The PAS Domain-Containing Protein HeuR Regulates Heme Uptake in Campylobacter jejuni

ABSTRACT Campylobacter jejuni is a leading cause of bacterially derived gastroenteritis. A previous mutant screen demonstrated that the heme uptake system (Chu) is required for full colonization of the chicken gastrointestinal tract. Subsequent work identified a PAS domain-containing regulator, termed HeuR, as being required for chicken colonization. Here we confirm that both the heme uptake system and HeuR are required for full chicken gastrointestinal tract colonization, with the heuR mutant being particularly affected during competition with wild-type C. jejuni. Transcriptomic analysis identified the chu genes—and those encoding other iron uptake systems—as regulatory targets of HeuR. Purified HeuR bound the chuZA promoter region in electrophoretic mobility shift assays. Consistent with a role for HeuR in chu expression, heuR mutants were unable to efficiently use heme as a source of iron under iron-limiting conditions, and mutants exhibited decreased levels of cell-associated iron by mass spectrometry. Finally, we demonstrate that an heuR mutant of C. jejuni is resistant to hydrogen peroxide and that this resistance correlates to elevated levels of catalase activity. These results indicate that HeuR directly and positively regulates iron acquisition from heme and negatively impacts catalase activity by an as yet unidentified mechanism in C. jejuni.

quiring iron from heme may be important during colonization of the chicken gastrointestinal tract (4).
Due to the prominent redox potential of Fe 2ϩ /Fe 3ϩ , most organisms require iron as a redox cofactor for enzyme complexes. Since the host often restricts iron availability to limit microbial growth-a process termed nutritional immunitymany bacterial species have developed methods for obtaining iron from host molecules, including heme (6). The heme utilization gene cluster of C. jejuni consists of genes encoding a predicted TonB-dependent heme receptor (chuA), a hemin ABC transporter permease (chuB), a hemin ABC transporter ATP-binding protein (chuC), and a periplasmic hemin-binding protein (chuD) (7). Divergently transcribed from the transport genes is a previously characterized heme oxygenase, chuZ, which directly binds heme and degrades it in the presence of an electron donor (7,8). All components of the heme utilization gene cluster are required for efficient use of heme as a sole iron source, and the ferric uptake repressor (Fur) protein directly binds to the promoter region of the gene cluster, implicating Fur as a chu locus regulator (7).
These data are consistent with observations that, in most organisms, intracellular iron levels are tightly regulated to avoid forming toxic hydroxyl radicals and superoxide anions via Haber-Weiss-Fenton catalysis (9). Iron homeostasis specifically impacts the C. jejuni response to oxidative stress in two ways: (i) excess iron is associated with repressed catalase production, thereby increasing hydrogen peroxide levels within the cell, and (ii) iron is a required cofactor for periplasmic reductases such as Mfr which interact with H 2 O 2 to generate reactive oxygen species (ROS) (10,11). However, C. jejuni, like most bacteria, requires nutrient iron to grow and proliferate, leading it to have evolved elaborate regulatory circuits to govern the expression of iron acquisition, metabolism, and efflux. While the above studies indicated that increased iron levels repress chu gene expression, likely via Fur binding, no positive regulators of heme and/or iron uptake have been identified. Due to the importance of iron in combating oxidative stress in C. jejuni, it would be surprising if Campylobacter did not have some mechanism for increasing iron uptake under unfavorable redox states. As such, we hypothesized that C. jejuni likely maintains intracellular iron levels using both negative regulation, as seen in previous work with Fur, and positive regulation.
In subsequent Tn-Seq experiments that aimed at finding additional chicken colonization determinants, a candidate regulator was identified, Cjj81176_1389 (unpublished data). This regulator is predicted to contain an N-terminal Per-Arnt-Sim (PAS) domain and a C-terminal helix-turn-helix DNA binding domain. The PAS domain is a ligand-binding domain that interacts with a chemically diverse array of small molecules (12). These domains, while highly divergent in primary sequence identity, maintain a conserved three-dimensional architecture (12). When ligand binds to a PAS domain, the protein either directly initiates a signaling response, or the protein-ligand complex becomes capable of responding to a secondary cue, including gas molecules, photons, or redox potential (12). We reasoned that since the heme utilization gene cluster and Cjj81176_1389 appear to be required for full colonization of the chicken gastrointestinal, and since Cjj81176_1389 may be sensing intracellular redox potential, Cjj81176_1389 could be directly regulating chu gene expression.
