strain whereas the A7436 strain showed only a slight delay in the growth ability (Figure 7a). Additionally, the effect of oxidative stress generated in a later stage of bacterial cultures was tested. Surprisingly, no effect was observed when bacteria lacking PgRsp protein were grown for 4 h and subsequently treated with H 2 O 2 ( Figure 7b). with different H2O2 concentrations, added at the initiation of bacterial cultures. We observed growth retardation of the TO11 strain compared to the A7436 strain during bacterial growth in rich culture media lacking cysteine (conditions inducing mild aerobic stress) and in rich culture media lacking cysteine and with the addition of 0.05 mM H2O2 (conditions inducing high oxidative stress). Higher concentration of H2O2 (0.25 mM) was lethal for the TO11 strain whereas the A7436 strain showed only a slight delay in the growth ability (Figure 7a). Additionally, the effect of oxidative stress generated in a later stage of bacterial cultures was tested. Surprisingly, no effect was observed when bacteria lacking PgRsp protein were grown for 4 h and subsequently treated with H2O2 (Figure 7b). Resistance of P. gingivalis to oxidative stress is an important factor not only for survival in the oral microbiome but also inside host immune cells. Using THP-1-derived macrophages as model cells, we studied the role of PgRsp protein in invasion of host immune cells. We observed that attachment of the mutant TO11 strain to macrophages did not differ from that observed for the wildtype A7436 strain ( Figure 8). However, lower invasion of and survival ability inside macrophages of the pgrsp mutant cells were observed, suggesting impaired capability to enter macrophages and/or resist antimicrobial mechanisms inside macrophages. Resistance of P. gingivalis to oxidative stress is an important factor not only for survival in the oral microbiome but also inside host immune cells. Using THP-1-derived macrophages as model cells, we studied the role of PgRsp protein in invasion of host immune cells. We observed that attachment of the mutant TO11 strain to macrophages did not differ from that observed for the wild-type A7436 strain ( Figure 8). However, lower invasion of and survival ability inside macrophages of the pgrsp mutant cells were observed, suggesting impaired capability to enter macrophages and/or resist antimicrobial mechanisms inside macrophages. retardation of the TO11 strain compared to the A7436 strain during bacterial growth in rich culture media lacking cysteine (conditions inducing mild aerobic stress) and in rich culture media lacking cysteine and with the addition of 0.05 mM H2O2 (conditions inducing high oxidative stress). Higher concentration of H2O2 (0.25 mM) was lethal for the TO11 strain whereas the A7436 strain showed only a slight delay in the growth ability (Figure 7a). Additionally, the effect of oxidative stress generated in a later stage of bacterial cultures was tested. Surprisingly, no effect was observed when bacteria lacking PgRsp protein were grown for 4 h and subsequently treated with H2O2 (Figure 7b). Resistance of P. gingivalis to oxidative stress is an important factor not only for survival in the oral microbiome but also inside host immune cells. Using THP-1-derived macrophages as model cells, we studied the role of PgRsp protein in invasion of host immune cells. We observed that attachment of the mutant TO11 strain to macrophages did not differ from that observed for the wildtype A7436 strain ( Figure 8). However, lower invasion of and survival ability inside macrophages of the pgrsp mutant cells were observed, suggesting impaired capability to enter macrophages and/or resist antimicrobial mechanisms inside macrophages. To further understand the involvement of PgRsp in the regulation of gene expression during exposure to oxidative stress, expression of selected genes under various redox conditions was tested. Surprisingly, the largest changes in gene expression were observed when culturing bacteria in BM medium without cysteine (BM-Cys + Hm) (Table 2), with a significant decrease in expression of genes involved in bacterial resistance to oxidative stress (e.g., sod, bcp, tpx). When bacteria were exposed to 0.25 mM H 2 O 2 , higher expression of oxyR, tpx, and rbr genes and decreased expression of bcp, pgdps, and ahpC genes were observed. When bacteria were cultured in rich culture medium supplemented with cysteine (BM + Cys + Hm), gene expression did not change as dramatically as in bacteria exposed to oxidative stress. This could suggest that PgRsp protein is an important factor responsible for the protection of P. gingivalis during exposure to agents disturbing redox homeostasis. Since the phenotype of the bacteria lacking this protein was significantly altered when exposed to oxidative stress, we termed this protein a P. gingivalis redox-sensing protein (PgRsp). Table 2. Analysis of gene expression in response to different redox conditions during growth in liquid culture medium of the P. gingivalis pgrsp mutant strain (TO11) in comparison to the wild-type A7436 strain. Data are marked as mean ± standard deviation. Changes statistically significant are marked with an asterisk. Besides the downregulation of genes responsible for protection against oxidative stress, we observed downregulation of hmuY and fimA genes, and genes encoding gingipains during exposure of bacteria to air, implying regulation of those genes by the PgRsp protein. During exposure of P. gingivalis to H 2 O 2 , the mutant TO11 strain showed over 7 times higher expression of the oxyR gene. This could suggest that increased production of OxyR protein may compensate the lack of PgRsp, which results in higher expression of genes connected with protection against oxygen, namely sod, rbr, and tpx.

