Functional characterization of a novel class A carbapenemase CAE-1 in carbapenem-resistant Pseudomonas aeruginosa clinical isolates

ABSTRACT

Two carbapenem-resistant Pseudomonas aeruginosa (CRPA) isolates were collected, which lacked known carbapenemases but produced a β-lactamase CAE-1. CAE-1 was recently found in Comamonas aquatica and conferred resistance to penicillins and cephalosporins. In this study, we aim to know whether CAE-1 was responsible for the carbapenem resistance in the CRPA isolates and its enzyme hydrolyzing characteristics. Both CRPA clinical isolates exhibited minimal inhibitory concentrations (MICs) of 8 µg/mL for imipenem and meropenem, with a positive result in the modified Hodge test for meropenem. The bla_CAE-1 and bla_KPC-2 were cloned and expressed in P. aeruginosa PAO1 and Escherichia coli DH5α, respectively. The bla_CAE-1-carrying PAO1 transformant also tested positive using the modified carbapenem inactivation method and was resistant to piperacillin-tazobactam, ticarcillin-clavulanate, and cefoperazone-sulbactam, but susceptible to ceftazidime-avibactam. Expression of bla_CAE-1 in P. aeruginosa PAO1 elicited a 64- to 128-fold increase in MICs for piperacillin, ceftazidime, cefepime, and aztreonam, and an eightfold increase for meropenem, exhibiting a broader resistance profile than in E. coli DH5α. Steady-state kinetic assays showed that CAE-1 had catalytic efficiency against all β-lactams tested, with comparatively lower efficiency against three carbapenems relative to KPC-2, while demonstrating approximately equivalent efficiency for the other β-lactam antibiotics tested. Whole-genome sequencing revealed that bla_CAE-1 was a class A β-lactamase and encoded on an integrative and conjugative element, which might facilitate its horizontal transfer. The class A β-lactamase CAE-1 is a carbapenemase posing a high risk for horizontal dissemination. Enhanced surveillance for bla_CAE-1-harboring isolates is needed.

INTRODUCTION

Pseudomonas aeruginosa causes common, life-threatening infections and exhibits rising resistance to carbapenem antibiotics (1–3). Carbapenem-resistant P. aeruginosa (CRPA) has become a global threat due to a high mortality rate of 20%–30% (4). An important mechanism for the development of CRPA is the presence of carbapenemases, which can hydrolyze many β-lactams (5). The types of carbapenemases present in P. aeruginosa vary by region and include class A serine-β-lactamases (KPC and GES), class B metallo-β-lactamases (IMP, NDM, and AFM), and class D serine-β-lactamases (OXA-48) (6). Historically, carbapenemase detection rates in CRPA ranged from 10% to 20%, predominantly involving class B metallo-β-lactamases (7, 8). Our research group previously determined the enzymatic hydrolysis activity of the metalloenzyme AFM-1 from a clinical CRPA isolate (9). In recent years, the carbapenemase detection rate in CRPA has risen to 22% (7). Notably, KPC-2, a class A serine β-lactamase commonly found in carbapenem-resistant Klebsiella pneumoniae, has been transferred via plasmids into CRPA strains and now represents the most prevalent carbapenemase, with a detection rate of 49% (7, 10). The porin OprD is the principal channel protein through which imipenem enters P. aeruginosa; therefore, its functional inactivation can mediate imipenem resistance (11).

Recently, a novel class A serine β-lactamase, designated CAE-1, was identified on a plasmid from a strain of Comamonas aquatica isolated from sewage. Cloning assay in Escherichia coli demonstrated that CAE-1 confers resistance to ampicillin, piperacillin, cefazolin, cefuroxime, and ceftriaxone, but not to carbapenems. Searching the bla_CAE-1 sequence in the NCBI Nucleotide database revealed that bla_CAE-1 could be detected in P. aeruginosa, C. aquatica, Comamonas thiooxydans, and Brevundimonas sp. The latter three species were recognized as opportunistic human pathogens and predominantly found in environmental sources, such as sewage (12–14).

