The bacterial strains and plasmids used in this study are listed in Supplementary Table S1. Escherichia coli and V. cholerae strains were cultured at 37°C in Luria-Bertani (LB) broth supplemented with appropriate antibiotics. V. vulnificus strains were cultured in LB broth or thiosulfate citrate bile salt sucrose (TCBS) agar at 30°C. When necessary, either ferrous sulfate (25 μM) as an iron source or 2,2′-dipyridyl (100 μM) as an iron chelator was added exogenously to the LB broth when the A₆₀₀ value of the culture reached approximately 0.1. Antibiotics were used at the following concentration: For E. coli, ampicillin 50 μg/mL, kanamycin 25 μg/mL, tetracycline 10 μg/mL, chloramphenicol 25 μg/mL; for V. vulnificus and V. cholerae, kanamycin 100 μg/mL, tetracycline 2 μg/mL, chloramphenicol 2 μg/mL. All media used in this study were purchased from Difco (MI, United States). All reagents and antibiotics were purchased from Sigma Aldrich (MO, United States).

To construct a fur deletion derivative, Δfur, the primers Δfur_FF_xbaI and Δfur_FR_speI (Supplementary Table S2) were used for amplification of the upstream region of fur, and Δfur_BF_speI and Δfur_BR_xhoI for the downstream region of fur. The PCR products were cloned to the predigested suicide vector pDM4. The resulting plasmid was mobilized from S17-1 λpir to the wild type V. vulnificus MO6-24/O strain by conjugation. A double crossover was selected in LB plated with 10% sucrose. Colonies that grew on sucrose plate but sensitive to chloramphenicol were selected. The mutation was confirmed through PCR and DNA nucleotide sequencing.

Derivatives of V. vulnificus MO6-24/O or its fur-deletion isotype, Δfur, harboring the luxAB reporter transcriptionally fused to each of leuO, vhuα, vhuβ, rpoS, and katG were described previously (Kim et al., 2018). Overnight cultures of tested strains harboring the luxAB reporter fusions grown in LB were inoculated into fresh LB medium. To make an iron-limiting condition, 100 μM 2,2′-dipyridyl was added exogenously to the LB broth when the A₆₀₀ value of the culture reached approximately 0.1, and samples were diluted 125-fold with LB broth. At various growth stages, 0.006% (v/v) n-decylaldehyde (in 50% ethanol) was added and luminescence was measured using a microplate reader (Mithras LB 940; Berthold, Bad Wildbach, Germany) as previously described (Wen et al., 2012). The specific transcription level was expressed as relative light units (RLU) normalized to cell density.

The 1,540-bp DNA fragment of the leuO upstream region (−975 to +565 with respect to the translation start site) was amplified by PCR using the primers leuO_DCO_F and leuO_DCO_R. The resulting product was ligated to the pGEM-T Easy vector to construct pGEM-LeuO. To introduce mutations into each of the four regions (SM1 ~ 4) containing putative Fur boxes, mutagenesis was performed using primer sets leuO_SDM_1_F and leuO_SDM_1_R for SM1, leuO_SDM_2_F and leuO_SDM_2_R for SM2, leuO_SDM_3_F and leuO_SDM_3_R for SM3, and leuO_SDM_4_F and leuO_SDM_4_R for SM4. The resulting four plasmids were named pGEM-SM1 through pGEM-SM4, respectively.

The Strep-tagged Fur was expressed in E. coli BL21(DE3) cells harboring the fur clone pASK-IBA7-Fur (Kim et al., 2013b) by induction with 0.2 μg/mL anhydrotetracycline. After centrifugation, bacterial pellets were resuspended in a buffer (100 mM Tris-Cl, 150 mM NaCl, and 1 mM EDTA, pH 7.5). The cells were then sonicated and centrifuged at 4,585 × g for 10 min. The resulting supernatant was subjected to purification using Strep-Tactin affinity resin (IBA BioTAGnology, Göttingen, Germany), and specifically bound protein was eluted with E buffer (100 mM Tris-Cl, 150 mM NaCl, and 1 mM EDTA, and 2.5 mM desthiobiotin, pH 7.5) according to the manufacturer’s instructions. The eluted protein was separated on a 12% SDS-PAGE to assess the purity. The purified Fur protein was dialyzed using Spectra/Por molecular porous membrane tubing (molecular weight cutoff of 10,000; Spectrum Laboratoried Inc., Rancho Dominguez, CA) with A buffer (50 mM Tris-Cl, 100 mM NaCl, 1 mM MgCl₂, and 2 mM dithiothreitol, pH 8.0). The protein was concentrated using the Vivaspin 6 instrument (Vivagen, Seoul, Korea). The protein concentration was determined by the Bradford method (Bradford, 1976).

