Identification and characterization of zinc importers in Corynebacterium diphtheriae

ABSTRACT

Corynebacterium diphtheriae is the causative agent of diphtheria, a severe respiratory disease in humans. C. diphtheriae colonizes the human upper respiratory tract, where it acquires zinc, an essential metal required for survival in the host. While the mechanisms for zinc transport by C. diphtheriae are not well characterized, four putative zinc ABC-type transporter loci were recently identified in strain 1737: iutABCD/E (iut), znuACB (znu), nikABCD1 (nik1), and nikABCD2 (nik2). A mutant deleted for all four loci (Δ4) exhibited similar growth to that of the wild-type strain in a zinc-limited medium, suggesting there are additional zinc transporters. Two additional gene loci predicted to be associated with metal import, mntABCD (mnt) and sidAB (sid), were deleted in the Δ4 mutant to construct a new mutant designated Δ6. The C. diphtheriae Δ6 mutant exhibited significantly reduced growth under zinc limitation relative to the wild type, suggesting a deficiency in zinc acquisition. Strains retaining the iut, znu, mnt, or sid loci grew to near-wild-type levels in the absence of the other five loci, indicating that each of these transporters may be involved in zinc uptake. Plasmid complementation with cloned iut, znu, mnt, or nik1 loci also enhanced the growth of the Δ6 mutant. Quantification of intracellular zinc content by inductively coupled plasma mass spectrometry was consistent with reduced zinc uptake by Δ6 relative to the wild type and further supports a zinc uptake function for the transporters encoded by iut, znu, and mnt. This study demonstrates that C. diphtheriae zinc transport is complex and involves multiple zinc uptake systems.

IMPORTANCE

Zinc is a critical nutrient for all forms of life, including human bacterial pathogens. Thus, the tools that bacteria use to acquire zinc from host sources are crucial for pathogenesis. While potential candidates for zinc importers have been identified in Corynebacterium diphtheriae from gene expression studies, to date, no study has clearly demonstrated this function for any of the putative transporters. We show that C. diphtheriae encodes at least six loci associated with zinc import, underscoring the extent of redundancy for zinc acquisition. Furthermore, we provide evidence that a previously studied manganese-regulated importer can also function in zinc import. This study builds upon our knowledge of bacterial zinc transport mechanisms and identifies potential targets for future diphtheria vaccine candidates.

INTRODUCTION

Corynebacterium diphtheriae, a Gram-positive bacterium, causes both a severe respiratory disease and cutaneous infections in humans. To successfully colonize and replicate within the host, C. diphtheriae encodes systems necessary to acquire critical nutrition, including metals such as zinc, which functions in numerous proteins as an enzymatic cofactor or a structural component (1, 2). Despite the abundance of zinc in the host, most zinc is bound to proteins, resulting in limited bioavailability for invading pathogens (2–4). During inflammation, zinc present in serum is depleted by host zinc transporters that result in storage and sequestration of zinc in hepatocytes (5). The host S100 proteins further sequester zinc (3, 4, 6, 7). Calprotectin, a heterodimer of S100A8 and S100A9, is highly abundant in neutrophils and binds several metals, including zinc, restricting their availability (8). “Nutritional immunity” describes the collective mechanisms by which the host restricts the availability of biologically critical metals, including iron, manganese, and zinc (2, 9).

Bacterial pathogens encode a variety of mechanisms to overcome nutritional immunity to survive within the host environment. Zinc import mechanisms encoded by Gram-positive bacterial pathogens include ATP-binding cassette (ABC) transporters, production of zinc-scavenging molecules, and ZIP family proteins (2). Gram-positive bacterial zinc ABC transporter systems comprise a lipid-anchored substrate-binding protein (SBP) that binds zinc on the cell surface, as well as permease and ATPase proteins, which allow for the active transport of zinc through the membrane (2, 10). Zinc-scavenging molecules (metallophores) are another mechanism for bacterial zinc acquisition (2). Like the iron-binding siderophores, metallophores are enzymatically synthesized and exported from the cell; upon binding zinc, they are imported through ABC transporters that recognize the metal-metallophore complex. Finally, ZIP family transporters are found in eukaryotes as well as bacteria, and while the system is not well understood, it is composed of a single membrane protein that controls the movement of zinc across the cytoplasmic membrane. Structural studies for the ZIP transporter from Bordetella bronchiseptica have recently been described (11).

Because zinc acquisition is crucial for survival, bacterial pathogens often encode more than one transporter. The ZnuABC/AdcABC family of high-affinity zinc ABC transporters is prevalent across pathogenic Streptococci, with some species encoding several SBPs for zinc import (12). Streptococcus pneumoniae encodes two redundant SBPs: AdcA and AdcAII (13). Individual deletions of the genes encoding the SBPs do not significantly impact growth, but a mutant deleted for both genes fails to grow under zinc limitation. Staphylococcus aureus encodes the AdcABC zinc transport system as well as the pathway to produce, secrete, and use the metallophore staphylopine (StP) (14). Three biosynthetic enzymes, CntKLM, are required to produce StP, while CntE is needed for secretion. The import of StP requires the CntABCDF transporter, which belongs to the NikA/Opp family of ABC transporters. Mycobacterium tuberculosis (Mtb) also secretes zinc-scavenging molecules (termed kupyaphores) to capture extracellular zinc and other metals (15). While the Mtb ZnuABC locus is uncharacterized, the ZnuABC transporters found in Mycobacterium smegmatis and Mycobacterium avium subsp. paratuberculosis are zinc importers (16). In Clostridioides difficile, the ZupT ZIP family protein contributes to zinc import (17). It is evident from these examples that bacterial pathogens are equipped with various tools to acquire zinc and overcome host sequestration. However, zinc in excess is toxic, likely through mis-metalation of enzymes (2); as such, many pathogens also encode zinc exporters that are induced under zinc stress to reduce intracellular zinc. Studies in Mtb, S. pneumoniae, and other bacterial pathogens indicate that such zinc efflux systems are critical for survival in in vivo and in vitro infection models (12, 18), suggesting important roles in pathogenesis.