PgRsp Regulates Hmu Operon, Bcp Genes and its own Expression
To confirm whether regulation of selected genes, whose expression was significantly changed in the mutant TO11 strain in comparison to the wild-type A7436 strain was direct, the interaction of the PgRsp protein with fragments of promoters of the hmu operon and bcp gene using EMSA analysis was analyzed. Retardation in migration of promoter fragments after addition of the PgRsp protein (Figure 9a,b) demonstrated that the protein may regulate expression of those genes directly through recognizing and binding to DNA. Although PgRsp binds heme, no significant changes in PgRsp-DNA migration were observed when heme was added to the samples (Figure 9c). was analyzed. Retardation in migration of promoter fragments after addition of the PgRsp protein (Figure 9a,b) demonstrated that the protein may regulate expression of those genes directly through recognizing and binding to DNA. Although PgRsp binds heme, no significant changes in PgRsp-DNA migration were observed when heme was added to the samples (Figure 9c).  Many proteins belonging to the Crp/Fnr superfamily autoregulate their own production [31], therefore we examined whether PgRsp protein can also regulate its own expression. For this purpose, the interaction of the purified PgRsp protein with the promoter region of the pgrsp gene was examined using EMSA analysis. Figure 9d showed that expression of the pgrsp gene could be autoregulated.

Discussion
The oral cavity is inhabited by microorganisms forming one of the most diverse microbiomes of the human body. Development of periodontitis is linked with environmental shift in the oral microbial consortium [58], leading to more reduced conditions and domination of Gram-negative bacteria over early colonizers, the latter mainly aerobic, Gram-positive bacteria (e.g., S. gordonii). P. gingivalis, as a secondary colonizer, has to interact with other bacteria and host cells to invade and colonize this niche and, importantly, to evade the host immune response. Therefore, precise regulation of gene expression is crucial for nutrient uptake and bacterial survival in a very hostile inflammatory environment within the oral microbiome and inside the host cells. Among a variety of changing environmental stimuli is redox status influenced by oxygen tension, nutrient availability (including iron and heme), and ROS and RNS levels. Oxidative stress is caused by metabolic processes occurring in bacteria and by agents present in the environment, and is generated by host cells, mostly neutrophils and macrophages, which produce ROS and H 2 O 2 to eliminate bacteria, as well as by redox-active drugs used to combat pathogens. P. gingivalis cannot grow under aerobic conditions but is highly aerotolerant. To resist, replicate, and cause pathogenic processes, P. gingivalis has developed a variety of mechanisms to successfully cope with those harmful conditions.