Here we demonstrate that Cjj81176_1389 mutants have significantly reduced ability for chicken gastrointestinal tract colonization, particularly during competition with wild-type C. jejuni. Subsequent transcriptomic analysis found that this regulator, here referred to as HeuR (for heme uptake regulator), positively influences chu gene expression, as well as other iron uptake systems. Purified HeuR specifically binds the chuZA promoter region, indicating it may regulate chu gene expression directly. Furthermore, heuR mutants are unable to efficiently use heme as the sole iron source and exhibit decreased levels of cell-associated iron. Finally, we show that loss of heuR leads to hydrogen peroxide resistance, in turn due to elevated levels of catalase activity.

RESULTS
C. jejuni strains lacking heuR are unable to efficiently colonize the chicken gastrointestinal tract. Colonization studies using defined insertion deletion mutants were performed using day-ofhatch chicks. Animals were inoculated with 2.5 ϫ 10 8 CFU of wild-type C. jejuni, 1.7 ϫ 10 8 CFU of the heuR mutant, or 1.9 ϫ 10 8 CFU of the chuA mutant. Following 7 days of colonization, viable bacteria present within cecal samples were enumerated, and the medians were found to be 5.8 ϫ 10 9 CFU/g for wild-type C. jejuni, 6.6 ϫ 10 8 CFU/g for the heuR mutant, and 5.6 ϫ 10 7 CFU/g for the chuA mutant (Fig. 1A). The chicks that were mock infected with phosphate-buffered saline (PBS) did not yield detectable C. jejuni. Statistical analysis of the medians revealed that the heuR mutant was significantly decreased for colonization, but the chuA mutant was not (P ϭ 0.1).
Competition analysis using the heuR mutant indicated a severe competitive defect of the mutant versus wild-type C. jejuni. The heuR mutant yielded a mean competitive index value of 0.004 Ϯ 0.007, representing a 250-fold decrease in colonization potential compared to wild-type C. jejuni (Fig. 1B). Statistical analysis of this competitive disadvantage was found to be significant. colonization of day-of-hatch white Leghorn chicks with either wild-type C. jejuni, the heuR mutant, or the chuA mutant. Cecal loads were determined using selective media following a 7-day colonization. The median values are noted and results compared using the Mann-Whitney U test. LoD, limit of detection. (B) Competition analysis of wild-type C. jejuni versus the heuR mutant. Following correction for inocula, the ratio of heuR mutant to wild type was determined and is presented as a competitive index. Statistical analysis was performed using a one-sample t test against a hypothetical value of 1.
Genes involved in inorganic ion transport are underexpressed in the heuR mutant. To determine the extent of the HeuR regulon, we employed transcriptomic analysis of wild-type C. jejuni and heuR mutant cultures from solid media containing sheep's blood. Since the conditions required for HeuR-mediated regulation are unknown, we chose these inherently more heterogeneous cultures given that the resolution of transcriptome sequencing (RNA-Seq) would still allow us to observe statistically significant changes. RNA-Seq analysis of the heuR mutant identified 43 genes whose transcripts were decreased in abundance by at least 2-fold compared to wild-type C. jejuni (Table 1). These values represented statistically significant decreases for all genes identified. The Microbial Genomic context Viewer (MGcV) identified the NCBI GI numbers for 31/43 loci and, using a cluster of orthologous group (COG) analysis, assigned a functional class to 19/31 loci. Of the 19 loci for which a function could be assigned, 8 (42%) were involved in inorganic ion transport and metabolism, including all five chu genes ( Fig. 2A). Also, amino acid transport and metabolism were well represented with four loci (21%). The MGcV analysis did not identify NCBI GI numbers for the enterochelin system, which was also identified-iron acquisition by this system has been described in Campylobacter and multiple organisms (13,14).
In addition to those genes that were found to exhibit reduced transcription in the heuR mutant, transcriptomic analysis identified 16 genes with elevated transcript abundance in the mutant ( Table 2). The majority of this set included genes belonging to two gene clusters: Cjj81176_1390 to -1394 and Cjj81176_1214 to -1217. MGcV identified NCBI GI numbers for 11/16 genes and assigned functions to 10/11 loci. This group is fairly diverse, with the largest representation belonging to the amino acid transport and metabolism COG (30%) (Fig. 2B).
