A. baumannii ATCC 17978 was purchased from The American Type Culture Collection (Manassas, VA, United States). Escherichia coli DH5α was obtained from Invitrogen (Carlsbad, CA, United States). The tetracycline-resistant and sucrose-sensitive plasmid pMJG42, apramycin-resistant pMJG120, and gentamicin-resistant pMJG125 plasmids were kept in our laboratory at the University of Chicago (Chicago, IL, United States). The bacteria were cultured in lysogeny broth (LB) medium at 37°C. When required, the antibiotics added for selection were tetracycline (10 μg/mL), apramycin (50 μg/mL), and gentamicin (10 μg/mL).

Gene deletion and complementation plasmids were generated as previously described (Jacobs et al., 2014b; Gebhardt et al., 2015; Gebhardt and Shuman, 2017). Briefly, gene deletions were performed using allelic exchange plasmid pMJG42. The resulting plasmids (pMJG42-ΔgigAB, pMJG42-ΔptsP) were transformed into ATCC 17978 via electroporation to obtain ATCC 17978 ΔgigAB and ATCC 17978 ΔptsP mutants. After tetracycline selection and sucrose counterselection, the clones were subjected to colony PCR. Gene deletions were confirmed by sequencing. For complementation of the deleted gigA/gigB, the entire open reading frames of gigA/gigB were amplified by PCR, cloned into pMJG120 or pMJG125 to obtain pMJG120-gigAB or pMJG125-gigAB, and transformed into ATCC 17978 ΔgigAB via electroporation to obtain ATCC 17978 ΔgigAB pMJG120-gigAB or ATCC 17978 ΔgigAB pMJG125-gigAB.

ATCC 17978 ΔgigAB was transformed with pMJG42-gigA/gigB to generate ATCC 17978’ with in situ complementation of gigA and gigB. ATCC 17978 ΔptsP was transformed with pMJG42-gigAB to generate ATCC 17978 ΔptsPΔgigAB. All bacterial strains, plasmids, and primers in this study were summarized in Supplementary Tables 1–3.

Eight strains of ATCC 17978 ΔgigAB were randomly selected from different batches for whole-genome sequencing. Genomic DNA was prepared using the QIAamp DNA Mini Kit (Qiagen, Germany) and then subjected to whole genome sequencing (WGS) using the Illumina Hiseq2500 platform (Illumina, CA, United States) following the 2 × 100 bp protocol. The average sequencing throughput was 1 Gb. Raw fastq reads were trimmed by Trimmomatic for quality control (Bolger et al., 2014) and subsequently mapped against the reference genome of ATCC 17978-mff (Accession No. CP012004) with Bowtie2 (Langmead and Salzberg, 2012). Variant calling was performed using the bcftools call function with the default parameters (Danecek et al., 2021). We had submitted all of these data to NCBI BioProject database under the BioProject ID PRJNA738724.¹

After antibiotics selection and sucrose counterselection, 24 clones were randomly selected for colony PCR. Gene deletions were confirmed by sequencing. The gene deletion efficiency was calculated as (the number of the clones with successful deletion mutation)/24 × 100% The experiment was repeated three times, and data were expressed as the mean ± standard deviation (SD).

Overnight cultures of the indicated strains were back-diluted into fresh LB and grown for 2 h. After outgrowth, aliquots of the cultures were serially diluted. Then, a 10-μL aliquot was spotted onto LB agar plates with or without stressors as follows: HCl (medium adjusted to pH 5.5), ZnCl₂ (final concentration = 1.25 mmol/L). Colony forming units (CFU) were counted at 12 h after incubation at 37 or 50°C. Efficiency of plating (EOP) was calculated as (CFU recovered on stress medium)/(CFU recovered on plain medium at 37°C).

ATCC 17978, ATCC 17978 ΔgigAB, ATCC 17978 ΔptsP, and ATCC 17978 ΔptsP ΔgigAB were cultured in LB medium without antibiotics. ATCC 17978 ΔgigAB pMJG120 and ATCC 17978 ΔgigAB pMJG120-gigAB were cultured in LB medium containing 50 mg/L apramycin. ATCC 17978 ΔgigAB pMJG125 and ATCC 17978 ΔgigAB pMJG125-gigAB was cultured in LB medium containing 10 mg/L gentamicin, in the presence or absence of 1% (w/v) arabinose.

Each strain was grown overnight on the appropriate LB agar plate, and a single colony was picked and expanded in 2 mL LB broth overnight. A 1 μL aliquot was diluted at 1:1,000, and the dilution was added into triplicate wells of a 96-well plate at 200 μL/well. LB medium without bacteria was used as a blank. The OD₆₀₀ was determined every 15 min using a Biotek plate reader (Winooski, VT, United States). Growth curves were generated using GraphPad Prism 5 (San Diego, CA, United States). Each experiment was performed in triplicate and repeated three times. The mean was calculated for each experiment, and data were presented as the mean of three experiments.

