Acinetobacter baumannii Coordinates Urea Metabolism with Metal Import To Resist Host-Mediated Metal Limitation

ABSTRACT During infection, bacterial pathogens must adapt to a nutrient metal-limited environment that is imposed by the host. The innate immune protein calprotectin inhibits bacterial growth in vitro by chelating the divalent metal ions zinc (Zn2+, Zn) and manganese (Mn2+, Mn), but pathogenic bacteria are able to cause disease in the presence of this antimicrobial protein in vivo. One such pathogen is Acinetobacter baumannii, a Gram-negative bacterium that causes pneumonia and bloodstream infections that can be complicated by resistance to multiple antibiotics. A. baumannii inhibition by calprotectin is dependent on calprotectin Mn binding, but the mechanisms employed by A. baumannii to overcome Mn limitation have not been identified. This work demonstrates that A. baumannii coordinates transcription of an NRAMP family Mn transporter and a urea carboxylase to resist the antimicrobial activities of calprotectin. This NRAMP family transporter facilitates Mn accumulation and growth of A. baumannii in the presence of calprotectin. A. baumannii is found to utilize urea as a sole nitrogen source, and urea utilization requires the urea carboxylase encoded in an operon with the NRAMP family transporter. Moreover, urea carboxylase activity is essential for calprotectin resistance in A. baumannii. Finally, evidence is provided that this system combats calprotectin in vivo, as deletion of the transporter impairs A. baumannii fitness in a mouse model of pneumonia, and this fitness defect is modulated by the presence of calprotectin. These findings reveal that A. baumannii has evolved mechanisms to subvert host-mediated metal sequestration and they uncover a connection between metal starvation and metabolic stress.

Calprotectin is important for host defense against A. baumannii in the lung. Calprotectin comprises 45% of the cytoplasmic protein in neutrophils (10), and A. baumannii infection of the murine lung leads to robust recruitment of neutrophils, which are necessary for bacterial clearance (11). Neutrophil recruitment causes a dramatic accumulation of calprotectin that colocalizes with sites of lobar inflammation and A. baumannii colonization (12). Calprotectin-deficient mice have increased bacterial burdens and mortality from A. baumannii pneumonia (13). Finally, recombinant calprotectin inhibits A. baumannii growth in vitro, and this is dependent on an intact hexahistidine Mn binding site within calprotectin; this finding suggests that A. baumannii requires Mn for full fitness (8,13).
Mn is an essential cofactor for life and is predominantly utilized as a redox-active cofactor for enzymes, including superoxide dismutase and ribonucleotide reductase (14). Several families of Mn transporters have been identified in bacteria. The most widely conserved of these are Mn ATP binding cassette (ABC) transporters and the natural resistance-associated macrophage protein (NRAMP) family of Mn transporters, which are important for the virulence of many bacterial pathogens (14). NRAMP family transporters are transmembrane proteins that utilize the proton motive force as an energy source for transport (14). For the pathogen Staphylococcus aureus, both an NRAMP family transporter and an ABC family Mn transporter are important for bacterial resistance to calprotectin (15). To date, no Mn transporters have been characterized in A. baumannii.
Calprotectin-mediated Mn deprivation restricts A. baumannii growth, presumably because Mn-dependent bacterial processes are rendered inactive without their cognate cofactor. However, exactly which bacterial processes are inhibited and how the bacterium responds to these alterations in physiology remain unknown. For instance, multiple metabolic enzymes involved in carbon metabolism, including phosphoglyceromutase (16) and pyruvate carboxylase (17), require Mn or are activated by Mn, but whether central metabolic processes are altered by calprotectinmediated Mn sequestration is unclear. We hypothesized that understanding the effects of calprotectin exposure on A. baumannii physiology in vitro may uncover bacterial processes essential for infection in niches where calprotectin is abundant.
The overall goal of this study was to identify mechanisms by which A. baumannii overcomes calprotectin-based nutritional immunity. An operon that contains a putative Mn transporter and urea catabolism enzymes from the urea amidolyase family was identified. Based on transcriptional regulation, we hypothesized that urea amidolyase is a component of the A. baumannii response to calprotectin-mediated Mn sequestration. Mn transport was demonstrated to be important for growth in the presence of calprotectin and colonization of the murine lung. Urea catabolism was found to be vital for growth in the presence of calprotectin and functionally linked to Mn acquisition in A. baumannii. Taken together, these results uncover that host-mediated metal sequestration restricts metabolism in bacterial pathogens, and they broaden the understanding of the bacterial factors required to survive this restriction.

