The Molecular Basis of Acinetobacter baumannii Cadmium Toxicity and Resistance

ABSTRACT

Acinetobacter species are ubiquitous Gram-negative bacteria that can be found in water, in soil, and as commensals of the human skin. The successful inhabitation of Acinetobacter species in diverse environments is primarily attributable to the expression of an arsenal of stress resistance determinants, which includes an extensive repertoire of metal ion efflux systems. Metal ion homeostasis in the hospital pathogen Acinetobacter baumannii contributes to pathogenesis; however, insights into its metal ion transporters for environmental persistence are lacking. Here, we studied the impact of cadmium stress on A. baumannii. Our functional genomics and independent mutant analyses revealed a primary role for CzcE, a member of the cation diffusion facilitator (CDF) superfamily, in resisting cadmium stress. We also show that the CzcCBA heavy metal efflux system contributes to cadmium efflux. Collectively, these systems provide A. baumannii with a comprehensive cadmium translocation pathway from the cytoplasm to the periplasm and subsequently the extracellular space. Furthermore, analysis of the A. baumannii metallome under cadmium stress showed zinc depletion, as well as copper enrichment, both of which are likely to influence cellular fitness. Overall, this work provides new knowledge on the role of a broad arsenal of membrane transporters in A. baumannii metal ion homeostasis.

IMPORTANCE Cadmium toxicity is a widespread problem, yet the interaction of this heavy metal with biological systems is poorly understood. Some microbes have evolved traits to proactively counteract cadmium toxicity, including Acinetobacter baumannii, which is notorious for persisting in harsh environments. Here, we show that A. baumannii utilizes a dedicated cadmium efflux protein in concert with a system that is primarily attuned to zinc efflux to efficiently overcome cadmium stress. The molecular characterization of A. baumannii under cadmium stress revealed how active cadmium efflux plays a key role in preventing the dysregulation of bacterial metal ion homeostasis, which appeared to be a primary means by which cadmium exerts toxicity upon the bacterium.

INTRODUCTION

The heavy metal cadmium is enriched in industrial settings and mining sites through anthropogenic activity. Although various metalloproteins retain their functions when bound to cadmium (1), cadmium is primarily associated with toxicity in biological systems (2). Despite cadmium being redox inactive, cadmium stress is commonly connected to the induction of oxidative stress in both humans and microbes (3). Recent advances in the field have revealed cadmium-metalloprotein interactions as a primary feature of cadmium toxicity (4, 5). In addition to dysregulation of central carbon metabolism and lipid homeostasis (6), cadmium stress in Streptococcus pneumoniae causes manganese and zinc starvation via interaction with noncognate metalloproteins, these being a manganese-recruiting protein and a zinc-responsive transcriptional regulator (5). Manganese is the primary cofactor for superoxide dismutase in S. pneumoniae; hence, cadmium-induced manganese depletion leads to oxidative stress susceptibility (5, 7). In Salmonella enterica, excess cadmium was found to affect zinc and manganese, but not iron, accumulation (4). Conversely, iron supplementation was shown to improve the growth of Bacillus subtilis when exposed to high levels of cadmium (8). These studies show that dysregulation of metal ion homeostasis and protein mismetallation by cadmium is central to its toxicity in cells.

Acinetobacter baumannii is rapidly emerging as one of the world’s most antimicrobial-resistant bacterial pathogens (9–11). Its highly effective intrinsic and acquired stress resistance features have allowed this human pathogen to survive in the hospital environment and thrive when colonizing immunocompromised individuals (12, 13). Various members of the genus Acinetobacter, including some strains of A. baumannii, have a strong environmental presence. This has resulted in the A. baumannii genome harboring diverse metabolic and stress resistance features that are typically not linked to pathogenicity (14, 15). Its environmental roots are also reflected in its extensive arsenal of putative heavy metal export mechanisms that may enable persistence under diverse metal ion stresses (16, 17). A. baumannii metal ion biology is a rapidly expanding field of research. In particular, its diverse array of metal ion acquisition systems has been studied intently (18–24). In contrast, our understanding of heavy metal resistance is limited, with only the roles of zinc and copper toxicity being studied in recent years (16, 17, 25–27). Metal ion toxicity analyses in A. baumannii have revealed widespread implications for cellular physiology, including dysregulation of metal ion homeostasis, oxidative stress susceptibility, and alterations in membrane biology. Similar to the impact of cadmium on the dysregulation of S. pneumoniae and S. enterica metal ion homeostasis, zinc stress in A. baumannii has been shown to result in copper depletion (16).

