Expression, Purification, and Characterization of the Recombinant, Two-Component, Response Regulator ArlR from Fusobacterium nucleatum

Fusobacterium nucleatum is associated with the incidence and development of multiple diseases, such as periodontitis and colorectal cancer (CRC). Until now, studies have proved only a few proteins to be associated with such pathogenic diseases. The two-component system is one of the most prevalent forms of bacterial signal transduction related to intestinal diseases. Here, we report a novel, recombinant, two-component, response regulator protein ArlR from the genome of F. nucleatum strain ATCC 25,586. We optimized the expression and purification conditions of ArlR; in addition, we characterized the interaction of this response regulator protein with the corresponding histidine kinase and DNA sequence. The full-length ArlR was successfully expressed in six E. coli host strains. However, optimum expression conditions of ArlR were present only in E. coli strain BL21 CodonPlus (DE3) RIL that was later induced with isopropyl β-D-1-thiogalactopyranoside (IPTG) for 8 h at 25 °C. The SDS-PAGE analysis revealed the molecular weight of the recombinant protein as 27.3 kDa with approximately 90% purity after gel filtration chromatography. Because ArlR was biologically active after its purification, it accepted the corresponding phosphorylated histidine kinase phosphate group and bound to the analogous DNA sequence. The binding constant between ArlR and the corresponding histidine kinase was about 2.1 μM, whereas the binding constant between ArlR and its operon was 6.4 μM. Altogether, these results illustrate an effective expression and purification method for the novel two-component system protein ArlR.


Introduction
Fusobacterium nucleatum is a typical, opportunistic, and pathogenic gram-negative bacterium related to the incidence and development of diseases such as periodontitis, head and neck cancer, poor pregnancy, and colorectal cancer (CRC) [1][2][3][4]. F. nucleatum can invade endothelial cells and adhere to several mammalian cells, including murine lymphocytes and macrophages, and human oral epithelial cells, thus participating in colonization and evasion of host defenses [5,6]. Notably, F. nucleatum has been reported to be present in a high abundance in CRC cells [7]. Many studies have shown that F. nucleatum plays a significant role in increasing the proliferation and migration of CRC cells because of its ability to form a pro-inflammatory microenvironment suitable for the development of CRC and enhancing the drug resistance of CRC cells [8][9][10]. In dental biofilms, F. nucleatum connects the early colonizing commensals and late pathogenic colonizers. Moreover, it functions as a connector between the primary and secondary colonizing organisms; therefore, F. nucleatum is commonly known as a "bridge organism" [11,12].
A two-component system (TCS) is a form of signal transduction system that mainly exists in microorganisms and plants [13]. Bacterial signal transduction systems mainly consist of two proteins: histidine kinase and transcription protein. Bacterial signal transduction systems use kinases with extracellular or periplasmic sensing domains for transferring the phosphate groups to their corresponding response regulator proteins, thus altering gene expression [14][15][16][17][18][19][20][21]. TCSs are involved in bacterial pathogenicity, biofilm formation, and physiological responses to osmotic changes in various bacteria [14]. Recently, many studies have found that the TCS is related to intestinal diseases [22][23][24][25][26][27][28]. Wakimoto et al. (2013) proved that the phosphate regulon transcriptional regulatory protein (PhoB) was necessary for the survival of Bacteroides fragilis in peritoneal abscesses [26]. Similarly, Massmig et al. (2020) suggested that the CntAB two-component system in the gut microbiome catalyzes the oxidative cleavage of L-carnitine into TMA related to cardiovascular diseases [23]. Another significant characteristic of the two-component system is its ability to promote colonization of bacteria in the mammalian intestine; for example, FusKR TCS is required for the robust colonization of enterohemorrhagic E. coli in the mammalian intestine [29]. The CpxRA TCS of Salmonella enterica plays a significant role in gut colonization in Salmonella-induced colitis [30]. Moreover, the two-component system affects the metabolic response of the intestinal bacteria [31][32][33]. Sonnenburg et al. (2006) demonstrated that hybrid TCSBT3172 functions as a metabolic reaction center by coupling the nutrient-sensing to the dynamic regulation of monosaccharide metabolism [32], and  proved that CitAB TCS in Vibrio cholerae contributes to the anaerobic citrate fermentation [31]. Although studies have shown that F. nucleatum is associated with a variety of intestinal diseases, there are only few studies on the virulence and pathogenicity of the two-component system in this bacterium. What is currently known is that Wu et al. (2021) demonstrated that F. nucleatum employs the CarRS two-component system for interspecies interaction, virulence, and nutrient acquisition [34].
TCSs are potential targets for antimicrobial drug design. Because histidine phosphorylation of TCS in bacteria differs from serine/tyrosine/threonine phosphorylation of signaling systems in mammalian cells, TCS inhibitors may exert less toxicity in the host [35,36]. TCSs have been well-studied in several pathogenic bacteria such as E. coli and Pseudomonas aeruginosa; however, they have not been well-studied in F. nucleatum. In this study, we have analyzed the two-component system of F. nucleatum ATCC 25,586, one of the recently sequenced genome strains, which further enhances our knowledge and understanding of its physiological function and pathogenic mechanism of virulence.
In summary, this study provides a method for the expression, purification, and characterization of the response regulator (RR), TCS protein ArlR. This study is the first to report the purification and characterization of the TCS-RR protein ArlR in F. nucleatum.