RESULTS
A. baumannii encodes an NRAMP family transporter that mediates resistance to calprotectin. We hypothesized that A. baumannii is able to overcome calprotectin-mediated Mn chelation by utilizing a metal transporter system that has high affinity for Mn. To identify predicted Mn transporters in A. baumannii, the KEGG database (18,19) was searched for Mn transporter orthologs in the A. baumannii ATCC 17978 genome. This search identified only one gene encoding a protein with similarity to NRAMP or ABC family Mn transporters, the gene A1S_1266. A1S_1266 encodes a potential NRAMP family member. NRAMP family members are integral membrane proteins that transport divalent cations, often having specificity for Mn(II) (14). A1S_1266 is in a predicted operon containing genes that catabolize urea to ammonia, which we named the manganese and urea metabolism (mum) operon (Fig. 1A). As A1S_1266 is predicted to encode a transporter, this gene was named mumT (Fig. 1A).
We hypothesized that mumT encodes an Mn importer that is important for growth under Mn-restricted conditions, such as upon exposure to calprotectin. Consistent with this, normalized mumT transcript abundance increased upon calprotectin treatment (Fig. 1B). Genetic deletion of mumT delayed A. baumannii growth in the presence of recombinant calprotectin, as did deletion of the Zn import gene znuB ( Fig. 1C and D). This growth lag could be complemented by the expression of mumT from a plasmid or by the addition of excess Mn (see Fig. S1A to C in the supplemental material). Importantly, growth inhibition of a ⌬znuB mutant in the presence of calprotectin was only fully rescued by the addition of excess Zn and was not rescued by excess Mn alone (20). We hypothesized that if MumT preferentially imports Mn, loss of mumT would increase resistance to toxic levels of Mn but not other divalent cations. In support of this, the ⌬mumT mutant was able to grow in 3 mM Mn, which is highly toxic to wild-type A. baumannii ( Fig. 1E and F), whereas the ⌬mumT mutant was more sensitive than wild-type A. baumannii to Fe toxicity and had similar sensitivity to Zn toxicity (see Fig. S1D and E). Finally, to determine whether mumT is required for Mn acquisition, cellular Mn concentrations were measured by inductively coupled plasma mass spectrometry (ICP-MS) (Fig. 1G). The ⌬mumT strain had lower Mn levels in cell pellets than did wildtype A. baumannii, and this defect was complemented by expression of mumT in trans. In contrast, cellular Zn and Fe levels in the ⌬mumT mutant were not significantly different than the wild-type A. baumannii levels (see Fig. S1F). Together, these results demonstrate that A. baumannii mumT is important for growth under Mn-restricted conditions and for accumulation of cellular Mn.
mumT upregulation in the presence of calprotectin requires the LysR family transcriptional regulator MumR. mumT is present immediately downstream from a gene that encodes a predicted LysR family transcriptional regulator, mumR (Fig. 1A). We hypothesized that mumR is required for transcriptional control of mumT. To understand the contributions of mumR to mumT regulation, a strain containing an in-frame deletion of mumR was generated. The abundance of mumT transcript was significantly decreased in the ⌬mumR strain relative to that with wild-type A. baumannii ( Fig. 2A). This suggested that MumR activates mumT expression. To confirm this finding, activity of the mumT promoter was investigated using a reporter system. The mumT promoter was cloned into a plasmid harboring the Photorhabdus luminescens luciferase operon, luxABCDE, such that activity of the mumT promoter resulted in luminescence. This vector was transformed into wild-type A. baumannii and ⌬mumR strain cells, and luminescence was measured under various growth conditions. In rich medium, the mumT promoter was active in wild-type A. baumannii but not in the ⌬mumR strain (Fig. 2B). When incubated with 125 g/ml calprotectin, the activity of the mumT promoter was significantly enhanced in wild-type A. baumannii, and this was dependent on mumR. This result demonstrated that calprotectin exposure induces expression from the mumT promoter. Furthermore, it showed that mumR is required for upregulation of mumT in the presence of calprotectin.
mumT is in an operon with mumC, which encodes a urea carboxylase that contributes to A. baumannii urea utilization.
Based on the small intergenic distances between the open reading frames (ORFs) for mumT (A1S_1266) and mumC (A1S_1270), we predicted that these genes constitute an operon. This was confirmed by performing PCR on cDNA prepared using RNA isolated from wild-type A. baumannii (see Fig. S2A in the supplemental material). PCR products were amplified across adjacent ORFs for mumT, mumL, mumU, mumH, and mumC, and no products were amplified for primers designed to amplify the mumR-mumT or mumC-A1S_1271 junctions, demonstrating that mumTLUHC form an operon.
