IscR Regulation of Type 3 Fimbriae Expression in Klebsiella pneumoniae CG43

In Klebsiella pneumoniae, we have previously shown that IscR, an Fe–S cluster-containing transcriptional factor, plays a dual role in controlling capsular polysaccharide biosynthesis and iron-acquisition systems by switching between its holo and apo forms. In this study, the effect of IscR on type 3 fimbriae expression and biofilm formation was investigated. We found that production of the major subunit of type 3 fimbriae, MrkA, was increased in the ΔiscR and iscR3CA strains, a strain expressing a mutant IscR that mimics apo-IscR, at both the translational and transcriptional levels. Based on the fact that type 3 fimbriae expression is the major factor affecting biofilm formation, increased biofilm formation was also found in ΔiscR or iscR3CA, suggesting that holo-IscR represses biofilm formation. However, the repression of type 3 fimbriae expression by IscR is indirect. To further understand the regulatory mechanism of IscR, the effect of IscR on the expression of mrkHIJ, which encodes cyclic di-GMP (c-di-GMP)-related regulatory proteins that control type 3 fimbriae expression, was studied. We found that holo-IscR could directly repress mrkHI transcription, indicating that MrkHI is required for IscR regulation of type 3 fimbriae expression. Finally, deletion of iscR attenuated K. pneumoniae virulence in a peritonitis model of mouse infection, while the absence of the [2Fe–2S] cluster of IscR had no effect on K. pneumoniae virulence during infection. Taken together, our results demonstrate the underlying mechanism of the [2Fe–2S] cluster of IscR in controlling type 3 fimbriae expression and its effect on K. pneumoniae pathogenesis.