Bioinformatics Analysis
The complete genome sequence of F. nucleatum ATCC 25,586 (NP_603483.1) was obtained in the FASTA format from the NCBI database (https:// www. ncbi. nlm. nih. gov/). The Stockholm format files from the conserved protein domain family HATPase_c (Pfam02518) of histidine kinase (HK) and the conserved domain family Response_reg (Pfam00072) of RR protein were downloaded from the Pfam database (http:// pfam. xfam. org/). The command hmmbuild in the HMMER 2.0 software (see Table 1 for commands) was used to construct the hidden Markov models for HATPase_c and Response_reg, which revealed HKs and RRs with HATPase_c and Response_reg domains. Subsequently, a comparison of obtained results with those available on the MiST website provides a comprehensive classification of the signal transduction systems [37]. Finally, we stated three points to determine a pair of two-component signal systems: (1) histidine kinase catalytic domain HATPase_c must be located in the HK C-terminus, (2) the upstream of HATPase_c should have the phosphate receptor domain HisKA, and (3) N-terminus of RR should be the phosphate receptor in the target TCSs. Multiple sequence alignments were performed using ClustalX and ESPript [38]. The molecular architecture of the TCS response regulator proteins and ArlR was determined by SMART and Pfam [39]. The structure was predicted from Phyre2 [40]. The promoter of ArlRS operon was predicted by BRPOM [41].

Construction of the Recombinant Plasmid
Using the restriction-free (RF) cloning method, we constructed the pET-28a-arlR plasmid. The RF-cloning.org webserver was used to design primers F1 (5′-CAG CCA TCA TCA TCA TCA TCA CAG CAGC ATG TTA TTA TTT TCT TGG GTG AGG -3 ′ , where the sequence in italics represents arlR gene-specific region) and R1 (5′-GGA GCT CGA ATT CGG ATC CGCG TTA ATC CTC TTT ATA TTG AAA TAT ATAG -3′, where the sequence in italics represent the arlR gene-specific region) complementary to the insert and vector [42]. GAT AAA TTT AAA AAT TCA CTTG -3 ′ , where sequence in italics represents arlSC gene-specific region) and R2 (5′-GGA GCT CGA ATT CGG ATC CGCG TTA AAA TAG TAG TGT TAT TTT TGT TCCC -3 ′ , where sequence in italics represents arlSC gene-specific region). The cloning steps of arlSC were carried out as described above for ArlR.

Expression of the Recombinant ArlR
Six different E. coli host strains (BL21 (DE3), BL21-CodonPlus (DE3) RIL, BL21 (DE3) pLysS, Tuner (DE3), C43 (DE3), Transetta (DE3)) were used to express ArlR. We selected a positive transformant from each of the host strains and cultured it overnight at 37 °C in 20 mL of Luria-Bertani (LB) medium supplemented with the corresponding antibiotics. Next, the liquid culture medium was transformed into 50 mL of LB medium with 1% culture inoculation and cultivated at 37 °C until the OD 600 reached 0.6 to 0.8. Subsequently, the protein expression was induced by the addition of 0.5 mM IPTG at different intervals of time (8 h, 20 h) and temperatures (16 °C, 25 °C). The cells were harvested by centrifugation, resuspended in 1 mL of buffer A (20 mM Tris-HCl pH 8.0, 150 mM NaCl), and disrupted by sonication for 5 min on ice. The lysate was then centrifuged at 18,200 × g for 10 min at 4 °C. The soluble and insoluble proteins were detected by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE).