Next, we sought to identify the functions of members of the mum operon. MumH is homologous to allophanate hydrolase, and MumC is a putative member of the biotin carboxylase family. In fungi, the enzyme urea amidolyase contains urea carboxylase and allophanate hydrolase domains. Urea amidolyase converts urea first to allophanate via biotin-mediated carboxylation (urea carboxolyase domain) and then converts allophanate to carbon dioxide and ammonia via hydrolysis (allophanate hydrolase domain) (21). Because mumH and mumC are adjacent ORFs, we posited that MumC is a biotin-dependent urea carboxylase and Significance was calculated using a Student's t test. (G) Mn levels of WT A. baumannii with empty vector, strain ⌬mumT with empty vector, and strain ⌬mumT with a vector containing mumT under control of the 16S promoter grown to mid-log phase. Cells were digested in metal-free nitric acid and analyzed by inductively coupled plasma mass spectrometry. Data are combined from three biological replicates that were measured in technical triplicates, with the means and standard deviations graphed. Significance was calculated using a Student's t test. For all panels, statistical significance is indicated as follows: *, P Ͻ 0.05; **, P Ͻ 0.01; ***, P Ͻ 0.001; n.s., not significant.
MumH and MumC mediate urea degradation (Fig. 3A). Urea amidolyase enzymatic activity has been characterized in enzymes cloned from Oleomonas sagaranensis (22,23), but the roles of urea amidolyase in bacterial physiology and pathogenesis remain unknown.
To evaluate the importance of each mum gene in urea degradation, in-frame deletions of mumL, mumU, mumH, and mumC were generated. Inactivation of mumC, but not of mumH, impaired growth in the presence of toxic levels of urea ( Fig. 3B and C) but not in LB alone (see Fig. S2B in the supplemental material). While A. baumannii cannot grow using urea as a sole carbon source (data not shown), A. baumannii can utilize urea as a sole nitrogen source in a manner dependent on the presence of mumC ( Fig. 3D and E). Importantly, the ⌬mumH mutant was capable of utilizing urea as a sole nitrogen source (Fig. 3E), which indicated that the single ammonia molecule produced by MumC-mediated urea carboxylation (Fig. 3A) is sufficient to support A. baumannii growth as a sole nitrogen source.
Urea carboxylase activity in Oleomonas sagaranensis and Candida utilis requires either Mg 2ϩ , Mn 2ϩ , or Co 2ϩ (23,24), suggesting that A. baumannii utilization of urea by MumC may require Mn 2ϩ . To address this, we chelated Mn by the addition of calprotectin to medium containing urea as the sole nitrogen source. The addition of calprotectin was sufficient to substantially impair growth in urea ( Fig. 3F and G). Importantly, growth was not inhibited when a variant of calprotectin (⌬S1) that is unable to bind Mn (8) was added to the medium. Consistent with these findings, mumT inactivation decreased growth in high concentrations of urea, which was complemented by the expression of mumT in trans, demonstrating that the MumT Mn transporter is important to enable urea degradation (see Fig. S2C and D in the supplemental material). Together, these results demonstrated that mumC is vital for catabolism of urea in A. baumannii and suggest that urea catabolism may be altered by Mn availability. mumC is important for A. baumannii growth in calprotectin. The entire mum operon is upregulated following exposure to calprotectin (Fig. 4A). Based on this observation, we hypothesized that additional mum genes may be important for growth in calprotectin. The sensitivities of A. baumannii ⌬mumL, ⌬mumU, ⌬mumH, and ⌬mumC mutant strains to calprotectin growth inhibition were evaluated (Fig. 4B). The strains lacking mumL and mumU exhibited sensitivity to calprotectin similar to that of wildtype A. baumannii ( Fig. 4C; see also Fig. S3A in the supplemental material). This demonstrated that mumL and mumU are not important for growth in the presence of calprotectin. In contrast, the ⌬mumH and ⌬mumC mutants, strains harboring in-frame deletions of allophanate hydrolase and urea carboxylase, respectively, had increased sensitivity to calprotectin ( Fig. 4B and C). The ⌬mumC mutant did not exhibit decreased transcription of mumT (see Fig. S3B). The growth deficit of the ⌬mumC mutant in calprotectin could be complemented by the addition of exogenous Mn to the medium (see Fig. S1C in the supplemental material) or by providing a copy of mumC in trans (see Fig. S3C and D). Importantly, the ⌬mumC strain did not exhibit resistance to Mn toxicity (see Fig. S3E), suggesting that its role in calprotectin resistance is not related to Mn transport. These data indicate that urea catabolism via mumH and mumC is important for A. baumannii resistance to calprotectin.
