We previously reported that NO-induced HmpA encoded by VVMO6_RS01375 is crucial for survival under host-derived nitrosative stress and pathogenesis of V. vulnificus during infection (Kim et al., 2019). Notably, we further found that the expression of VVMO6_RS01380, which is divergently transcribed from hmpA (Figure 1A), is also induced by NO (Kim et al., 2019). VVMO6_RS01380 encodes an Rrf2-family transcriptional regulator showing an amino acid sequence homology to E. coli NsrR, S. Typhimurium NsrR, and S. coelicolor NsrR (61, 62, and 35% identity, respectively) (Figure 1B). Moreover, the protein encoded by VVMO6_RS01380 contains three conserved cysteine residues, C91, C96, and C102, which are known to be essential for the Fe-S cluster ligation of Rrf2-family transcriptional regulators (Figure 1B) (Volbeda et al., 2017). This observation led us to designate the VVMO6_RS01380 gene product as an Fe-S cluster-containing transcriptional regulator NsrR.

For the comprehensive identification of NsrR-regulated genes in V. vulnificus, the transcriptomes of the wild-type and ΔnsrR strains were compared by RNA-seq. The transcriptome analysis revealed that, in total, 47 genes were differentially expressed by the nsrR deletion: 44 genes were up-regulated and 3 genes were down-regulated (Figure 1C, Supplementary Table 1). The overall fold changes of the up-regulated genes were greater than those of the down-regulated genes. This result implies that NsrR serves mainly as a repressor rather than as an activator. Intriguingly, the up-regulated genes included several genes that are predicted to encode proteins involved in the defense against nitrosative stress such as NO dioxygenase HmpA, NO detoxification protein NnrS, NO2- reductase large subunit, NO2- reductase small subunit, cytochrome c NO2- reductase subunit c₅₅₂ NrfA, and NO reductase transcriptional regulator NorR (Stern et al., 2012; Kim et al., 2019). Among them, hmpA was the most highly up-regulated gene in the ΔnsrR strain (Figure 1C), suggesting that NsrR is a strong repressor of hmpA expression. Meanwhile, the down-regulated genes, iscR, iscS, and iscU, constitute the isc operon (iscRSUA-hscBA-fdx) encoding proteins required for the biogenesis of the Fe-S cluster (Lim and Choi, 2014). Taken together, this result shows that NsrR controls the expression of multiple genes involved in nitrosative stress responses, especially hmpA.

To validate the RNA-seq results and examine whether NsrR mediates the induction of hmpA in response to NO, the hmpA transcript levels in the wild-type and ΔnsrR strains were compared under nitrosative stress in vitro and ex vivo. The hmpA transcript level in the wild-type strain was significantly elevated upon exposure to an in vitro NO donor, NO/PPNPs (NO-releasing poly(lactic-co-glycolic acid)-polyethylenimine nanoparticles) (Figure 2A) (Nurhasni et al., 2015). This result confirms our previous observation that hmpA is induced by NO (Kim et al., 2019). Additionally, the hmpA transcript level was dramatically increased in the ΔnsrR strain compared with that in the wild-type strain even in the absence of NO/PPNPs (Figure 2A), verifying that NsrR negatively regulates hmpA. Strikingly, the hmpA transcript level in the ΔnsrR strain was not affected by the addition of NO/PPNPs (Figure 2A), indicating that NsrR recognizes NO and alleviates the repression of hmpA expression in vitro.

The role of NsrR in hmpA expression was further investigated ex vivo using NO-producing murine macrophage RAW 264.7 cells. As shown in Figure 2B, the hmpA transcript level in the wild-type strain exposed to NO-producing RAW 264.7 cells was considerably elevated compared with that exposed to Dulbecco's modified Eagle's medium (DMEM; negative control). The extent of the increase in the hmpA transcript level upon exposure to the RAW 264.7 cells diminished by the addition of the NO synthase inhibitor L-N^G-monomethyl arginine citrate (L-NMMA) (Figure 2B). This result suggests that the hmpA induction upon exposure to RAW 264.7 cells is attributable to NO produced by the murine macrophages. In contrast, the highly increased hmpA transcript level in the ΔnsrR strain was not altered by the RAW 264.7 cells and L-NMMA (Figure 2B), confirming that NsrR mediates the derepression of hmpA under nitrosative stress derived from host immune cells. The combined results show that NsrR has a critical role to sense NO and to induce the hmpA expression both in vitro and ex vivo.