INTRODUCTION
Iron is essential to most bacteria for growth and reproduction. It plays a key role as a cofactor in the electron transport chain and for various enzymes in the tricarboxylic acid (TCA) cycle and oxygen metabolism (Neilands, 1981). While limited iron concentrations abolish bacterial growth, high intracellular iron concentrations may damage bacteria owing to the formation of undesired reactive oxygen species (ROS) (Touati, 2000;Andrews et al., 2003). In addition, bacteria must alter the expression of various virulence genes to adapt to iron availability during infection, such as those involved in siderophore production, fimbriae expression, and biofilm formation (Lin et al., 2010;Rohmer et al., 2011;Wu et al., 2012;Porcheron and Dozois, 2015). Therefore, it is important for bacteria to develop a tight regulatory circuit in response to iron availability in order to perform infection processes successfully.
Klebsiella pneumoniae is a gram-negative facultative anaerobe that causes community-acquired diseases including pneumonia, bacteremia, septicemia, and urinary and respiratory tract infections in patients with underlying diseases (Podschun and Ullmann, 1998). In addition, hypervirulent K. pneumoniae is also a major pathogen for pyogenic liver abscesses in diabetic patients (Han, 1995;Lau et al., 2000;Yang et al., 2009). Heavy capsular polysaccharide (CPS) is a characteristic of isolates causing pyogenic liver abscesses, protecting the bacteria from serum killing (Sahly et al., 2000;Lin et al., 2004). Apart from CPS, fimbriae are considered another crucial virulence factor in K. pneumoniae pathogenesis (Livrelli et al., 1996). Type 3 fimbriae, which are encoded by the mrkABCDF operon, are often expressed in heavily encapsulated K. pneumoniae isolates and play an important role in biofilm formation (Di Martino et al., 2003;Jagnow and Clegg, 2003;Wu et al., 2012). Biofilm formation is considered to be a key factor in the development of nosocomial infections, increasing bacterial tolerance to antibiotics, which causes problems during medical treatment (Murphy and Clegg, 2012). Therefore, a majority of K. pneumoniae isolates from catheter-associated urinary tract or hospital-acquired respiratory tract infections can produce functional type 3 fimbriae (Ong et al., 2010). Furthermore, the ability of K. pneumoniae colonization and subsequent persistence in mice was reduced when the type 3 fimbriae expression was abolished . In addition, immunization of mice with purified type 3 fimbriae confers protection against following challenge with virulent K. pneumoniae (Lavender et al., 2005;Wang et al., 2016). Even though these findings indicate the significance of type 3 fimbriae to the virulence of K. pneumoniae, the regulation of the type 3 fimbrial gene expression in K. pneumoniae still remains largely unknown (Murphy and Clegg, 2012).
In bacteria, c-di-GMP is a bacterial second messenger that modulates biofilm formation and controls expression of the virulence genes (Tamayo et al., 2007). In K. pneumoniae, several studies revealed that a gene cluster adjacent to the type 3 fimbriae, mrkHIJ, was involved in the modulation and sensing of cyclic di-GMP (c-di-GMP) and also regulated the type 3 fimbriae expression (Wilksch et al., 2011;Murphy and Clegg, 2012;Wu et al., 2012;Yang et al., 2013). MrkH is a PilZ-domain protein that is able to bind to c-di-GMP and is a central positive regulator in type 3 fimbriae expression (Wilksch et al., 2011;Yang et al., 2013). MrkI is a LuxR-type transcriptional regulator that activates type 3 fimbriae expression . MrkJ possesses an EAL domain that allows it to serve as a functional c-di-GMP phosphodiesterase (PDE) for hydrolysis of c-di-GMP to further repress type 3 fimbriae expression and biofilm formation (Johnson and Clegg, 2010). Similarly, YfiN and YjcC are also c-di-GMP PDE proteins and play a negative role in control of type 3 fimbriae expression in K. pneumoniae (Wilksch et al., 2011;Huang et al., 2013). In addition to c-di-GMP-related proteins, several transcriptional regulators have reported to be involved in control of type 3 fimbriae expression in K. pneumoniae, such as histone-like nucleoid-structuring protein (H-NS), CRP, and ferric uptake regulator (Fur) Ares et al., 2016;Lin et al., 2016). Thus, the regulation of type 3 fimbriae in K. pneumoniae in response to different environmental stimulus is more complicated than previously expected.
During the infection, iron availability is an important environmental signal affecting bacterial pathogenesis. In K. pneumoniae, CPS biosynthesis, type 3 fimbriae expression, and bacterial biofilm formation in K. pneumoniae were controlled by Fur and iron availability (Lin et al., 2010;Wu et al., 2012). Like Fur, we found that IscR also affects the expression of genes involved in the iron-acquisition system and CPS biosynthesis in K. pneumoniae (Wu et al., 2014). IscR acts as a crucial transcriptional regulator that controls iron-sulfur (Fe-S) cluster biosynthesis in bacteria. Fe-S clusters are important cofactors of multiple proteins involved in various cellular processes in bacteria (Andrews et al., 2003;Outten et al., 2004). Switching between the [2Fe-2S] holo and apo forms of IscR can affect the regulatory specificity of IscR to target genes in response to iron availability and oxidative stress (Outten et al., 2004;Yeo et al., 2006;Pullan et al., 2007;Giel et al., 2013). In several bacteria, IscR has been demonstrated to be implicated in pathogenesis (Wu and Outten, 2009;Lim and Choi, 2013;Miller et al., 2014;Haines et al., 2015). In Escherichia coli, expression of type 1 fimbriae is inhibited by apo-IscR to further reduce biofilm formation (Wu and Outten, 2009). Furthermore, IscR can induce the production of CAF/I fimbriae in response to iron depletion in enterotoxigenic E. coli (Haines et al., 2015). In Vibrio vulnificus, the iscR mutant reduces the expression of multiple virulence factors to further affect mouse mortality (Lim and Choi, 2013). In Yersinia pseudotuberculosis, IscR plays a critical role in the control of type 3 secretion and virulence (Miller et al., 2014). However, the regulatory role of IscR in K. pneumoniae pathogenesis remains largely unknown.
In this study, we investigated if IscR regulates type 3 fimbriae expression, biofilm formation, and virulence in K. pneumoniae; the role of the [2Fe-2S] cluster in IscR regulation was also evaluated. Therefore, we found that holo-IscR is able to inhibit type 3 fimbriae expression to further reduce biofilm formation. Although IscR repression of type 3 fimbriae expression occurs indirectly, we demonstrated that holo-IscR directly represses mrkHI expression to further reduce type 3 fimbriae expression. Furthermore, we also found that deletion of iscR reduces the survival rate of mice, while the iscR 3CA mutant maintains K. pneumoniae virulence. Taken together, our results show that IscR plays an important role in K. pneumoniae pathogenesis.

Bacterial Strains, Plasmids, and Media
All bacterial strains and plasmids used in this study are listed in Table 1. Primers used in this study are list in Table 2. Bacterial were routinely cultured at 37 • C in Luria-Bertani (LB) medium supplemented with appropriate antibiotics including ampicillin (100 µg/ml), kanamycin (25 µg/ml), and streptomycin (500 µg/ml).

Western Blotting
The total proteins of exponential phase K. pneumoniae cultures were separated by SDS-PAGE (approximately 5 µg per lane) and transferred to PVDF membrane. Western analysis was followed as previously described (Lin et al., 2016). Rabbit anti-MrkA antibody was used as the primary antibody. Goat antirabbit immunoglobulin G antibody conjugated to horseradish peroxidase (Abcam) was used as the secondary antibody. After incubation with corresponding antibody, the signal in the membranes was collected by ImageQuant LAS 4000 mini (GE Health, United States) after the visualization with an enhanced chemiluminescence ECL western blotting luminal reagent (PerkinElmer, Wellesley, MA, United States).