The A. baumannii genome harbors a broad arsenal of both highly conserved and recently acquired efflux systems that have putative functions in metal resistance (16). The P-type ATPase CopA is the primary mediator of cytoplasmic copper efflux in A. baumannii and plays a role in its virulence potential (26). The complexity of copper homeostasis in the periplasm has also been studied (27), but strain-to-strain variation complicates defining its significance across the species (16, 26). Zinc resistance is mediated by an efficient efflux pathway that consists of the cation diffusion facilitator (CDF) family member CzcD, a periplasmic zinc chaperone named CzcI, and the heavy metal efflux (HME) system CzcCBA (25).

Current models propose that bacterial resistance to cadmium toxicity is mediated via various mechanisms. Bacterial cell wall components such as polysaccharides, proteins, and lipids can prevent cadmium from entering the cell (28). Once cadmium passes the cell wall layer, intercellular metal sequestration and efflux pumps play a significant role in detoxifying the cell from cadmium toxicity. Metallothioneins are widely known metal chelators involved in copper and cadmium resistance, especially in eukaryotes (29), while their role in bacterial microorganisms has been less explored until recently (30). Bacterial cadmium efflux can be mediated by HME, CDF, or P-type ATPase family members. Examples include CadA, a P-type ATPase from Streptococcus thermophilus (31), the CDF member YiiP from Escherichia coli (32, 33), and the CznABC HME system from Helicobacter pylori (34). These systems may display overlapping substrate profiles within a bacterial species, which requires experimental verification. Although a recent study has provided insights into the role of the C-terminal domain in CDF metal specificity (35), overall, the specificity for transition metals varies greatly between members within the CDF, HME, and P-type ATPase families and is difficult to predict based on protein sequence analyses.

In this study, we defined the genome-wide cadmium stress adaptation mechanisms in A. baumannii. This work identified both CDF and HME systems that play key roles in cadmium export in A. baumannii, which were confirmed by single-gene mutant analyses. We also show that the bacterium encodes a highly efficient cadmium-sensing regulator for the transcriptional activation of the primary cadmium resistance mechanism. By using global metallomic analyses under cadmium stress, this study provides novel insights into cadmium toxicity.

RESULTS AND DISCUSSION

A. baumannii employs multiple transport systems to resist cadmium stress.

To delineate the molecular basis of cadmium toxicity and the mechanisms that aid in cadmium stress resistance, we created a Tn5-based random transposon mutant library in the laboratory reference strain A. baumannii ATCC 17978. Transposon-directed insertion site sequencing (TraDIS) sequencing of the base library in rich medium revealed a density of at least 113,000 unique mutants. We subjected 10⁹ mutant cells to a subinhibitory concentration (60 μM) of cadmium for 16 h. This concentration was found to present a minor but significant impact on proliferation of this strain (Fig. S1) and was thus deemed appropriate for TraDIS analysis. Cadmium exposure resulted in 67 genes displaying significantly differential numbers of insertions (fold change of ≥2; Q value, ≤0.05) (Table S3): decreased mutant fitness (decreased transposon insertions) in 23 genes and enhanced fitness (increased transposon insertions) in 44 genes.

The TraDIS analyses identified ACX60_13165 (CadR), which encodes a putative MerR-like transcriptional regulator, as most affected by cadmium treatment (−4.1-fold change) (Fig. 1). The regulatory gene is transcribed divergently from the CDF family transport gene czcE (ACX60_13160), which our analyses defined as the gene with the second greatest loss of represented transposon mutants under cadmium stress (−3.8-fold change). The previously characterized HME-RND zinc efflux system CzcCBA (ACX60_01405-15) and its metal chaperone, encoded by czcI (ACX60_01400) (25), were also shown to be important in conferring cadmium resistance, where mutant abundance decreased between −2.4- and −3.0-fold. The CDF member czcD, previously revealed to play a role in zinc efflux, was not found to be significantly impacted by cadmium treatment. We identified that transposon mutants of a putative DmeF-like cobalt-exporting CDF (ACX60_18165) were enriched under cadmium stress (2.1-fold change). Interestingly, independent mutant analyses revealed a marginal role for the A. baumannii dmeF member during cadmium stress but a primary role in cobalt resistance (Fig. S2).