Antibiotic sensitivity testing was performed as previously described (Gebhardt et al., 2015). The antibiotics used in this study are summarized in Table 1. Data were expressed as minimum inhibitory concentration (MIC).

Bone marrow-derived macrophages (BMDMs) were obtained from 8 to 12 week old female C57BL/6J (Jackson Laboratories) mice as previously described (Toda et al., 2021). Briefly, bone marrow cells were collected from the femur and tibia of mice and maintained in RPMI 1640 medium (Gibco, Thermo Fisher Scientific, Waltham, MA, United States) supplemented with 10 ng/mL mouse macrophage colony-stimulating factor (mMCSF; Gibco), 10% fetal bovine serum (Gibco), and 1% Pen/Strep (Gibco) at 37°C in a humidified atmosphere of 5% CO₂ for 7 days.

Mouse BMDMs were plated in a 96-well plate at a density of 50,000 cells/well and cultured overnight. Cells were infected with wild-type ATCC 17978, ATCC 17978 ΔgigAB, or ATCC 17978 ΔgigAB pMJG120-gigAB at 5 × 10⁵ CFU/mL. The plate was centrifuged at 2,170 rpm for 30 min at room temperature, followed by incubation at 37°C for 30 min. After replacing the medium with RPMI 1640 containing 100 mg/L gentamicin to kill extracellular bacteria, the infected cells were incubated for an additional 1 h (t = 0 h). Then, the infected cells were cultured in RPMI 1640 supplemented with 25 mg/L gentamicin (wild type and ΔgigAB strains) or 1.5 mg/L polymyxin (ΔgigAB pMJG120-gigAB strain). Cell lysates were collected at 0, 2, and 6 h post infection using phosphate buffered saline (PBS) containing 1% Triton-X100, serially diluted, and plated on LB agar plates. CFU were enumerated after 18 h of growth at 37°C. Each experiment was performed in triplicate and repeated three times. The mean CFU of surviving bacteria was calculated for each experiment, and data were presented as the mean of three experiments.

Infection of Galleria mellonella larvae (Knutson’s LiveBait, Brooklyn, MI) was performed as described previously (Jacobs et al., 2014a; Gebhardt and Shuman, 2017). Briefly, the bacteria were grown overnight in an orbital shaker (37°C, 200 rpm), and overnight cultures were resuspended in PBS to a final OD₆₀₀ of 1.0. G. mellonella larvae were randomly divided into three groups (n = 10/group). A total of 10 μL cultures (5 × 10⁶ CFU/mL) were inoculated into the last left proleg of each larva. After injection, larvae were incubated at 37°C. The number of dead larvae was recorded hourly. Each experiment was performed in triplicate and repeated three times. The mean larval survival was calculated for each experiment, and data were presented as the mean of three experiments.

Data were expressed as the mean ± SD. Statistical analysis was performed using GraphPad Prism 5. Differences among groups were compared using one-way ANOVA followed by Dunnett’s post-hoc test. Killing curves were plotted using the Kaplan-Meier method. A P-value of < 0.05 was considered statistically significant.

In our preliminary study, we noticed that the gigA/gigB deletion efficiency in wild-type 17978 was only 4.2%, suggesting that loss of gigA/gigB inhibits the growth of 17978 (Table 2). To further explore how the genetic background affects the efficiency of gigA/gigB deletion, we assessed ΔgigAB mutation efficiency in various 17978 genetic backgrounds. All gene deletions and complementation were confirmed by sequencing. The results of these analyses are shown in Table 2, and indicate that strains harboring either a ptsP deletion or in trans-complementation of gigA/gigB greatly increased the frequency of isolating the gigA/gigB double deletion mutation, suggesting that ptsP deletion and gigA/gigB complementation can compensate for the apparent growth defect caused by loss of gigA/gigB.

In the arabinose-inducible pMJG125 vector-based complementation of gigA/gigB background, we observed that the ΔgigAB colonies were smaller than wild-type colonies in the absence of arabinose; this phenotype was eliminated by the supplementation with 1% arabinose (Table 2 and Supplementary Figure 1). This finding further confirms that gigA/gigB are important for the growth of 17978 and that complementation of gigA/gigB with arabinose supplementation promotes the growth of ΔgigAB mutant to the wild-type level.

To determine if the loss of gigA/gigB required the generation of suppressing mutations, we performed whole genome sequencing on eight 17978 ΔgigAB clones isolated from different batches of gene knockout experiments. We found that, other than the gigA/gigB deletion, the genome of each of the sequenced ΔgigAB clones was 100% identical to the genome of ATCC 17978-mff reference strain, suggesting that deletion of the gigA/gigB genes does not require suppressing/compensatory mutations and that ATCC 17978 can survive without gigA/gigB. Thus, gigA/gigB are important for the growth but not required for the survival of ATCC 17978 under our routine laboratory culturing conditions.