Quantitative Reverse-Transcription PCR (qRT-PCR)
Total RNA extraction, reverse transcription of isolated mRNA to cDNA, qRT-PCR, and data analysis were performed according to the previous study (Lin et al., 2013). Primers and probes were designed for selected target sequences using Universal ProbeLibrary Assay Design Center (Roche-applied science) and shown in Table 2. Relative gene expressions were quantified using the comparative threshold cycle 2 − C T method with 23S rRNA as the endogenous reference.

Measurement of Promoter Activity
To generate the promoter region of mrkHI that lacked #2 type 1 IscR box, the plasmid, pmrkHIZ15, was used as the template for the inverse-PCR with the primer pair GT342/GT345 to generate the DNA fragment of P mrkHI 2 . Subsequently, the purified DNA fragment was treated with DpnI for 2 h, and then subject to T4 PNK treatment and self-ligation. The ligation product was transformed into E. coli DH5α and confirmed by DNA sequencing. Then, the DNA fragment containing P mrkHI 2 was subcloned into placZ15 to generate pmrkHIZ15 2. The DNA fragment of P mrkHI 1 , which lacked #1 type 1 IscR box, was also generated by PCR-amplified with primer pair GT348/GT289 and the amplicon was then cloned into placZ15 to generate pmrkHIZ15 1. Finally, the promoterreporter plasmids, pmrkAZ15, pmrkHIZ15, pmrkHIZ15 2, and pmrkHIZ15 1, were mobilized into K. pneumoniae strains by electroporation, respectively. The β-galactosidase activity of logarithmic phase bacteria was measured as previously described (Lin et al., 2006).

Biofilm Formation
Biofilm formation was assessed by the ability of the cells to adhere to the walls of 96-well microtitre dishes made of PVC (TPP 96 flat) with some modification of the reported protocol (Lembke et al., 2006). The plate contained an aliquot of 1:10 diluted overnight bacteria culture and then was incubated at 37 • C statically for 24 h for biofilm formation. The unadherent bacteria was washed triply with 200 µl PBS and then adherent bacteria was stained with 200 µl of 0.1% safranin solution at room temperature for 30 min. The plates was rinsed twice with deionised water to remove excess stain. Finally, the safranin stained biofilm was solubilized in 200 µl of 95% ethanol and the absorbance determined at a wavelength of 492 nm. Purification of IscR::His 6 and IscR 3CA ::His 6 The plasmids, pET30b-IscR and pET30b-IscR 3CA , in E. coli BL21(DE3)[pLysS] (Invitrogen, United States) was used to overexpress the recombinant proteins IscR::His 6 and IscR 3CA ::His 6 , respectively. The detail of expression and purification of the recombinant proteins was followed as previously described (Wu et al., 2014).

Electrophoretic Mobility Shift Assay (EMSA)
DNA fragments of the putative promoter region of mrkHI were amplified with Pfu polymerase using the indicated primer sets to generate DNA probes for EMSA (P mrkHI−1 and P mrkHI−3 ). To obtain DNA fragment of P mrkHI−2 , the plasmid pmrkHIZ15 2 was as a template to be amplified with GT349/GT290 for DNA probe in EMSA. EMSA was performed as previously described (Wu et al., 2014). The assay was repeated in at least 3 independent experiments.

The Peritonitis Model of Mouse Infection
To evaluate the role of iscR in K. pneumoniae virulence, the 8weeks male BALB/c mice (National Laboratory Animal Center, Taiwan) were injected intraperitoneally with 100 µl of bacterial suspension containing 5 × 10 4 CFU of mid-log K. pneumoniae strains. The survival rate of the infected mice was monitored daily for 10 days.

Ethics Statement
All the animal experiments were followed as the recommendation in the Guide for the Care and Use of Laboratory Animals of the National Laboratory Animal Center (NLAC, Taiwan). The animal protocols were approved by China Medical University Experimental Animal Center (Permit number: 2016-212).

Statistical Method
The results of qRT-PCR analysis and promoter activity were performed at least triplicate. The results are showed as the mean and standard deviation. Differences between groups were evaluated by an unpaired t-test. The survival rate was determined by log-rank test using GraphPad Prism 5.0. Values of P < 0.05 and P < 0.01 were considered statistically significant difference.

