The transcriptional regulator IscR integrates host-derived nitrosative stress and iron starvation in activation of the vvhBA operon in Vibrio vulnificus

For successful infection of their hosts, pathogenic bacteria recognize host-derived signals that induce the expression of virulence factors in a spatiotemporal manner. The fulminating food-borne pathogen Vibrio vulnificus produces a cytolysin/hemolysin protein encoded by the vvhBA operon, which is a virulence factor preferentially expressed upon exposure to murine blood and macrophages. The Fe-S cluster containing transcriptional regulator IscR activates the vvhBA operon in response to nitrosative stress and iron starvation, during which the cellular IscR protein level increases. Here, electrophoretic mobility shift and DNase I protection assays revealed that IscR directly binds downstream of the vvhBA promoter PvvhBA, which is unusual for a positive regulator. We found that in addition to IscR, the transcriptional regulator HlyU activates vvhBA transcription by directly binding upstream of PvvhBA, whereas the histone-like nucleoid-structuring protein (H-NS) represses vvhBA by extensively binding to both downstream and upstream regions of its promoter. Of note, the binding sites of IscR and HlyU overlapped with those of H-NS. We further substantiated that IscR and HlyU outcompete H-NS for binding to the PvvhBA regulatory region, resulting in the release of H-NS repression and vvhBA induction. We conclude that concurrent antirepression by IscR and HlyU at regions both downstream and upstream of PvvhBA provides V. vulnificus with the means of integrating host-derived signal(s) such as nitrosative stress and iron starvation for precise regulation of vvhBA transcription, thereby enabling successful host infection.

Upon entering the host, enteropathogenic bacteria inevitably encounter drastic environmental changes and cope with harsh conditions raised by the host immune defense system. The ability to recognize environmental changes and produce appropriate virulence factors is essential for the pathogens to survive and develop diseases in the host (1,2). Accordingly, pathogens have evolved sophisticated mechanisms to regulate the expression of specific virulence genes in response to various hostderived signals (1,2). In an effort to understand the virulence gene regulation, numerous transcriptional regulators have been characterized (3). Notably, a number of transcriptional regulators use an iron-sulfur (Fe-S) cluster as a cofactor to sense environmental signals. Because the Fe-S cluster is easily disrupted under host-like conditions such as oxidative and nitrosative stress and iron starvation, it allows pathogens to promptly recognize and adapt to the host environment (4,5).
The opportunistic human pathogen Vibrio vulnificus is a causative agent of food-borne diseases ranging from gastroenteritis to life-threatening septicemia (15,16). It has been reported that V. vulnificus exploits various transcriptional regulators such as the cAMP receptor protein (CRP) 4 , NanR, SmcR, and HlyU for the well-coordinated expression of its virulence factors (17)(18)(19)(20). CRP and NanR recognize depletion of specific nutrients (18,21), whereas SmcR senses increased cell density in the host to regulate virulence genes (19,22). In addition, IscR and HlyU are preferentially induced by host cells or in septicemic patients, respectively, and also involved in activa-tion of virulence genes in V. vulnificus (20,23,24). Particularly, IscR also regulates numerous virulence genes in other pathogens such as Pseudomonas aeruginosa, Yersinia pseudotuberculosis, and Erwinia chrysanthemi, highlighting the importance of IscR for bacterial pathogenesis (25)(26)(27).
Among the virulence factors of V. vulnificus, a cytolysin/hemolysin VvhA is an extracellular pore-forming toxin essential for its hemolytic activity (28,29). VvhA is a product of the vvhA gene, which is cotranscribed with vvhB encoding a chaperonelike protein required for the production of active VvhA (17,30). In the present study, we discovered that the vvhBA expression is highly induced in V. vulnificus exposed to murine blood and murine macrophage RAW 264.7 cells. To elucidate the regulatory mechanisms by which V. vulnificus increases vvhBA expression in host-like conditions, the exact role of IscR in vvhBA regulation was investigated. The vvhBA transcript and VvhA protein levels were compared in the WT and the isogenic iscR-deletion mutant (⌬iscR) under nitrosative stress and iron starvation. In addition, the combined effects of IscR, HlyU, and the histone-like nucleoid-structuring protein (H-NS) on vvhBA regulation were analyzed at the molecular levels (31, 32). Consequently, this study demonstrated that IscR is a sensor of hostderived nitrosative stress and iron starvation, and activates vvhBA transcription along with HlyU by relieving H-NS repression, contributing to the precise regulation of VvhA production during infection, which is essential for fitness and pathogenesis of V. vulnificus in the host.