In this study, we identified two clinical CRPA isolates, PA56381 and PA56391, from two hospitalized patients. Whole-genome sequencing (WGS) did not detect any known carbapenemase genes; however, both isolates were found to carry bla_CAE-1. Both isolates tested positive in the modified Hodge test. Cloning of bla_CAE-1 from the clinical CRPA isolate into P. aeruginosa PAO1 resulted in resistance to piperacillin, ceftazidime, cefepime, and aztreonam, with minimal inhibitory concentrations (MICs) ranging from 64 to 256 µg/mL, and reduced susceptibility (intermediate) to meropenem (MIC = 4 µg/mL). Kinetic studies showed CAE-1 had catalytic efficiency against all β-lactams tested, including carbapenems. The above results suggest that CAE-1 is a carbapenemase and it contributed to reduced carbapenem susceptibility in combination with porin loss in P. aeruginosa clinical isolates. Comparative genomic analysis suggested that bla_CAE-1 could be horizontally transferred through integrative and conjugative elements (ICEs) or plasmids and is already disseminated among clinical P. aeruginosa isolates.

MATERIALS AND METHODS

Bacterial strains and antimicrobial susceptibility testing

Clinical isolates of CRPA, PA56381 and PA56391, were recovered from a tertiary teaching hospital in Shanghai in 2019. Species identification was performed using MALDI-TOF MS. The MICs of 15 antibiotics were determined by the broth microdilution method, and the results were interpreted following the CLSI guidelines (15). P. aeruginosa ATCC 27853 was used as the quality control strain. Phenotypic detection of carbapenemase activity was conducted using a modified Hodge test (16). To verify carbapenemase production, the modified carbapenem inactivation method (mCIM) was performed according to CLSI guidelines (15). The test was conducted on the P. aeruginosa PAO1 transformant harboring bla_CAE-1, utilizing the empty vector transformant P. aeruginosa PAO1 (pUCP24) and K. pneumoniae ATCC BAA-1705 as negative and positive controls, respectively. Porin phenotype experiments of two clinical strains were analyzed against the complete OprD porin of PAO1 reference, which served as the wild-type control (17).

Cloning of bla_CAE-1

In order to express bla_CAE-1 in P. aeruginosa PAO1 and E. coli DH5α, the full-length 909 bp bla_CAE-1 and its upstream 142 bp promoter region were co-amplified from the chromosome by PCR (primers shown in Table S1). The PCR products were digested with HindIII and BamHI and then ligated into the E. coli–P. aeruginosa shuttle vector pUCP24. The resulting recombinant plasmid, pUCP24-bla_CAE-1, was initially transformed into E. coli DH5α. Transformants were selected on Luria-Bertani agar plates containing 50 µg/mL gentamicin and confirmed by PCR and Sanger sequencing. Subsequently, pUCP24-bla_CAE-1 was electroporated into P. aeruginosa PAO1 and selected on LBA plates with 50 µg/mL gentamicin. To compare the resistance levels of bla_CAE-1 to carbapenems and other antibiotics, bla_KPC-2, a common class A carbapenemase in P. aeruginosa and Enterobacterales, was selected as a reference (18). The bla_KPC-2 with the 142 bp upstream promoter was cloned into plasmid pUCP24, and the resulting recombinant plasmid pUCP24-bla_KPC-2 was transferred into P. aeruginosa PAO1 and E. coli DH5α (Table S1). These transformants were then used for antimicrobial susceptibility testing. E. coli ATCC 25922 and P. aeruginosa ATCC 27853 were used as quality control strains. To compare the expression levels of bla_CAE-1, total RNA was extracted from logarithmic-phase cultures of E. coli DH5α and P. aeruginosa PAO1 transformants. Quantitative real-time PCR (RT-qPCR) was performed to determine the transcriptional levels. The housekeeping genes mdh and rpsL were used as internal controls for E. coli and P. aeruginosa, respectively. Relative expression levels were analyzed to assess transcriptional differences between the species.

Kinetic measurements of CAE-1

The bla_CAE-1 and bla_KPC-2 genes were cloned into the plasmid pET28a (+) vector, and the resulting plasmids were transformed into E. coli BL21 (DE3) (Table S1). CAE-1 and KPC-2 proteins were purified using His60 Ni Superflow Resin (Takara, Japan), with the protein purity (>90% homogeneity) confirmed by SDS-PAGE. Steady-state kinetic parameters kcat and Km were measured at 25°C with a UV-2700 spectrophotometer (Shimadzu, Kyoto, Japan) in 0.5 mL of 50 mM sodium phosphate buffer containing 50 μM ZnSO₄ at pH 7.4 (9). The data are presented as the mean ± standard deviation based on three independent measurements.