Effect of the State of the Fe-S Cluster in IscR on Type 3 Fimbriae Expression
To study whether IscR regulates K. pneumoniae type 3 fimbriae expression, we determined the levels of MrkA (the major subunit of type 3 fimbriae) in CG43S3 (WT) and iscR strains. Compared with the levels in the WT, iscR produced higher amounts of MrkA ( Figure 1A), suggesting that IscR represses type 3 fimbriae expression. To further investigate the role of the [2Fe-2S] cluster in the IscR regulation of type 3 fimbriae, an iscR mutant (named iscR 3CA ), created by replacing three cysteines with alanines to mimic apo-IscR, was used to observe type 3 fimbriae expression. Likewise, we also found that levels of MrkA were higher in the iscR 3CA strain than in the WT strain, suggesting that the [2Fe-2S] cluster is required for the repression of type 3 fimbriae expression by IscR. However, we noted that higher levels of MrkA production were found in the iscR 3CA strain than in the iscR mutant, implying that apo-IscR acts as an activator in controlling type 3 fimbriae expression. To further confirm the role of the [2Fe-2S] cluster in IscR regulation of fimbriae expression, the empty vector (pACYC184) and complement plasmids pIscR FIGURE 1 | IscR affects the type 3 fimbriae expression in K. pneumoniae. K. pneumoniae CG43S3 WT, iscR, and iscR 3CA strains was grown to mid-log phase at 37 • C in LB broth to observe the type 3 fimbriae expression by (A) western blot analysis against MrkA (the upper panel) and GAPDH antiserum (the lower panel, for internal control). The MrkA and GAPDH proteins are indicated by an arrow, respectively, and (B) qRT-PCR analyses of mrkA gene expression. (C) β-galactosidase activities of K. pneumoniae CG43S3 lacZ and the isogenic strains ( lacZ-iscR and lacZ-iscR 3CA ) carrying the reporter plasmid pmrkAZ15 (P mrkA ::lacZ) were determined using log-phase cultures grown in LB medium. The results are representative of three independent experiments. Error bars indicate standard deviations. * P < 0.05 and * * P < 0.01 compared to the indicated group.
and pIscR 3CA were introduced into the iscR strain to observe MrkA production. Compared to the MrkA expression in iscR [pACYC184], the introduction of the complement plasmid pIscR into iscR repressed the MrkA production level, whereas a slight increase in MrkA production was found in iscR [pIscR 3CA ]. These results suggest that IscR plays a dual role in the regulation of type 3 fimbriae expression and that the status of the [2Fe-2S] cluster in IscR is critical for this regulation. Furthermore, to further study the activity of IscR as a transcriptional regulator in control of type 3 fimbriae expression, mRNA levels of mrkA were measured in the WT, iscR, and iscR 3CA strains by qRT-PCR. As shown in Figure 1B, the mRNA levels of mrkA were higher in the iscR and iscR 3CA strains than in the WT strain. In addition, the mRNA levels of mrkA in the iscR 3CA strain were higher than those in the iscR strain. Furthermore, to further identify IscR could affect the promoter activity of mrkA, a plasmid carrying P mrkA fused with the lacZ reporter gene was constructed and introduced into the lacZ, lacZ-iscR, and lacZ-iscR 3CA strains, respectively. As shown in Figure 1C, P mrkA activity was higher in the lacZ-iscR strain than in the lacZ strain. In addition, P mrkA activity was significantly higher in the lacZ-iscR 3CA strain than in the lacZ-iscR strain. Taken together, this confirms that the status of the [2Fe-2S] cluster in IscR is critical for regulating type 3 fimbriae expression at the transcriptional level.

IscR Represses Biofilm Formation
Based on the fact that type 3 fimbriae are a major mediator of biofilm formation, we further speculated that IscR could affect K. pneumoniae biofilm formation. As shown in Figure 2, we iscR [piscR 3CA ], were grown at 37 • C for 24 h in LB, and bacterial biofilm formation was quantified as described in "Materials and Methods." The results are representative of three independent experiments. Error bars indicate standard deviations. * P < 0.05 and * * P < 0.01 compared to the indicated group.
found that the iscR [pACYC184] strain resulted in a slight increase in biofilm formation compared with that in theWT [pACYC184] strain, confirming that the repression of type 3 fimbriae activity by IscR was also reflected in biofilm formation. To further investigate the role of the [2Fe-2S] cluster in the IscR regulation of type 3 fimbriae, iscR 3CA [pACYC184] was used to observe biofilm formation. Biofilm formation activity was elevated in the iscR 3CA strain compared with that in the WT [pACYC184] strain. This indicates that the [2Fe-2S] cluster of IscR is required for the repression of biofilm formation. Furthermore, the introduction of the complement plasmid pIscR into the iscR strain repressed biofilm formation as compared with that in iscR [pACYC184] and iscR [pIscR 3CA ], confirming that holo-IscR acts as a negative regulator in controlling biofilm formation. In addition, we noted that biofilm formation was stronger in iscR [pIscR 3CA ] than in iscR [pACYC184], implying that apo-IscR plays a positive role in biofilm formation.