To assess the binding of Fur to upstream region of leuO, the 299-bp regions (−288 to +11 with respect to the translation start site) of the gene containing wild type and mutated bases at each of Fur box candidates (SM1 ~ 4) were amplified by PCR with the primers leuO_EMSA_F and ³²P-labeled leuO_EMSA_R using pGEM-leuO and pGEM-SM1 ~ 4 as template DNAs. Similarly, to assess the binding of Fur on the region upstream to vhuα, the 303-bp regions (−261 to +42 with respect to the translation start site) were amplified by PCR with the primers HU_alpha_EMSAF and ³²P-labeled HU_alpha_EMSAB. In the same way, for vhuβ used the 257-bp regions (−212 to +45 with respect to the translation start site); for rpoS, the 650-bp regions (−556 to +94 with respect to the translation start site); and for katG, the 495-bp regions (−401 to +94 with respect to the translation start site) were amplified by PCR with the each primer set HU_beta_EMSAF and ³²P-labeled HU_beta_EMSAB, rpoS_EMSA_longF and ³²P-labeled rpoS_EMSA_R, katG_EMSA_F and ³²P-labeled katG_EMSA_R, respectively.

For gel shift assays, 10 ng of the labeled probe was incubated with increasing amounts of purified Fur protein in a 20 μL reaction mixture in binding buffer (Alice et al., 2008) containing 10 mM Tris-borate (pH 7.5), 100 μg/mL bovine serum albumin, 5% (v/v) glycerol, 40 mM KCl, 1 mM MgCl₂, and 1 μg poly(dI-dC). The reaction mixture was supplemented with either 1 mM MnCl₂ or 1 mM EDTA for 30 min at 30°C. The resulting mixtures were resolved in a 6% neutral polyacrylamide gel. Each of the labeled probes (10 ng) was incubated with increasing amounts of purified Fur protein, and gel shift assays were performed as described above. The gels were exposed to a BAS_MP 2040s imaging plate (Fujifilm, Tokyo, Japan) and scanned using a BAS-1500 instrument (Fujifilm).

An end-labeled 405-bp DNA fragments of the leuO upstream region (−376 to +29 with respect to the translation start site) was amplified using the primers leuO_fp_F2 and 6-FAM labeled leuO_fp2_R_FAM. To determine the Fur binding site, 200 ng of the amplified leuO upstream region was incubated with purified Fur (2 μM) in 50 μL of binding buffer containing 10 mM Tris-borate (pH 7.5), 100 μg/mL bovine serum albumin, 5% (v/v) glycerol, 40 mM KCl, 1 mM MgCl₂, 1 μg poly(dI-dC) for 30 min at 30°C. After incubation, 0.01 unit of DNase I (Promega, Madison, WI) was added, and the reaction mixture was incubated at 37°C for 1 min. The reaction was terminated by the addition of 5 μL of RQ1 DNase stop solution (Promega, Madison, WI), and inactivated at 65°C for 10 min. To precipitate the samples, 55 μL phenol-chloroform was added, and the mixture was precipitated at room temperature for 5 min. After centrifugation at 11,323 × g at 4°C for 5 min sodium acetate (NaCOOH, pH 5.2) was added to 50 μL of supernatant. Following the addition of 100 μL of 100% ethanol, the mixture was incubated for 1 h at −80°C and then centrifuged in 11,323 × g at 4°C for 15 min. The sample was then washed with 70% ethanol, dried in water bath at 60°C, and dissolved in 10 μL of distilled water. DNA sequencing of the sample was carried out in GBST (Green-Bio Science and Technology, Seoul National University). Raw data obtained through ABI 3730xl were analyzed using the Peak Scanner software (Applied Biosystems, Waltham, MA, United States).