Bioinformatic analysis

Short-read sequencing of isolates PA56381 and PA56391 was conducted at the Beijing Genomics Institute (Illumina HiSeq X platform). Additionally, isolate PA56381 was subject to long-read sequencing at Shanghai Personal Biotechnology Co. Ltd. (Oxford Nanopore, Oxford, U.K.). Genomes were assembled using Unicycler v0.5.0 (19). We determined the cgMLST of 13 CAE-1 producers (11 from public database) with the homemade pipeline pyMLST (https://github.com/bvalot/pyMLST) using a core genome of 3,831 previously defined genes (20). Core genome alignment of 19 ST360 isolates was generated using snippy v4.6.0 (https://github.com/tseemann/snippy) and Gubbins v3.3.0 (https://github.com/nickjcroucher/gubbins), and a maximum likelihood phylogenetic tree was constructed using RAxML v8.2.4 (21).

RESULTS AND DISCUSSION

Properties of P. aeruginosa PA56381 and PA56391 and carbapenemase detection

Two CRPA clinical isolates (PA56381 and PA56391) were consecutively recovered from sputum specimens of two hospitalized patients in a medical ward at a tertiary care hospital in Shanghai, China. PA56381 was isolated on 15 March 2019, followed by PA56391 on 20 March 2019. Both patients had a history of invasive mechanical ventilation. Following invasive mechanical ventilation, both patients received empirical antibiotic therapy with either meropenem or piperacillin-tazobactam. The antibiotic regimen was subsequently shifted to a combination of a cephalosporin and amikacin upon the availability of the initial antimicrobial susceptibility testing report for CRPA (Fig. S1). The clinical isolate P. aeruginosa PA56381 exhibited low-level resistance to carbapenems with meropenem and imipenem MIC 8 µg/mL and to most β-lactams. Ceftazidime showed intermediate resistance, but the isolate was susceptible to ciprofloxacin, polymyxin B, amikacin, and ceftazidime-avibactam, with ticarcillin-clavulanate being resistant (Table 1). The modified Hodge test was positive for meropenem and ertapenem but negative for imipenem, suggesting the potential presence of a carbapenemase. Notably, the bla_CAE-1-carrying PAO1 transformant tested positive in the CLSI-recommended mCIM assay, yielding an inhibition zone diameter of 12 mm (within the positive range of 6–15 mm), whereas the empty vector control tested negative (22 mm) and the positive control K. pneumoniae ATCC BAA-1705 functioned as expected (Fig. S2).

TABLE 1.

In vitro antimicrobial susceptibility for clinical isolates and recombinant strains producing CAE-1 and KPC-2

Antibiotic	MICs (mg/L)
	P. aeruginosa (PA56381)	P. aeruginosa (PA56391)	E. coli DH5α (pUCP24)	E. coli DH5α (pUCP24-blaCAE-1)	E. coli DH5α (pUCP24-blaKPC-2)	P. aeruginosa PAO1 (pUCP24)	P. aeruginosa PAO1 (pUCP24-blaCAE-1)	P. aeruginosa PAO1 (pUCP24-blaKPC-2)
Piperacillin	64	64	<4	128	256	<4	256	512
Piperacillin-tazobactam	16/4	16/4	<2/4	4/4	128/4	<2/4	128/4	>256/4
Ticarcillin-clavulanate	256/2	>512/2	<4/2	<4/2	256/2	8/2	>512/2	>512/2
Ceftriaxone	NAa	NA	<0.25	16	8	NA	NA	NA
Ceftazidime	16	16	<0.125	0.25	4	1	64	64
Ceftazidime-avibactam	1/4	1/4	0.06/4	0.125/4	0.06/4	1/4	1/4	2/4
Cefepime	16	16	<0.06	<0.06	2	1	128	>128
Cefoperazone-sulbactam	16	16	<0.25	2	8	2	128	256
Aztreonam	16	16	<0.25	<0.25	32	2	128	512
Imipenem	8	8	<0.06	0.25	4	1	1	64
Meropenem	8	8	<0.016	0.03	1	0.5	4	>32
Ertapenem	32	32	<0.03	<0.03	1	4	8	>64
Levofloxacin	0.25	0.5	<0.125	<0.125	<0.125	0.25	0.25	0.25
Amikacin	4	4	<1	<1	<1	2	2	2
Polymyxin B	1	1	0.25	0.5	0.5	1	1	1

^a NA, not applicable.