For sensing and responding to the changes in the environment, pathogenic bacteria produce many transcription regulators [59]. They can act directly and/or indirectly on gene expression through the regulatory mechanism(s), often encompassing a multilayer regulatory network. In our previous studies we showed that PgFur regulates expression of virulence factors of P. gingivalis and is involved in invasion of human cells [13,16]. We proposed that PgFur is a part of a multilayer regulatory network, which also regulates expression of other transcription factors, including the gene encoding PgRsp protein [16]. Therefore, in this study we initiated characterization of PgRsp in regard to its involvement in potential regulatory mechanisms. Our results suggest that expression of the pgrsp gene is iron/hemeand growth phase-dependent. A similar gene expression pattern was observed for the pgfur gene [16]. Iron/heme homeostasis is closely linked to the oxidative stress response in P. gingivalis [60], which could explain the similar influence of external environmental conditions on the mechanism of action of PgFur and PgRsp proteins. Our findings showed that deletion of the pgrsp gene changes the interaction between P. gingivalis cells and other periodontopathogens. Recently, we reported similar phenotype for the pgfur mutant strain constructed in the A7436 strain [13,16]. We assume that these two proteins mutually regulate expression of genes involved in virulence in an iron/heme-dependent manner.
Besides changes in expression of genes selected in this study, we showed direct interaction of PgRsp protein with the promoter of the hmu operon, encoding six proteins, which are associated with heme uptake and its transport through the outer membrane [12,55]. It has been shown that HmuY protein is an important factor of P. gingivalis virulence [12,55,56]. It is expressed during bacterial invasion and proliferation inside the host cells [15]. Many studies have shown that HmuY protein is produced at higher levels in biofilm structures, thus enabling efficient uptake of heme [35,61,62]. PgRsp regulation of the Hmu heme acquisition system components and proteases involved in degradation of the host proteins suggests that it could be directly involved in regulation of virulence, allowing the bacterium to survive in the diverse microbiome of the oral cavity, as well as inside host cells. In addition, HmuY and hemagglutinin domains of Kgp and RgpA gingipains may provide more efficient heme accumulation on the bacterial cell surface, thus participating in formation of oxidative buffer to neutralize ROS [63,64]. Here we demonstrated that in the pgrsp mutant strain this ability was disturbed.
Sensing oxygen species and the redox state of the external environment is very important for anaerobic bacteria. One of the best known and characterized bacterial proteins involved in oxidative stress response and aerotolerance is OxyR [30,[65][66][67]. It has been shown that also heme limitation may induce expression of the oxyR gene and genes regulated by this transcription factor, such as sod, pgdps, and ahpC [60,67]. Our study showed that also PgRsp is important for the P. gingivalis response to oxidative stress. We observed that heme limitation together with deletion of the pgrsp gene significantly altered expression of several genes involved in the oxidative stress response. The largest changes in the phenotype, as well as in gene expression, were observed when bacteria were grown in nutrient rich conditions and exposed to mild aerobic stress in the absence of cysteine.
It has been shown that Bcp and VimA proteins are important for anaerobic bacteria in regard to their resistance to air and H 2 O 2 exposure; they are linked with gingipain expression [68], autoaggregation, and virulence [69]. Similar observations were made for the pgrsp mutant strain, demonstrating higher biofilm formation and coaggregation ability with other oral pathogens. Importantly, we also found that the pgrsp-deficient strain was less virulent in the macrophage infection model. Such regulation is possible since we observed direct interaction of PgRsp protein with the promoter of the bcp gene. All these data suggest that proteolytic activity together with heme accumulation on the bacterial cell surface could be coordinated to cope with harmful oxidative effects.
UstA protein is an important factor regulating gene expression during the stationary phase of bacterial growth. Interestingly, P. gingivalis ustA expression was higher in the stationary phase and under oxygen exposure [70]. It has been shown by others that deletion of the ustA gene resulted in impaired growth in rich culture medium and in changes in expression of genes involved in the response to oxidative stress [70]. In addition, exposure of the TO11 strain to mild oxidative conditions resulted in two times higher expression of the ustA gene, whereas H 2 O 2 had an opposite impact on expression of the gene. Overexpression of the ustA gene could explain the growth of the TO11 strain, which produces significantly higher bacterial biomass compared to the A7436 strain. Based on results obtained in this study and observations presented by Abaibou and coworkers [68] and Johnson and coworkers [67] we assume that the operon encoding Bcp and VimA proteins, as well as the ustA gene are important targets for PgRsp regulatory activity.