Purification of the Recombinant Protein
The E. coli host strain with the highest protein expression was selected for large-scale protein expression through the subsequent purification steps. Five grams of wet cells, harvested by centrifugation, was resuspended in 35 mL of buffer A. Aliquots of 1 mL were taken from the 35 mL resuspended buffer. Each aliquot was centrifuged for precipitation. Cells were resuspended in buffer A containing different additives (5% glycerol, 50 mM L-Arg, 50 mM betaine, 1% Triton X-100, 50 mM L-Arg, and 50 mM L-Glu) separately and respectively. The cells were then disrupted by sonication and analyzed by 15% SDS-PAGE, which revealed buffer A with 5% glycerol to be the most suitable buffer. The crude extracts were precipitated by centrifugation at 18,200 × g for 45 min at 4 °C. The soluble supernatant of the lysate was mixed with 5 mL of Ni-NTA affinity resin (GE Healthcare, USA), and the solution was incubated for 30 min in a rotating shaker at 4 °C. We further loaded the resulting slurry onto a 50 mL column, where the resin was washed away with buffer B (20 mM Tris-HCl, pH 8.0, 150 mM NaCl, 5% glycerol, 20 mM imidazole). The target protein was eluted from the column by buffer C (20 mM Tris-HCl, pH 8.0, 150 mM NaCl, 5% glycerol, 250 mM imidazole). The eluted protein was concentrated and loaded onto the HiLoad Superdex 200 26/60 column (GE Healthcare, USA) equilibrated with buffer A containing 5% glycerol. We used 15% SDS-PAGE and Coomassie staining to analyze the samples obtained from each purification step. The expression and purification of ArlSC were performed as described above for ArlR.

Circular Dichroism Spectroscopy
Circular dichroism (CD) spectroscopy was performed on a Chirascan™ spectrometer (Applied Photophysics Ltd.) with a 0.1 cm path length of the quartz cuvette. The samples were prepared using buffer A (20 mM Tris-HCl, pH 8.0, 150 mM NaCl) and 0.15 mg/ mL (55 μM) ArlR. The sample was 200 μL in volume with 1 nm bandwidth length, 185-260 nm scanning range, and 0.5 s time per point. The spectrum of the target protein at 191 to 260 nm was obtained over a range of temperatures starting from 20 °C and incrementally increasing to 94 °C, which we used to measure the ArlR melting temperature (Tm). Ultimately, the Tm value was calculated using the Global 3 software, whereas the secondary structure was visualized and analyzed using the Deconvolution software that comes with the instrument.

Autophosphorylation and Phosphoryl Transfer Assays
The purified ArlSC was pre-equilibrated with the phosphorylation buffer (20 mM Tris-HCl at pH 8.0, 50 mM KCl, 5 mM MgCl 2 ) in a final volume of 100 μL. Subsequently, an autophosphorylation reaction was initiated by adding the aliquots of ATP to a final concentration of 10 μM and lasting for 30 min at 25 °C. The intrinsic kinase activity of ArlSC was measured using the Promega Kinase-Glo Luminescent Kinase Assay Kit. As for the phosphoryl transfer assays, the purified ArlR was added to the phosphorylated ArlSC to initiate a phosphotransfer reaction in the phosphorylation buffer with 10 μM ATP for 10 min at 25 °C. We selected 3 μM phosphorylated ArlSC in this phosphoryl transfer assays, and the concentration of ArlR protein ranged from 0 to 15 μM. Lastly, the persistent ATP was measured using the Promega Kinase-Glo Luminescent Kit.