mumT contributes to the fitness of A. baumannii in a murine pneumonia model. To investigate the contribution of mumT to A. baumannii fitness in the murine lung, C57BL/6 mice were inoculated intranasally with a 1:1 mixture of wild-type A. baumannii and the ⌬mumT mutant. After 36 h, bacterial burdens were quantified from the lungs and the liver, a site of systemic dissemination ( Fig. 5A and B). Strain ⌬mumT burdens were significantly lower than wild-type burdens in both the lungs and the livers of C57BL/6 mice, indicating that mumT contributes to fitness in this infection model. To define the role of calprotectin in the fitness defect of the ⌬mumT strain, calprotectin-deficient mice (S100A9 Ϫ/Ϫ ) were also coinfected with wild-type A. baumannii and the ⌬mumT strain. As in C57BL/6 mice, ⌬mumT strain burdens were significantly lower than those in the wild type in the lungs of calprotectin-deficient mice. However, the fitness deficit of the ⌬mumT strain in the liver was completely rescued in the absence of calprotectin. These findings indicate that calprotectin is vital for limiting A. baumannii ⌬mumT strain dissemination to the liver. To elucidate whether the differential rescue of the ⌬mumT mutant in the liver and the lung of calprotectin-deficient mice correlates with Mn concentrations, A. baumannii-infected livers and lungs were subjected to ICP-MS analysis. Mn abundance in the liver was over 25-fold greater than Mn abundance in the lung (Fig. 5C). The high level of Mn in the liver correlated with increased fitness of the ⌬mumT strain in the liver relative to that in the lung in calprotectin-deficient mice; this indicated that the liver of a calprotectin-deficient mouse is Mn replete. The result indicating that the ⌬mumT strain is attenuated in the livers of wildtype mice suggests that calprotectin is sufficient to Mn starve A. baumannii even in the Mn-abundant liver. Taken as a whole, these data reveal that mumT is important for A. baumannii fitness during infection and demonstrate that calprotectin is important for preventing ⌬mumT strain dissemination to the murine liver.
The mum system is broadly conserved across bacteria. Calprotectin has antimicrobial activity against numerous pathogens (8). Because of the importance of the mum operon in A. bauman- nii resistance to calprotectin, the conservation of this system was investigated (Fig. 6). The mum operon was present in all A. baumannii strains queried and other Acinetobacter species. A mumR homolog was present adjacent to the mum operon in all Acinetobacter species interrogated but was not present outside the Acinetobacter genus, suggesting that Acinetobacter has evolved with a unique regulatory mechanism for this operon. The mum operon is present with at least four of the five genes retained in some other Gammaproteobacteria, including the urinary pathogen Proteus mirabilis, and more distantly related Proteobacteria, including Agrobacterium tumefaciens. Portions of the mum operon, including NRAMP family transporters and allophanate hydrolase genes, are also present in diverse bacterial phyla, including Actinobacteria and Firmicutes. Importantly, portions of the mum operon are present in diverse bacterial pathogens, including S. aureus and Neisseria meningitidis. These observations indicate that the mum

DISCUSSION
A. baumannii colonization of the murine lung generates a robust immune response, which ultimately results in copious amounts of calprotectin being present at the host-pathogen interface (12). Here, we demonstrated that the mum operon responds to calprotectin and contributes to A. baumannii calprotectin resistance (Fig. 7). MumT was established as a Mn transporter, the unique transcriptional regulation of mumT by mumR was determined, and a link between mumT and urea catabolism via mumC was identified. Additionally, urea catabolism was identified as the first metabolic pathway linked to calprotectin resistance, an important step in identifying the mechanisms by which calprotectin disrupts bacterial physiology and inhibits bacterial growth during infection. Finally, the mum system was demonstrated to be important for A. baumannii fitness in the murine lung and liver, and calpro-tectin was found to be required by Mn-starved A. baumannii in the liver.