Then, we examined whether the introduction of recombinant nsrR can reduce the increased hmpA transcript level in the ΔnsrR strain. Introduction of a nsrR-expressing plasmid significantly decreased the hmpA transcript level, although it was not comparable with that in the wild-type strain (Supplementary Figure 1A). One possible explanation for this lack of complementation is that the recombinant NsrR expressed from the exogenous plasmid is less functional for unknown reasons. On the other hand, ectopic expression of nsrR on the chromosome effectively reduced the hmpA transcript level comparable with that in the wild-type strain (Supplementary Figure 1B). Similarly, the HmpA protein levels in the ΔnsrR strain were highly increased compared with those in the wild-type strain and significantly decreased by complementation (Supplementary Figures 1C,D). Altogether, the results suggest that NsrR is a major transcriptional regulator that recognizes NO and regulates hmpA expression mainly at the transcription level.

As shown in Figure 1B, V. vulnificus NsrR contains three conserved cysteine residues (C91, C96, and C102) that are predicted to act as ligands of the NO-responsive Fe-S cluster (Tucker et al., 2008; Volbeda et al., 2017). To investigate the role of these three cysteine residues, three different strains were constructed: a parent strain GR204 chromosomally encoding 3 × FLAG-tagged NsrR (NsrR^FLAG), an isogenic nsrR-deletion mutant, and an isogenic nsrR_3CS mutant chromosomally encoding apo-locked NsrR^FLAG (NsrR3CSFLAG) (see see Materials and Methods for a detailed description). The hmpA transcript and HmpA protein levels in the ΔnsrR strain were highly elevated compared with those in the parent strain (Figures 3A,B), indicating that NsrR^FLAG in the parent strain is still functional as a repressor of hmpA. Notably, the hmpA transcript and HmpA protein levels in the nsrR_3CS strain were comparable with those in the ΔnsrR strain (Figures 3A,B). Moreover, a similar effect of the mutation in the three cysteine residues and the nsrR deletion on hmpA expression was observed in the wild-type background (Supplementary Figure 1E). These results that NsrR_3CS cannot repress the hmpA transcription reveal that coordination of the Fe-S cluster by the three cysteine residues is essential for the NsrR activity to repress hmpA.

Furthermore, the NsrR3CSFLAG protein level in the nsrR_3CS strain was significantly elevated compared with the NsrR^FLAG protein level in the parent strain (Figure 3B). This observation prompted us to examine the activity of the nsrR promoter (P_nsrR, determined in Figures 4B,C) in the wild-type, ΔnsrR, and nsrR_3CS strains using the P_nsrR-luxCDABE transcriptional fusion reporter. The P_nsrR activity in the ΔnsrR strain was higher than that in the wild-type strain (Figure 3C), demonstrating that NsrR represses its own transcription. Additionally, the increased P_nsrR activity in the nsrR_3CS strain was comparable with that in the ΔnsrR strain (Figure 3C). This result suggests that NsrR relieves the repression of its own transcription by the mutation in the three cysteine residues and the consequent loss of the Fe-S cluster. Combined with the previous data (Figure 2), we propose a model in which holo-NsrR containing the Fe-S cluster represses both hmpA and nsrR transcription, shifts to the clusterless apo-form under nitrosative stress, and then alleviates the repression of hmpA and nsrR transcription.

To map the hmpA promoter, the transcription start site (TSS) of hmpA was determined by primer extension analysis. A single reverse transcript was produced from the primer extension of RNA isolated from the ΔnsrR strain grown to an A₆₀₀ of 0.5 (Figure 4A). This reverse transcript observed in the ΔnsrR strain was not detected in the wild-type strain (Figure 4A), confirming that the hmpA transcription is strongly repressed by NsrR. The 5′-end of hmpA was located 57-bp upstream of the translation start codon of hmpA. Next, the TSS of nsrR was determined in a similar way. A single reverse transcript was produced from the primer extension of RNA isolated from the wild-type strain grown to an A₆₀₀ of 0.5 (Figure 4B). The 5′-end of nsrR was located 18-bp upstream of the translation start codon of nsrR. The putative promoters constituting the TSSs were named P_hmpA and P_nsrR to represent the hmpA promoter and the nsrR promoter, respectively. The sequences for putative −10 and −35 regions of each promoter were assigned based on the similarity to the consensus sequences of E. coli σ⁷⁰ promoters (Figure 4C). Strikingly, these results show that P_hmpA and P_nsrR overlap with each other. This overlapping promoter region was termed the nsrR-hmpA regulatory region for our further research.