β-galactosidase activity was measured as described previously (Miller, 1972). Briefly, V. vulnificus strains were cultured overnight in LB medium, harvested, and diluted to an A₆₀₀ of 0.005, and β-galactosidase activities from cells harboring the genes transcriptionally fused with lacZ as described above was measured.

RNA was isolated from V. vulnificus using the easy-BLUE™ total RNA extraction Kit (iNtRON Biotechnology, Seongnam, Korea) and treated with the RNase-free DNase set (Promega, Madison, WI, United States) to remove any residual DNA. The purified RNA was quantified using a Biophotometer (Eppendorf, Hamburg, Germany). Subsequently, cDNA was synthesized from 500 ng of RNA using the CellScript^™ All-in-One cDNA Master Mix (Cellsafe, Yongin, Korea) following the manufacturer’s instructions. One microliter of cDNA was used for RT-PCR analysis on a Stratagene Mx3000p qPCR machine (Agilent Technologies, Santa Clara, CA, United States) using QGreenBlue 2 × Green qPCR Master Mix (Cellsafe, Yongin, Korea). The RT-PCR reactions were performed in triplicate in a 96-well plate using primer shown in Supplementary Table S2. The PCR conditions used to amplify all genes were: 10 min at 95°C and 40 cycles of 95°C for 15 s and 64°C for 40 s. The genes encoding type I glyceraldehyde-3-phosphate dehydrogenase (RS_10395) and DNA-directed RNA polymerase subunit alpha (RS_13660) of V. vulnificus were used as endogenous loading controls. Quantification was carried out using the Light Cycle 480 II real-time PCR system software program.

V. vulnificus strains (MO6-24/O, Δfur, Δfur complemented with pRK415-fur, ΔkatG) were cultured overnight in LB medium, washed, and sub-cultured in fresh LB medium. All cells were harvested at exponential phase (A₆₀₀ value of approximately 1.0) and sonicated using Ultrasonic Homogenizer (KUS-650, KBT, Seongnam, Korea) in 10 cycles with 1 s of sonication followed by 2 s of rest for each cycle. After centrifugation, the supernatant was concentrated using Amicon Ultra-0.5 mL Centrifugal Filter Units (10,000 NMWL, UFC501024, Merck-Millipore, Germany). Each of 50 μL sample was mixed with 10 mM hydrogen peroxide (Duksan, Ansan, Korea). The reaction tubes were vortexed, incubated for 2 min at 37°C, and 600 μL of working solution was added. The working solution consisted of 100 mL cobalt (II) solution (20.3 g / 1 L DIW), 100 mL sodium hexametaphosphate solution (10 g / 1 L DIW), 800 mL sodium bicarbonate solution (180 g/ 2 L DIW) (Hadwan, 2018). The tubes were vortexed for 5 s and then kept at room temperature for 10 min in the dark. After 10 min, the catalase activities were measured at 440 nm using a Multimode Plate Reader (PerkinElmer, Waltham, MA, United States).

The flow rate was set to 1.0 mL/min, and the cFP peak was detected at 256 nm. As a reference, a 1 mM cFP solution (from a company) dissolved in 30% methanol was also prepared and analyzed using the same HPLC method. The amount of cFP in each sample was estimated based on the area of the cFP peak observed in the 1 mM cFP reference sample.

After culturing cells in LB broth for 24 h, a 20 mL supernatant of each sample was collected and mixed with an equal volume of ethyl acetate, and then finally dissolved in 150 μL of 30% methanol. The cFP in each sample was resolved using high-performance liquid chromatography (HPLC) using a C₁₈-reverse phase column Mightysil RP-18 GP (Kanto, Tokyo, Japan) with 30% methanol as the mobile phase. The flow rate was set to 1.0 mL/min, and the cFP peak was detected at 256 nm. As a reference, 1 mM of cFP (Bachem, Bubendorf, Switzerland) dissolved in 30% methanol were also prepared in the same way, and analyzed using the same HPLC methods. The amount of cFP in each sample was estimated based on the area of the cFP peak observed in the 1 mM cFP reference sample.

We initially examined the effect of Fur-iron on the transcription of toxR encoding inner membrane porin, which serves as receptor for cFP (Park et al., 2006, 2019). Using a lacZ-fusion, we quantitatively compared the transcription level of ToxR in wild type and Δfur (fur-deletion isotype) strains in the presence or absence of iron. However, no significant difference was observed between two groups of cells (data not shown). Furthermore, through a gel-shift assay using 363-bp region upstream to toxR as a probe (Supplementary Figure S1A) and purified ToxR, we determined that Fur does not bind to the upstream region, regardless of iron availability (Supplementary Figure S1B). These results indicate that the expression of ToxR is not influenced by iron.