In C. diphtheriae, we previously identified several candidate zinc transporters through studies examining zinc-dependent regulation of gene expression in C. diphtheriae strain 1737 (19–21). The expression studies showed that C. diphtheriae encodes two Zn-regulated ZnuABC-like transporters: znuACB (znu) and iutABCD/E (iut) (20, 21). Deletion of znu and iut (22) did not result in growth limitation in low zinc medium, suggesting that additional zinc import mechanisms are present. C. diphtheriae also encodes two zinc-regulated peptide/nickel transporters, dip2125-28 (nikABCD1, nik1) and dip2162-65 (nikABCD2, nik2), which are related to the NikA/Opp family of ABC transporters. Both of these uptake systems show similarities to the S. aureus StP importer, which is required to import the zinc-StP complex and not the metal alone. Additionally, growth in zinc-limited medium results in the expression of the sidA and sidB genes, which encode a putative non-ribosomal peptide synthase and a polyketide synthase, respectively (19); these types of enzymes are required for the biosynthesis of metallophores such as the iron- and zinc-binding metallophore yersiniabactin, which was first identified in Yersinia pestis (14, 23). Genes encoding ZIP family transporters have not been identified in the genomes of any C. diphtheriae strains.

In this study, we explored the mechanisms by which C. diphtheriae transports zinc. We showed that the deletion of six putative zinc uptake systems in C. diphtheriae (strain Δ6) resulted in a growth defect in a zinc-limited medium. The Δ6 mutant was deleted for the putative transport loci iut, znu, nik1, nik2, mntABCD (mnt), and sidAB (sid). We show that chromosomally expressed iut, znu, sid, or mnt operons can independently support wild-type growth in a zinc-limited medium, suggesting that each of these transport systems is involved in zinc uptake. The function of nik1 and nik2 in zinc transport appears complex, and the role in zinc transport for these systems was not determined. This is the first report of zinc transport systems in C. diphtheriae and will facilitate future studies of metal acquisition in this important human pathogen.

RESULTS

Zinc-regulated ABC transporters in C. diphtheriae

We previously identified four zinc-regulated loci that encode ABC transporters with potential roles in C. diphtheriae zinc import: iut, znu, nik1, and nik2 (Fig. 1A). We deleted each of these loci in C. diphtheriae strain 1737 and tested the mutants’ growth in a semi-defined zinc-limited medium (mPGT) (Fig. 1B). All strains grew to a similar density to that of the wild type, indicating that the loss of individual transport systems does not significantly affect growth under low zinc conditions. To address the possibility that the systems are functionally redundant, we deleted all four loci in one strain, designated Δ4 (Δiut Δznu Δnik1 Δnik2). The growth of the Δ4 mutant was not significantly different from that of the wild-type strain in the mPGT medium (Fig. 1B), which suggests that additional zinc uptake systems are likely present in the Δ4 strain.

Fig 1.

Deletions of iut, znu, nik1, and nik2 do not impact growth. (A) Genetic map of zinc-regulated loci: iut, znu, nik1, nik2, and sid. Transcriptional regulation of the various loci by Zn, Mn, or Fe is indicated: upward arrows indicate induction and downward arrows indicate repression in response to the indicated metal. (B) Overnight growth in zinc-limited medium (mPGT) for wild-type C. diphtheriae strain 1737 (WT), single locus deletion mutants, and the Δ4 mutant. The data shown are the mean and standard deviation of four biological replicates; no significant differences in growth were found between the deletion strains and the WT by two-way ANOVA.

Characterization of zinc transport through suppression of zinc toxicity in a cztA mutant

Since the Δ4 strain did not display a growth defect under low zinc growth conditions, we utilized an alternative approach to demonstrate the function of the four putative ABC-type zinc transport systems deleted in the Δ4 mutant. We also used this alternate method to identify additional potential zinc transporters. This approach involved the introduction of a cztA deletion into the Δ4 strain; the cztA gene encodes a cation efflux system, and deletion of the cztA gene (ΔcztA) results in impaired growth in high-zinc conditions but does not affect growth at low zinc levels (19). We hypothesized that the intracellular accumulation of zinc and associated toxicity observed in the ΔcztA strain could be diminished by the deletion of zinc importers. As previously observed, the wild-type strain was unaffected under high zinc conditions, while the cztA deletion mutant grew poorly at elevated zinc levels (Fig. 2) (19). Compared to the cztA mutant, the Δ4 ΔcztA strain required higher levels of zinc to inhibit growth; this apparent reduced toxicity is consistent with the hypothesis that deletion of the zinc transport system(s) in the Δ4 mutant reduces the intracellular levels of zinc and, thus, reduces the zinc-associated toxicity. However, the Δ4 ΔcztA strain is still more sensitive to zinc than the wild type, which is consistent with the conclusion from the growth studies in Fig. 1B that the Δ4 mutant contains additional zinc importers. The Δ4 strain showed similar growth to the wild-type strain in high zinc conditions as expected (Fig. 2).

Fig 2.

Suppression of zinc toxicity by deletion of ABC transporters in ΔcztA. Culture densities (OD₆₀₀) for the indicated strains were measured following overnight growth (16–18 h) in HIBTW with zinc supplementation as indicated (ZnCl₂). The data shown are the mean and standard deviation of six biological replicates.