To investigate whether NsrR directly binds to the nsrR-hmpA regulatory region, electrophoretic mobility shift assays (EMSAs) were performed. The addition of NsrR to 6-carboxyfluorescein (6-FAM)-labeled DNA probe encompassing the nsrR-hmpA regulatory region resulted in a single retarded band in an NsrR concentration-dependent manner (Figure 5A). The same unlabeled DNA fragment competed for NsrR binding in a dose-dependent manner (Figure 5A), confirming the specific binding of NsrR. Then, the binding of NsrR_3CS to the nsrR-hmpA regulatory region was compared with that of NsrR. The amount of the retarded band of the DNA-NsrR_3CS complex was reduced compared with that of the DNA-NsrR complex (Figure 5B). This result implies that the DNA-binding affinity of NsrR_3CS is considerably lower than that of NsrR, which leads to the derepression of hmpA and nsrR under nitrosative stress.

To determine the precise location of NsrR-binding site(s) in the nsrR-hmpA regulatory region, DNase I protection assays were performed using the same DNA probe. When NsrR was added to the DNA probe, NsrR protected a single region extending from −2 to +18 (NSRRB, centered at +8.5 from the TSS of hmpA) from DNase I digestion (Figures 4C, 5C). The sequence of NSRRB showed about 87% similarity to the 11-1-11 bp consensus NsrR-binding sequence in E. coli (Bodenmiller and Spiro, 2006; Partridge et al., 2009). Combined with the EMSA data (Figure 5A), these results indicate that NsrR concurrently represses hmpA and its own transcription by directly binding to the single specific sequence in the nsrR-hmpA regulatory region.

To determine other factors involved in the hmpA regulation, we further explored various known transcriptional regulators in V. vulnificus. Among them, the role of Lrp in the hmpA regulation was evaluated. The hmpA transcript and HmpA protein levels in the lrp-deletion mutant (Δlrp) were significantly increased compared with those in the parent strain and restored by complementation (Figures 6A,B). The hmpA transcript level in the wild-type strain was not altered by exogenous leucine (Supplementary Figure 2A), suggesting that Lrp negatively affects the hmpA transcription in a leucine-independent manner. To investigate the regulatory relationship between NsrR and Lrp, the lrp-deleted nsrR_3CS mutant (nsrR_3CSΔlrp), in which both NsrR and Lrp are not functional, was constructed from the parent strain. Interestingly, the hmpA transcript and HmpA protein levels in the nsrR_3CSΔlrp strain were comparable with those in the nsrR_3CS strain (Figures 6C,D). The observation that Lrp was not able to affect hmpA transcription in the absence of functional NsrR indicates that the negative effect of Lrp on hmpA is mediated by NsrR. This result led us to examine whether Lrp positively regulates the cellular level of NsrR to repress hmpA. However, both the P_nsrR activity and NsrR^FLAG protein level were not affected by the lrp deletion (Supplementary Figure 2B, Figure 6B).

Next, EMSAs were performed to investigate whether Lrp directly binds to the nsrR-hmpA regulatory region. The addition of Lrp to the DNA probe resulted in a single retarded band in an Lrp concentration-dependent manner (Figure 6E). The same unlabeled DNA fragment showed competition for Lrp binding in a dose-dependent manner (Figure 6E), demonstrating the specific binding of Lrp. DNase I protection assays revealed that Lrp largely protected a single region extending from −75 to −11 (LRPB, centered at −43 from the TSS of hmpA) from DNase I digestion (Figures 4C, 6F). Combined with the EMSA data (Figure 6E), these results indicate that Lrp binds directly and specifically to the nsrR-hmpA regulatory region. Notably, within the region protected by Lrp, a periodic pattern of reduced cleavage followed by short regions of enhanced cleavage was observed (Figure 6F). This pattern known as phased hypersensitivity implies DNA bending by a multimeric Lrp (Pul et al., 2007), suggesting that the Lrp multimer induces a conformation change of the nsrR-hmpA regulatory region. Moreover, EMSA with both NsrR and Lrp showed that NsrR and Lrp simultaneously bind to the nsrR-hmpA regulatory region, rather than displace each other (Supplementary Figure 2C). Altogether, the combined results propose that direct binding of Lrp to the nsrR-hmpA regulatory region does not alter the nsrR transcription but represses hmpA transcription presumably through the modification of the DNA conformation enhancing the hmpA repression by NsrR.