No Binding Activity of IscR in the Promoter of mrkA
For further investigation of the mechanism of IscR regulation of mrkA transcription, sequences of putative IscR binding sites were manually analyzed in the promoter region of mrkA. We found a putative type 1 IscR box (5 -ATAACTTAATGAAACGTGAACAAAT-3 ) with 60% (15/25 bp) homology to the consensus sequence of E. coli located between −15 and −39 bp relative to the transcriptional start codon of mrkABCDF (type 3 fimbriae gene cluster). Therefore, an electrophoretic mobility shift assay (EMSA) was performed to demonstrate whether IscR could directly bind to the promoter region of mrkABCDF. We found that the purified recombinant IscR::His 6 protein was unable to bind to the promoter region of mrkABCDF (data not shown). In addition, we also noted a putative type 2 IscR box (5 -ACCACCCTCGCGTTTTCATCTATCAA-3 ) with about 69% (18/26 bp) homology to the consensus sequence of E. coli located between −75 and -100 bp relative to the transcriptional start codon of mrkABCDF. However, the purified recombinant IscR 3CA ::His 6 protein did not bind to this sequence in the promoter region of mrkABCDF either (data not shown). Therefore, these results indicate that IscR represses type 3 fimbriae expression indirectly.

MrkH and MrkI Are Involved in IscR Regulation of Type 3 Fimbriae
A c-di-GMP-related gene cluster (mrkHIJ) adjacent to the type 3 fimbriae operon has been demonstrated to regulate type 3 fimbriae expression (Murphy and Clegg, 2012;Wu et al., 2012). To further investigate whether mrkHIJ are involved in the IscR regulon, the effect of iscR deletion on the mRNA levels of these genes was determined by qRT-PCR. Compared with levels in the WT strain, the mRNA levels of mrkH and mrkI were markedly increased in the iscR and iscR 3CA strains, while no apparent effect was found on the mRNA level of mrkJ ( Figure 3A). This indicates that the [2Fe-2S] cluster of IscR is required for repressing the expression of mrkH and mrkI. Furthermore, we also noted that the expression levels of mrkH and mrkI in the iscR 3CA strain were higher than those in the iscR strain. This implies that apo-IscR activates mrkH and mrkI expression. Likewise, the activities of P mrkHI were FIGURE 3 | Effect of IscR on mrkH, mrkI, and mrkJ expression. (A) qRT-PCR analyses of the mrkH, mrkI, and mrkJ expressions for WT, iscR, and iscR 3CA strains in LB medium. (B) β-galactosidase activities of K. pneumoniae CG43S3 lacZ and the isogenic strains ( lacZ-iscR and lacZ-iscR 3CA ) carrying the reporter plasmid pmrkHZ15 (P mrkHI ::lacZ) were determined using log-phase cultures grown in LB medium. The results are representative of three independent experiments. Error bars indicate standard deviations. * P < 0.05 and * * P < 0.01 compared to the indicated group.
also higher in the lacZ-iscR and lacZ-iscR 3CA strains than in the lacZ strain ( Figure 3B), but no apparent difference was found between the lacZ-iscR and lacZ-iscR 3CA strains, indicating that holo-IscR acts a transcriptional repressor of mrkHI expression.

IscR Directly Binds the Promoter of mrkHI
For further investigation of the mechanism of IscR regulation of mrkHI transcription, sequences of putative IscR binding sites were manually analyzed in the promoter region of mrkHI. As shown in Figure 4A, we found two predicted type 1 IscR boxes (#1 and #2) located at −277 to −254 and −211 to −187 relative to the translation start site of mrkHI, respectively. The #1 and #2 sites are 76% (19/25 bp) and 68% (17/25 bp) homologous, respectively, with the consensus sequence of E. coli. We hypothesized that IscR would bind directly to the #1 or #2 site in the promoter region of mrkHI to repress gene transcription, and we confirmed this by performing an EMSA. As shown in Figure 4B, we found that IscR::His 6 could directly bind to PmrkHI-1, PmrkHI-2, and PgalF-1 (as a positive control), whereas no specific binding between IscR::His 6 , and PmrkH-3 or PgalF-2 (as a negative control) was found. Although the binding activity of IscR::His 6 to PmrkHI-1 and PmrkHI-2 seemed to be weaker than that to PgalF-1, the intensity of the shifted bands slightly increased when the protein concentration was increased ( Figure 4B). This result supports that IscR could specifically bind to the #1 site in the promoter region of mrkHI to control mrkHI expression. In addition, compared with that of IscR::His 6 , the recombinant [2Fe-2S]-clusterless IscR, IscR 3CA ::His 6 , was unable to bind PmrkHI-1 (Data not shown). These results indicate a direct interaction between IscR and the #1 site of the type 1 IscR box in the mrkHI promoter and confirm that the [2Fe-2S] cluster of IscR plays a crucial role in this interaction.