Identification of additional zinc importers

To identify additional zinc transport systems, we investigated the sid and mnt loci as possible candidates for zinc import. The sid locus encodes a non-ribosomal peptide synthetase and a polyketide synthase similar to enzymes required for the biosynthesis of metallophores such as yersiniabactin from Y. pestis and, to a lesser extent, StP from S. aureus and kupyaphores from Mtb (14, 15, 23). The sid locus is adjacent to the nik2 operon (Fig. 1A), and both sid and nik2 are optimally expressed in response to zinc limitation; the sid genes are also repressed by iron (19, 24). A previous study showed that a C. diphtheriae strain 1737 sidAB deletion mutant had growth similar to the wild type in a medium limited for both iron and zinc, which suggests that deletion of the sid genes had no effect on the cell’s ability to transport sufficient iron and zinc to maintain wild-type levels of growth (24).

The mnt locus encodes a manganese-regulated ABC transporter that was initially proposed to function in manganese import; however, deletion of the mnt locus had no impact on growth in manganese-limited medium (22). This medium was also limited for zinc, indicating that zinc limitation also has no effect on the growth of the mnt deletion mutant. The MntA SBP is annotated as a component of a putative zinc/manganese transport system and belongs to the ZnuA protein family, which includes the C. diphtheriae IutA, IutE, and ZnuA SBPs.

To examine the function of sid and mnt in the cztA model, we introduced deletion mutations for sid and mnt separately into the Δ4 and Δ4 ΔcztA strains and assessed the growth of these mutants in the presence of increasing zinc levels (Fig. 3A). The Δsid Δ4 ΔcztA strain showed nearly identical growth to the Δ4 ΔcztA parent at all zinc levels tested, indicating that the sidAB system does not affect zinc sensitivity under these conditions. In contrast, the Δmnt Δ4 ΔcztA was not sensitive to zinc and exhibited wild-type growth even at the highest levels of zinc, suggesting that the MntA-D transporter likely contributes substantially to zinc import. Furthermore, a strain deleted for only mnt and cztA (Δmnt ΔcztA) tolerated higher levels of zinc than did the Δ4 ΔcztA strain (Fig. 3B). These results clearly indicate that the loss of the MntA-D system has a critical role in the reduction of zinc-dependent toxicity in the ΔcztA strain and strongly suggest a zinc import function for MntA-D.

Fig 3.

The Mnt transporter contributes to zinc toxicity in a cztA-deficient strain. Culture densities for the indicated strains were measured following overnight growth (16–18 h) in zinc-supplemented (ZnCl₂) HIBTW medium. (A) Growth of WT, Δ4, ΔmntΔ4, and ΔsidΔ4 and respective ΔcztA derivatives are compared. The data shown are the mean and standard deviation of four biological replicates. (B) Growth of WT and Δmnt strains with and without cztA are compared. The data shown are the mean and standard deviation of at least three biological replicates.

The strong contribution of mnt to the reduction of zinc toxicity relative to the other putative transport systems may partially be due to how these various transport loci are expressed under the high zinc conditions used in these studies. Expression of the zinc-regulated loci (iut, znu, nik1, nik2, and sid) is repressed to various levels in response to the high zinc concentrations used in the cztA experiments (19), but mnt expression is repressed only by manganese and not by zinc (22, 25). Thus, the mnt genes are expected to be expressed under the high zinc and low manganese conditions used for the cztA studies (Fig. 3). Despite limitations of the cztA approach, this system has provided indirect evidence for zinc transport by the loci deleted in the Δ4 mutant and suggests a role for mnt in zinc import.

Deletion of six transport loci results in deficient growth in zinc-limited medium

Above, we assessed whether the mnt and sid loci could contribute to zinc import in a C. diphtheriae cztA mutant (Fig. 3). The results suggested a role for mnt in zinc import but not for sid under the high zinc conditions used in the study. To assess the requirement for the mnt and sid loci for growth in a zinc-limited medium, we deleted both the mnt and sid loci in the Δ4 strain to construct the Δ6 mutant, deleted for iut, znu, mnt, nik1, nik2, and sid. Compared to the wild type, the Δ6 mutant showed a striking reduction in overall growth (Fig. 4). This finding suggests that multiple transporters targeted in the Δ6 strain contribute to the growth in zinc-limited medium. The growth defect for the Δ6 mutant was partially rescued by supplementation with zinc (Fig. 4A and B). In the growth studies shown in Fig. 4 to 7, we assayed for growth by continuously monitoring culture density using a microplate reader to observe growth trends over an extended period (instead of using single-point growth assays as was done in Fig. 1B).

Fig 4.

Deletion of six loci results in reduced growth in zinc-limited medium. C. diphtheriae strains (indicated) were grown in mPGT or mPGT with 1 µM ZnCl₂ (+Zn). (A) Measurements were taken at 5-min intervals over 23 h. The data shown are the mean (solid line) and standard deviation (shaded area) of 12 biological replicates. (B) The highest measured OD₆₀₀ mean and standard deviation from the data in panel A are shown for statistical comparison. Note that peak density for each strain/condition occurred at different times. A two-way ANOVA Holm-Sidak’s multiple comparisons test was used to compare among samples as indicated. ***P < 0.001 and ****P < 0.0001; n = 12.

Fig 7.