The role of CRP in the hmpA regulation was also explored. The hmpA transcript and HmpA protein levels in the crp-deletion mutant (Δcrp) were considerably decreased compared with those in the parent strain and restored by complementation (Figures 7A,B). In addition, the hmpA transcript level in the wild-type strain was decreased by exogenous glucose while that in the Δcrp strain was not affected (Supplementary Figure 3A). These results indicate that CRP has a positive effect on the hmpA transcription which is relieved in the presence of exogenous glucose. Then, we compared the hmpA transcript and HmpA protein levels in the parent strain, the nsrR_3CS strain, and the crp-deleted nsrR_3CS mutant (nsrR_3CSΔcrp). Similar to Lrp, the hmpA transcript and HmpA protein levels in the nsrR_3CSΔcrp strain were comparable with those in the nsrR_3CS strain (Figures 7C,D), suggesting that the positive effect of CRP on the hmpA transcription is also mediated by NsrR. Thus, we further examined whether the effect of CRP on hmpA expression results from the increased cellular level of NsrR. Notably, the P_nsrR activity and NsrR^FLAG protein level were significantly increased by the crp deletion (Supplementary Figure 3B, Figure 7B), showing that CRP acts as a repressor of nsrR transcription. Moreover, the Lrp protein level in the Δcrp strain was elevated compared with that in the parent strain as we observed previously (Figure 7B) (Lee et al., 2020). Accordingly, we hypothesized that CRP indirectly activates hmpA through the repression of both nsrR and lrp in a sequential manner.

To investigate whether CRP directly binds to the nsrR-hmpA regulatory region, EMSAs were performed. As shown in Figure 7E, the addition of CRP to the DNA probe resulted in a single retarded band in a CRP concentration-dependent manner. The same unlabeled DNA fragment competed for CRP binding in a dose-dependent manner (Figure 7E), confirming the specific binding of CRP. DNase I protection assays determined three regions protected by CRP extending from −10 to +10 (CRPB1, centered at −0.5 from the TSS of nsrR), −69 to −31 (CRPB2, centered at −50 from the TSS of nsrR), and −99 to −88 (CRPB3, centered at −93.5 from the TSS of nsrR) from DNase I digestion (Figures 4C, 7F). Combined with the EMSA data showing a single retarded band by CRP (Figure 7E), this result implies that CRP binds to CRPB1, CRPB2, and CRPB3 with similar DNA-binding affinities. Taken together, the combined results propose that CRP directly and specifically binds to the nsrR-hmpA regulatory region to repress nsrR as well as lrp, and consequently induces the hmpA transcription in a sequential cascade. In conclusion, the results in this study suggest that NsrR tightly regulates the hmpA transcription in response to NO, which could be elaborated by Lrp and CRP.

The strains and plasmids used in this study are listed in Supplementary Table 2. Unless otherwise noted, the V. vulnificus strains were grown aerobically in Luria-Bertani (LB) medium supplemented with 2% (w/v) NaCl (LBS) at 30°C, and their growth was monitored spectrophotometrically at 600 nm (A₆₀₀). When required, 3 μg/ml chloramphenicol was added to the media. To visualize the cellular NsrR protein levels, V. vulnificus GR204, which carries 3 × FLAG-coding sequence fused to the 3′-end of nsrR ORF on the chromosome, was constructed as a parent strain (Supplementary Table 2). The parent strain and its isogenic mutants were used to quantify the cellular NsrR protein levels. The murine macrophage RAW 264.7 cells were grown in DMEM containing 10% fetal bovine serum (VWR, Radnor, PA) and the antibiotics [100 units/ml penicillin G and 100 μg/ml streptomycin (Gibco-BRL, Gaithersburg, MD)] in air supplemented with 5% CO₂ at 37°C. To induce NO production, the RAW 264.7 cells were suspended in fresh DMEM containing 500 ng/ml E. coli O111:B4 lipopolysaccharide (Sigma, St. Louis, MO) and 1 mM L-arginine (Sigma) (Walker et al., 1997; Choi et al., 2020).