Microscale Thermophoresis (MST) Assay
As previously reported [43], the affinity of ArlR and ArlSC was determined by MST using Monolith NT.115 (NanoTemper Technologies, Germany). Briefly, ArlSC was labeled with NHS fluorescent dye and centrifuged at 18,200 × g for 10 min to avoid precipitation. A 16 step 1:1 (v/v) ArlR serial stock, with twofold dilution, was prepared with the MST buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 10 mM MgCl 2 , 0.05% Tween-20), such that each dilution step reduced the protein concentration by 50%. Equal volumes of the labeled ArlSC and unlabeled ArlR were mixed, incubated for 5 min at room temperature, and added to capillaries for measurement. In this assay, the initial concentration of ArlR protein was 60 μM, and the concentration of labeled ArlSC protein was 20 nM. The data were analyzed using the NanoTemper Analysis software.
According to the protocol, as described previously [44], the binding of ArlR to its operon was determined by MST using Monolith NT.115 (NanoTemper Technologies, Germany). Briefly, the operon of ArlR was amplified with primers of F2 (Cy5-5′-TTT GAA ATA TTT GTA AAA GGAG-3′) and R2 (5' -GTT ATT ATT TTC TTG GGT GAG GAA GATG -3′). A 16-step 1:1 (v/v) serial ArlR stock, with twofold dilution, was prepared with the MST buffer (50 mM Tris-HCl at pH 8.0, 150 mM NaCl, 10 mM MgCl 2 , 0.05% Tween-20). Equal volumes of the labeled DNA and unlabeled protein solutions were mixed and incubated for 5 min at room temperature. Lastly, the samples were loaded into silica capillaries and measured by MST instrument. The initial concentration of ArlR protein was 50 μM, and the concentration of labeled operon was 20 nM.

Two-Component System in F. nucleatum Strain ATCC 25,586
The two-component system of F. nucleatum strain ATCC 25,586 was searched by the software HMMER and compared with the results on the website MiST. Four pairs of the typical TCSs identified were CarRS, ArlRS, LytRS, and YesNM, respectively. The response regulator (RR) proteins in F. nucleatum strain ATCC 25,586 contained a conserved receiver domain and a variable effector domain (Fig. 1a). Previous studies have reported that the response regulatory ArlR contributes to adhesion, autolysis, and multidrug resistance [45,46]. The alignment analysis between the ArlR in F. nucleatum ATCC 25,586 and other reported response regulation ArlR showed a high sequence similarity. By comparing with the sequence of the ArlR protein in Staphylococcus aureus, it can be said that the conserved aspartate phosphate group receptor in the 61st residue is most likely the phosphate acceptor site (Fig. 1d). These results indicated ArlR protein in F. nucleatum strain ATCC 25,586 as a typical response regulatory protein in the two-component system.
ArlR is a member of the OmpR/PhoB superfamily. The domains predicted by Pfam show that the receiver domain of ArlR contains 103 amino acids, and the effector domain contains 97 amino acids (Fig. 1b). The structure predicted by the Phyre 2 (Fig. 1c) matches the results with Pfam but shows an extended flexible region between these two domains. Moreover, the gene of the ArlR is composed of a large percentage of AT base pairs (~ 74.3%). All of these factors may cause certain difficulties in the expression or purification of ArlR; hence, we optimized its expression and purification conditions.

Expression and Purification of the Recombinant ArlR
The arlR gene from F. nucleatum strain ATCC 25,586 was inserted into the pET-28a plasmid to create a recombinant plasmid and then separately transformed into the six different host strains. The expression of ArlR protein varied significantly with the varying host strains after inducing with 0.5 mM IPTG for 8 h or 20 h, at 16 °C or 25 °C. Host strains had the most significant impact on the ArlR expression; however, the best expression conditions were present in E. coli host strain BL21 CodonPlus (DE3) RIL induced with 0.5 mM IPTG for 8 h at 25 °C (Fig. 2a). The ArlR protein was degraded significantly during the subsequent gel experiment. To increase the stability of ArlR protein, several additives such as glycerin, L-arginine, 50 mM L-Arg, 50 mM L-Glu, betaine, and 1% Triton X-100 were added into the buffer before sonication, respectively. SDS-PAGE revealed that the recombinant ArlR was expressed both in the soluble and inclusion body forms, and the solubility of ArlR could increase slightly by adding 5% glycerol to the additive (Fig. 2b). Therefore, a buffer containing 5% glycerol was selected for protein purification.
Further, ArlR was purified by Ni-NTA affinity chromatography and gel filtration chromatography. We obtained ArlR with a high concentration and purity (Fig. 3a) using Ni-NTA affinity chromatography. However, the ArlR protein was slightly degraded using gel filtration chromatography (Fig. 3b). ArlR was collected in a 50-mL concentration tube after gel filtration chromatography and was concentrated to 500 μL by centrifugation at 4 °C. The yield of ArlR was about 10 mg/L of the culture solution. The secondary structure and thermal stability of the purified ArlR were further determined by circular dichroism spectroscopy. The circular dichroic spectrum revealed two minimum peaks at 208 nm and 222 nm of ArlR, which were the characteristic structural peaks of α-helix (Fig. 3c) . 3d). Consequently, the height of the characteristic peak of the ArlR α-helix gradually decreased with the increase in temperature, which may be due to the dissociation and the changes in the protein secondary structure. Calculations showed the Tm of ArlR as 55.8 °C.