Purified HeuR binds to the chuZA promoter region. The putative structure of HeuR includes a C-terminal helix-turn-helix DNA binding domain. Given the transcriptome data demonstrating that chu RNA abundance is dependent on HeuR, we hypothesized that HeuR directly regulates chu gene expression. Purification of 6ϫHis-tagged HeuR was successful as soluble protein was obtained at concentrations of 4.3 mg/ml. Incubation of purified 6ϫHis-HeuR with radiolabeled chuZA promoter and subsequent resolution by native polyacrylamide gel electrophoresis indicated binding of the protein to the chuZA probe using 30 g of 6ϫHis-HeuR (Fig. 3). The shifted species increased in intensity until 125 g of HeuR was used, and the fragment representing unbound chuZA promoter was not detected. These results were in contrast to those obtained for a radiolabeled fragment that represented the unrelated intergenic region between C. jejuni 81-176 mapA and ctsW, which are not regulated by HeuR (based on our RNA-Seq results). Mobility of this probe was relatively unaffected until 500 g of HeuR was used, when unbound probe intensity was modestly decreased. These results indicate that HeuR binds to a sequence corresponding to the chuZA promoter, but not indiscriminately to one other unrelated sequence.
Mutants of heuR are unable to use heme as a source of iron. To determine whether decreased expression of the Chu system affected the ability of the heuR mutant to use heme as an iron source, cultures of each strain were grown under iron-chelated conditions and supplemented with heme, hemoglobin, or ferric chloride. Both the heuR and chuA mutants were hypersensitive to iron-limiting conditions compared to wild-type C. jejuni (see Fig.  S1 in the supplemental material). At 160 M desferrioxamine (DFOM), the heuR mutant with empty vector and the chuA mutant with empty vector exhibited greatly reduced growth compared to their behavior in iron-replete medium (Fig. 4A). Heme added to these cultures was unable to rescue growth of the heuR and chuA mutants. Complementation of the heuR mutant with cloned heuR restored the ability to use heme as an iron source. In contrast, cloned heuR introduced into the chuA mutant did not restore heme-dependent growth. Similar results were also seen using whole hemoglobin (Fig. 4B). That iron restriction is the explanation for these phenotypes was supported by the restoration of growth in all strains through the addition of ferric chloride (Fig. 4C).
Mutation of heuR results in decreased cell-associated iron. The heuR mutant expressed a lower abundance of mRNA from several iron uptake systems (Table 1), suggesting it may play a Identified COGs for genes in Table 1 that were found to be underexpressed in the heuR mutant. (B) COGs identified for genes in Table 2 that were overexpressed in the heuR mutant.  general role in regulating iron acquisition. To test this, we examined cell-associated iron levels using inductively coupled plasma mass spectrometry (ICP-MS) of wild-type C. jejuni, heuR mutant, and chuA mutant lysates. This analysis demonstrated that iron levels of the heuR and chuA mutants significantly decreased compared to those of wild-type C. jejuni (Fig. 5A). Iron levels in samples of the heuR mutant complemented with cloned heuR were similar to those of wild-type C. jejuni (Fig. 5B) (P ϭ 0.28). Similar to the results of the mutation alone, iron levels in the chuA mutant with pECO102 were significantly decreased compared to those of wild-type C. jejuni. Consistent with our hypothesis that HeuR regulates iron acquisition through chu expression, when heuR was expressed in the chuA mutant, cellassociated iron remained significantly decreased compared to that in wild-type C. jejuni (Fig. 5B).
The heuR mutant exhibited increased resistance to hydrogen peroxide. We observed a lack of alpha-hemolysis on blood plates on which the heuR mutant was grown (data not shown). As alphahemolysis is due to the formation of methemoglobin in the presence of hydrogen peroxide, we hypothesized that the heuR mutant may degrade hydrogen peroxide more readily than wild-type C. jejuni. Support for this came by examining whether the heuR mutant could detoxify hydrogen peroxide-which is often done through catalase-mediated degradation-and grow in the presence of the molecule. The heuR mutant was incubated in the presence of hydrogen peroxide and found to be resistant. Growth of wild-type C. jejuni was generally limited by hydrogen peroxide at concentrations above 0.15 mM and exhibited a 50% inhibitory concentration (IC 50 ) value of 0.233 mM (Fig. 6A). In contrast, the heuR mutant was unaffected by hydrogen peroxide, with significantly increased growth compared to wild-type C. jejuni at concentrations above 0.15 mM. As a result, the IC 50 value was not experimentally determined and by extrapolation was found to be 5.  which was significantly more growth than that of wild-type C. jejuni. As such, the chuA mutant exhibited an IC 50 of 0.526 mM, a more than 2-fold increase compared to wild-type C. jejuni but approximately 10-fold lower than the heuR mutant.