For construction of the isogenic deletion mutants, target genes were inactivated in vitro by deletion of each ORF using the PCR-mediated linker-scanning mutation method as described previously (Jang et al., 2016; Choi et al., 2020). Briefly, the deleted ORF fragment was amplified by PCR with appropriate primer pairs (Supplementary Table 3), and the resulting fragment was ligated into SphI-SpeI-digested pDM4 (Milton et al., 1996). E. coli S17-1 λpir (Simon et al., 1983) containing pDM4 with the desired insert was used as a conjugal donor to an appropriate V. vulnificus strain to generate the deletion mutant (Supplementary Table 2). The conjugation and isolation of the transconjugants were conducted using a method described previously (Choi et al., 2020). The lrp-deletion mutant ZW181 and the crp-deletion mutant DI0201 were constructed previously (Choi et al., 2002; Lee et al., 2020).

For construction of the parent strain GR204 encoding NsrR^FLAG on the chromosome, the 3 × FLAG-coding sequence was fused to the 3'-end of nsrR ORF by PCR using the primer pairs NSRR01-F and NSRR01F-R, or NSRR02F-F and NSRR02-R (Supplementary Table 3). The amplified fragment was cloned into pDM4, resulting in pGR2008 (Supplementary Table 2). E. coli S17-1 λpir containing pGR2008 was used as a conjugal donor to the ΔnsrR strain as described above to generate GR204 (Supplementary Table 2).

The three cysteine residues in NsrR (C91, C96, and C102) were replaced with serine to examine their regulatory function with the minimal structural change of NsrR. For construction of the nsrR_3CS strain DY192, three cysteine residues were substituted with serine in vitro by using the QuikChange® site-directed mutagenesis kit (Agilent Technologies, Loveland, CO) (Bang et al., 2012; Lim et al., 2014a). The complementary mutagenic primers listed in Supplementary Table 3 were used to create pDY1907 carrying the nsrR_3CS gene on pDM4 (Supplementary Table 2). E. coli S17-1 λpir containing pDY1907 was used as a conjugal donor to the ΔnsrR strain as described above to generate DY192, and the nsrR_3CS mutation in DY192 was confirmed by DNA sequencing. For construction of the 3 × FLAG-tagged nsrR_3CS strain GR217 (Supplementary Table 2), a similar method was adopted except using pGR2016 carrying 3 × FLAG-coding sequence fused to the 3′-end of nsrR_3CS ORF on pDM4 instead of pDY1907.

To complement the nsrR mutation with a plasmid-based system, the nsrR gene was amplified by PCR using the primer pair NSRRC-F and -R (Supplementary Table 3). The amplified fragment was cloned into the broad-host-range vector pJH0311 (Goo et al., 2006) to create pDY1702 (Supplementary Table 2). To complement the lrp and crp mutation, pZW1818 and pKK1502 carrying the lrp and crp gene on pJH0311, respectively, were used in this study (Supplementary Table 2) (Jang et al., 2017; Lee et al., 2020). The plasmids were transferred into appropriate mutants by conjugation as described above.

To complement the nsrR mutation by ectopic expression of nsrR on the chromosome, the nsrR regulatory region and its ORF was integrated into a cryptic lacZ gene by PCR using specific primer pairs listed in Supplementary Table 3 (Hall, 1999; Chodur et al., 2017). The amplified fragment was cloned into pDM4, resulting in pGR2007 (Supplementary Table 2). E. coli S17-1 λpir containing pGR2007 was used as a conjugal donor to the ΔnsrR strain as described above to generate GR203 (Supplementary Table 2).