Phosphotransferase Activity of the Phosphorylated ArlSC to ArlR
Kinase-Glo™ Luminescent Kinase assay detected the presence of the kinase activity of the cytoplasmic domain of histidine kinase (ArlSC). As shown in Fig. 4a, the luminescence intensity of the reaction system gradually decreased; that is, the remaining ATP content in the kinase buffer gradually decreased with the increase in the concentration of ArlSC, thereby suggesting that the ArlSC protein has histidine kinase activity. Using the microscale thermophoresis (MST) experiment, we investigated an interaction between the phosphorylated histidine kinase cytoplasmic domain (P-ArlSC) and its corresponding response regulator (ArlR). The microscale thermophoresis (MST) assay revealed that P-ArlSC can interact with ArlR with high affinity (Kd = 2.1 μM, Fig. 5a). We further investigated the phosphotransferase activity of P-ArlSC to ArlR by adding different concentrations of ArlR to the P-ArlSC (3 μM) solution containing ATP. The remaining ATP content in the kinase buffer gradually decreased with an increase in the concentration of ArlR, suggesting that the P-ArlSC protein has the phosphotransferase activity to ArlR (Fig. 4b). The kinase activity of ArlSC and the phosphotransferase activity of P-ArlSC to ArlR. ATP as a substrate by the Ultra-Glo™ Luciferase to catalyze producing photon of light. Hence, luminescence intensity is directly proportional to the amount of ATP persisting in the reaction solution and inversely proportional to histidine kinase activity or phosphotransferase activity. a The luminescence intensity of kinase reaction buffer. As the ArlSC protein concentration gradually increases, the luminescence intensity of the buffer gradually decreases, i.e., the remaining ATP content in the reaction solution gradually decreases, indicating that the ArlSC protein can bind ATP and has histidine kinase activity. b The luminescence intensity of phosphotransferase reaction buffer. As the ArlR protein concentration gradually increases, the luminescence intensity gradually decreases, i.e., the remaining ATP content in the reaction solution gradually decreases, indicating that the P-ArlSC protein can transfer the phosphate group to ArlR. The experiment was repeated thrice. Error bars represent the average difference between the data points and their mean. **, p < 0.01, ***, p < 0.001

Binding of Phosphorylated ArlR to Its Operon
The response regulatory protein of the TCS can usually be in combination with its operon to regulate its expression. To test this hypothesis, BPROM was used to predict the promoter of ArlR; the results showed the presence of a promoter between the arlR gene (Fn1260) and its upstream gene (Fn1259) in F. nucleatum strain ATCC 25,586. Subsequently, the approximately 150 bp sequence between the arlR gene and its upstream gene was amplified by a pair of primers labeled with cy-5. MST assay was used to confirm the interaction between P-ArlR and the corresponding DNA sequence. The intermolecular Kd value of P-ArlR and its operon was 6.4 μM (Fig. 5b).