Plasmid complementation of the Δ6. (A) The Δ6 strain carrying various plasmids was grown in zinc-limited mPGT medium. The data shown are the mean (solid line) and standard deviation (shaded area) of 18 biological replicates. The growth curves for pKN-mnt and pKN-znu partially overlap due to similar growth. (B) The highest measured OD₆₀₀ mean and standard deviation from the data in panel A are shown for statistical comparison. Note that peak density for each strain/condition occurred at different times. An ordinary one-way ANOVA Holm-Sidak’s multiple comparisons test was used to compare among samples, each comparing against the vector. ***P < 0.001 and ****P < 0.0001; n = 18.

Growth analysis of putative zinc transport systems

Having established that deletion of all six loci (iut, znu, nik1, nik2, sid, and mnt) in the Δ6 mutant resulted in a growth defect in zinc-limited medium, we tested the function of the individual loci by creating mutant strains lacking all combinations of five of the six loci (e.g., iut⁺ indicates a strain deleted for znu, nik1, nik2, mnt, and sid while retaining the wild-type chromosomal iut locus). The function of the wild-type locus retained on the chromosome was tested by assessing the ability of the strain to grow in a zinc-limited medium. First, we examined systems with SBPs predicted to bind directly to metal ions: iut, znu, and mnt. Strains harboring only the respective systems, iut⁺ (Fig. 5A), znu⁺ (Fig. 5B), and mnt⁺ (Fig. 5C), were able to grow to wild-type density without zinc supplementation (Fig. 5D). Similarly to the wild type, the addition of zinc slightly improved the growth of the mutant strains. The data suggest that each of these systems can import sufficient zinc to support robust growth under the test conditions and illustrate the functional redundancy of zinc import in C. diphtheriae.

Fig 5.

The presence of the Iut, Znu, or Mnt transporters restores wild-type growth to the Δ6 mutant. C. diphtheriae strains (A) iut⁺, (B) znu⁺, and (C) mnt⁺ (see text for explanation) were grown in mPGT or mPGT with 1 µM ZnCl₂ (+Zn). The data shown are the mean (solid line) and standard deviation (shaded area) of 12 biological replicates. For WT and Δ6, only mean results (dotted/dashed lines) are shown, and the data are replicated from Fig. 4A for comparison. Measurements were taken at 5-min intervals over 23 h. (D) The highest measured OD₆₀₀ mean and standard deviation from the data in panels A–C are shown for statistical comparison. Note that peak density for each strain/condition occurred at different times. A two-way ANOVA Holm-Sidak’s multiple comparisons test was used to compare among mPGT (left) and mPGT +Zn (right) samples, each comparing against the WT in the same condition. ***P < 0.001 and ****P < 0.0001; n = 12.

Although the sidAB genes are predicted to encode enzymes involved in the synthesis of a metallophore, neither the metallophore nor the transport system required for the uptake of the metallophore has been identified (24, 26). Nevertheless, the sid⁺ strain showed similar growth to the wild type in a zinc-limited medium, suggesting a role in zinc uptake for the sidAB genes (Fig. 6A and D). In contrast, the growth of nik1⁺ and nik2⁺ was markedly different from that of the wild-type strain. The growth of nik1⁺ was highly variable regardless of zinc supplementation, as evident from the standard deviation of multiple replicate experiments (Fig. 6B and D). While the overall growth for nik1⁺ was higher than Δ6, the peak growth density was not statistically different from Δ6 (Fig. 6B and D); the addition of zinc also did not improve the growth of nik1⁺. The growth of nik2⁺ was comparable to the Δ6 strain and supplementation with zinc did not affect its growth (Fig. 6C and D). The inability of supplemental zinc to rescue the growth defects of the nik1⁺ and nik2⁺ strains is in contrast to what is observed for Δ6, in which growth is significantly enhanced by the addition of zinc (Fig. 4). The reason for the unusual growth phenotype for the nik1⁺ and nik2⁺ strains is not known, and their role in zinc uptake is also unclear.

Fig 6.

Nik1 and Nik2 importers do not support wild-type growth when expressed from the chromosome. C. diphtheriae strains (A) sid⁺, (B) nik1⁺, and (C) nik2⁺ were grown in mPGT or mPGT with 1 µM ZnCl₂ (+Zn). The data shown are the mean (solid line) and standard deviation (shaded area) of 12 biological replicates. For WT and Δ6, only mean results (dashed lines) are shown, and the data are replicated from Fig. 4A for comparison. Measurements were taken at 5-min intervals over 23 h. (D) The highest measured OD₆₀₀ mean and standard deviation from the data in panels A–C are shown for statistical comparison. Note that peak density for each strain/condition occurred at different times. A two-way ANOVA Holm-Sidak’s multiple comparisons test was used to compare among mPGT (left) and mPGT +Zn (right) samples, each comparing against the WT in the same condition.**P < 0.01, ***P < 0.001, and ****P < 0.0001; n = 12. Growth of the sid ⁺ strain was not significantly different from wild type in either condition. Growth of the sid ⁺ strains is significantly different from the Δ6 strain in zinc-limited medium (P value = 0.0007).