Divergent resistance phenotypes conferred by CAE-1 in P. aeruginosa and E. coli

WGS revealed that the two CRPA isolates carried a class A β-lactamase CAE-1. To examine the functionality of bla_CAE-1, the gene was cloned and transformed into P. aeruginosa PAO1 and E. coli DH5α. The P. aeruginosa PAO1 (pUCP24-bla_CAE-1) strain conferred resistance or reduced susceptibility to all β-lactams tested, including piperacillin-tazobactam, ticarcillin-clavulanate, and cefoperazone-sulbactam, but was susceptible to ceftazidime-avibactam. Compared with P. aeruginosa PAO1 (pUCP24), MICs of cefepime, piperacillin, ceftazidime, and aztreonam had a 64- to 128-fold increase, while meropenem showed an eightfold increase. The P. aeruginosa PAO1 (pUCP24-bla_CAE-1) demonstrated resistance to all three β-lactamase inhibitor combinations (tazobactam, clavulanate, and sulbactam), while remaining susceptible to ceftazidime-avibactam, which indicates the above three β-lactamase inhibitors exhibit no inhibitory activity against CAE-1, whereas avibactam effectively inhibits this β-lactamase. The observed inhibition profiles provide additional evidence that CAE-1 is a new class A carbapenemase.

Cloning of bla_CAE-1 into P. aeruginosa PAO1 conferred only an intermediate MIC to meropenem, with no change in imipenem susceptibility. Both clinical isolates showed resistance to meropenem and imipenem. Genomic analysis of the clinical isolates revealed a frameshift mutation (Gly388fs) in the oprD gene compared to that of PAO1. The loss of OprD porin production was confirmed by SDS-PAGE. As OprD loss is a well-established mechanism associated with imipenem resistance in P. aeruginosa (22). Therefore, the imipenem resistance in these clinical isolates is primarily attributable to OprD inactivation.

The MICs for E. coli and P. aeruginosa expressing CAE-1 were quite different. Compared to the P. aeruginosa PAO1 (pUCP24-bla_CAE-1), the E. coli DH5α (pUCP24-bla_CAE-1) strain exhibited significantly lower MICs for all β-lactams, with MIC elevations observed only for ampicillin, piperacillin, cefazolin, cefuroxime, and ceftriaxone (Table 1). To investigate whether the differential resistance profiles between E. coli and P. aeruginosa transformants were due to variations in gene expression, RT-qPCR was performed (Table S2). Interestingly, bla_CAE-1 expression was lower in P. aeruginosa PAO1 compared with E. coli DH5α. This indicates that the higher carbapenem MICs observed in P. aeruginosa PAO1 are not attributable to overexpression of the carbapenemase but rather to species-specific expression variation (23). The observed resistance profile may result from the host genetic background, where CAE-1 functions synergistically with P. aeruginosa’s intrinsic low outer membrane permeability and active efflux systems to prevent antibiotic accumulation more effectively than in the E. coli host (24).

Enzyme kinetics

Kinetic parameters confirmed that CAE-1 hydrolyzed piperacillin, aztreonam, cephalosporins, and carbapenems (Table 2). CAE-1 demonstrated higher catalytic efficiency (kcat/Km) against piperacillin, aztreonam, and cefepime than against carbapenems and ceftazidime.

TABLE 2.

Kinetic parameters of CAE-1 and KPC-2

Antibiotic	CAE-1			KPC-2
	Km (μM)a	kcat (s−1)a	kcat/Km (μM−1s−1)	Km (μM)a	kcat (s−1)a	kcat/Km (μM−1s−1)
Meropenem	23.53 ± 3.74	0.35 ± 0.02	0.02	7.91 ± 1.78	4.90 ± 1.16	0.66
Imipenem	30.52 ± 9.34	0.23 ± 0.05	0.01	63.05 ± 7.39	80.51 ± 10.36	1.28
Ertapenem	24.27 ± 3.82	1.71 ± 0.67	0.07	22.07 ± 6.35	11.85 ± 2.00	0.56
Piperacillin	167.38 ± 3.56	95.29 ± 15.36	0.57	21.26 ± 4.61	13.51 ± 2.39	0.64
Ceftazidime	138.72 ± 6.63	3.56 ± 0.38	0.03	61.15 ± 21.75	0.69 ± 0.19	0.01
Cefepime	104.83 ± 10.08	20.07 ± 1.24	0.19	173.53 ± 18.23	17.35 ± 1.57	0.10
Aztreonam	42.07 ± 7.50	8.14 ± 0.73	0.20	92.13 ± 22.95	57.86 ± 16.69	0.62

^a kcat and Km values were calculated as the mean + SD of three independent measurements with three different enzyme purifications.