Expression of vvhBA is induced upon exposure to murine blood and macrophages
In an effort to identify virulence genes significantly induced in V. vulnificus upon invasion of the host bloodstream, transcriptomes of the bacteria exposed to murine blood or M9 minimal medium supplemented with 0.4% (w/v) glucose (M9G; negative control) were analyzed by RNA-seq. V. vulnificus exposed to murine blood differentially expressed 942 genes compared with that exposed to M9G; 491 genes were up-regulated and 451 genes were down-regulated (Dataset S1). Among the genes encoding extracellular toxins, expression of the vvhBA operon was the most elevated upon exposure to murine blood (about 26.0-fold; Fig. S1A), which was confirmed by quantitative RT-PCR (qRT-PCR) (about 81.1-fold; Fig. 1A). This result suggested that vvhBA is preferentially expressed in response to host-derived signals which exist in the murine bloodstream.
Because VvhA-dependent cytotoxicity of V. vulnificus toward murine peritoneal macrophages has been reported (33), we questioned whether the expression of vvhBA increases upon exposure to murine macrophage RAW 264.7 cells. As shown in Fig. 1B, V. vulnificus that was exposed to nitric oxide (NO)producing RAW 264.7 cells increased vvhBA transcription compared with that exposed to Dulbecco's modified Eagle's medium (DMEM; negative control). Strikingly, the extent of increase in the vvhBA transcript level upon exposure to RAW 264.7 cells diminished significantly by addition of the NO synthase inhibitor L-N G -monomethyl arginine citrate (L-NMMA) (Fig. 1B). This result indicated that NO is one of the murine macrophage-derived signals that V. vulnificus senses to induce vvhBA expression. Taken together, the combined results showed that vvhBA expression is induced under host-like conditions in response to certain host-derived signals.

IscR positively regulates vvhBA transcription
Our previous microarray analysis predicted that the expression of vvhBA is up-regulated by IscR (24). Consistent with this, vvhBA transcript and VvhA protein levels were reduced in ⌬iscR and restored by complementation (Fig. 2, A and B). This result confirmed that IscR activates vvhBA expression mostly at the transcription level. Because IscR exists in two forms, holoand apo-IscR, whose regulatory characteristics are distinct from each other (12), we investigated whether apo-IscR activates vvhBA transcription. For this purpose, vvhBA transcript and VvhA protein levels were determined in the iscR 3CA mutant of which the iscR coding region on the chromosome was replaced with iscR 3CA encoding apo-locked IscR (34). Notably, vvhBA transcript and VvhA protein levels significantly increased in the iscR 3CA mutant compared with the WT and ⌬iscR (Fig. 2, C and D), demonstrating that apo-IscR is able to activate vvhBA transcription in vivo. Moreover, the IscR 3CA protein level in the iscR 3CA mutant was higher than the IscR protein level in the WT, possibly due to derepression of the isc Figure 1. Expression of vvhBA upon exposure to murine blood and RAW 264.7 cells. A, the genes induced in V. vulnificus upon exposure to murine blood were identified by RNA-seq analysis. vvhBA was selected as the most highly induced extracellular toxin-encoding gene, and its induction was confirmed by qRT-PCR. Each column represents the vvhBA transcript level in V. vulnificus exposed to murine blood relative to M9G (negative control). Error bars represent the S.E. calculated using DeSeq2 for RNA-seq and the S.D. for qRT-PCR. B, V. vulnificus was exposed to DMEM (negative control) or RAW 264.7 cells in the presence or absence of L-NMMA. The vvhBA transcript levels were determined by qRT-PCR, and the vvhBA transcript level in the cells exposed to DMEM without L-NMMA was set to 1. Error bars represent the S.D. *, p Ͻ 0.05; ***, p Ͻ 0.0005; ns, not significant.