In our previous study (Park et al., 2019), we investigated the regulatory elements in the region upstream to leuO, which encodes the master regulator for the cFP-signaling pathway in V. vulnificus. Within this region, we identified a putative Fur-binding sequence, suggesting that Fur may be involved in the regulation of leuO expression. To examine this possibility, we constructed the luxAB reporter gene fusions with leuO in both wild-type MO6-24/O strain and the Δfur, and quantitatively measured luciferase activities in the presence or absence of iron. When iron was supplied, the RLU values in the Δfur mutant were approximately three times higher than these in the wild-type strain (Figure 1). Under iron-limited conditions, there was no significant difference in leuO expression was observed between the wild-type strain and Δfur mutant. It is noteworthy that, even in Δfur cells, the expression of leuO was still repressed to some extent by iron, albeit significantly less compared to the wild type cells. This suggests the possible involvement of unidentified factor(s) in the iron-dependent regulation of leuO.

The leuO gene encodes a key regulatory component in cFP – signaling pathway which is closely associated with virulence of pathogenic Vibrio species, and hence its regulation is important for the pathogenicity. Therefore, we defined in detail the cis-acting elements for Fur in the gene. There exist four regions with nucleotide sequences homologous to the Fur box (Wen et al., 2016) in the leuO upstream region, named PF (putative Fur box) sites 1 ~ 4 (Figure 2A). Meanwhile, gel shift assay demonstrated that purified Fur binds to the upstream region of leuO and it appears that at least two shifts occurred (Supplementary Figure S2A). To determine which sites of these putative Fur-binding sites are actually bound by Fur, we performed competition gel-shift assay using a radiolabeled 299-bp oligomer, as described in Materials and Methods, as a probe and purified Fur. The probe was synthesized by PCR using leuO_EMSA_F and leuO_EMSA_R as primers (Supplementary Table S2). We then prepared four DNA oligomers of the same size with mutations in each of those four PF sites (named SM1 ~ 4) as shown in Figure 2A. Each of these mutated oligomers, without radiolabeling, was added as a competitor at about 10-fold higher concentration than the radiolabeled probe oligomer. As shown in the lane 3 of Figure 2B, the competitor without any mutation completely outcompeted the binding of Fur for the probe. The competitor without any mutations in each of the PF1 and PF3 sites also hindered the binding to Fur to the probe (lanes 4 and 6). In contrast, the competitor oligomer with mutations in each of PF2 or PF4 sites did not significantly interfere in the binding of Fur onto the probe oligomer, while the competitor with mutations in PF2 slightly interfered in the binding, but much less than the one with mutation in PF4. In addition, footprinting experiments also show high binding affinity to the PF2 and PF4 regions (Supplementary Figure S2B). These results suggested that Fur strongly binds to the PF4 region, and weakly to the PF2 region.

In a previous study, it was demonstrated that LeuO has a negative regulatory effect on two genes, vhuα and vhuβ, which encode histone-like proteins that are involved in the regulation of numerous genes (Balandina et al., 2002; Grove, 2011; Kim et al., 2018). To investigate the potential regulation exerted by Fur-iron on these genes, we examined their transcriptional levels using lacZ transcriptional fusions in both wild-type and Δfur cells under iron-rich and iron-depleted conditions (Figure 3). In wild-type cells, the transcription of both vhuα and vhuβ genes was repressed in the presence of iron. However, in the fur-deletion mutant cells, the expression of these genes was derepressed under iron-depleted conditions. In the presence of iron, the expression was still repressed, but to a significantly lesser extent than in wild-type cells. To further investigate the binding of Fur to the upstream regions of these two genes, we performed gel-shift assays using purified Fur in the presence or absence of iron. The results showed that Fur only binds to the upstream regions when iron is present (Supplementary Figure S3). These findings suggest that Fur-iron complex plays a role in the regulation of vhuα and vhuβ genes, and its binding is dependent on the availability of iron.