IscR Regulates the Expression of Type 3 Fimbriae through MrkH and an Unknown Factor
Although MrkH is a well-known regulator of type 3 fimbriae (Wilksch et al., 2011;Yang et al., 2013), we wanted to identify whether MrkH is the sole transcriptional regulator in the regulation of type 3 fimbriae expression by IscR. An iscR mrkH strain was generated, and the mRNA expression of mrkA was analyzed in the iscR, iscR mrkH, and mrkH strains. As shown in Figure 5A, the mRNA expression of mrkA in the iscR mrkH strain was significantly higher than that in the mrkH strain. Similarly, the promoter activity of mrkA also confirmed this finding; although the effect was weak, it was still significant (Figure 5B). This suggests that in addition to MrkH, another transcriptional factor is also involved in the repression of type 3 fimbriae expression by IscR.

Role of IscR in K. pneumoniae Virulence
To understand the effect of the deletion of iscR on K. pneumoniae virulence, the survival of mice that were intraperitoneally inoculated with the WT, iscR, or iscR 3CA strains was monitored for 10 days. As shown in Figure 6A, when mice were inoculated with the WT, iscR, and iscR 3CA strains, survival decreased to 16.7, 66.7, and 16.7% by day 10, respectively. The deletion of iscR showed a trend to attenuate the bacterial virulence; however, the difference did not reach statistical significance. In addition, this result also implied that the absence of the [2Fe-2S] cluster of IscR did not eliminate K. pneumoniae virulence during infection. To further confirm the result, the mice were intraperitoneally inoculated with iscR [pACYC184], iscR [pIscR], and iscR [pIscR 3CA ] to monitor the survival rate. As shown in Figure 6B, we found that mice inoculated with iscR [pACYC184], iscR [pIscR], or iscR [pIscR 3CA ] exhibited survivals of 100, 33.3, and 50% by day 10, respectively. The complementation of iscR significantly decreased the survival of mice, suggesting that IscR is involved in the K. pneumoniae virulence. Besides, the complementation of iscR 3CA appeared to partly restore the bacterial virulence; however, the difference is not statistically different.
FIGURE 5 | Role of MrkH in IscR regulation of type 3 fimbriae expression. (A) qRT-PCR analyses of mrkA expression in iscR, iscR-mrkH, and mrkH strains which was grown in LB medium. (B) β-galactosidase activities of K. pneumoniae CG43S3 lacZ-iscR, lacZ-iscR-mrkH, and lacZ-mrkH carrying the reporter plasmid pmrkAZ15 (P mrkA ::lacZ) were determined using log-phase cultures grown in LB medium. The results are representative of three independent experiments. Error bars indicate standard deviations. * P < 0.05 and * * P < 0.01 compared to the indicated group. ] (dark triangle) infected mice (n = 6 per group) was monitored daily for 10 days and determined by log-rank test using GraphPad Prism 5.0; P-value of <0.05 was considered statistically significant.