mumT encodes an NRAMP family homolog. mumT is unique compared to previously identified NRAMPs, as MumT shares low sequence homology with reported NRAMPs (Ͻ25%), and NRAMPs are typically monocistronic (25). Inactivation of the S. aureus NRAMP family member MntH increases sensitivity to calprotectin, although a second Mn transporter (MntABC) must be deleted to see a dramatic growth difference in the presence of calprotectin (15). Similar to this finding, deletion of mumT delays A. baumannii growth in the presence of calprotectin, and this growth difference is reversed by the addition of excess Mn to the medium. In contrast to results obtained with S. aureus, A. baumannii growth is significantly decreased by inactivation of mumT alone, suggesting that A. baumannii may not encode another high-affinity Mn import system. Previously reported NRAMP family transporters have varied specificities for Mn, Fe, and other divalent cations (26)(27)(28)(29)(30)(31). The metal specificity of MumT was eval-  A. baumannii and strain ⌬mumT. Following 36 h of infection, mice were euthanized and lungs (primary infection) and livers (dissemination) were harvested, and bacteria were enumerated by dilution plating on nonselective medium and medium containing kanamycin. (A) Bacterial burdens of wild-type and strain ⌬mumT recovered from lungs of C57BL/6 and CP Ϫ/Ϫ mice. The mean results and standard deviations are indicated by horizontal lines. Significance was calculated using one-way ANOVA with Tukey's multiple-comparisons test. (B) Bacterial burdens of wild-type and strain ⌬mumT recovered from livers of C57BL/6 and CP Ϫ/Ϫ mice. Median and interquartile range data are indicated by horizontal lines. Significance was calculated using a Kruskal-Wallis test with Dunn's multiplecomparisons test. For panels A and B, each symbol represents the burden recovered from an individual mouse, and the results of two independent experiments were combined. The limit of detection was 100 CFU per organ and is indicated by the dashed line in both panels. (C) Mn levels in A. baumannii-infected livers and lungs were measured by ICP-MS. Organs harvested from three mice were used for this analysis. Significance was calculated using Student's t test. *, P Ͻ 0.05; **, P Ͻ 0.01; ***, P Ͻ 0.001; ****, P Ͻ 0.0001. uated by determining sensitivity to toxic levels of Mn, Fe, and Zn. The ⌬mumT strain was less sensitive than wild-type A. baumannii to Mn toxicity. This is consistent with the decreased ability of this strain to accumulate Mn, as measured by ICP-MS. Together, the increased sensitivity of the ⌬mumT mutant to calprotectin and decreased sensitivity of the ⌬mumT mutant to toxicity of Mn strongly suggest that mumT encodes a transporter that imports Mn. Interestingly, the ⌬mumT strain had increased sensitivity to  Fe toxicity. We hypothesize that the enhanced sensitivity of the ⌬mumT strain to Fe toxicity stems from disruption of the Mn:Fe ratio. This is consistent with the finding that the Mn:Fe ratio is important for Neisseria meningitidis to survive metal toxicity (32).
The ⌬mumT mutant was less fit than wild-type A. baumannii for colonizing the lung and dissemination to the liver. Of note, the ⌬mumT strain was not attenuated in dissemination to the livers of mice lacking calprotectin, consistent with the model showing that calprotectin is required to starve A. baumannii of Mn and prevents colonization of the liver. S. aureus inactivated for Mn transporters mntH and mntABC is also attenuated in the livers of wild-type but not calprotectin-deficient mice (15). Similarly, A. baumannii inactivated for the Zn transporter znuABC is significantly attenuated in the livers of wild-type mice but not significantly attenuated in calprotectin-deficient mice (13). Together, these findings implicate calprotectin metal sequestration as particularly important in host defense of the liver relative to other organs. The heightened efficacy of calprotectin in the liver may be because the liver, the site of Mn absorption into the systemic circulation and Mn excretion into bile, is the most Mn-replete organ in the body (33). The concentrations of Mn in the A. baumannii-infected liver are approximately 25-fold higher than Mn concentrations in the A. baumannii-infected lung. In this setting of excess Mn, the ⌬mumT mutant appeared to be capable of importing sufficient Mn through other transport systems, unless calprotectin was present to sequester Mn. The ⌬mumT strain was less fit than wild-type A. baumannii in the lungs of mice lacking calprotectin, suggesting that additional stresses beyond calprotectin exist in the lung that decrease the fitness of this mutant strain.