FIG 1.

The A. baumannii cadmium resistome. The effect of cadmium treatment (60 μM for 16 h) on the abundance of gene mutations (Tn5 insertions) mapping to the A. baumannii ATCC 17978 chromosome and plasmid pAB3 was determined by TraDIS analysis. The data are means from biological triplicates (also presented in Table S3). The Q value was determined using a previously described method (53).

A solute-binding protein (SBP; ACX60_10710) with similarity to methionine and metal ion binding SBPs such as NlpA (COG1464.1) was impacted by cadmium stress (−3.2-fold change). The gene encoding this periplasmic protein is located in an operon that also encodes its putative ATP-binding cassette (ABC) transporter for cargo delivery and subsequent translocation across the inner membrane, but none of these genes were affected by cadmium stress. Hence, the SBP may play a role in periplasmic metal buffering instead of facilitating transport.

The relatively large number of genes in which insertional disruption proved to impart a positive effect included various transcriptional regulators. H-NS (12.6-fold change) is one of A. baumannii’s most prominent global transcriptional regulators (36). Although no role in metal ion stress has been indicated in previous transcriptomic analyses of an ATCC 17978 hns mutant, in Salmonella, the ferric uptake regulator (Fur) regulates the expression of hns, and H-NS competes with Fur for some gene targets in Salmonella (37, 38). The GacAS regulatory system, mutation of which is beneficial to survival during cadmium stress (11.7- and 12.1-fold changes, respectively), plays a key role during infection and in lifestyle changes, including biofilm formation (39). Other regulators include those with putative roles in sensing saturated fatty acids, including ACX60_05275 (3.3-fold change) and ACX60_17890 (2.0-fold change). This may indicate that A. baumannii adapts to cadmium stress by changing the structure and subsequently the biophysical properties of its membrane. Analysis of the total fatty acid profile of A. baumannii under cadmium stress revealed minor but significant changes in 16:1 fatty acids (1.5% decrease) and 18:0 fatty acids (0.9% increase) (Fig. S3). These changes may indicate a slight increase in the membrane rigidity following cadmium exposure (40).

A. baumannii employs a dedicated inner membrane cadmium exporter.

To confirm the roles of the putative cadmium transporters in cadmium stress resistance as observed by TraDIS (Fig. S4), we examined mutants with mutations in czcA (ABUW_0268), czcD (ABUW_0269), czcE (ABUW_2851), and czcF (ABUW_3664). A. baumannii AB5075_UW and its mutant derivatives were grown with increasing concentrations of cadmium (1, 5, 10, 20, or 30 μM CdCl₂) (Fig. 2A to E). We found that inactivation of czcE rendered the cells hypersusceptible to cadmium stress, with significant growth perturbation seen at concentrations as low as 1 μM cadmium, which is approximately 30-fold lower than the AB5075_UW parental strain. In addition to a primary role in zinc resistance (25), CzcCBA contributes to cadmium resistance, with the czcA::T26 mutant showing increased sensitivity at cadmium concentrations of 20 μM or greater. Consistent with the examination of these efflux systems in strain AB50875_UW, the ATCC 17978 ΔczcA and ΔczcE mutants displayed increased susceptibility to cadmium stress (Fig. S5). CzcD and CzcF were found not to play a major role in cadmium resistance. The concerted efforts from this study and previous work (25) have identified significant roles for CzcA, CzcD, and CzcE in zinc and/or cadmium resistance, but no major role has been identified for CzcF. Hence, we examined the response of CzcF to stress induced by other transition metals (zinc, copper, nickel, cobalt, and iron), which revealed a minor role in zinc and nickel resistance (Fig. S6).

FIG 2.

Cadmium resistance in A. baumannii. The effects of cadmium (1, 5, 10, 20, or 30 μM CdCl₂) on growth of (A) wild-type, (B) czcA::T26, (C) czcD::T26, (D) czcE::T26, and (E) czcF::T26 cells were compared to that of untreated (UT) cells. The growth was examined by measuring the OD₆₀₀. The data are the means from at least biological triplicates (with standard errors of the means [SEM]). The visibility of error bars may be occluded by the symbols.

The impact of cadmium on the A. baumannii metallome.