DISCUSSION
In many bacteria, iron-responsive regulators play important roles in controlling the expression of several virulence factors according to iron availability (Wu and Outten, 2009;Wu et al., 2012;Lim and Choi, 2013;Miller et al., 2014). Previously, we demonstrated that iron availability affects CPS, iron acquisition systems, type 3 fimbriae expression, and biofilm formation in K. pneumoniae (Lin et al., 2010;Wu et al., 2012Wu et al., , 2014. In addition, while the regulatory role of IscR in CPS biosynthesis and the iron-acquisition system have been demonstrated (Wu et al., 2014), the role of IscR in K. pneumoniae pathogenesis had not been entirely elucidate. In this study, we found that type 3 fimbriae expression and biofilm formation were also affected by the status of [2Fe-2S] cluster of IscR, playing a critical role in mediating the expression of virulence factors for successful infection.
In K. pneumoniae, Fur has been demonstrated to play a central role in directly activating mrkA and mrkH expression to increase biofilm formation . In this study, we found that, compared to the WT strain, both the K. pneumoniae iscR and iscR 3CA strains harbored increased the type 3 fimbriae expression and biofilm-forming activities, which could be reversed by the complementation of pIscR, but not pIscR 3CA into the iscR strain (Figures 1, 2). Likewise, overexpression of pIscR in K. pneumoniae CG43S3 also repressed the mRNA expression of mrkA, while pIscR 3CA in K. pneumoniae CG43S3 activated the mrkA expression (Supplementary Figure S1). These results suggest that IscR represses type 3 fimbriae expression and biofilm formation in a Fe-S-cluster-dependent manner. Although Fur and IscR exert positive and negative regulatory effects, respectively, on K. pneumoniae biofilm formation under iron-repleted conditions, we speculated that Fur played a major role, since iscR expression is repressed in response to environmental iron (Wu et al., 2014). Furthermore, although type 3 fimbriae are critical in K. pneumoniae biofilm formation, it has also been demonstrated that the function of fimbriae is hindered by the concomitant expression of a thick capsule on the bacterial surface (Schembri et al., 2005;Goncalves Mdos et al., 2014). Previously, we found that holo-IscR but not apo-IscR could activate CPS production (Wu et al., 2014). In Figure 1, the repressive effect of IscR on MrkA production at the translational level was more apparent than that at the transcriptional level. We hypothesized that the reduced CPS production in the iscR strain would promote the assembly of type 3 fimbriae on the bacterial surface. Moreover, we also found that type 3 fimbriae expression is slightly increased in the iscR 3CA strain as compared to that in the WT strain (Figure 1). In contrast to pIscR, the introduction of pIscR 3CA into the iscR strain increased K. pneumoniae biofilm formation (Figure 2B). It implies that apo-IscR acts as an activator to increase biofilm formation. However, the recombinant protein IscR 3CA ::His 6 did not bind to the promoter regions of mrkA or mrkH, suggesting that apo-IscR indirectly regulates the expression of type 3 fimbriae. In K. pneumoniae, c-di-GMP is a critical second messenger that influences biofilm formation and type 3 fimbriae expression (Johnson and Clegg, 2010;Wilksch et al., 2011;Wu et al., 2012). To identify possible IscR-regulated genes that are involve in c-di-GMP signaling, we analyzed the upstream sequences of c-di-GMP-related genes in K. pneumoniae CG43, (Lin et al., 2016). We found a typical type 2 IscR box exhibiting more than 73% homology with the E. coli consensus sequence upstream of the c-di-GMP-related genes D364_06025 and D364_22720 in K. pneumoniae CG43 (data not shown). Apo-IscR may affect biofilm formation and type 3 fimbriae expression via the regulation of c-di-GMP-related gene expression, which awaits further investigation.
Holo-IscR directly repressed mrkHI expression to affect type 3 fimbriae expression, while apo-IscR appeared to indirectly activate mrkHI expression (Figures 3, 4). In many bacteria, small non-coding RNAs play a critical role in post-transcriptional regulation, allowing bacteria to adapt to various environmental stimuli (Waters and Storz, 2009). RyhB is a well-known ironresponsive small RNA in bacteria (Oglesby-Sherrouse and , and we previously demonstrated that apo-IscR can directly activate several iron uptake systems that affect intracellular iron homeostasis (Wu et al., 2014). In K. pneumoniae, RyhB is involved in the regulation of CPS biosynthesis and the iron-acquisition system . Because small RNA regulates its target mRNAs via base pairing (Masse et al., 2003;Geissmann and Touati, 2004), the 5 untranslated region (5 UTR) of mrkHI mRNA was analyzed for sequences complementary to the RyhB sequence. No apparent potential interacting site was found by using RNAhybrid in BiBiServ (Rehmsmeier et al., 2004). Therefore, whether other small RNAs are involved in IscR regulation of type 3 fimbriae expression requires further investigation. Recently, H-NS was also reported as an activator in control of type 3 fimbriae expression and biofilm formation in K. pneumoniae (Ares et al., 2016). Deletion of hns increased the mRNA expression of mrkH, mrkI, and mrkJ. Thus, H-NS may affect the c-di-GMP concentration, through activation of mrkJ expression, to further influence MrkH and MrkI activity and type 3 fimbriae expression (Ares et al., 2016). As shown in Figure 3A, deletion of iscR increased the mRNA levels of mrkH and mrkI but not mrkJ; thus, IscR and H-NS seemed to regulate the expression of type 3 fimbriae in different manners.