Plasmid complementation restores growth to the Δ6 strain

To support the growth results observed using the chromosomally expressed genes described in Fig. 5 and 6, we cloned the iutA-D, znuACB, nikA-D1, nikA-D2, and mntA-D loci into a shuttle vector (pKN2.6Z) in which the cloned genes are expressed from a constitutive promoter. Cloning of the sidAB genes was not done due to the large size of the coding region (13 kb) (24). Plasmids were transformed into the Δ6 strain and tested for their ability to restore growth under zinc-limited conditions. Plasmids encoding the iut, znu, nik1, or mnt locus significantly improved the growth of the Δ6 strain, suggesting that the cloned genes encode functional zinc importers under the conditions tested (Fig. 7A and B). While the results for plasmids encoding iut, znu, and mnt are consistent with the chromosomal expression of the genes (Fig. 5), plasmid-expressed nik1 supported more robust and consistent growth, which may be due to greater expression resulting from multiple copies expressed from a constitutive promoter. A substrate for the nik1 uptake system has not been identified, but the system appears to restore growth in the zinc-limited medium in the absence of the sidAB genes, indicating that the product of sidAB is not required for zinc acquisition by the ABC transporter encoded by the nik1 locus. Neither the chromosomal nor plasmid-encoded copies of nik2 restored growth to the Δ6 strain. Surprisingly, expression of the nik2 loci from the plasmid further reduced the growth of the Δ6 strain (Fig. 7A and B).

The data from the plasmid complementation studies using the cloned loci support a zinc import function for iut, znu, nik1, and mnt. The results using the cloned nik2 locus are less clear and will require additional studies.

Quantification of cellular zinc content in C. diphtheriae strains by ICP-MS

The growth defect of the Δ6 strain may reflect reduced intracellular zinc content due to a lack of functional high-affinity zinc transporters. The introduction into the Δ6 strain of uptake systems that support wild-type growth indicates the potential for the respective systems to transport sufficient zinc into the cell (Fig. 5). We used inductively coupled plasma mass spectrometry (ICP-MS) to measure zinc content in various C. diphtheriae strains. Identical media conditions but higher culture volumes were used to grow bacteria for ICP-MS analysis than those used for the growth analyses (Fig. 8A). The Δ6 strain showed significantly reduced levels of intracellular zinc compared to WT when grown in 1 µM zinc (+Zn) (Fig. 8B). The data indicate that the Δ6 strain is unable to accumulate wild-type levels of zinc when grown in 1 µM zinc. Strains containing iut, znu, or mnt also showed wild-type levels of intracellular zinc when grown in 1 µM zinc, which is consistent with a zinc transport function. No significant differences in intracellular zinc content were observed among the strains under zinc-limited growth (−Zn). The reason for similar intracellular zinc levels between Δ6 and all of the other strains in the zinc-limited medium might be associated with the very low levels of zinc uptake under these conditions. No significant differences in manganese content were observed across samples or conditions, despite the absence of mnt in the Δ6, iut⁺, and znu⁺ strains (Fig. 8C).

Fig 8.

ICP-MS analysis of select strains. For the analysis of cellular metal content, 30 mL cultures grown in 125 mL flasks were harvested after overnight growth in mPGT or mPGT with 1 µM ZnCl₂ supplementation. Cell culture density was measured (A) at the time of harvest; zinc (B) and manganese (C) content of cell pellets were assessed by ICP-MS. The data shown are the mean and standard deviation of three biological replicates; metal content was normalized by wet cell mass. For panel A, a two-way ANOVA Holm-Sidak’s multiple comparisons test was used to compare WT (mPGT) with other strains in the same condition and WT (+Zn) with other strains in the same condition. For panels B and C, an ordinary one-way ANOVA Holm-Sidak’s multiple comparisons test was used for the same comparisons. *P < 0.05 and **P < 0.01; n = 3.

DISCUSSION

Zinc is the second most abundant trace metal in the human host and is an essential nutrient for microorganisms due to its critical function in numerous metalloproteins (1, 2). Bacterial pathogens require zinc for survival in the human host and many of the high-affinity zinc uptake systems are considered essential virulence factors (2, 27). In this study, we examined six independent systems in C. diphtheriae that are associated with zinc transport; these include five ABC-type transporters and a putative metallophore encoded by the sidAB genes. This is the first report identifying a growth phenotype for zinc transporters in C. diphtheriae. A surprising finding in this study was the redundancy of the transport systems. Deletion of all six loci was required to detect a phenotype that was typical of a zinc transport mutant, which is exemplified by diminished growth in a zinc-limited medium that can be partially restored by zinc supplementation. Redundancy in metal transport systems is common in bacterial pathogens; however, to the authors’ knowledge, this level of redundancy in zinc transport systems observed in C. diphtheriae has not been described for other bacteria. Two to three transport systems have typically been identified as zinc uptake systems in other pathogens, including S. aureus (14), M. smegmatis (16), and S. pneumoniae (13). Each of these pathogens, like C. diphtheriae, encodes more than one tool for zinc acquisition, which poses a challenge to understanding the specific function of each system.

In a previous report that examined gene expression in C. diphtheriae, we identified two zinc/Zur-regulated loci that showed homology to Mn/Zn ABC transport systems: znuABC (znu) and iutA-D/E (iut) (19, 21). The iut gene cluster is composed of the complete ABC transporter encoded by iutABCD and a separately transcribed SBP encoded by iutE (Fig. 1A). A function of the iutABCD/E locus was not previously determined, and both SBP proteins, IutA and IutE, were shown to bind Zn²⁺ and Mn²⁺ with similar high affinity (21). Expression of iutABCD is repressed under iron-replete conditions (21) and induced weakly in response to manganese (22). There is also evidence for weak transcriptional repression of the iutABCD locus in response to zinc (19). Expression of iutE was repressed in response to zinc in a Zur-dependent manner and activated in high iron conditions (19, 21). In the present study, the iut⁺ strain is deleted for the znu, nik1, nik2, mnt, and sid loci but retains the wild-type chromosomal iut locus. Our growth studies with iut⁺ show that iutABCD/E can enhance the growth of C. diphtheriae in zinc-limited medium in the absence of the other five loci and that plasmid-encoded iutABCD can rescue the growth defect of the Δ6 strain in the absence of iutE. The difference in the expression of iutA-D and iutE in response to zinc and iron suggests that the IutA and IutE SBPs are produced in different environments within the host. While IutA and IutE may have partially redundant functions in zinc uptake, additional studies will be needed to better define the roles of each SBP in metal transport.