Bacterial cells tightly balance metal ion abundances, as hyperaccumulation or deficit beyond the cellular set points can have detrimental impacts on metalloprotein function and homeostasis of other metals. TraDIS analyses provided some insight into the impact of cadmium stress on cobalt homeostasis (Fig. 1 and Fig. S2). First, we examined cadmium accumulation in A. baumannii strain AB5075_UW and the czcA::T26 and czcE::T26 mutant strains. These analyses revealed that czcA inactivation did not impact accumulation following supplementation with 1 μM cadmium. In contrast, the czcE::T26 mutant accumulated cadmium at nearly 8-fold-higher levels than the wild type (Fig. 3A). Since growth of the czcE::T26 mutant was dramatically perturbed at cadmium concentrations greater than 1 μM, only the wild type and czcA::T26 mutant were examined in 15 μM cadmium. The czcA::T26 mutant accumulated cadmium at greater levels than those observed for the wild type (144 versus 86 μg cadmium·g cells⁻¹) (Fig. 3A).

FIG 3.

Impact of cadmium exposure on A. baumannii metal ion homeostasis. The accumulation of (A) cadmium, (B) iron, (C) zinc, and (D) copper in wild-type, czcA::T26, or czcE::T26 cells with or without cadmium supplementation (1 μM or 15 μM CdCl₂), as determined by inductively coupled plasma mass spectrometry. Metal ion levels were quantified as the weight of the metal (μg) per the weight of the dry cell pellet (g). For all panels, the data are the mean of at least biological triplicates (with SEM). Statistical analyses were performed using a two-tailed Student's t test; n.s., not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. The detection limit of zinc is indicated by the dotted line (b.d., below detection).

Examination of the impact of cadmium on the most abundant A. baumannii metal ions revealed that zinc and copper levels were affected, but not iron (Fig. 3B to D). Hence, an association between the enrichment of transposon mutants in the global regulator HNS and cadmium-induced changes in cellular iron could not be made. Therefore, the role of HNS during cadmium stress requires further study in A. baumannii. Zinc levels in the wild type were significantly lower upon exposure to 1 μM cadmium and below the limit of detection upon exposure to 15 μM cadmium. Consistent with these findings, zinc levels in the czcA::T26 and czcA::T26 mutants decreased to a greater extent upon exposure to 1 μM cadmium than those in the wild type (Fig. 3B). Interestingly, the wild type and czcA::T26 mutant hyperaccumulated copper when exposed to 15 μM cadmium (Fig. 3D). Despite the similar cellular cadmium levels observed in the czcE::T26 mutant following treatment with 1 μM cadmium and the wild type with 15 μM cadmium (Fig. 2A), the accumulation of copper was not affected in the czcE::T26 mutant (Fig. 2D).

Cadmium stress influences the A. baumannii metalloregulator responses.

Transcriptional analyses revealed that czcE is upregulated 480-fold by cadmium stress (Fig. 4A), while czcI was significantly downregulated by cadmium stress. Analysis of CadR (ABUW_2852), a MerR-type regulator divergently transcribed from czcE, suggested that it was responsible for activation of czcE, which is consistent with the TraDIS analysis (Fig. 4B). We found that czcF was significantly upregulated upon treatment with cadmium, despite it not playing a major role in cadmium resistance (Fig. 4C and Fig. S6). Similar to czcE, we found that czcF is regulated by a divergently transcribed MerR-type regulator, ABUW_3665 (Fig. 4C).

FIG 4.

Regulation of cadmium resistance. (A) The transcriptional responses of czcI, czcC, czcA, czcD, czcE, and czcF in wild-type AB5075_UW cells treated with 30 μM CdCl₂ were examined by qRT-PCR. The transcription levels of (B) czcE and (C) czcF were examined in regulator mutant and AB5075_UW wild-type cells by qRT-PCR following treatment with 30 μM CdCl₂. (D) The transcription levels of znuA, oprC, and copA in wild-type cells were examined by qRT-PCR following treatment with 30 μM CdCl₂ for 30 min. For all panels, the data are means (SEM) for at least biological triplicates. The visibility of error bars may be occluded by the symbols. Statistical analyses were performed using a two-tailed Student's t test. n.s., not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Our metallomic analyses determined that cadmium toxicity impacts zinc and copper homeostasis (Fig. 2). To ascertain the impact of cadmium stress on the expression of metal ion homeostasis contributors, we analyzed transcription of znuA, encoding a component of the major zinc acquisition system ZnuABC (18), oprC, encoding a putative copper acquisition system (41), and copA, encoding the primary copper efflux system (25). Under cadmium stress, A. baumannii increases expression of znuA (3.3 log₂-fold) (Fig. 4D), which may suggest a role in attempting to compensate for zinc starvation. Interestingly, cadmium-induced copper overaccumulation may be a result of the significant upregulation of oprC (8.2 log₂-fold) (Fig. 4D), although further experiments are required to conclusively define this mechanism.