To further analyze the DNA sequence of P mrkH to observe whether other transcriptional regulator is also involved in mrkH regulation, we found a putative binding site of phosphorylated ArcA is overlapped with the #2 site that displays 80% (12/15 bp) homology with the consensus sequence of E. coli (Liu and De Wulf, 2004). ArcA acts as the transcriptional regulator in the ArcA/B two-component system, regulating gene expression to adapt to aerobic and microaerobic conditions (Iuchi and Weiner, 1996;Alexeeva et al., 2003). Oxygen-limited conditions may trigger the kinase activity of ArcB, which then phosphorylates and activates the DNA binding activity of ArcA (Carpenter and Payne, 2014). Therefore, we suggest that the expression of mrkH and type 3 fimbriae could be affected by phosphorylated ArcA in response to redox growth conditions. This possibility should be further investigated. Furthermore, a slightly higher expression level of mrkA was found in iscR mrkH as compared with that in mrkH (Figure 5). This indicates that IscR may affect other transcriptional regulator(s) besides MrkH in order to influence type 3 fimbriae expression. In addition to MrkH, MrkI is a LuxR-type transcriptional regulator containing a conserved aspartate residue (D56) able to receive a phosphorylated group for affecting its regulatory activity. In our previous study, we have demonstrated that a D56A site-directed MrkI mutant, which loss the phosphorylated state in MrkI, exhibited to decrease MrkA production as compared to that in a D56E site-directed MrkI mutant, which mimic the phosphorylate state of MrkI . Therefore, whether IscR could also affect the phosphorylation of MrkI to further influence type 3 fimbriae expression remains to be studied.
During infection, Fe-S cluster homeostasis in bacteria is deeply affected by iron starvation and oxidative stress conditions (Miller and Britigan, 1997;Wilks and Burkhard, 2007;Py et al., 2011). These stimuli may affect the functions of IscR in controlling gene expression during infection. In Figure 6A, we found that IscR is required for regulating the virulence of K. pneumoniae during infection in a mouse peritonitis model. Furthermore, the iscR 3CA mutant exhibited a similar virulence to that of the WT strain. In K. pneumoniae, CPS and ironacquisition systems are critical and important virulence factors for successful infection (Lin et al., 2004;Regueiro et al., 2006;Russo et al., 2015). We have previously demonstrated that CPS can be activated by holo-IscR, while the three iron-acquisition systems (fhuA, sitA, and iucA) are directly activated by apo-IscR (Wu et al., 2014). It may be described that the deletion of iscR in K. pneumoniae reduces CPS production and ironacquisition system expression to further diminish virulence during infection, while the introduction of pIscR into the iscR strain increases virulence compared with that of the vector alone or pIscR 3CA in the iscR strain. Furthermore, aerobactin (iuc) plays a critical role in the growth and survival of highly virulent K. pneumoniae strains (Russo et al., 2015). It may be that high expression of aerobactin biosynthesis is one of the factors used by the iscR 3CA mutant for maintaining a similar virulence to that of the WT strain. In addition, overexpression of suf, which is another Fe-S assembly gene cluster, in the iscR 3CA mutant leads to a defect in the proton motive force in E. coli and Y. pseudotuberculosis (Ezraty et al., 2013;Miller et al., 2014). Such a defect could influence the expression of multiple virulence factors, such as antibiotic resistance, the type 3 secretion pathway, and bacterial mobility (Ezraty et al., 2013;Miller et al., 2014). In K. pneumoniae CG43S3, we also found that the mRNA level of sufA was apparently increased in the iscR 3CA mutant as compared to that in the WT and iscR strains (Supplementary Figure S2A). This means that apo-IscR could activate the suf operon, as it does in E. coli and Y. pseudotuberculosis. However, compared to the WT strain, no apparent effect on growth rate was found in the iscR and iscR 3CA strains grown in LB medium (Supplementary Figure  S2B). Therefore, whether overexpression of suf in the iscR 3CA mutant affects the proton motive force in K. pneumoniae remains to be investigated.
Aside from the role of the CPS and iron-acquisition systems during infection, type 3 fimbriae play crucial roles in adhesion to host cells, persistence, and biofilm formation (Hornick et al., 1992;Tarkkanen et al., 1997;Sebghati and Clegg, 1999;Jagnow and Clegg, 2003). Furthermore, type 3 fimbriae are key factors that affect the ability of K. pneumoniae to colonize and subsequently persist in mice . Therefore, we suggest that IscR is also involved in K. pneumoniae colonization and adherence with eukaryotic cells and catheters through the regulation on the type 3 fimbriae expression and the bacterial biofilm forming activity. The role of IscR in K. pneumoniae virulence according to different infectious routes and host interactions needs to be further evaluated. In addition, IscR also plays an important role in mediating proper levels of Fe-S cluster biosynthesis in bacteria via transcriptional repression of the iscRSUA operon (Yeo et al., 2006;Giel et al., 2013), indicating that the manner in which IscR impacts K. pneumoniae virulence is dependent on the conditions that K. pneumoniae encounters during the course of infection.
Taken together, our results show that IscR can directly repress mrkHI expression to affect type 3 fimbriae expression and biofilm formation. Furthermore, the deletion of iscR decreases K. pneumoniae virulence during infection, demonstrating that IscR is implicated in K. pneumoniae pathogenesis.

AUTHOR CONTRIBUTIONS
Conceived and designed the experiments: T-HL, CY-T, and C-TL. Performed the experiments: T-HL, Y-CL, C-CW, and C-TL. Analyzed the data: C-YT, Y-CL, and CL. Contributed reagents/materials/analysis tools: C-YT, Y-CL, C-CW, and C-FH. Wrote the paper: T-HL and C-TL. All authors read and approved the final manuscript.

FUNDING
The work is supported by the grant from Ministry of Science and Technology (MOST 105-2320-B-039-051-) and Taichung Tzuchi Hospital, The Buddhist Tzu Chi Medical Foundation (TTCRD103-10 and TTCRD104-11). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

ACKNOWLEDGMENTS
We are grateful to Mr. Jing-Ciao Lin and Miss. Yi-Min Hong for their technical assistance during the study.