Complementation of the heuR mutant with cloned heuR resulted in a slight hypersensitivity compared to wild-type C. jejuni, with an IC 50 value of 0.203 mM (Fig. 6B).
Resistance of the heuR mutant to hydrogen peroxide is correlated with increased catalase activity. We hypothesized that the elevated resistance of the heuR mutant to hydrogen peroxide may result from increased catalase activity. To determine whether this was the case, suspensions of strains were normalized and added to 30% H 2 O 2 . In this standard assay, the presence of bacterial catalase leads to conversion of two H 2 O 2 molecules to two H 2 O molecules and one O 2 molecule. As such, generation of O 2 leads to the formation of bubbles, the extent to which correlates to the amount of catalase activity for each strain. The rate of bubble formation can then be qualitatively reported. This assay revealed that wildtype C. jejuni carrying empty vector was normally weakly positive (Table 3). In contrast, the heuR mutant transformed with empty vector was strongly positive for catalase activity. When heuR was reintroduced into the heuR background, catalase activity decreased, and the activity of the cells resembled wild-type activity. Also, similar to wild-type C. jejuni, the chuA mutant remained weakly positive, although the chuA mutant did appear to exhibit more catalase activity than wild-type C. jejuni.

DISCUSSION
Using a Tn-Seq approach, we previously identified several determinants required by C. jejuni for efficient colonization of the chicken gastrointestinal tract (4). One determinant was a component of the Campylobacter heme utilization gene cluster chuB. In this work, we describe a regulator, which we have named HeuR, that is also required for wild-type levels of colonization of the chicken gastrointestinal tract. Additionally, because HeuR is predicted to contain a PAS domain-a protein signature often associated with gas molecule and redox potential sensing-and because it and chu are both required for chicken colonization, we hypothesized that HeuR regulates heme uptake through chu gene expression.
Recently, HeuR was also found to control expression of Cjj81176_1390, as the gene product was overexpressed in a heuR mutant (15). The group that identified this phenotype named HeuR the Campylobacter "flagella interaction regulator" (CfiR) due to flagellar interactions that occurred in response to gene disruption. This group further found that the gene product of Cjj81176_1390 forms dimers linked by a disulfide bridge that may be responsive to the redox status of the cell (15). Our observations, recounted below, show that HeuR controls more genes than just those that promote flagellar interactions, including many that may be involved in combating oxidative stress (Fig. 7). As such, we suggest that the designation CfiR does not accurately represent the full regulatory role in the bacterium, and it is our recommendation that the name be changed to HeuR.
By RNA-Seq analysis of a heuR mutant strain, we identified several genes whose mRNA products were significantly less abundant, including those in the previously identified Campylobacter heme utilization (chu) gene cluster. Purified 6ϫHis-HeuR was found to bind specifically to the chu promoter, albeit at relatively high levels (125 g). Previously, regulators containing a PAS domain were found to be inhibited for DNA binding by the PAS domain itself (16). While the reason behind this observation is unknown, it may be due to insufficient copurification of ligand that is required for a higher-efficiency conformation. This may be similar for HeuR and may explain why more protein was required. We are currently working to determine whether this is the case, as well as identifying the specific binding site of HeuR relative to the chu promoters.