The substrate affinities (Km value) of CAE-1 to the three carbapenems were similar, but the substrate turnover rates (kcat value) varied, with ertapenem having the highest (1.71 s⁻¹) and imipenem the lowest (0.23 s⁻¹), leading to significant differences in catalytic efficiency (kcat/Km): ertapenem (0.07 μM⁻¹s⁻¹), meropenem (0.02 μM⁻¹s⁻¹), and imipenem (0.01 μM⁻¹s⁻¹). This indicates that CAE-1 exhibits high binding affinity for carbapenems but demonstrates limited hydrolytic activity, thereby failing to confer significant resistance (MIC elevation) in vivo. According to the catalytic efficiency, piperacillin was the optimal substrate for CAE-1, with a low substrate affinity (Km) of 167.38 μM and the highest turnover rates (kcat) of 95.29 s⁻¹. The catalytic efficiencies of aztreonam and cefepime were similar, at 0.20 μM⁻¹s⁻¹ and 0.19 μM⁻¹s⁻¹, respectively. The difference lay in that cefepime had poor affinity (higher Km) but high turnover rates, whereas aztreonam had the opposite characteristics. CAE-1 also hydrolyzed ceftazidime, but with a lower substrate affinity (138.72 μM), resulting in a low hydrolysis efficiency of 0.02 μM⁻¹s⁻¹.

The enzyme kinetic parameters of KPC-2 were simultaneously measured under the same experimental conditions as a reference. CAE-1 demonstrated lower catalytic efficiency against the three carbapenems compared to KPC-2, while displaying approximately comparable efficiency for the other β-lactams tested. This difference in enzymatic activity accounts for the distinct MICs of bla_CAE-1 and bla_KPC-2 in P. aeruginosa PAO1.

Dissemination of bla_CAE-1 among P. aeruginosa

WGS and multilocus sequence typing suggested that the two CRPA isolates carrying bla_CAE-1 belong to ST360. To investigate the evolutionary origins of the two ST360 isolates, we conducted phylogenetic reconstruction using a data set comprising 17 additional ST360 P. aeruginosa genomes obtained from the Pseudomonas aeruginosa genome database (25). A recombination-filtered core genome phylogeny revealed that PA56381 and PA56391 are closely related (differing by 0 core SNPs), suggesting nosocomial transmission, and form a clade with four Vietnamese isolates collected between 2012 and 2014, with pairwise core SNP distances within the clade ranging from 7 to 23 (Fig. 1A).

Fig 1.

Phylogeny of ST360 isolates and genetic characterization of CAE-1 positive P. aeruginosa isolates. (A) Phylogeny of 19 ST360 isolates (17 from public database). Two isolates from this study were marked by red dots on the tips. (B) Minimum spanning tree based on the cgMLST analysis of 13 CAE-1 positive P. aeruginosa isolates. The node colors correspond to the sequence types to which the strains belong. Allele differences were marked adjacent to the edges. (C) Comparison of the genetic context of the bla_CAE-1 gene in different P. aeruginosa strains (ST360, ST964, and ST463). Strain PAO1 was used as a reference to identify the integration site of the ICE carrying bla_CAE-1. Genes are represented by boxes and colored based on gene function classification. Boxes at the top of the sequence map depict genes on the forward strand, while those at the bottom depict genes on the reverse strand. Gray shadows indicate link regions that share at least 90% sequence identity.

A BLASTN search of bla_CAE-1 sequence against P. aeruginosa database identified eleven clinical P. aeruginosa isolates harboring bla_CAE-1. This cohort included ten ST463 isolates and one ST964 isolate. Subsequent cgMLST analysis revealed that the ten ST463 isolates, along with two ST360 isolates from this study, demonstrated evidence of clonal dissemination (≤4 allele difference) (Fig. 1B).