We further investigated the impact of iron on the expression of rpoS, which encodes an alternative sigma factor and is involved in the cFP pathway. The activity of β-galactosidase, measured from a lacZ fusion to rpoS, was found to be strongly repressed by iron in wild-type cells. However, in the fur-deletion mutant, the activity was derepressed regardless of iron availability (Figure 4). To confirm the direct binding of Fur to the upstream region of rpoS, we performed a gel-shift assay, and the results are shown in Supplementary Figure S4.

Our previous study demonstrated that katG, which encodes a catalase, is a member of the cFP-signaling regulon, and its transcription is regulated by RpoS (Kumamoto and Vukich, 1998; Kim et al., 2013b). Consequently, we investigated the direct regulation of the katG gene by Fur. To assess the impact of Fur on katG transcription, we utilized a katG-lacZ transcription fusion. The β-galactosidase activity from the reporter fusion in the Δfur mutant was higher than that in the wild-type strain under iron-depleted conditions (Figure 5). In the fur-deletion mutant, the expression of katG was derepressed in the absence of iron; however, in the presence of iron, it remained repressed. Fur-iron repression occurs through direct binding to the upstream region of the gene (Supplementary Figure S5).

RpoS is an alternative sigma factor known to initiate the transcription of numerous genes, forming a regulon. This suggests that these downstream genes may also be regulated by Fur-iron through modulation of RpoS expression. To investigate this, we measured the expression of three representative genes, aldA (encoding an aldehyde dehydrogenase), vvpE (encoding a metalloprotease), and gabD (encoding a succinate-semialdehyde dehydrogenase), known to be transcribed by RpoS in V. vulnificus (Jeong et al., 2001; Vijayakumar et al., 2004; Kim et al., 2018). The expression levels of these genes were semi-quantitatively measured using qRT-PCR. As we predicted, the results showed that all three genes were repressed by Fur in the presence of iron. In Δfur isotype cells, the repression was relieved, and the introduction of exogenous fur in plasmid restored the repression (Figure 6). The nucleotide sequences in the regions upstream of these three genes did not show any apparent Fur box (Supplementary Figure S7), suggesting that the Fur-dependent repression of these genes was exerted via RpoS.

We expanded our investigation to examine the production of cFP, which serves as the signal molecule of the ToxR-LeuO-dependent QS system (Kim et al., 2018). To compare the levels of cFP in the culture supernatants, we employed HPLC. We found that the amounts of cFP detected in both wild-type cells and fur-deletion cells did not exhibit a significant difference (Supplementary Figure S6).

The katG gene is a target virulence factor regulated by the cFP-mediated QS pathway. As shown in Figure 6, the transcription of katG, which encodes a catalase, is negatively controlled by Fur-iron. To quantitatively measure catalase activity, we performed assays on MO6-24/O, Δfur, and Δfur(pRK-fur) cells following the procedure outlined in the Materials and methods section. In Δfur cells, a higher level of catalase activity was observed compared to wild-type cells. However, the introduction of the fur gene on a plasmid reduced the activity to the wild-type level (Figure 7). This finding is consistent with the results presented in Figure 5, suggesting that Fur represses the catalase activity of V. vulnificus by regulating katG expression.

The cFP-mediated QS pathway is also present in the human pathogen V. cholerae (Park et al., 2006; Bina and Bina, 2010; Bina et al., 2013). In this pathogen, LeuO, which receives the cFP signal through ToxR similar to V. vulnificus, positively regulates the expression of ctxAB, the genes encoding the exotoxin cholera toxin, as well as ToxT, which induces the expression of various virulence-associated genes (Silva and Benitez, 2016; Bhandari et al., 2021; Bina and Bina, 2023). To explore whether Fur-iron also regulates these genes involved in the cFP pathway in V. cholerae, we compared the transcription levels of leuO, toxT, and ctxAB in the V. cholerae O1 El Tor strain N16961 and its fur-deletion derivative using qRT-PCR. As shown in Figure 8A, the expression levels of these four genes were significantly higher in the fur-deletion mutant compared to wild-type cells. Additionally, the transcription level of ctxA, which encodes the effector subunit of the cholera toxin, was quantitatively assessed using the luxAB reporter fusion (Figure 8B). Deletion of fur significantly enhanced ctxA transcription, and the introduction of the fur gene in trans complemented the mutant phenotype, confirming the repression of ctx by Fur.