Conclusions.

Cadmium resistance in A. baumannii was found to be primarily mediated by the CDF member CzcE. However, mutation of czcA, the critical inner membrane component of an HME system, resulted in A. baumannii growth perturbation in the presence of cadmium, which suggests that cytoplasmic cadmium is translocated into the periplasm by CzcE and then exported into the extracellular space by CzcCBA. The combination of a highly attuned cadmium sensing regulator (CadR) and dedicated inner-membrane cadmium transporter illustrates the significance of cadmium efflux in the organism. The seemingly opportunistic role of CzcCBA in cadmium resistance indicates that cadmium exerts greater toxicity in the cytoplasm than in the periplasm, which is consistent with previous reports (42, 43). However, the less significant role of CzcCBA than CzcE in cadmium resistance may also be due to the lack of its upregulation, rather than an inability to carry out export from either the cytoplasm or periplasm. To date, the transcriptional regulator(s) of czcCBA has not been identified.

Cadmium stress resulted in copper overaccumulation with transcriptional analyses revealing oprC as a possible copper import system gene. Although oprC has been shown to be regulated by OxyR in Acinetobacter oleivorans (44), recent analyses of OxyR in A. baumannii did not identify such an association (45). Hence, the physiological role of OprC warrants further investigation. Zinc homeostasis was also affected by cadmium stress, resulting in a reduction in cellular zinc. However, the molecular basis behind cadmium-induced zinc depletion in A. baumannii remains unclear. These results show that cadmium stress results in an inverse dysregulation of copper and zinc homeostasis, which is at least in part consistent with our previous analysis of these metals (16).

The impact of toxic heavy metals on cell membranes is well established (46), but how its modulation aids in resistance is poorly understood. Cadmium stress induced a change in the composition of the A. baumannii fatty acids via an overall reduction in double bonds and by extension membrane fluidity. Although regulation of fatty acids biosynthesis is vastly distinct in S. pneumoniae, a change in the biophysical properties of the membrane similar to that in A. baumannii under cadmium stress was observed (6). The unique means of membrane-mediated resistance to heavy metal stress was also evidenced in the analysis of Cupriavidus metallidurans and Pseudomonas putida, which identified a role for a fatty acid cis-trans isomerase and a cyclopropane synthase (47). These types of enzymes allow rapid biochemical change to facilitate a major biophysical shift but are lacking in most A. baumannii strains, including ATCC 17978, and were therefore not identified by TraDIS. However, international clone I isolates, which include AB5075_UW, harbor a genetic locus that encodes a cyclopropane synthase as well as a putative two-component regulatory system for heavy metal sensing. Here, we found that A. baumannii ATCC 17978 is likely to employ fatty acid desaturases to combat membrane-mediated cadmium stress, but the link between heavy metal resistance and membrane hemostasis may be stronger in some clonal lineages, such as those of the international clone I group.

In addition to a link with lipid homeostasis, heavy metals are known to influence antibiotic resistance, including in Acinetobacter species (48, 49), but this is not well understood from a genetic perspective. Genomic analyses did not identify any putative antibiotic resistance markers in close proximity to the primary cadmium efflux system, czcE, or the cadmium/zinc resistance locus (czcICBAD). However, czcF was found to be part of a mobile genetic element that also harbors a putative arsenate resistance mechanism (ABUW_3667 to ABUW_3671), which supports the coexistence of distinct antimicrobial resistance mechanisms.

In summary, this study has identified the previously unknown cadmium efflux pathways of A. baumannii. These systems are highly conserved across this species and genus; hence, this work is broadly applicable to Acinetobacter biology, exemplified by our parallel analysis of two distinct strains. We have garnered new knowledge of environmental stress adaptation strategies, with systems that have unique and shared metal ion specificities. The analyses of cadmium stress across a range of distinct metal ion export systems also provided novel insights into the alignment of metal sensors and efflux systems, which is of major significance to bacterial fitness in a broad range of environments.

MATERIALS AND METHODS

Bacterial strains, chemicals, media, and growth conditions.