Proteins from the Crp/Fnr superfamily play an important role in sensing stress conditions and are crucial for bacterial virulence [31,71]. To date, only one protein from this superfamily has been characterized in P. gingivalis. Lewis and coworkers [71] showed that HcpR protein (an Fnr homolog) regulates expression of genes responsible for the resistance of P. gingivalis to nitrosative stress by sensing NO through its interaction with heme bound to the HcpR protein [33]. We demonstrated that PgRsp protein belongs to the Crp/Fnr superfamily of transcription factors and could be classified as a relatively close homolog to the CooA family. Proteins homologous to CooA can sense CO [31,53], employing interaction of CO with heme bound to the protein [51,53], thus regulating genes important for utilization of CO as an energy source [72,73]. The molecular mechanism of the best-characterized homolog of CooA protein from R. rubrum, RrCooA, implies binding of Fe(III)heme by Pro2 and Cys75, when the protein is not active. Reduction of iron bound to heme causes replacement of coordinating Cys75 by His77. For protein activation, the CO molecule is bound to heme by a Pro2 residue, resulting in conformational changes of the protein, leading to interaction with DNA. This kind of protein activation is unique for RrCooA protein [51][52][53], since in the CooA homolog from Carboxydothermus hydrogenoformans, the heme molecule is coordinated by Ala2 and His82 residues and no transfer of the heme group occurs during the change of iron redox state [52,73].
Different redox state sensing mechanisms would be of great importance for bacteria for a fast response to changes in redox state of the external environment of the oral cavity or periodontal pockets, as well as inside the host cells. For example, PerR proteins sense hydrogen peroxide by metal-catalyzed His oxidation, which leads to conformational changes and dissociation from DNA and activation of expression of oxidative stress response genes [74,75]. On the other hand, in the classic OxyR paradigm, H 2 O 2 oxidizes Cys residues, influencing the DNA-OxyR interaction, which leads to the activation of genes encoding proteins with antioxidant function. Similar feature of PgRsp could be important for P. gingivalis not only to resist oxidative stress generated by host immune cells during progression of periodontal diseases but importantly during metabolic disorders connected with chronic periodontitis [10]. However, our analysis of PgRsp point mutation variants affected in Cys76, Met78, or His85, did not allow us to confirm heme coordinating ligands since all protein variants examined bound heme with similar ability. Therefore we assume that the regulation mechanism employed by PgRsp differs compared to those used by other members of the Crp/Fnr superfamily of proteins.
The P. gingivalis response to the oxidizing environment is considered to be regulated mostly by OxyR protein. We assume that the mechanism of PgRsp protein response to the redox state could be important for sensing the redox state. It could imply activation by heme-iron-catalyzed oxidation similar to that observed in PerR protein or more likely by an activation process similar to that characterized for CooA protein with the exception that instead of an CO molecule, an O 2 molecule could be bound to the heme iron. This would cause protein structure rearrangements and changes in the interaction of the PgRsp protein with DNA and/or PgFur. The proposed mechanism is also supported by studies demonstrating that some proteins from the CooA family are able to bind an oxygen molecule through heme and that they may regulate resistance to oxidative stress in anaerobic bacteria [34,75]. Based on our results, one may assume that PgRsp could be a redox-sensing protein, utilizing for this purpose heme bound to the protein. However, the heme-based mechanism of this regulation requires further investigation. Nevertheless, results gained in this study broaden knowledge on P. gingivalis gene regulation mechanisms, focused on its adaptation to different redox state environments.

Conclusions
PgRsp protein belongs to a novel family of heme-binding proteins, the Crp/Fnr superfamily of transcription regulators. Based on our data, one may conclude that this protein is a heme-based redox sensor which plays a role in the oxidative stress response and P. gingivalis virulence. However, at this stage of study we are unable to state whether PgRsp senses redox conditions (a more complex process) and/or oxygen (direct binding of an oxygen molecule). The amino acid identity and similarity of the PgRsp protein to other heme-binding proteins of the Crp/Fnr family is low; we assume that the heme-binding motif and transcriptional activation mechanism may differ from those known for CooA proteins. PgRsp as a redox-sensing protein could alter gene expression alongside or cooperate with OxyR protein, as well as with PgFur protein in the response and subsequent resistance to oxidative stress.