To analyze the effect of the nsrR deletion on the V. vulnificus transcriptome, total RNA was isolated from biological duplicates of the wild-type and ΔnsrR strains, grown aerobically to an A₆₀₀ of 0.5 in M9 minimal media supplemented with 0.4% (w/v) glucose (M9G) and then exposed to PPNPs for 10 min (Nurhasni et al., 2015; Kim et al., 2019). The RNA was further purified by removing DNA using TURBO DNase (Ambion, Austin, TX), and mRNA was selectively enriched by depleting rRNA using a Ribo-Zero rRNA removal kit (Epicenter, Madison, WI) according to the manufacturer's instructions. Strand-specific cDNA libraries were constructed and sequenced using HiSeq 2500 (Illumina, San Diego, CA) as described previously (Lee et al., 2019). The raw sequencing reads were mapped to the V. vulnificus MO6-24/O reference genome (GenBank^TM accession numbers: CP002469 and CP002470, www.ncbi.nlm.nih.gov), and the expression level of each gene was calculated as a reads per kilobase of transcript per million mapped sequence reads (RPKM) value using EDGE-pro v1.3.1 (Estimated Degree of Gene Expression in PROkaryots) (Magoc et al., 2013). The RPKM values were normalized and analyzed statistically using DeSeq2 v1.26.0 to identify the differentially expressed genes (fold change ≥ 2 with p < 0.05) (Love et al., 2014). A heat map was generated by the Matplotlib python package using the RPKM-fold change for each gene (Hunter, 2007).

Relative transcript levels in the total RNA isolated from the V. vulnificus strains grown under various environmental conditions were determined by quantitative RT-PCR (qRT-PCR). In detail, V. vulnificus was grown to an A₆₀₀ of 0.5 in M9G and then exposed to either 0.15 mg/ml PPNPs (negative control) or NO/PPNPs for 10 min (Nurhasni et al., 2015; Kim et al., 2019). Additionally, V. vulnificus grown to an A₆₀₀ of 0.5 in LBS was exposed to DMEM (negative control) or RAW 264.7 cells at a multiplicity of infection 10 for 10 min in the presence or absence of 500 μM L-NMMA (Sigma), which is a known NO synthase inhibitor (Nathan and Hibbs, 1991; Choi et al., 2020). When necessary, V. vulnificus was grown to an A₆₀₀ of 0.5 in LBS with various amounts of L-leucine (Sigma) or 1% glucose (Sigma). Total RNA from the V. vulnificus cells was isolated and quantified using a RNeasy® Mini Kit (Qiagen, Valencia, CA) and a NanoDrop One^c Microvolume UV-Vis Spectrophotometer (Thermo Fisher Scientific, Waltham, MA), respectively. cDNA was synthesized from 500 ng of the total RNA with the iScript^TM cDNA synthesis kit (Bio-Rad, Hercules, CA). Real-time PCR amplification of the cDNA was performed with the Chromo 4 real-time PCR detection system (Bio-Rad) and specific primer pairs (Supplementary Table 3) as described previously (Jang et al., 2017). Relative expression levels were calculated with the 16S rRNA expression level as an internal reference for normalization (Jang et al., 2017).

For primer extension analysis, primers HMPAUP-R and NSRRUP-R (Supplementary Table 3) complementary to the coding region of hmpA and nsrR, respectively, were end-labeled with [γ-32P]-ATP and added to the RNA. The primers were extended with SuperScript II reverse transcriptase (Invitrogen, Carlsbad, CA). The cDNA products were purified and resolved on a sequencing gel alongside sequencing ladders generated from pDY1706 and pDY1707 (Supplementary Table 2) with the same primers, respectively. The plasmid pDY1706 was constructed by cloning the 219-bp hmpA upstream region extending from −120 to +99, amplified by PCR using a primer pair HMPAUP-F and -R (Supplementary Table 3), into pGEM-T Easy (Promega, Madison, WI). Similarly, pDY1707 carrying the 198-bp nsrR upstream region extending from −113 to +85 on pGEM-T Easy was constructed using a primer pair NSRRUP-F and -R (Supplementary Table 3). The primer extension product was visualized with the Typhoon FLA 7000 phosphorimager (GE healthcare, Menlo Park, CA).

To overexpress NsrR and NsrR_3CS, each ORF of nsrR and nsrR_3CS was amplified by PCR using specific primer pairs (Supplementary Table 3). The amplified fragments were cloned into pET-28a(+) (Novagen, Madison, WI) to create pEJ1902 and pEJ1903, respectively (Supplementary Table 2). The resulting His₆-tagged NsrR and NsrR_3CS were expressed in E. coli BL21 (DE3) and purified by affinity chromatography according to the manufacturer's instructions (Qiagen). The buffers used for NsrR and NsrR_3CS are as follows: 20 mM Tris-Cl (pH 8.0), 500 mM NaCl, and 5 mM β-mercaptoethanol; additional 10% glycerol for a lysis buffer; additional 20 mM imidazole for a wash buffer; additional 250 mM imidazole for an elution buffer; additional 50% glycerol for a dialysis buffer. To overexpress Lrp and CRP, pZW1903 carrying the lrp gene on pET-28a(+) and pHK0201 carrying the crp gene on pRSET A (Invitrogen) were used in this study (Supplementary Table 2) (Choi et al., 2002; Lee et al., 2020). The His₆-tagged Lrp and CRP were purified as described previously (Lee et al., 2020).