Acinetobacter baumannii AB5075_UW mutant derivatives were purchased from the Manoil laboratory (50), and T26 transposon insertions were confirmed by PCR. The A. baumannii ATCC 17978 czcA and czcE mutants were generated previously in our laboratory (25) (see Tables S1 and S2 for strains and oligonucleotides used in the study, respectively). All chemicals were purchased from Sigma-Aldrich (Australia) unless otherwise indicated.

A. baumannii strains were routinely grown in Luria-Bertani broth (LB) containing 1% tryptone (BD Bacto), 0.5% yeast extract (BD Bacto), and 0.5% sodium chloride. For overnight culturing, a single colony from LB agar (1.5%) was used to inoculate 4 ml of LB medium. Overnight cultures were diluted to an optical density at 600 nm (OD₆₀₀) of 0.01 in 200 μl LB for growth assays or 20 ml LB for all other analyses. For growth assays, cultures in LB medium were incubated at 37°C with shaking in a FLUOstar Omega spectrophotometer (BMG Labtech), with the OD₆₀₀ values presented. The 20-ml cultures used for all other analyses were incubated at 37°C in an Innova 40R shaking incubator (Eppendorf) at 230 rpm until they reached an OD₆₀₀ of 0.7.

Construction of transposon mutant library.

A. baumannii ATCC 17978 was used to generate a dense transposon library. The library was constructed using the protocol previously described (51). Briefly, transposomes were prepared using an EZ-Tn5 transposase (Epicentre Biotechnology) and a custom transposon carrying a kanamycin resistance cassette amplified from the pUT_Km plasmid (52) using the primer sets listed in Table S2. The transposomes (0.25 μl) were electroporated into 60 μl of freshly prepared electrocompetent A. baumannii cells using a Bio-Rad GenePulser II (1.8 kV, 25 μF, and 200 Ω) in a 1-mm electrode gap cuvette (Bio-Rad). Cells were recovered by resuspension in 1 ml of SOC (super optimal broth with catabolite repression) medium and incubated at 37°C with shaking (200 rpm) for 2 h. Tn5 insertion recombinants were selected on Mueller-Hinton (MH) agar (BD Bacto) supplemented with 7 μg·ml⁻¹ kanamycin. Approximately 12 to 16 transformations were performed and pooled for each batch, with 10,000 to 50,000 resulting transformants. Approximately 250,000 mutants were collected and pooled from a total of 10 batches, snap-frozen, and stored in 20% glycerol at −80°C.

Time-kill assay for the selection of cadmium concentration for mutant library treatment.

Approximately 10⁹ CFU from an overnight ATCC 17978 culture was subcultured into 10 ml fresh MH broth spiked with different CdCl₂ concentrations (0, 3, 60, 80 160, and 320 μM) and incubated at 37°C with shaking. At 0, 1, 2, 4, 5 and 24 h, 100-μl samples were taken and 10-fold serially diluted in sterile phosphate-buffered saline (PBS), and 10 μl of each dilution was spotted on MH agar. Colonies were enumerated to determine the number of surviving cells after overnight incubation at 37°C.

TraDIS of mutant library and data analysis.

Approximately 10⁹ viable Tn5 mutant cells were inoculated into 10 ml of MH broth (BD Bacto) and grown at 37°C for 8 h with shaking (200 rpm). A total of 500 μl containing approximately 10⁹ cells was seeded into 10 ml fresh MH broth with or without 60 μM CdCl₂ in duplicate and grown for 16 h at 37°C with shaking (200 rpm). Genomic DNA was then extracted from approximately 10¹⁰ cells using the DNeasy UltraClean microbial kit (Qiagen) according to the manufacture’s protocol. Sequencing and analysis of the transposon mutant library were performed using the transposon-directed insertion site sequencing (TraDIS) protocol as described previously (53). The primer sets used for PCR amplification of TraDIS fragments (Pf5_PCR) and sequencing (Pf5_Seq) are listed in Table S2. Samples were sequenced on a HiSeq2500 Illumina sequencing platform, generating approximately 2 million 50-bp single-end reads per sample, as previously described (54). TraDIS sequence reads were analyzed using the BioTraDIS pipeline with default parameters (as described in reference 53).

Mutant genome analyses.