IscR-mediated vvhBA activation in host environments
operon ( Fig. 2D) (6,7). This result suggested that the elevated IscR 3CA level contributes to the increased vvhBA expression, because both holo-and apo-IscR bind to the Type 2 DNA motif with similar affinities in Escherichia coli (12).

IscR activates vvhBA by sensing nitrosative stress and iron starvation
Because the Fe-S cluster is sensitive to nitrosative stress (4), we examined whether IscR mediates the induction of vvhBA expression in response to NO. The vvhBA transcript and VvhA protein levels were significantly elevated by an NO donor DEA NONOate, diethylammonium (Z)-1-(N,N-diethylamino)diazen-1-ium-1,2-diolate, in the WT (Fig. 3, A and B), consistent with Fig. 1B. In contrast, vvhBA transcript and VvhA protein levels were not increased in response to DEA NONOate in ⌬iscR (Fig. 3, A and B), suggesting that IscR responds to hostderived nitrosative stress and activates vvhBA transcription.
Meanwhile, our observation that many genes essential for V. vulnificus to survive under iron starvation were up-regulated in murine blood (Fig. S1B) led us to investigate the effect of iron starvation on vvhBA expression. As shown in Fig. 3, C and D, vvhBA transcript and VvhA protein levels significantly increased in the presence of an iron chelator, 2,2Ј-dipyridyl (DP) in the WT, which was not apparent in ⌬iscR. This result revealed that IscR recognizes iron starvation as another environmental change in the host and induces vvhBA expression. Strikingly, the IscR level in the WT increased upon exposure to DEA NONOate and DP (Fig. 3, B and D), implying that induction of vvhBA expression is attributed to increased IscR levels. Combined with the previous observations (Fig. 2, C and D), we propose a model in which IscR senses nitrosative stress and iron starvation and shifts to the apo-form, which leads to release of repression of the isc operon, elevation of apo-IscR level, and subsequent activation of vvhBA transcription.

IscR binds downstream of P vvhBA to activate vvhBA
To examine whether IscR directly binds to the vvhBA promoter P vvhBA , electrophoretic mobility shift assays (EMSAs) were performed. Because IscR was purified and used under aerobic conditions, most purified IscR would be in the apo-form (11,34). Addition of IscR to the radiolabeled DNA probe encompassing the P vvhBA regulatory region (Ϫ353 to ϩ148 from the transcription start site of vvhBA (17)) resulted in two retarded bands in a concentration-dependent manner (Fig. 4A). This result indicated that at least two binding sites of IscR with different DNA-binding affinities are present in the P vvhBA regulatory region. The same unlabeled DNA fragment competed for IscR binding in a dose-dependent manner (Fig. 4A), confirming the specific binding of IscR to the P vvhBA regulatory region.
To determine the precise location of the IscR-binding sites in the P vvhBA regulatory region, DNase I protection assays were performed using the same DNA probe labeled with 6-carboxyfluorescein (6-FAM). IscR protected two regions extending from ϩ48 to ϩ75 (ISCRB2, centered at ϩ61.5) and ϩ89 to ϩ118 (ISCRB3, centered at ϩ103.5), respectively, from DNase I digestion (Fig. 4B). Increasing IscR levels revealed an additional protected region extending from Ϫ5 to ϩ24 (ISCRB1, centered at ϩ10) (Fig. 4B). The sequence of ISCRB1, ISCRB2, and ISCRB3 showed about 83, 93, and 93% identity, respectively, to the consensus IscR-binding sequence of Type 2 DNA motif, which both holo-and apo-IscR can bind ( Fig. 4C) (12), supporting apo-IscR-mediated vvhBA activation (Fig. 2, C and D). Combined with the EMSA data ( Fig. 4A), this result indicated that IscR binds directly to ISCRB2 and ISCRB3 with similar binding affinities, but binds relatively weakly to ISCRB1. It is noteworthy that all three binding sites of IscR are located downstream of P vvhBA (Fig. 4C), which is unusual for a positive regulator.  . Specific binding of IscR to P vvhBA and sequences of the P vvhBA regulatory region. A, a 501-bp DNA of the P vvhBA regulatory region (5 nM) was radiolabeled and then incubated with increasing amounts of IscR as indicated. For competition analysis, various amounts of the unlabeled DNA fragment were used as a self-competitor and added to a reaction mixture containing 5 nM radiolabeled DNA and 30 nM IscR. B1, a DNA-IscR complex; F, free DNA. B, the same DNA of the P vvhBA regulatory region (32.3 nM) was labeled with 6-FAM, incubated with increasing amounts of IscR as indicated, and then digested with DNase I. The regions protected by IscR are indicated by black boxes (ISCRB1, ISCRB2, ISCRB3). Nucleotide numbers shown are relative to the transcription start site of vvhBA, which was determined previously (17). C, sequence analysis of the P vvhBA regulatory region. The transcription start site of vvhBA and the putative translational initiation codon of VvhB are indicated by solid and dashed bent arrows, respectively. The putative Ϫ10 and Ϫ35 regions are underlined and the putative ribosome-binding site (AGGA) is boldface. The binding sequences of IscR are shown with the black boxes as described above. The binding sequences of HlyU (HLYUB; a white box) and H-NS (HNSB1, HNSB2, HNSB3, HNSB4, HNSB5, HNSB6; gray boxes) were determined later in this study (Fig. 6, C and D). The consensus sequences of the IscR-binding Type 2 DNA motif are indicated above the V. vulnificus DNA sequences. W, A or T; Y, C or T; R, A or G; x, any nucleotide.

