In Vivo Development of Aztreonam Resistance in Meropenem-Resistant Pseudomonas aeruginosa Owing to Overexpression of the blaPDC-16

ABSTRACT The rapid acquisition of antibiotic resistance of Pseudomonas aeruginosa has been a complex problem in clinics. Two meropenem-resistant P. aeruginosa isolates were collected from the same patient on May 24, 2021, and June 4, 2021, respectively. The first was susceptible to aztreonam, while the second displayed resistance. This study aimed to identify the genetic differences between two P. aeruginosa isolates and uncover alterations formed by the within-host bacterial evolution leading to aztreonam resistance during therapy. Strains were subjected to antimicrobial susceptibility testing using the broth microdilution method. Genomic DNAs were obtained to identify their genetic differences. The relative mRNA levels of β-lactam-resistance genes were determined by real-time PCR. Both isolates belonged to ST 773 high-risk clones with the same antibiotic resistance genes, eliminating the possibility of horizontally obtaining resistance genes. Reverse transcription (RT)-PCR results showed that the blaPDC-16 mRNA level in the second one was about 1,500 times higher than that in the first one. When 3-aminophenyl boronic acid was added, the second strain recovered its susceptibility to aztreonam, which confirmed that the overexpression of blaPDC-16 was the main reason for the isolate’s resistance to aztreonam. Compared to the first strain, the second showed a single amino acid substitution in AmpR located upstream of blaPDC-16, which may contribute to the upregulation of blaPDC-16 and lead to aztreonam resistance. AmpR plays an essential role in regulating antibiotic resistance in P. aeruginosa, and there is a need to be alert to clinical treatment failures associated with mutations in ampR. IMPORTANCE Pseudomonas aeruginosa is notorious for being highly resistant to antimicrobial agents. In this study, two P. aeruginosa strains isolated from the same patient with different susceptibility to aztreonam were used to illustrate the within-host resistance evolution process of P. aeruginosa. Both isolates, which belonged to a ST773 high-risk clone, had the same β-lactam resistance genes (blaPDC-16, blaIMP-45, blaOXA-1, and blaOXA-395), which means the second isolate might have been derived from the first isolate by gaining aztreonam resistance via mutations associated with aztreonam resistance relative genes. Subsequently, we found that mutation in ampR may be the cause of aztreonam resistance in the second isolate. Mutation in ampR leads to its loss of control over blaPDC-16, allowing overexpression of blaPDC-16 and further resistance to aztreonam. This study revealed that ampR plays an essential role in regulating antibiotic resistance in P. aeruginosa. There is a need to be alert to clinical treatment failures associated with mutations in ampR.

of the protective skin barrier and numerous injury-induced immune alterations that impair the ability to control the spread of disease (3,4). According to China Antimicrobial Surveillance Network (CHINET) data in China 2021 (http://www.chinets.com/Data/AntibioticDrugFast), CRPA isolated from hospitals has reached 24.7%, making treatment challenging. Moreover, the growing prevalence of nosocomial infections caused by CRPA is associated with increased morbidity and mortality (5,6).
The hyperproduction of the intrinsic inducible cephalosporinase AmpC is the primary mechanism used by P. aeruginosa to cope with b-lactams (7,8). Meanwhile, AmpC hyperproduction is intimately linked to the peptidoglycan recycling pathway, regulated by an AmpG-AmpD-NagZ-AmpR regulatory mechanism in Gram-negative bacteria (9). AmpR, a LysR-type transcriptional regulator usually located upstream of the b-lactamase gene, was classified as regulating the expression of the chromosomally encoded b-lactamase gene via an AmpG-AmpD-NagZ-AmpR regulatory mechanism (10). The emergence of a D135N AmpR mutant resistant to ceftolozane-tazobactam in vitro was reported during evolution experiments performed with a hypermutable PAO1 strain exposed to ceftazidime (11,12). A similar mutant expressing fully derepressed expression levels of the Pseudomonas-derived cephalosporinase (PDC) b-lactamase was selected under ceftazidime treatment from a cystic fibrosis strain in a murine model of chronic lung infection (11,13). In addition, several extensively drug-resistant strains from various geographical origins have also been detected to carry a D135N mutation, emphasizing the importance of AmpR in b-lactam resistance acquisition (14). However, the exact evolution processes of a susceptible strain to a resistant clone are unknown. This study aimed to illustrate the D135G substitution of AmpR, leading to the switch from susceptibility to the resistance of P. aeruginosa harboring bla IMP-45 to aztreonam through a clinical case, highlighting the importance of the global transcriptional regulator ampR.
Expression of the b-lactam resistance genes and mexAB-OprM efflux pump genes of P. aeruginosa HS90 and P. aeruginosa HS110. The b-lactam resistance genes (bla PDC-16 , bla IMP-45 , bla OXA-1 , bla OXA-395 ) and efflux pump genes (including MexA, MexB, MexR, and OprM) expression of the two isolates was compared at the mRNA levels. Interestingly, the bla PDC-16 mRNA level in P. aeruginosa HS110 was about 1,500 times higher than in P. aeruginosa HS90. Meanwhile, the other b-lactam resistance genes (bla IMP-45 , bla OXA-1 , and bla OXA-395 ) mRNA levels in P. aeruginosa HS110 were not obviously different than that in P. aeruginosa HS90. In addition, the mRNA levels of the efflux pump genes (including MexA, MexB, MexR, and OprM) were not very different between P. aeruginosa HS90 and P. aeruginosa HS110 (Fig. 1).
When the nonspecific efflux pump inhibitor carbonyl cyanide m-chlorophenylhydrazone (CCCP) and the AcrAB-TolC pump inhibitor phenylalanine-arginine b-naphthylamide (PAbN) were added, the MIC of aztreonam in the two strains had a 2-fold change. When adding CCCP or PAbN, the aztreonam MIC of P. aeruginosa HS90 decreased from 4 to 2 mg/liter (Table 1). When adding CCCP or PAbN, the aztreonam MIC of P. aeruginosa HS110 decreased from 64 to 32 mg/liter. Based on the above results, we speculate that efflux pumps do not play the most important role in aztreonam resistance.
Subsequently, we compared the aztreonam susceptibility using aztreonam disc and aztreonam disc plus 3-aminophenyl boronic acid (APB) (300 mg/disc). The results showed that P. aeruginosa HS110 had restored aztreonam susceptibility in vitro when adding APB. Meanwhile, the inhibitory zone diameter around aztreonam plus APB of P. aeruginosa HS90 was not obviously different from aztreonam alone (Fig. 2).