Unlike other reported NRAMP family transporters, mumT is in an operon. The other genes in this operon were not previously described in A. baumannii and lacked an obvious link to Mn homeostasis. A predicted function was not identified for mumU or nml, but based on sequence homology, mumH and mumC are predicted to encode allophanate hydrolase and urea carboxylase, enzymes that catalyze the biotin-and ATP-dependent two-step catabolism of urea to ammonia and carbon dioxide. Homologs of these genes in O. sagaranensis have been cloned and their enzymatic activities verified (22,23); however, allophanate hydrolase in Pseudomonas functions in cyanuric acid metabolism, not urea metabolism (34). Therefore, the physiological role of these enzymes in bacteria can vary. A. baumannii utilized urea as a nitrogen source but not a carbon source, and urea nitrogen utilization depends on mumC. These results demonstrate that urea carboxylase has a physiological role in urea catabolism in this bacterium.
Since mumC is in an operon with mumT, we investigated whether mumC requires mumT-delivered Mn for activity. Previous reports suggested that urea carboxylase activity requires divalent cations (23,24). A strain inactivated for mumT was impaired for growth in urea as a sole nitrogen source and in the presence of toxic levels of urea. Furthermore, this effect was specific to inactivation of mumT and does not occur when other genes in the operon are inactivated. Therefore, this finding suggests that mumC-mediated urea catabolism is Mn dependent.
It is well established that calprotectin inhibits bacterial growth in vitro and hampers the growth of some bacteria during infection (35). Ostensibly, calprotectin inhibits growth by suppressing metal-dependent bacterial processes. However, it is unclear what specific bacterial processes are inhibited and how this affects bacterial physiology; currently, the only bacterial process known to be inhibited by calprotectin is S. aureus superoxide dismutase activity (36). Because evolutionary conservation of genomic organization can suggest similar function, we hypothesized that other genes in the mum operon may be important for resistance to calprotectin. In keeping with this, inactivation of mumC significantly decreased growth in the presence of calprotectin. This result demonstrated that urea degradation increases the ability of A. baumannii to combat calprotectin metal limitation. Furthermore, calprotectin, but not calprotectin lacking the ability to tightly bind Mn, completely inhibited growth of A. baumannii utilizing urea as a sole nitrogen source. One interpretation of these results is that urea degradation is a Mn-, Zn-, or Fe-dependent process that is inhibited by calprotectin.
The question remains: why is urea degradation important for growth in the presence of calprotectin? Urea is generated as a by-product of metabolism in rich medium, and calprotectinmediated metal starvation may cause a metabolic strain by inhibiting metal-dependent metabolic processes. This could lead to a buildup of urea that requires mumC-mediated breakdown. Future work to query this hypothesis will also help define metabolic pathways in A. baumannii. Alternatively, urea and/or ammonia could serve as a signaling molecule within the bacterial cell. Our findings emphasize the importance of improving the understanding of A. baumannii metabolism and the role of metabolism in A. baumannii virulence. In this regard, we report that calprotectinmediated metal starvation and urea catabolism are linked in A. baumannii.
The finding that urea carboxylase is important for defense against the antimicrobial protein calprotectin in vitro extends the known role of urea in microbial pathogenesis. There are two described pathways for catabolizing urea in bacteria: urease and urea amidolyase (37). Urease is a key virulence factor for Helicobacter pylori, as it is required for local alkalinization and chemotaxis in the stomach (38)(39)(40). P. mirabilis also utilizes urease as a virulence factor in the bladder, the site of host urea excretion; urease activity of P. mirabilis alters the pH and causes calculus formation in urine (41). Urease is also required for virulence of the fungal pathogens Cryptococcus neoformans (42) and Coccidioides posadasii (43). The only reported virulence role for urea amidolyase systems is for C. albicans, which uses urea-produced ammonia to regulate pH and induce the yeast-to-hypha transition; this system is important for escape from macrophages and colonization of the kidney (44,45). The present study indicates that urea amidolyase systems are also important for defense against the human antimicrobial protein calprotectin.
The mum operon is conserved across many, but not all, bacteria. The mum operon is present in many nonpathogenic organisms, including Acinetobacter baylyi, suggesting this operon did not evolve exclusively as a virulence factor. However, it is present in many pathogens, including S. aureus. Therefore, a better understanding of the genes within this operon may reveal drug targets for the treatment of multidrug-resistant infections.