HlyU and H-NS regulate vvhBA by directly binding to the P vvhBA regulatory region
To further understand how IscR activates vvhBA transcription despite binding downstream of P vvhBA , we investigated whether IscR interacts with other transcriptional regulator(s) in the P vvhBA regulatory region. Among the previously proposed transcriptional regulators of vvhBA, the exact regulatory mechanisms of HlyU and H-NS have not been elucidated yet (23,32). Hence, the role of these two transcriptional regulators in vvhBA regulation was further examined. The vvhBA transcript and VvhA protein levels were reduced in the hlyU-deletion mutant (⌬hlyU), but elevated in the hns-deletion mutant (⌬hns) (Fig. 5,  A and B). The varied levels of vvhBA transcript and VvhA protein in the deletion mutants were restored by complementation (Fig. 5, C and D). This result confirmed that HlyU is a positive regulator, whereas H-NS is a negative regulator of vvhBA transcription.
Next, EMSAs revealed that each transcriptional regulator binds specifically to the P vvhBA regulatory region (Fig. 6, A and  B). Based on this result, the precise binding sites of HlyU and H-NS in the P vvhBA regulatory region were determined by DNase I protection assays. HlyU clearly protected the region extending from Ϫ128 to Ϫ114 (HLYUB, centered at Ϫ121) from DNase I digestion (Fig. 6C). In addition, the regions protected by H-NS extended from Ϫ119 to Ϫ109 (HNSB1, centered at Ϫ114), Ϫ98 to Ϫ76 (HNSB2, centered at Ϫ87), ϩ16 to ϩ26 (HNSB3, centered at ϩ21), ϩ60 to ϩ74 (HNSB4, centered at ϩ67), ϩ87 to ϩ99 (HNSB5, centered at ϩ93), and ϩ105 to ϩ129 (HNSB6, centered at ϩ117) (Fig. 6D). Notably, the binding sites of both IscR and HlyU overlapped with those of H-NS, which were located throughout the P vvhBA regulatory region (Fig. 4C). In addition, IscR and HlyU bound to the P vvhBA regulatory region at 30 nM, whereas H-NS showed binding at a much higher concentration of 200 nM (Figs. 4A, and 6, A and B). The combined results suggested that IscR and HlyU compete with H-NS for binding to the P vvhBA regulatory region and exhibit higher DNA-binding affinities.