Additionally, transcript levels of several genes involved in enterochelin transport were decreased in the mutant strain, but were not identified by COG analysis as being involved in ion transport. As a result, this COG analysis is not comprehensive as the number of genes in the HeuR regulon that are involved in iron transport and metabolism is likely underreported. Nevertheless, based on  these results, it becomes apparent that HeuR is primarily involved in maintaining sufficient levels of intracellular iron by upregulating systems that can scavenge for iron from diverse sources, which is supported by our ICP-MS data. In addition to those genes that were underexpressed, transcripts of several were more abundant in the heuR mutant. Most of these genes are in two gene clusters: Cjj81176_1390 to -1394 and Cjj81176_1214 to -1217. The former, Cjj81176_1390 to -1394, includes the overexpressed gene (Cjj81176_1390) noted above whose product promotes flagellar interactions, supporting both the previous study and our work (15). The latter gene cluster that is upregulated includes one gene that encodes an additional ion transporter and multiple genes predicted to be involved in methionine biosynthesis. The methionine biosynthetic enzymes encoded by these genes have been shown in other organisms to be sensitive to oxidation/oxidative stress (17). In an oxidative environment, these enzymes may need to be replenished, and their inverse expression with that of iron-scavenging systems may be common. Whether upregulation of these genes is a direct or indirect effect of HeuR is currently unknown.
Growth assays using an iron chelator found that both the heuR and chuA mutants are reduced for growth under iron-limiting conditions. That this was observed in media without hemoglobin may indicate that the Chu system is partially responsible for the transport of free iron. As such, it would likely be an accessory system since growth was restored when iron was supplemented back, indicating the presence of another iron transport system. When heme and hemoglobin were added to these mutants, growth was not restored, supporting the conclusion that both are required for using heme as an iron source.
Analysis of cell-associated levels of iron showed that the heuR mutant is significantly decreased for iron and that these levels resemble those of a chuA mutant. Most of the iron reduction observed in these experiments is likely due to decreased heme utilization, as cultures were obtained from media containing sheep's blood. This is further supported by the observation that complementation with heuR restores iron levels in an heuR mutant background, but was unable to do so in a chuA mutant background. As our data indicate a minor role for HeuR and the Chu system in acquiring iron independently of heme, we are currently determining whether decreased iron is apparent in the heuR and chuA mutants grown in media without heme or hemoglobin. This will help us determine to what extent other iron sources are impacting cellular iron levels in these mutants.
When the heuR mutant was grown on blood plates, differences in the hemolytic pattern relative to wild-type C. jejuni became apparent. Typically, C. jejuni exhibits alpha-hemolysis on blood plates, likely due to the production of hydrogen peroxide that oxidizes the hemoglobin in the blood to methemoglobin (18). Alpha-hemolysis was not observed in the heuR mutant, indicating that this mutant either produces less hydrogen peroxide or exhibits increased catalase activity. The mutant was highly resistant to hydrogen peroxide, which was associated with elevated levels of catalase activity; reintroduction of heuR led to both decreased resistance to hydrogen peroxide and catalase activity. Of the genes known to regulate catalase activity in C. jejuni-including ahpC, fur, katA, and perR-none was found to be significantly up-or downregulated in the heuR mutant (data not shown). Also, since iron levels are similar in the chuA mutant, it is unlikely that iron levels are responsible for the elevated catalase activity. Thus, we hypothesize that HeuR regulates the production of an unknown factor that perhaps posttranscriptionally regulates the activity of C. jejuni catalase. We are currently working to determine what this factor is.
The heuR mutant exhibited significant decreases in colonization of the chicken cecum. This leads to the following two questions. (i) Is this system mainly required for combating oxidative stress in the cecum, or is it used for other cellular processes? (ii) Is the system acquiring iron from a heme source, or does it serve the dual purpose of acquiring free iron from the environment?
Oxidative stress in the cecum may come from a variety of sources, like metabolism by the intestinal microbiota or from the generation of reactive oxygen species due to intestinal inflammation. While the Campylobacter-chicken relationship is often considered to be commensal in nature, C. jejuni colonization of the chicken gastrointestinal tract may result in low levels of transient inflammation (19 -21; L. M. Davis and V. J. DiRita, unpublished observation). Also, iron acquisition systems like Chu may be more important for acquiring iron that can be used in other cellular processes, including energy generation and DNA replication (6). Based on the results from the colonization studies, either scenario is possible. During monoinfection, the heuR mutant does not need to compete for iron with wild-type C. jejuni and should colonize to relatively high levels, which is what we observe. Conversely, during competition, the heuR mutant would need to compete with wild-type C. jejuni for resources and would be at a disadvantage for colonization-the observed phenotype. We are currently working to determine whether decreasing oxidative stress in the chicken rescues colonization of the heuR mutant or whether other processes, like energy production or DNA replication, are affected. Since they are not mutually exclusive, it is possible that both scenarios are supported.