Discussion
The ArlR is a member of the OmpR/PhoB subfamily, the largest subfamily of bacterial RRs [47]. These RRs can bind to DNA to activate transcription. Studies have reported a strong correlation of the ArlRS two-component system with survival and virulence in gram-positive bacteria, such as Staphylococcus aureus and Staphylococcus epidermidis [48,49]. Yet, the function of the ArlRS two-component system in gram-negative bacteria such as F. nucleatum is unclear. Here, we identified a two-component system of F. nucleatum, from which we selected the ArlRS TCS that was highly homologous to the ArlR TCS of S. aureus to characterize the interacting relationship between ArlR protein and the corresponding histidine kinase and the binding constant of ArlR with its operon. Studies have reported that the two-component system is related to bacterial pathogenicity and intestinal diseases [22][23][24][25][26][27][28]. Thus, this study attempts to identify all typical two-component systems in F. nucleatum ATCC 25,586, which may be critical to its pathogenicity. Therefore, we combined and analyzed the results obtained by HMMER 2.0 software and MiST website and obtained four typical two-component systems (CarRS, ArlRS, LytRS, and YesNM). The The interaction between P-ArlSC and its corresponding response regulator protein ArlR, and binding of P-ArlR to its promoter measured by MST assay. a MST showing the interaction between P-ArlSC and ArlR with Kd = 2.1 μM. b MST showing the binding of P-ArlR to its own promoter with Kd = 6.4 μM. The experiment was repeated thrice. Error bars represent the average difference between the data points and their mean CarRS two-component system has recently been proved to be necessary for the coaggregation of F. nucleatum ATCC 25,586 [51]. In addition, none of the other three two-component systems in F. nucleatum has been reported. We next used multiple sequence alignments to identify other critical two-component system proteins of this strain. The results showed that the ArlR in F. nucleatum ATCC 25,586 has a high degree of sequence similarity to other reported ArlR response regulatory proteins (Fig. 1). To investigate the function of ArlRS from F. nucleatum, we successfully transformed the recombinant plasmid into the E. coli host strain BL21 (DE3) competent cells; however, the expression of ArlR protein was low (Fig. 2a, line  4). To find the optimal expression conditions of ArlR protein in vitro, the recombinant plasmid was transformed into five different host strains. The highest expression condition of ArlR was reported in E. coli BL21-CodonPlus (DE3) RIL host strain (Fig. 2). This may be due to E. coli BL21-CodonPlus (DE3) RIL possessing rare tRNAs for expressing AT-rich genomes [50]. In the process of protein purification, we found that the ArlR protein degraded rapidly and through adding 5% glycerol to the buffer can effectively prevent protein degradation. In addition, we recommend using smaller gel filtration columns and faster flow rates, which can significantly reduce the degradation and improve the purity of ArlR.
It is known that the kinase including the dimerization and histidine phosphotransfer (DHp) domain and the catalytic ATP-binding (CA) domain can bind the soluble REC domain [51]. Similarly, we observed that phosphorylated histidine kinase ArlSC can interact with its corresponding response regulator ArlR (Fig. 4b). We selected a fixed concentration of ArlSC protein and ATP for the kinase reaction and added equal volumes of different concentrations of ArlR protein to these kinase reaction systems for the phosphotransferase reaction. The luminescence intensity of the reaction solution gradually decreased with the increasing amount of the ArlR protein added, indicating that the ArlR response regulator protein accepted the phosphate group of the phosphorylated ArlSC histidine kinase protein, which caused the reaction to proceed in the direction of ATP consumption. To further confirm the interaction between phosphorylated ArlSC and ArlR protein, we conducted an MST experiment (Fig. 5a), and the results were close to previous reports in which the binding affinities of cognate HK-RR pairs are approximately 1-2 μM [51].
Lastly, this study conducted preliminary research on the function of the ArlR response regulator protein. Because some of the two-component system response regulator proteins can bind their own operons to regulate the transcription of their own genes [52,53], we explored whether the ArlR protein could bind its own operons. We used MST to determine the affinity constants of the regions on both sides of the promoter to the ArlR protein, and the ArlR protein can only bind to the operon on the side close to the ArlR two-component signal system (Fig. 5b). This value is consistent with the affinity constant of the ArlR protein and its corresponding DNA sequence in Staphylococcus aureus [54]. Overall, we concluded that the ArlR protein can regulate the transcription of the ArlRS two-component system; however, the related regulatory mechanisms and other functions of this protein, such as whether to regulate the transcription of other genes and its impact on bacterial virulence and survival, still need to be studied further.

Conclusion
Altogether, we identified four groups of TCS in F. nucleatum strain ATCC 25,586 and successfully cloned the arlR gene into pET-28a by restriction-free (RF) cloning method. We obtained ArlR with a purity higher than 90% and concentration above 15 mg/mL using