IscR and HlyU outcompete H-NS by simultaneously binding to the P vvhBA regulatory region
To determine whether IscR and HlyU compete with H-NS for binding, EMSAs were performed using reaction mixtures containing a fixed concentration of H-NS with various amounts of IscR. As the concentration of IscR increased, the DNA-H-NS complex (B3) was completely replaced by DNA-IscR complexes (B1) (Fig. 7A). This result indicated that IscR outcompetes H-NS that binds to the P vvhBA regulatory region. Similarly, EMSAs with a fixed concentration of H-NS and increasing amounts of HlyU revealed that HlyU also relieves H-NS binding to the P vvhBA regulatory region (Fig. 7B). Then, the effect of IscR and HlyU on the binding of each other to the P vvhBA regulatory region was examined. When IscR and HlyU were incubated together in the reaction mixture, the DNA-IscR-HlyU complexes (B4), as well as DNA-IscR (B1) and DNA-HlyU complexes (B2), were observed (Fig. 7C). This result showed that IscR and HlyU can bind simultaneously to the P vvhBA regulatory region, and they do not compete for binding sites. Taken together, the combined results suggested that IscR and HlyU bind simultaneously with strong DNA-binding affinities to alleviate H-NS binding from the P vvhBA regulatory region. Because these three transcriptional regulators did not regulate the expression of one another (Fig. S2), we hypothesized that IscR, together with HlyU, relieves H-NS repression of vvhBA transcription by binding downstream of P vvhBA .

IscR and HlyU alleviate H-NS repression to activate vvhBA additively in vivo
To ascertain whether IscR and HlyU activate vvhBA by relieving H-NS repression in vivo, vvhBA transcript and VvhA protein levels were compared in the WT and various deletion mutants. Interestingly, vvhBA transcript and VvhA protein levels were comparable in ⌬hns, the iscR hns double-deletion mutant (⌬iscR⌬hns), and the h-ns-deleted iscR 3CA mutant (iscR 3CA ⌬hns) (Fig. 8, A and B). This observation that IscR did not affect vvhBA transcription in the absence of H-NS sug-

IscR-mediated vvhBA activation in host environments
gested that IscR merely relieves H-NS repression. Furthermore, this result showed that IscR does not act as a roadblock for RNA polymerases in vivo despite its unusual binding sites on the P vvhBA regulatory region. Similarly, vvhBA transcript and VvhA protein levels in the hlyU hns double-deletion mutant (⌬hlyU⌬hns) were comparable with those in ⌬hns (Fig. S3), indicating that HlyU also does not positively affect vvhBA transcription in the absence of H-NS. These results supported our hypothesis that IscR and HlyU relieve repression of vvhBA transcription by outcompeting H-NS for binding to the P vvhBA regulatory region.
The relationship between IscR and HlyU in vvhBA regulation was also investigated in vivo. The vvhBA transcript and VvhA protein levels in the iscR hlyU double-deletion mutant (⌬iscR⌬hlyU) were significantly lower than those in either ⌬iscR or ⌬hlyU (Fig. 8, C and D). This result revealed that these two transcriptional regulators have an additive effect in activating vvhBA transcription in vivo. The elevated IscR 3CA level did not increase vvhBA transcript and VvhA protein levels in the hlyU-deleted iscR 3CA mutant (iscR 3CA ⌬hlyU) compared with those in either ⌬iscR or ⌬hlyU (Fig. 8, C and D), indicating that IscR alone cannot induce vvhBA expression to the WT level in the absence of HlyU. Taken together, the combined results demonstrated that both IscR and HlyU are required for complete release of H-NS repression, leading to full activation of vvhBA transcription in vivo.

Discussion
The V. vulnificus vvhBA operon encodes a well-studied extracellular pore-forming toxin VvhA, which exhibits powerful hemolytic and cytolytic activities (28,33). VvhA is actually expressed during infection and contributes to the severe intestinal tissue damage, subsequent invasion of V. vulnificus into the bloodstream, and dissemination of the pathogen to other organs in a murine infection model (29,33). In this study, RNAseq analysis identified vvhBA as the most highly induced extracellular toxin-encoding gene in V. vulnificus upon exposure to murine blood ( Fig. 1A and Fig. S1A). In addition, vvhBA expression increased upon exposure to NO-producing murine macrophage RAW 264.7 cells (Fig. 1B). These observations indicate that V. vulnificus recognizes certain host-derived signal(s) to induce vvhBA expression during infection.
Our next concern was with the host-derived signal(s) that V. vulnificus senses to activate vvhBA. Interestingly, the Fe-S cluster containing transcriptional regulator IscR activated vvhBA transcription in response to nitrosative stress and iron starvation (Figs. 2 and 3). This IscR-mediated activation of VvhA could be beneficial for V. vulnificus because the pathogen inevitably encounters host-derived nitrosative stress and iron starvation during infection (35,36). In the host, NO is produced by immune cells, particularly by macrophages, as a primary antimicrobial agent (35). Simultaneously, free iron is depleted by the host iron-sequestering system to prevent outgrowth of