MATERIALS AND METHODS
Bacterial strains and reagents. The strains used in this study are described in Table S1 in the supplemental material. All strains are derivatives of the human clinical isolate A. baumannii ATCC 17978. Cloning was performed in Escherichia coli DH5␣. Bacteria were routinely grown in lysogeny broth (LB) at 37°C unless otherwise noted. Solid medium contained 1.5% agar. Antibiotics were added at the following concentrations for A. baumannii and E. coli, respectively: 500 g ml Ϫ1 and 100 g ml Ϫ1 ampicillin, 40 g ml Ϫ1 kanamycin, 10 g ml Ϫ1 , and 5 g ml Ϫ1 tetracycline. All antibiotics were purchased from Sigma (St. Louis, MO). Inframe deletion strains (⌬mumR, ⌬mumT, ⌬mumL, ⌬mumU, ⌬mumH, and ⌬mumC mutant strains) were generated via homologous recombination utilizing the suicide plasmid pFLP2 and screened by PCR and Southern blotting as previously described (46). Some constructs were generated by ligating the stitched PCR product directly into pFLP2 (for the ⌬mumR and ⌬mumT mutants) or by using Gibson recombineering (for the ⌬mumL, ⌬mumU, ⌬mumH, and ⌬mumC mutants) (New England Biolabs, Ipswich, MA). Primers used to generate in-frame deletion strains, complementation plasmids, and reporter plasmids are listed in Table S2 in the supplemental material. Complementation vectors for the ⌬mumT and ⌬mumC strains were constructed in pWH1266 under control of the 16S promoter (r01) as previously described (46), except that complementation vectors did not include a c-Myc tag and the mumC complementation vector was cloned between EcoRV and BamHI sites. p.r01.WH1266 was used as the empty-vector control. Antibiotic selection of strains containing the pWH1266 plasmid used 500 g ml Ϫ1 ampicillin. Luciferase promoter reporter constructs were generated in pMU368 and derivatives (47). The Photorhabdus luminescens luciferase operon lux-ABCDE was PCR amplified from pXen1 (48) with primers that included 5= BamHI and 3= SpeI restriction sites; the products were then digested and ligated into pMU368 to generate p.luxABCDE.MU368. To permit selection of the pMU368 plasmid in kanamycin-resistant deletion strains, a tetracycline resistance cassette was PCR amplified from AB0057 genomic DNA and ligated into p.luxABCDE.MU368 at the KpnI restriction site to create p.luxABCDE.MU368.tet. The ligation product was transformed into DH5␣ and selected on 5 g ml Ϫ1 tetracycline-LB agar, and the plasmid was purified. An approximately 300-bp segment of the mumT promoter was PCR amplified with primers that included 5= SacI and 3= BamHI restriction sites, restriction digested, and ligated into p.lux.ABCDE.MU368.tet to generate p.P mumT .luxABCDE.MU368.tet. The ligation product was transformed in DH5␣, plasmid purified by using a miniprep kit, and transformed into wild-type A. baumannii or ⌬mumR mutant cells. Antibiotic selection of strains containing the p.MU368.tet plasmid was achieved by using 10 g ml Ϫ1 tetracycline. Recombinant human calprotectin was expressed and purified as previously described (35).
Bacterial growth assays. Unless otherwise stated, all growth assays were carried out in 96-well flat-bottom plates in 100-l volumes following inoculation with 1-l aliquots of overnight cultures. The optical density at 600 nm (OD 600 ) was measured as a proxy for growth.
(i) Urea and metal toxicity assays. All urea and metal toxicity assays were carried out in LB medium with the addition of freshly prepared, sterile metal or urea stocks. Stock solutions (10 M) of urea were prepared in LB, and the final volume of each well was 90 l. Stocks containing 100 mM MnCl 2 , 100 mM ZnCl 2 , or 100 mM FeSO 4 were prepared in water, filter sterilized, and used immediately.
(iii) mumT-reporter luminescence assay. For reporter luminescence assays, A. baumannii strains harboring luminescence reporter plasmids were grown in medium containing a titration of calprotectin in 60% LB and 40% calprotectin buffer. Cultures were grown in black-sided 96-well plates (Corning), and luminescence was measured using a plate reader (BioTek, Winooski, VT).