Second, the question arises as to whether C. jejuni uses the Chu system to acquire iron from either/or a heme or nonheme source. It is currently unknown how much heme chickens consume in a commercial diet, and we could find no data on heme availability in their natural diet. With the exception of higher plants that can produce heme in the plastids, the absence of heme molecules in most grains, plants, and insects makes it less obvious how chickens would have access to heme-containing molecules (22). The likelihood then becomes that C. jejuni uses this system to scavenge heme either from a microbial source or from the host. The C. jejuni 81-176 genome encodes all heme biosynthetic enzymes, with the exception of homologs for protoporphyrinogen oxidase (hemY or hemG). As such, in an iron-rich environment, C. jejuni can likely synthesize its own heme. It could then use the heme uptake system to recover heme from other Campylobacter cells or from heme-producing residents of the microbiota. Finally, it re-mains possible that this system is able to transport free iron. While the heme receptor and heme oxygenase (ChuA and ChuZ, respectively) would likely not be needed for free iron transport, the accompanying ABC transport system (ChuBCD) may be promiscuous and function similar to other iron transporters. It should be noted that growth of both the heuR and chuA mutants could be rescued in chelator-treated media by the addition of iron. This suggests that other iron transport systems are active, at least in vitro, so the requirement of the Chu system for in vivo iron transport appears unlikely. Regardless, both scenarios are supported by the colonization assays since a mutant deficient for uptake of either heme or nonheme iron would be at a disadvantage in the host. We are currently investigating whether the ChuBCD transporter can translocate nonheme iron into C. jejuni.
Campylobacters, like most organisms, have evolved tight regulation on iron acquisition. This is largely due to the observation that iron is a required cofactor for numerous cellular processes but also has the capacity to participate in the formation of reactive oxygen species, leading to oxidative stress within the bacterial cell (23). HeuR participates in controlling the utilization of alternate iron sources. The canonical ferric uptake regulator (Fur) represses iron acquisition under conditions where iron is abundant and readily available and also promotes resistance to both acid and oxidative stress (24,25). Fur exists as a monomer under low-iron conditions, but when iron is present, Fur dimerizes and binds to discrete locations, termed "Fur boxes" to alter gene expression. PerR is another iron-responsive regulator that contributes to peroxide and superoxide stress defense (26)(27)(28). Thus, C. jejuni has evolved separate but overlapping regulons controlled by HeuR, Fur, and PerR to aid the cellular response to stimuli, including changes in iron availability and cognate perturbations in cellular oxidative stress (29).
Finally, the question also arises as to the impact of HeuR and the Chu system on dissemination of C. jejuni to extraintestinal sites like the spleen and/or liver in poultry (30). Similarly, this system may also be important in human infection as C. jejuni has been shown to infect the blood, liver, and pancreas of patients, causing bacteremia, hepatitis, and pancreatitis, respectively (31)(32)(33). These sites, compared to the gut, are presumably more restricted for free iron, and the ability of Campylobacter to acquire iron from a heme source may be important during these infections.
In conclusion, our work demonstrates that HeuR is a transcriptional regulator that binds to the promoter region of chuZA and induces expression of bacterial factors which play a role in inorganic ion transport. As a result of HeuR activity, C. jejuni elaborates proteins that aid in the utilization of both heme and hemoglobin as alternate sources of nutrient iron, thereby elevating cell-associated iron in response to iron restriction via nutritional immunity. Lack of HeuR promotes resistance to peroxide, a phenotype that is correlated with elevated catalase production. Together, these results reveal that HeuR is critical for colonization of the gastrointestinal tract of an animal host by promoting metal homeostasis and regulating cellular processes, which are critical for pathogenesis.

MATERIALS AND METHODS
Day-of-hatch chick colonization. Day-of-hatch white Leghorn chickens were inoculated by oral gavage with 100 l of an approximately 1 ϫ 10 9 -CFU/ml bacterial suspension in sterile phosphate-buffered saline (PBS) (10 8 CFU). Colonized chickens were housed for 7 days and euthanized, and cecal contents were collected. Cecal contents were homogenized in sterile PBS, and serial dilutions were plated on Campylobacter selective media: Mueller-Hinton (MH) agar containing 10% sheep's blood, cefoperazone (40 g/ml), cycloheximide (100 g/ml), trimethoprim (10 g/ml), and vancomycin (100 g/ml). Plates were grown for 48 h under microaerobic conditions, and colonies were enumerated. Statistical analysis of the medians was performed using a Mann-Whitney test.