To explore the roles of ptsP and gigA/gigB in the antibiotic resistance of ATCC 17978, we performed antibiotic susceptibility tests in the wild type and gene deletion strains. As shown in Table 1, although the MIC of apramycin, chloramphenicol, and kanamycin were decreased in at least two deletion mutation strains compared with those in wild-type ATCC 17978, the results did not reach statistical significance. These data suggest that, in contrast to our previous findings in the A. baumannii AB5075 strain background, ptsP and gigA/gigB are not required for antibiotic resistance of ATCC 17978.

To explore the involvement of ptsP in gigA/gigB-mediated growth of ATCC 17978, we performed growth curve analyses. As shown in Figure 1A, 17978 ΔgigAB exhibited remarkably suppressed growth compared with the wild-type, whereas 17978 ΔptsP exhibited a comparable growth rate to the wild-type, suggesting that gigA/gigB contribute to 17978 growth. Interestingly, 17978 ΔptsPΔgigAB showed comparable growth to the wild-type strain, indicating that loss of ptsP alleviates the growth defect associated with the loss of gigA/gigB. In addition, pMJG125-based complementation of gigA/gigB also restored the growth of 17978 ΔgigAB to the wild-type level in the presence of arabinose (Figure 1B).

We next sought to explore any additional roles of gigA/gigB and ptsP in stress resistance of ATCC 17978. Although the wild-type 17978 grown at 50°C showed a moderately reduced colony size phenotype when compared with those grown at 37°C, no significant loss of CFU was observed (EOP = 1; Figure 2A, left panel). However, loss of both gigA and gigB resulted in a dramatic reduction in CFU at 50°C (EOP = 10^–5; Figure 2A, right panel), suggesting that gigA and gigB contribute to high-temperature resistance of 17978. In the absence of arabinose, complementation of both gigA and gigB partially restored the growth of ΔgigAB mutant at 50°C (EOP = 10^–3; Figure 2B, left panel). Importantly, arabinose supplementation further restored the growth of ΔgigAB mutant with gigA/gigB complementation to the wild-type level at 50 °C (EOP = 1; Figure 2B, right panel), despite the small sizes of the colonies. This finding suggests that gigA/gigB contribute to high-temperature resistance of 17978 on solid media. As observed in the growth curves described above, 17978 ΔptsP ΔgigAB and 17978 ΔptsP exhibited comparable growth at 50°C (EOP = 1; Figure 2C), suggesting that loss of ptsP restores the growth of ΔgigAB mutant under high-temperature stress. Taken together, these results suggest that gigA/gigB mediate high-temperature resistance of 17978 on LB agar plates, whereas ptsP negatively regulates this response.

When we examined the ability of the ΔgigAB strain to survive acid stress (pH = 5.5), we did not observe significant differences in the growth between the wild-type and ΔgigAB mutant strains (EOP = 1; Figure 3A), suggesting that the 17978 strain is not sensitive to pH stress as measured herein, and that the loss of gigA and gigB does not confer an acid stress sensitivity on the 17978 strain.

When cultured on LB containing Zn²⁺, both wild-type and ΔgigAB mutant demonstrated significantly suppressed growth compared with those cultured on LB without Zn²⁺ (EOP = 10^–4, Figure 3B). No major difference was observed in the growth between the wild-type and ΔgigAB mutant. This finding suggests that factors other than gigA and gigB mediate zinc resistance of ATCC 17978.

Evading macrophage phagocytosis is critical for the survival of pathogens in vivo (Rosales and Uribe-Querol, 2017). To investigate the roles of gigA/gigB in macrophage killing evasion of ATCC 17978, we infected murine BMDMs with the wild-type, ΔgigAB mutant, and gigAB complementation strains and monitored their survival and replication. When BMDMs were infected with wild-type 17978, we observed a 10-fold reduction of intracellular live bacteria at 2 h after infection (Figure 4). On the other hand, when BMDMs were infected with ΔgigAB mutant, we observed a 300-fold reduction of live bacteria in BMDMs at 2 h after infection (Figure 4), suggesting a decreased replication ability of the ΔgigAB mutant. Importantly, the ΔgigAB pMJG120-gigAB complementation strain exhibited a similar trend of survival and replication and comparable CFU at different time points to the wild-type 17978 (Figure 4). These results suggest that gigA and gigB promote macrophage killing evasion of ATCC 17978.

To examine the roles of gigA/gigB in the virulence of ATCC 17978, we performed a G. mellonella killing assay. As shown in Figure 5, inoculation of G. mellonella larvae with wild-type 17978 resulted in a rapid killing of the larvae starting 8 h after inoculation. No killing was observed in the larvae that received ΔgigAB mutant within 48 h after inoculation. Complementation of both gigA and gigB restored the virulence of bacteria to nearly wild-type level. Thus, gigA/gigB are required for the virulence of ATCC 17978. Much like for the growth and temperature studies described above, both the 17978 ΔpstP and 17978 ΔptsP ΔgigAB strains killed larvae with similar kinetics as the wild-type 17978 strain, suggesting that the loss of ptsP restores the virulence defect caused by the ΔgigAB deletion.