(iv) Sole nitrogen source growth assays. Sole nitrogen source growth assays were performed in modified E medium lacking nitrogen sources (49) and containing the following: 28  Mouse infections. A mouse pneumonia model of A. baumannii infection was employed as previously described (13). Briefly, wild-type A. baumannii or the ⌬mumT mutant strain was freshly streaked from frozen stocks onto LB agar or LB agar containing 40 g ml Ϫ1 kanamycin, respectively, 2 days prior to infection. Overnight cultures were grown in LB without antibiotic selection. On the day of the infection, overnight cultures were subcultured to 1:1,000 in 10 ml of LB and grown to midexponential phase. Cells were then harvested by centrifugation, washed twice in phosphate-buffered saline (PBS), and resuspended in PBS to a final concentration of 1 ϫ 10 10 CFU ml Ϫ1 . Wild-type and ⌬mumT strain suspensions were then combined in a 1:1 ratio, mixed thoroughly, and immediately utilized for infection. Mice were anesthetized via intraperitoneal injection of 2,2,2-tribromoethanol diluted in PBS. Anesthetized mice were inoculated intranasally with 5 ϫ 10 8 CFU in 50-l volumes. Infections proceeded for 36 h. Mice were then euthanized by use of CO 2 , and lungs and livers were removed and placed on ice. Organs were homogenized in 1 ml PBS, serially diluted in PBS, and dilutions were spot plated onto LB agar and LB agar containing 40 g ml Ϫ1 kanamycin. Strain Inductively coupled plasma mass spectrometry. To prepare bacterial samples for ICP-MS analysis, bacterial cultures were grown overnight in LB containing 500 g ml Ϫ1 ampicillin and subcultured 1:50 in LB containing 500 g ml Ϫ1 ampicillin. Bacteria were subcultured 1:50 for 1 h, and then cultures were diluted 1:100 into 10 ml of 60% LB-40% calprotectin buffer containing 500 g ml Ϫ1 ampicillin and grown for 8 h. Bacterial cultures were then transferred to preweighed metal-free 15-ml conical tubes (VWR, Radnor, PA). Pellets were harvested by centrifugation, washed twice with Milli-Q deionized water, and dried. The pellet weight was then recorded using an analytical balance (Mettler-Toledo, Columbus, OH). Pellets were digested with 1 ml 50% HNO 3 (optima-grade metal-free; Fisher, Waltham, MA) at 50°C overnight, diluted with 9 ml Milli-Q deionized water, weighed using an analytical balance, and subjected to mass spectrometry. Whole organs from A. baumannii-infected mice were homogenized in 1 ml PBS and digested in 2 ml HNO 3 and 500 l H 2 O 2 (optima-grade metal-free; Fisher, Waltham, MA) at 90°C overnight in metal-free Teflon jars for digestion. Digested samples were then diluted with 9 ml Milli-Q deionized water and submitted for ICP-MS analysis at the Vanderbilt Mass Spectrometry Research Center. Levels of 66 Zn, 55 Mn, and 56 Fe were measured, concentrations were determined by utilizing a standard curve for each metal, and results were normalized by dilution factor.
Determining the conservation of the mum operon in silico. The "compare region" feature of the SEED viewer (51) was used to identify genomic regions similar to the mum operon, with mumT set as the focus gene. From the 58 Archaea, 962 Bacteria, and 562 Eukarya genomes in the SEED database at the time of query, 88 were found to include sets of genes with similar sequences. Genomic regions from 11 organisms were selected for protein alignments. Protein sequences were downloaded from the SEED database and aligned by using ClustalW2 (52,53) for comparison to the 17978 homolog.
Quantitative RT-PCR. RNA isolation from A. baumannii, cDNA generation, and quantitative reverse transcription-PCR (qRT-PCR) using SYBR green (Bio-Rad) were performed as previously described (46). The threshold cycle (C T ) values for each transcript were normalized based on 16S rRNA levels.
Statistical analyses. All raw numerical data were saved in Excel files and imported into GraphPad Prism for statistical analysis. Specific statistical tests employed for each experiment are specified in the figure legends.

ACKNOWLEDGMENTS
We thank the members of the Skaar laboratory for their critical review of the manuscript and Sophie Jouan for providing recombinant calprotectin proteins. We particularly thank Brittany Mortensen Nairn for her assistance with animal experiments and the construction of p.luxABCDE.MU368. This work was supported by the NIH (grant AI101171 to E.P.S. and W.J.C.), VA (grant INFB-024-13F to E.P.S.), AHA (grant 15PRE25060007 to L.J.J.), and NIH (grant T32 GM07347 to the Vanderbilt Medical Scientist Training Program).
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

FUNDING INFORMATION
This work, including the efforts of Lillian Johnson Juttukonda, was funded by HHS | National Institutes of Health (NIH) (AI101171 and T32 GM07347). This work, including the efforts of Eric P. Skaar, was funded by U.S. Department of Veterans Affairs (VA) (INFB-024-13F). This work, including the efforts of Lillian Johnson Juttukonda, was funded by American Heart Association (AHA) (15PRE25060007).