For the competition experiment, equal numbers of wild-type C. jejuni cells and the heuR mutant were mixed, and approximately 1 ϫ 10 8 of each were used to inoculate day-of-hatch chicks by oral gavage. Harvest of cecal contents was done as described above, except serial dilutions were plated on Campylobacter selective media with and without kanamycin (100 g/ ml). Competitive index values were calculated following correction for inoculum amounts by dividing the number of heuR mutants by the number of wild-type bacteria. Statistical analysis was done using a one-sample t test against a hypothetical value of 1.
All chicken experiments and protocols were approved by the Institutional Animal Care and Use Committee in the Animal Care Program at Michigan State University.
RNA-Seq analysis. Triplicate cultures of wild-type DRH212 and the heuR mutant were grown on MH agar supplemented with 10% whole sheep's blood and 10 g/ml trimethoprim overnight at 37°C under microaerobic conditions. Cells were harvested and grown on new media for a second night using the same conditions. Following growth, cells were harvested from each plate, and RNA was extracted using a previously described protocol (34). DNA was degraded using Turbo DNA-free (Thermo) according to the manufacturer's instructions, and removal was confirmed by conventional PCR. RNA quality was assessed using an Agilent 2100 Bioanalyzer, and libraries were generated using the TruSeq RNA library preparation kit (Illumina). Libraries were sequenced using an Illumina HiSeq 2000 formulated for single-end, 50-nucleotide reads. Analysis of sequence data was performed using SPARTA (Simple Program for Automated reference-based bacterial RNAseq Transcriptome Analysis) (35). Briefly, raw reads were trimmed and assessed for quality using Trimmomatic and FastQC, respectively (36). Trimmed reads were mapped to the C. jejuni 81-176 reference genome using Bowtie, and expression levels were determined by HTSeq (37,38). Finally, differential expression analysis was performed using edgeR (39). Cluster of orthologous group (COG) identification of selected genes was performed using the Microbial Genomic context Viewer (MGcV).
Protein purification. The gene heuR was amplified by PCR from C. jejuni 81-176 genomic DNA using primers 5= Cjj1389 BamHI and 3= Cjj1389 SalI. The PCR fragment was subcloned into pGEM-T Easy using the manufacturer's protocol and sequenced using primers T7 and SP6. Fragments that were confirmed to have maintained the full nucleotide sequence of heuR were excised using BamHI and SalI and ligated into the expression vector pQE-30 (Qiagen). Ligation products were transformed into Escherichia coli NEB C3013, and insertion was confirmed using a restriction digest of purified plasmids. Induction of 6ϫHis-HeuR using 1 mM IPTG (isopropyl-␤-D-thiogalactopyranoside) was confirmed for six constructs by SDS-PAGE and Coomassie staining. One strain was used for protein purification where a 1-liter culture was grown to the mid-log phase and induced for 3 h at 37°C using 1 mM IPTG. Cells were pelleted at 8,000 rpm for 20 min at 4°C and resuspended in 20 ml lysis buffer (pH 8.0) (50 mM NaH 2 PO 4 , 300 mM NaCl, 10 mM imidazole). The cell suspension was passed three times through a French press, and cell debris was removed by centrifugation at 8,000 rpm for 20 min at 4°C. The cleared lysate was added to 5 ml of Ni nitrilotriacetic acid (NTA) agarose that was preequilibrated with lysis buffer and rocked overnight at 4°C. This resinprotein mixture was packed by gravity into a chromatography column, and lysate was passed through and collected. Nonspecific proteins were removed by washing the resin bed twice with 5 column volumes (25 ml) of wash buffer (pH 8.0) (50 mM NaH 2 PO 4 , 300 mM NaCl, 20 mM imida-zole). Bound protein was eluted four times using 1 ml of elution buffer (pH 8.0) (50 mM NaH 2 PO 4 , 300 mM NaCl, 250 mM imidazole). Successful purification of 6ϫHis-HeuR was verified by SDS-PAGE and Coomassie staining. Fractions containing the largest amount of HeuR were pooled and dialyzed against storage buffer (50 mM Tris [pH 7.5], 150 mM NaCl, 0.5 mM EDTA, 0.02% Triton X-100, 2 mM dithiothreitol [DTT], 50% glycerol). Concentrations of 6ϫHis-HeuR were determined using a bicinchoninic acid (BCA) protein assay (Pierce), and aliquots of protein were stored at Ϫ80°C.