IscR-mediated vvhBA activation in host environments
invading pathogens (36). Under these hostile conditions, the IscR-mediated induction of VvhA could disrupt the macrophages and components of the iron-sequestering system (33,37), thereby contributing to survival of V. vulnificus in the host.
Then, we tried to figure out the molecular mechanism by which IscR activates vvhBA transcription. Intriguingly, IscR directly bound downstream of P vvhBA (Fig. 4), which is unusual for a positive regulator. In addition to IscR, HlyU and H-NS acted as positive and negative regulators for vvhBA transcription, respectively, by directly binding to the P vvhBA regulatory region (Figs. 5 and 6). We verified that IscR and HlyU effectively relieve the H-NS binding with higher DNA-binding affinities (Fig. 7, A and B) and do not further affect vvhBA transcription in the absence of H-NS (Fig. 8, A and B, and Fig. S3). These results demonstrate that IscR and HlyU function as antirepressors against H-NS rather than direct activators for vvhBA transcription.
H-NS binds to multiple AT-rich regions throughout the promoter, which possibly results in the formation of nucleoprotein filaments and subsequent RNA polymerase trapping (31,38). In a similar way, H-NS represses rtxA in V. vulnificus, but HlyU can disrupt the whole H-NS nucleoprotein complex by binding to a single site located far upstream of the rtxA promoter, resulting in antirepression of rtxA (38). However, our study revealed that IscR and HlyU bind simultaneously to the P vvhBA regulatory region (Fig. 7C) and additively induce vvhBA transcription in vivo (Fig. 8, C and D). These results indicate that both antirepressors are required to bind upstream as well as downstream of P vvhBA for complete disruption of a possible H-NS nucleoprotein complex and antirepression of vvhBA. To our knowledge, this is the first report describing the additive mode of antirepression for gene activation.
Because the Fe-S cluster is one of the most abundant enzymatic cofactors in fundamental cellular processes, the isc operon, iscRSUA-hscBA-fdx, is highly conserved in many bacteria (10). IscR binds downstream of the iscR promoter in V. vulnificus, and iscR, yadR, and yhgI promoters in E. coli to act as a negative regulator (39,40). In contrast, IscR binds upstream of the target promoters to positively regulate prx3 and gbpA in V. vulnificus, the suf operon and ydiU in E. coli, and the yscW-lcrF operon in Y. pseudotuberculosis (27,34,39,41). However, unlike these conserved and typical regulatory roles of IscR according to its binding sites, our results showed that V. vulnificus IscR atypically binds downstream of P vvhBA to activate vvhBA as an antirepressor.
In general, transcriptional regulators that bind downstream of target promoters prevent the binding and progress of RNA polymerases as negative regulators (18). Nevertheless, a few positive regulators such as MetR, PhoP, DnaA, and Rns still bind downstream of their target promoters (42)(43)(44)(45)(46). Likewise, IscR binds downstream of P vvhBA and does not prevent vvhBA transcription even in the absence of H-NS (Fig. 8, A and B). One possible mechanism can be explained by the effect of other transcriptional regulators participating in vvhBA regulation. We previously reported that CRP activates vvhBA as a Class I activator ( Fig. 9) (17). Thus, CRP may strongly recruit RNA polymerases to P vvhBA , even though the transcription start site of vvhBA overlaps with ISCRB1 (Fig. 4C). Moreover, once RNA polymerases switch to elongation complexes, trailing elongation complexes could overcome an IscR-mediated roadblock by pushing the leading elongation complexes forward (47). Taken together, the combined results led us to conclude that alleviation of H-NS binding from the P vvhBA regulatory region by IscR is much more effective for vvhBA transcription than the possible roadblock effect of IscR against the action of RNA polymerases.
Besides IscR, HlyU, and CRP, ferric uptake regulator (Fur) represses vvhBA transcription in the presence of iron by directly binding to P vvhBA (Fig. 9) (48). Consistent with this, the vvhBA transcript level greatly increased in the fur-deletion mutant (⌬fur) compared with that in the WT (Fig. S4). Although Fur might be the major regulator responsive to iron availability, the vvhBA transcript level in the iscR fur doubledeletion mutant (⌬iscR⌬fur) significantly decreased compared with that in ⌬fur (Fig. S4), demonstrating the distinct role of IscR in the elaborate regulation of vvhBA.
Our current understanding of the transcriptional regulation of V. vulnificus vvhBA in the host environment is described in