The znu locus encodes components for an ABC transporter that includes an SBP, an ATP binding protein, and a permease component, encoded by znuA, znuC, and znuB, respectively. The operon also contains dip0439, which encodes a predicted membrane protein of unknown function (19). Similarly to the iut locus, the znu locus was also able to enhance the growth of the Δ6 mutant in zinc-limited medium, suggesting these two transport systems may have similar roles in zinc uptake.

Based on their high similarity to known zinc transporters, it was not surprising that the znu and iut systems are involved in zinc uptake. It was, however, unexpected that the mnt system is involved in zinc import. While homologs to the C. diphtheriae MntA-D ABC transporter in other Gram-positive bacteria have not been shown to transport zinc, the SBPs in many of these systems are known to bind both Mn and Zn with high affinity. While the Bacillus anthracis (Ba) mntBC-A importer is essential for growth under manganese limitation (28), in vitro binding studies with BaMntA indicate the capacity to bind both Mn²⁺ and Zn²⁺ with similar affinities (29). Additional studies suggested that zinc is not transported through the BaMnt system and that Zn²⁺ binding inhibits the transport of Mn²⁺. In S. pneumoniae, the Mn SBP PsaA can also bind to both Mn²⁺ and Zn²⁺, and growth data suggest that Zn²⁺ binding interferes with Mn²⁺ import (30). In S. pneumoniae, Zn²⁺ has not been shown to be imported through the associated Mn²⁺ permease, PsaB, even though PsaB shares a highly similar metal binding site to the zinc permease AdcB (31). It is believed that PsaA does not release the bound Zn²⁺ to the permease PsaB (32). Whether zinc can be transported by directly binding to the PsaB permease and bypassing the PsaA SBP has not been examined (31, 32).

Expression of the genes encoding the C. diphtheriae MntA-D ABC transporter, which shows homology with Mn/Zn ABC transporters, is repressed by manganese and is not affected by zinc or iron levels (19, 22). Deletion of the mntA-D locus showed no growth defect in manganese-limited medium (22); however, deletion of the gene encoding the manganese-dependent metalloregulator MntR resulted in sensitivity to high levels of manganese. The sensitivity was alleviated by deletion of mntA-D, which supports a role for MntA-D in Mn transport (22). In contrast to what has been described in other bacterial pathogens, we provide evidence in this study that the C. diphtheriae MntA-D system has the capacity for zinc transport. The evidence that supports a role for the mnt system in zinc uptake includes the following: (i) the mnt system was able to restore growth to the zinc transport mutant Δ6 in zinc-limited medium, either when present on the chromosome (mnt⁺) or cloned on a plasmid (Δ6/pKN-mnt), (ii) the mnt system resulted in an increase in the intracellular zinc content in the mnt⁺ strain when compared to the Δ6 mutant, and (iii) the deletion of the mnt locus was shown to reduce the zinc toxicity in the cztA mutant. These findings collectively suggest MntA-D can function in zinc transport. While the data indicate that the MntA-D system has a role in both manganese and zinc uptake, additional studies are required to better understand the precise mechanism of how this uptake system transports metals.

Yersiniabactin is a siderophore that was initially described in Y. pestis and was shown to be important in iron acquisition and virulence (33). Yersiniabactin was later shown to function as a metallophore that could also transport zinc in Y. pestis (23, 34). The sidAB locus (sid) was originally identified as an iron- and DtxR-regulated region in C. diphtheriae that contained two genes, sidA and sidB, which show similarity to the siderophore yersiniabactin biosynthetic genes irp2 and irp1, respectively (24). Expression of the sid operon was later shown to be regulated by zinc and Zur in C. diphtheriae (19). Deletion of C. diphtheriae sidAB had no effect on iron-siderophore production or growth in iron-limited medium, and a product for the sidAB genes was not identified (24). In this study, we show that retention of the sid locus in the sid⁺ strain significantly enhanced growth in a zinc-limited medium as compared to the Δ6 mutant, suggesting a role in zinc import (Fig. 6A and D).

Two additional ABC transporters found in this study, encoded by nik1 and nik2, showed similarities to uptake systems associated with nickel or peptide transport; however, uptake systems similar to those encoded by nik1 and nik2 have also been associated with the transport of metallophores, such as S. aureus StP (14). The nik2 and sid loci are divergently transcribed from unique promoters located in the intergenic region between the sid and nik2 operons (Fig. 1A); both loci are repressed in response to zinc (19). Because of shared zinc regulation and proximity, we have theorized that the nik2 transporter may function to import the product of the sidAB genes. However, the growth results with the sid⁺ strain suggest that neither nik2 and nik1 nor any of the other zinc uptake systems identified in this study (iut, znu, or mnt) are required for sidAB function in zinc uptake. While a transport system for the sidAB product has not been identified, there are several ABC nickel/peptide transporters and iron-siderophore transport systems in C. diphtheriae that may facilitate import.