Electrophoretic mobility shift assays. A specific DNA probe was generated from the chuZA promoter region, and a nonspecific probe was generated from the mapA-ctsW intergenic region by PCR of C. jejuni 81-176 genomic DNA. Amplicons were purified using a PCR purification kit (Qiagen) and subsequently end labeled with [␥-32 P]ATP using T4 polynucleotide kinase (PNK) (New England Biolabs). DNA binding reaction mixtures containing radiolabeled specific (chuZA promoter) and nonspecific (mapA-ctsW intergenic) probes at approximately 0.25 nM, DNA binding buffer (10 mM Tris, pH 7.5, 50 mM KCl, 1 mM EDTA, 1 mM DTT, 5% glycerol), 25 ng/l poly(dA-dT), and 100 g/ml BSA were incubated for 5 min at room temperature. Purified 6ϫHis-HeuR was diluted 2-fold in binding buffer, and 1 l containing 500, 250, 125, 60, 30, 15, 7.5, or 3.75 g of protein was added to individual 19-l binding reaction mixtures (20 l total). These reaction mixtures-and a control reaction mixture without 6ϫHis-HeuR-were incubated for 15 min at room temperature and subsequently resolved on a 5% polyacrylamide native Tris-glycine gel at 4°C. Gels were imaged using a Typhoon phosphoimager.
Hemoglobin and heme growth assays. For bacterial growth assays, C. jejuni strains were grown overnight on tryptic soy agar plates supplemented with 5% sheep blood and appropriate antibiotics (chloramphenicol and/or kanamycin). The following day, plate cultures were utilized to inoculate 5 ml of Mueller-Hinton broth supplemented with appropriate antibiotics. These liquid cultures were incubated under shaking conditions at 37°C overnight. The following day, cultures were diluted 1:10 in fresh Mueller-Hinton broth (40) supplemented with appropriate antibiotics (chloramphenicol and/or kanamycin) alone (medium alone) or supplemented with a 40, 80, or 160 M concentration of the iron chelator desferrioxamine mesylate (Sigma Aldrich) alone or in combination with 0.1 g/ml of purified human hemoglobin (1) or 0.1 g/ml of hemin at 37°C in room air supplemented with 5% CO 2 . Bacterial growth was measured using a spectrophotometric reading at 600 nm (OD 600 ). Statistical analysis was performed using a Student's t test.
ICP-MS analysis. Cultures of wild-type C. jejuni DRH212, the heuR mutant, and a chuA mutant were grown overnight at 37°C under microaerobic conditions on MH agar with 10% whole sheep's blood containing either 10 g/ml trimethoprim or 100 g/ml kanamycin (for mutant strains). For complementation analysis, the wild-type strain with pECO102, the heuR mutant with pECO102, the heuR mutant with pECO-heuR, the chuA mutant with pECO102, and the chuA mutant with pECO-heuR were grown on MH agar with 10% sheep's blood containing either chloramphenicol (30 g/ml) alone (for the wild type) or chloramphenicol and kanamycin (100 g/ml) (for the plasmid-containing mutants). Cultures were grown overnight under microaerobic conditions at 37°C. Strains were passed for a second night on fresh media and harvested for iron quantification. Approximately 1,010 cells were washed three times with 5 ml and resuspended in 500 l of molecular-grade, distilled water. Cells were solubilized overnight at 50°C in 1 ml 50% nitric acid, followed by the addition of 9 ml of water.Extracts were analyzed for 57Fe using a Thermo Scientific ICAP Q quadrupole inductively coupled plasma mass spectrometer in kinetic energy dispersion mode. Instrument calibration was done with aliquots of an Fe synthetic standard solution diluted at 2.5, 10, and 100 ppb and natural water standards NIST SRM 1640 and 1643e. A solution of 100 ppb Fe was analyzed every 5 to 6 samples to monitor instrumental drift; all samples were corrected for linear drift. The oxide formation rate was determined to be Ͻ2% by measuring CeO/Ce, and the