IscR-mediated vvhBA activation in host environments
the legend to Fig. 9. IscR, together with HlyU, relieves H-NS repression of vvhBA by sensing nitrosative stress and iron starvation. Although the exact environmental signal affecting the regulatory activity of HlyU is still unknown, it is verified that HlyU is preferentially expressed in the host (23). CRP activates vvhBA possibly by sensing the depletion of specific nutrients (17). At the same time, H-NS repression is weakened at elevated temperatures in the host (32), and the iron-Fur complex repression is also alleviated under iron starvation (48), leading to increased vvhBA expression during infection. These transcriptional regulators take advantage of the multiple binding sites encompassing both downstream and upstream of P vvhBA for tight regulation of vvhBA in response to various host-derived signals.

IscR-mediated vvhBA activation in host environments
In summary, this study demonstrated that vvhBA, encoding a cytolysin/hemolysin of V. vulnificus, is preferentially expressed upon exposure to murine blood and macrophages. The positive regulator IscR activates vvhBA transcription by sensing hostderived nitrosative stress and iron starvation. IscR exerts its effect additively with another positive regulator HlyU by binding downstream of P vvhBA to relieve binding of the negative regulator H-NS from the P vvhBA regulatory region. The collaborative regulation by multiple global regulators allows more precise tuning of vvhBA expression through integrating various host-derived signals encountered during infection, and thereby enhances the overall success of V. vulnificus pathogenesis.

Strains, plasmids, and culture conditions
The strains and plasmids used in this study are listed under Table S1. 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 (A 600 ). When required, 100 g/ml of ampicillin was added to the medium. The murine macrophage RAW 264.7 cells were grown in DMEM containing 10% fetal bovine serum (VWR, Radnor, PA) and appropriate antibiotics (100 units/ml of penicillin G and 100 g/ml of streptomycin (Gibco-BRL, Gaithersburg, MD)) in air supplemented with 5% CO 2 at 37°C. To induce NO production, the RAW 264.7 cells were suspended in fresh DMEM containing 500 ng/ml of E. coli O111:B4 lipopolysaccharide (Sigma) and 1 mM L-arginine (Sigma) (49,50).

RNA purification and RNA-seq analysis
To analyze transcriptomes differentially expressed upon exposure to murine blood, about 8 ϫ 10 7 CFU of V. vulnificus were inoculated to 800 l of murine blood or M9G (negative control). For two biological replicates, the V. vulnificus cells and murine blood were prepared from two independent colonies or mice, respectively, on the same day. The mixture was incubated at 37°C with rolling for 1 h and then centrifuged at 250 ϫ g for 3 min to harvest supernatant containing bacterial cells. Total RNA from the V. vulnificus cells was isolated and quantified using an RNeasy Mini Kit (Qiagen, Valencia, CA) and a NanoDrop One c Microvolume UV-visible Spectrophotometer (Thermo Scientific, Waltham, MA), respectively. Strand-specific cDNA libraries were constructed and sequenced using HiSeq 2500 (Illumina, San Diego, CA) by Lab-Genomics (Seongnam, Gyeonggi, South Korea) as described previously (20). The raw sequencing reads were mapped to the V. vulnificus MO6 -24/O reference genome (GenBank TM accession numbers: CP002469 and CP002470), and the expression level of each gene was calculated as reads per kilobase of transcript per million mapped sequence reads (RPKM) value using EDGE-pro version 1.3.1 (Estimated Degree of Gene Expression in PROkaryots) (51). The RPKM values were normalized and analyzed statistically using DeSeq2 version 1.26.0 to identify differentially expressed genes (greater than 2-fold change with p Ͻ 0.05) when exposed to murine blood (52). All manipulations for murine blood sampling were performed following the National Institutes of Health Guidelines for Humane Treatment and approved by the Animal Care and Use Committee of Seoul National University (SNU-170116-1).