How the ABC transporters encoded by nik1 and nik2 function in zinc transport is not known. The nik1⁺ and nik2⁺ strains showed no growth enhancement in zinc-supplemented medium (Fig. 6B and C), suggesting that the presence of the transporter encoded by nik2 (and possibly nik1) appears to interfere with zinc uptake in zinc-supplemented medium. In similar studies of bacterial zinc import, loss of high-affinity zinc transporters is typically rescued by supplementation of growth media with zinc to allow for import through low affinity or non-specific transporters, as was observed with the Δ6 mutant (Fig. 4). When the nik1 locus was expressed from a plasmid, it supported the growth of the Δ6 mutant like that observed for the other zinc uptake systems, suggesting that the contradictory growth trends observed with the chromosomal copy of nik1 may be due to poor expression (Fig. 7A). In contrast, plasmid-expressed nik2 did not rescue the Δ6 growth phenotype and was detrimental to the growth of the Δ6 mutant. These results indicate that weak expression of the nik2 locus was unlikely to be the cause of the poor growth of the nik2⁺ strain but rather that the Nik2 transporter might be interfering with zinc acquisition or zinc metabolism. Understanding the function of the products encoded by nik1, nik2, and sidAB in zinc acquisition will require additional analysis beyond the scope of the current study.

MATERIALS AND METHODS

Strains, media, and growth conditions

Bacterial strains and plasmids used in this study are listed in Table 1. C. diphtheriae strains were routinely grown in heart infusion broth with 0.2% (vol/vol) Tween 80 (HIBTW) or on heart infusion agar (1.5%, wt/vol) (HIA) at 37°C. C. diphtheriae strains were stored at −80°C in HIBTW with 20%, vol/vol glycerol. mPGT medium was prepared as described previously with changes in the quality of reagents (35). Modified mPGT medium is prepared by adding the following components to ddH₂O to the final concentration indicated: D,L-tryptophan (0.49 mM), L-glutamate (3.4 mM), maltose (40.9 mM), calcium pantothenate (1.9 µM), and magnesium sulfate (1 mM, trace metal grade). A solution of beta-alanine (26 µM), nicotinamide (18.8 µM), and pimelate (0.949 µM) is prepared in 1.85% (vol/vol) concentrated HCl (trace metal grade) and added to the medium to the indicated final concentrations. L-cystine is prepared in 7.4% (vol/vol) concentrated HCl (trace metal grade) and added to the medium to a final concentration of 0.83 mM. Casamino acids (Gibco) are prepared as a 20% solution (wt/vol) and treated with 5% (wt/vol) Chelex100 (Bio-Rad Laboratories, Inc.) at 4°C with constant stirring for at least 18 h. Chelex100-treated Casamino acids are filter sterilized and added to the medium to a final concentration of 0.5% (vol/vol). Following the addition of all components, the pH of the medium is adjusted to 7.0 using KOH (trace metal grade) and/or concentrated HCl (trace metal grade), as needed. The medium is filter sterilized and stored at room temperature until use. Prior to use, mPGT medium is supplemented with 1 µM FeCl₃ prepared as a 10 mM stock in 10 mM HCl (trace metal grade). Antibiotics were used at 25 µg/mL for kanamycin and 10 µg/mL for nalidixic acid.

TABLE 1.

Bacterial strains and plasmids

Strain	Genotype or description	Source
C. diphtheriae strains
1737	Wild type, Gravis biotype, tox+	(36)
Δiut	ΔiutABCD/E (dip0169-73)	This study
Δznu	ΔznuA-dip0439-znuCB (dip0438-41)	This study
Δnik1	ΔnikABCD1 (dip2128-25)	This study
Δnik2	ΔnikABCD2 (dip2162-65)	This study
Δ4	ΔiutABCD/E ΔznuA-dip0439-znuCB ΔnikABCD1 ΔnikABCD2	This study
ΔcztA	ΔcztA (dip1101)	(19)
Δ4 ΔcztA	ΔiutABCD/E ΔznuA-dip0439-znuCB ΔnikABCD1 ΔnikABCD2 ΔcztA	This study
Δmnt Δ4	ΔiutABCD/E ΔznuA-dip0439-znuCB ΔnikABCD1 ΔnikABCD2 ΔmntABCD (dip0615-18)	This study
Δmnt Δ4 ΔcztA	ΔiutABCD/E ΔznuA-dip0439-znuCB ΔnikABCD1 ΔnikABCD2 ΔmntABCD ΔcztA	This study
Δsid Δ4	ΔiutABCD/E ΔznuA-dip0439-znuCB ΔnikABCD1 ΔnikABCD2 ΔsidAB (dip2161-60)	This study
Δsid Δ4 ΔcztA	ΔiutABCD/E ΔznuA-dip0439-znuCB ΔnikABCD1 ΔnikABCD2 ΔsidAB ΔcztA	This study
Δ6	ΔiutABCD/E ΔznuA-dip0439-znuCB ΔnikABCD1 ΔnikABCD2 ΔmntABCD ΔsidAB	This study
iut+	ΔznuA-dip0439-znuCB ΔnikABCD1 ΔnikABCD2 ΔmntABCD ΔsidAB	This study
znu+	ΔiutABCD/E ΔnikABCD1 ΔnikABCD2 ΔmntABCD ΔsidAB	This study
nik1+	ΔiutABCD/E ΔznuA-dip0439-znuCB ΔnikABCD2 ΔmntABCD ΔsidAB	This study
nik2+	ΔiutABCD/E ΔznuA-dip0439-znuCB ΔnikABCD1 ΔmntABCD ΔsidAB	This study
Escherichia coli strains
NEB 5-alpha competent	Cloning strain	New England Biolabs, Inc.
S17-1 λpir	Mating strain	(37)
Plasmids
pKΔiut	Suicide vector for the deletion of the iutABCD/E (dip0169-0173) gene cluster	(21)
pKΔtro (znu)	Suicide vector for the deletion of the znuA-dip0439-znuCB (dip0438-41) operon	(21)
pKΔnik1	Suicide vector for the deletion of the nikABCD1 (dip2125-28) operon	This study
pKΔnik2	Suicide vector for the deletion of the nikABCD2 (dip2162-65) operon	This study
pKΔmntAD	Suicide vector for the deletion of the mntABCD (dip0615-18) operon	(21)
pKΔ1101 (cztA)	Suicide vector for the deletion of dip1101 (cztA)	(19)
pKN2.6z	C. diphtheriae shuttle vector; kanR	(38)
pKN-iutAD	pKN2.6z carrying the truncated mntA promoter region and iutABCDE genes	GenScript
pKN-znu	pKN2.6z carrying the truncated mntA promoter region and znuA-dip0439-znuCB genes	This study
pKN-nik1	pKN2.6z carrying the truncated mntA promoter region and nikABCD1 genes	This study
pKN-nik2	pKN2.6z carrying the truncated mntA promoter region and nikABCD2 genes	GenScript
pKN-mnt	pKN2.6z carrying the truncated mntA promoter region and mntABCD genes	This study