For Western blot analysis, V. vulnificus cells were lysed using B-PER^TM Bacterial Protein Extraction Reagent with Enzymes (Thermo Fisher Scientific), and residual cell debris was removed by centrifugation to obtain clear cell lysates. The protein levels of HmpA, Lrp, CRP, and DnaK in the clear cell lysates were determined as described previously (Kim et al., 2019; Lee et al., 2020). Similarly, cellular NsrR^FLAG protein was detected using Monoclonal ANTI-FLAG® M2 antibody produced in mouse (Sigma).

A 393-bp nsrR-hmpA regulatory region (−262 to +131 from the TSS of nsrR) was amplified with the primer PnsrR-F carrying a SacI restriction site and PnsrR-R carrying a SpeI restriction site (Supplementary Table 3). The resulting DNA fragment was cloned into the SacI-SpeI-digested pBBR-lux carrying the promoterless luxCDABE genes to create pGR2025 (Supplementary Table 2) (Lenz et al., 2004). pGR2025 was transferred into the V. vulnificus strains by conjugation as described above. The cellular luminescence and growth (A₆₀₀) of each strain grown to an A₆₀₀ of 0.5 in LBS were measured using a microplate reader (Infinite^TM microplate reader, Tecan, Männedorf, Switzerland), and RLUs were calculated by dividing the luminescence with the A₆₀₀ (Lee et al., 2019).

For the EMSAs, a 393-bp nsrR-hmpA regulatory region (−186 to +207 from the TSS of hmpA, equivalent to −262 to +131 from the TSS of nsrR) was amplified by PCR using 6-FAM-labeled PnsrRhmpA-F and -R as primers (Supplementary Table 3). Similarly, a 321-bp isc operon regulatory region [−194 to +127 from the TSS of isc operon (Lim et al., 2014b)] was amplified by PCR using 6-FAM-labeled Pisc-F and -R as primers (Supplementary Table 3). The 6-FAM-labeled DNA probe (10 nM) was then incubated with purified NsrR or CRP for 30 min at 30°C in a 20-μl reaction mixture containing 1 × NsrR binding buffer (10 mM Tris-Cl (pH 8.0), 10 mM KCl, 1 mM DTT, and 100 μg BSA; additional 1 mM cAMP only for CRP) and 0.1 μg of poly(dI-dC) (Sigma) as a non-specific competitor. Similarly, the DNA probe was incubated with purified Lrp or both NsrR and Lrp for 30 min at 30°C in a 20-μl reaction mixture containing 1 × Lrp binding buffer (50 mM Tris-Cl (pH 8.0), 20 mM KCl, 1 mM DTT, and 100 μg BSA, and 10% glycerol) and 0.1 μg of poly(dI-dC) (Sigma) as a non-specific competitor. For the competition analysis, various concentrations of unlabeled DNA fragment were added as a self-competitor to the reaction mixture before incubation. Electrophoretic analysis of the DNA-protein complexes was performed as described previously (Lee et al., 2020).

The same 393-bp nsrR-hmpA regulatory region was amplified by PCR using unlabeled PnsrRhmpA-F and 6-FAM-labeled PnsrRhmpA-R as primers for the DNase I protection assays (Supplementary Table 3). The binding of NsrR, Lrp, and CRP to the DNA probe (40 nM) was performed as described above, and DNase I digestion of the DNA-protein complexes followed the procedures described previously (Jang et al., 2017). The digested DNA products were precipitated with ethanol and eluted in sterilized H₂O, and then analyzed using an ABI 3730xl DNA analyzer (Applied Biosystems, Foster City, CA) with Peak Scanner^TM Software v1.0 (Applied Biosystems) (Hwang et al., 2019).

Average and standard deviation (SD) values were calculated from at least three independent experiments. Statistical analysis was performed by the Student's t-test using GraphPad Prism 7.0 (GraphPad Software, San Diego, CA).