qRT-PCR
Relative vvhBA transcript levels in the total RNA isolated from V. vulnificus grown under various environmental conditions were determined by qRT-PCR. In detail, V. vulnificus was grown in LBS to an A 600 of 0.5 and then exposed to RAW 264.7 cells at a multiplicity of infection of 10 or DMEM (negative control) for 1 h in the presence or absence of 500 M L-NMMA (Sigma), which is a known NO synthase inhibitor (50,53). Moreover, V. vulnificus was exposed to 25 M DEA NONOate (Cayman Chemical, Ann Arbor, MI) for 20 min or 50 M DP (Sigma) for 10 min when necessary. cDNA was synthesized from 1 g of the total RNA by using the iScript TM cDNA synthesis kit (Bio-Rad). Real-time PCR amplification of the cDNA was performed by using the Chromo 4 real-time PCR detection system (Bio-Rad) with pairs of specific primers (Table  S2) as described previously (54). Relative expression levels of the vvhBA transcript were calculated by using the 16S rRNA expression level as an internal reference for normalization (54).
To complement the mutations, pZW1510 carrying the hlyU gene on the broad host-range vector pJH0311 was used in this study (Table S1) (54,57). Similarly, the iscR and hns genes were amplified by PCR using pairs of specific primers listed in Table  S2 and cloned into pJH0311 to create pKK1531 and pGR1713, respectively (Table S1). The plasmids were transferred into the appropriate mutants by conjugation as described above.
The purified IscR, HlyU, H-NS, OmpU, and truncated VvhA were used to raise rabbit polyclonal antibodies against the respective V. vulnificus proteins (AB Frontier, Seoul, South Korea). For Western blot analysis, V. vulnificus exposed to various environmental conditions were harvested and fractionated into cells and supernatants by centrifugation. The 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 supernatants were filtered through a Puradisc TM 25-mm syringe filter (pore size 0.2 m; GE Healthcare, Chicago, IL) and concentrated using Amicon Ultra-15 (cut-off 30 kDa; Millipore, Burlington, MA). The protein levels of IscR, HlyU, H-NS, and DnaK in the clear cell lysates or VvhA and OmpU in the supernatant concentrates were determined as described previously (20,41).

EMSA and DNase I protection assay
For EMSAs, a 501-bp vvhBA promoter region (Ϫ353 to ϩ148 from the transcription start site of vvhBA) was amplified by PCR using unlabeled VVHBA02-F and [␥-32 P]ATP-labeled VVHBA02-R as primers (Table S2). The radiolabeled probe DNA was then incubated with purified IscR, HlyU, and H-NS for 30 min at 30°C in a 20-l reaction mixture containing binding buffer (40 mM Tris-Cl (pH 7.9), 70 mM KCl, 1 mM DTT, and 100 g of BSA) and 0.1 g of poly(dI-dC) (Sigma) as a nonspecific competitor. For competition analysis, the same but unlabeled 501-bp DNA fragment was used as a self-competitor DNA. Electrophoretic analysis of the DNA-protein complexes was performed as described previously (20).
Similarly, the same 501-bp vvhBA promoter region was amplified by PCR using unlabeled VVHBA02-F and 6-FAMlabeled VVHBA02-R as primers for DNase I protection assays (Table S2). The binding of IscR, HlyU, and H-NS to the labeled DNA was performed as described above, and DNase I digestion of DNA-protein complexes followed the procedures as described previously (54). The digested DNA products were precipitated with ethanol and eluted in sterilized H 2 O, and then analyzed using an ABI 3730xl DNA analyzer (Applied Biosystems, Foster City, CA) with Peak Scanner TM Software version 1.0 (Applied Biosystems) (59).

Data analysis
Averages and S.D. 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).

Data availability
All data presented in this paper are contained within the manuscript and supporting information. The raw data of RNAseq analysis were deposited in NCBI Sequence Read Archive (SRA) database under accession numbers PRJNA560127.