Genomic DNA of czcA, czcE, and cadR mutants was analyzed to confirm that no mutations other than the target gene were present, using previously published methods (40). Briefly, DNA was extracted on a QIASymphony SP system with a QIASymphony DSP virus/pathogen kit (Qiagen, Hilden, Germany). DNA concentrations were quantified using a Quant-IT double-stranded-DNA (dsDNA) high-sensitivity kit (Thermo Fisher Scientific, MA, USA). Sequencing libraries were generated using the Illumina Nextera XT library preparation kit (Illumina Inc., CA, USA) following the manufacturer’s instructions with minor modifications. Half volume was used for tagmentation reagents, amplification reagents, and input DNA. Library cleanup was performed using an AxyPrep MAG PCR cleanup kit (Corning Inc., NY, USA), and libraries were pooled manually and sequenced on a NextSeq 550 platform with a NextSeq 500/550 midoutput kit v2.5 (300 cycles) (Illumina Inc.). The Snippy pipeline version 4.3.6 was employed for single nucleotide polymorphism (SNP) identification (https://github.com/tseemann/snippy).

Cellular metal ion content analysis.

Untreated and metal-stressed bacteria (LB supplemented with 1 μM CdCl₂ or 15 μM CdCl₂) were harvested at mid-log phase (OD₆₀₀, 0.7) and washed, by resuspension and centrifugation at 7,000 × g for 8 min, first with PBS containing 5 mM EDTA and then with PBS as described previously (25, 26, 55, 56). Bacterial pellets were desiccated at 95°C overnight, followed by determination of dry cell weight, resuspension of pellets in 35% HNO₃, and boiling at 95°C for 1 h. Samples were diluted to a final concentration of 3.5% HNO₃ and analyzed by inductively coupled plasma mass spectrometry (ICP-MS) on an Agilent Solution 8900 QQQ ICP-MS (Adelaide Microscopy, University of Adelaide).

qRT-PCR.

For RNA extraction and qRT-PCR analysis, untreated or metal-stressed bacteria (LB supplemented with 30 μM CdCl₂) were harvested when they reached mid-log phase (OD₆₀₀, 0.7), and cells were lysed in QiaZol (Qiagen) as described previously (57, 58). RNA was extracted and purified using a PureLink RNA minikit (Thermo Fisher Scientific), with on-column DNase I treatment, according to the manufacturer's instructions. qRT-PCR was performed using the SuperScript III one-step RT-PCR kit (Thermo Fisher Scientific) on a QuantStudio 7 Flex System (Thermo Fisher Scientific). Transcription levels were corrected to those obtained for GAPDH prior to normalization to the transcription levels observed for untreated A. baumannii cultures. Oligonucleotide sequences are listed in Table S2.

Bacterial lipid extraction and gas chromatography-mass spectrometry analysis.

Overnight cultures of A. baumannii AB5075_UW were diluted to an OD₆₀₀ of 0.05 in fresh LB medium (20 ml) and grown to mid-log phase (OD₆₀₀, 0.7), with treated cultures supplemented with 15 μM CdCl₂, as per the metal accumulation analyses. Cells were harvested by centrifugation at 7,000 × g for 10 min and washed once with PBS; processed pellets were then resuspended in 50 μl of 1.5% NaCl buffer. For lipid extraction, 1 ml of chloroform-methanol (2:1 [vol/vol]) was added to the cell suspension, mixed vigorously for 2 min, and then incubated at room temperature for 10 min. Following the addition of 200 μl 1.5% NaCl, the cell suspension was mixed vigorously for 1 min and centrifuged at 6,000 × g for phase separation. The lower phase was recovered and concentrated via nitrogen evaporation. All samples were stored at −20°C prior to gas chromatography-mass spectrometry (GC-MS) analysis.

To generate fatty acid methyl esters (FAMEs), concentrated lipid samples were resuspended in 1:1 chloroform–trimethylsulfonium hydroxide (TMSOH) and subsequently analyzed using an Agilent 7890A GC system with a 30-m Agilent DB-FastFAME column (Agilent Technologies). Mass spectrometry was completed using a coupled Agilent 5975C MSD system (Agilent Technologies). FAME species were differentiated and determined by comparing to the FAME mix c4-24 standard (Sigma-Aldrich). Data analysis was completed using the Agilent MassHunter Qualitative Navigator software (Agilent Technologies).

Data availability.

TraDIS sequence reads were deposited in the European Nucleotide Archive under accession number ERP118051. The data are accessible via BioProject accession number PRJNA746061.