Cloning and C. diphtheriae mutant generation

All DNA sequences used in this study were derived from C. diphtheriae strain 1737 and amplified by PCR from genomic DNA or synthesized based on the published sequence of C. diphtheriae NCTC 13129 (39) (GenScript).

Deletions of targeted genes were introduced into C. diphtheriae by allelic exchange using Escherichia coli strain S17-1 λpir for conjugation [described in reference (40)] and the suicide plasmids noted in Table 1. Suicide plasmids were designed by incorporating the flanking genomic regions of each operon or cluster to leave <18 total amino acids from the first and last gene of the target into pK18mobsacB (41). The deletion of genes was verified by PCR across the gene locus using primers external to the deletion constructs.

Complementation plasmids in pKN2.6z were created using the pO5 promoter of the mntA promoter region (25); this region has strong promoter activity and is not regulated by MntR (Mn) due to the removal of the binding site. The pO5 promoter was fused to the start codons of cloned genes.

Growth assays

For terminal growth assays, freezer stocks were used to inoculate HIA plates and incubated at 37°C overnight. Resultant colonies were then used to inoculate 1 mL of mPGT in culture tubes, and cultures were incubated overnight (16–20 h) at 37°C with shaking. Overnight cultures were diluted with 1 mL of fresh mPGT and incubated for 4–6 h at 37°C with shaking. Culture density was measured by OD₆₀₀, and cultures were used to inoculate culture tubes containing 1 mL mPGT supplemented with 1 µM ZnCl₂ where indicated to a final OD₆₀₀ of 0.03. Cultures were incubated for 18–22 h, and OD₆₀₀ was measured using a 1:10 dilution of the cultures.

For kinetic growth assays, a similar procedure was used to prepare cultures from the HIA. However, after dilution of cultures into 1 mL mPGT without or with 1 µM ZnCl₂, the cell suspension was mixed by pipetting and 200 µL aliquots were distributed into a 96-well microtiter plate for technical replicates. Microtiter plates were sealed using a gas-permeable sealing film, and OD₆₀₀ was measured over 23 h in 5-min increments with shaking and temperature control at 37°C using a Varioskan LUX microplate reader. Baseline subtraction was performed for all samples to remove background absorbance introduced by the gas-permeable sealing film.

Zinc toxicity assays

For zinc toxicity assays, freezer stocks were used to inoculate HIA plates and incubated at 37°C overnight. Isolated colonies from the agar plates were used to inoculate 1 mL HIBTW, and cultures were incubated overnight (16–20 h) at 37°C with shaking. Culture density was measured by OD₆₀₀, and cultures were used to inoculate fresh HIBTW at an OD₆₀₀ of 0.06. The cell suspensions were further diluted 1:1 (100 µL with 100 µL medium) supplemented with ZnCl₂ at twice the final indicated concentrations to reach a final initial OD₆₀₀ of 0.03 and indicated ZnCl₂ concentration in a 96 deep-well plate with 2 mL capacity (sample volume of 200 µL). The plates were sealed using gas-permeable sealing film and incubated at 37°C with shaking; after 16–18-h incubation, OD₆₀₀ was measured.

Inductively coupled plasma mass spectrometry

For ICP-MS, freezer stocks were used to inoculate HIA plates and incubated at 37°C overnight. Resultant colonies were then used to inoculate 1 mL of mPGT in culture tubes, and cultures were incubated overnight (16–20 h) at 37°C with shaking. Overnight cultures were diluted with 1 mL of fresh mPGT and incubated for 4–6 h at 37°C with shaking. Culture density was measured by OD₆₀₀, and cultures were used to inoculate 30 mL of mPGT medium without or with 1 µM ZnCl₂ in 125 mL vented flasks at an OD₆₀₀ of 0.03. Flasks were incubated for 18–20 h at 37°C with shaking, and cells were harvested by centrifugation. Cell pellets were washed three times using double-distilled water, aspirated to remove excess liquid, weighed, and stored at −80°C. Cell pellets were boiled for 20 min prior to sending for elemental analysis.

Elemental analysis was performed at the Oregon Health & Science University Elemental Analysis Core. Briefly, cell pellets were digested by adding concentrated HNO₃ (trace metal grade, Thermo Fisher Scientific) and heating at 90°C for 1 h. After digestion, 1% HNO₃ was added to each sample and diluted as needed for analysis. ICP-MS was performed using an Agilent 8900 triple quadrupole equipped with an SPS4 autosampler. For each sample, data were acquired in triplicate and averaged. Data are normalized by wet cell mass.

Statistical analysis

GraphPad Prism version 10.1.2 was used for the analysis of data. Specific tests and P values are noted in the figure legends.