To gain insight into the mobile genetic elements contributing to the acquisition of bla_CAE-1, the genetic context of bla_CAE-1 in isolates PA56381 (ST360), PA99 (ST964), and ZY94 (ST463) was compared. Notably, bla_CAE-1 was located on an uncharacterized ICE approximately 120 kbp in size, which might result in the potential horizontal transfer of bla_CAE-1 gene among P. aeruginosa species (Fig. 1C).

Specifically, bla_CAE-1 is oriented in reverse, positioned immediately upstream of the gene encoding the LysR family transcriptional activator, ampR. This chromosomal bla_CAE-ampR configuration is structurally analogous to the ampC-ampR system observed in Gram-negative bacteria. However, despite this structural similarity, the expression of bla_CAE-1 and ampR was not significantly induced by cefoxitin challenge in both clinical isolate PA56381 and P. aeruginosa PAO1 transformant harboring bla_CAE-1 and ampR (data not shown). Additionally, two ISCaq2 elements consistently flank this bla_CAE-ampR configuration.

Inter-species dissemination of bla_CAE

A BLASTN homology analysis of the bla_CAE-1 against the NCBI GenBank database revealed five additional species beyond P. aeruginosa that also carried bla_CAE-1 or bla_CAE-like. The identified genetic elements were distributed across chromosomal locations in C. aquatica isolates BB1454 and NY8661 (LR813086.1 and CP096918.1), the plasmid p1_SCLZS63 (CP104280.1) from isolate SCLZS63, and the chromosome of C. thiooxydans isolate ZDHYF418 (CP063057.1). In addition, bla_CAE-like was found on plasmid pBH3a of Brevundimonas sp. BH3. The bla_CAE-like gene on pBH3a exhibited a frameshift mutation compared to bla_CAE-1, resulting from a deletion of an adenine at the 10th nucleotide position, located within the N-terminal signal peptide region. Comparative genomic analysis suggests that the ISCaq2 element downstream of the bla_CAE-ampR configuration may facilitate the inter-species transfer of bla_CAE (Fig. 2).

Fig 2.

Comparison of the genetic context of the bla_CAE gene across species. Genes are depicted as boxes (left panel) or arrows (right panel), with colors corresponding to functional classifications. The conserved IS-bla_CAE-ampR-IS unit is highlighted by a yellow dashed box in the left panel and further magnified in the right panel.

Similar to the well-characterized plasmid-encoded AmpC β-lactamase CMY-2 (26), bla_CAE-1 and its derivatives can be horizontally transferred via plasmids and ICEs. Copy number variation of bla_CAE-harboring plasmids/ICEs may amplify resistance levels, thereby exacerbating clinical threats. Notably, plasmids carrying bla_CAE-1 could also harbor metallo-beta-lactamase gene bla_AFM-1 simultaneously, such as p1_SCLZS63 (27), which might lead to pan-β-lactam resistance, leaving few therapeutic alternatives.

Conclusion

In conclusion, we identified and characterized the class A carbapenemase CAE-1, which contributed to carbapenem resistance in clinical P. aeruginosa isolates. Despite showing moderate hydrolytic efficiency against carbapenems compared to KPC-2, the carbapenemase activity is supported by a positive Hodge test in clinical isolates, a positive mCIM result in a PAO1 transformant, and its characteristic inhibitor profile, which shows resistance to tazobactam, clavulanate, and sulbactam, but strong inhibition by avibactam. The bla_CAE-1-mediated antibiotic resistance differed significantly between P. aeruginosa and E. coli, indicating species heterogeneity in the expression of CAE-1. CAE-1 has been identified on a conjugative plasmid in C. aquatica, which may promote rapid propagation of bla_CAE-1. Potential copy number variation of bla_CAE-harboring plasmids/ICEs might confer high-level β-lactam resistance, thereby exacerbating clinical threats.

DATA AVAILABILITY

The complete genome sequence of clinical isolate PA56381 and the draft genome sequence of PA56391 have been submitted to the DDBJ/EMBL/GenBank databases under BioProject accession number PRJNA1290494.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/aac.01362-25.

Supplemental material. aac.01362-25-s0001.pdf.

Fig. S1 and S2; Tables S1 and S2.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.