DISCUSSION
Infections caused by metallo-b-lactamase (MBL)-producing Enterobacterales and P. aeruginosa are increasingly reported worldwide and are usually associated with high mortality rates (.30%) (15). According to reports, most bla IMP-45 -positive P. aeruginosa currently occurs in China, with minor differences in antibiotics resistance profiles between strains, and are resistant to b-lactams (except for aztreonam), b-lactamb-lactamase inhibitor combinations (including ceftazidime-avibactam), aminoglycosides, and quinolones (16). Aztreonam, an old antibiotic strongly limited by the spread of extended-spectrum b-lactamase (ESBL) and AmpC during the last three decades, is gradually back in the public eye because of its inability to be hydrolyzed by MBL (17). Aztreonam is essential in treating infections caused by MBL-producing carbapenem-resistant Enterobacterales (CRE) and CRPA (15). Nevertheless, we compared two CRPA strains harboring bla IMP-45 isolated from a single patient but exhibiting different aztreonam susceptibility. Both isolates, which belonged to ST773 high-risk clone, had the same b-lactam resistance genes (bla PDC-16 , bla IMP-45 , bla OXA-1 , and bla OXA-395 ), which means P. aeruginosa HS110 might have been derived from the P. aeruginosa HS90 clinical isolate by gaining aztreonam resistance via mutations associated with aztreonam resistance.
According to the results of reverse transcription (RT)-PCR, the bla PDC-16 mRNA level in P. aeruginosa HS110 was about 1,500 times higher than that in P. aeruginosa HS90. In addition, HS110 restored aztreonam susceptibility in vitro when adding APB, inhibiting  AmpC activity. These findings suggested that overexpression of bla PDC-16 is the primary resistance mechanism to aztreonam in P. aeruginosa HS110. Chromosomal expression of PDC is tightly regulated by the peptidoglycan (PG) recycling pathway and is triggered by the presence of b-lactams (18). Under normal circumstances, the uridine 59-pyrophosphoryl-N-acetylmuramic acid-pentapeptide (UDP-NAM-P5), classified as AmpC repressor, binds to AmpR, leading to inhibiting ampC expression to basal levels. However, when bacteria are exposed to b-lactamase inducers (such as imipenem), large amounts of muropeptides (including N-acetylglucosamine-1,6-anhydro-N-acetylmuramyl-peptides and 1,6anhydro-N-acetylmuramyl-peptides) are generated and accumulate in the cytoplasm by increasing NagZ and AmpD activities, leading to AmpR-mediated induction of ampC expression by substituting UDP-NAM-P5 (19,20). Mutations in ampD and inactivation of PBP4 have also been reported in P. aeruginosa and Enterobacterales to induce overproduction of AmpC, leading to pathogen resistance to b-lactams (21). A previous study in 2017 suggested that a substitution causes the resistance of CFE-1 to cephalosporins in AmpR.
The aspartic acid at position 135 is modified to alanine to allow the constitutive highlevel expression (derepression) of CFE-1 (22). However, we did not detect the mutations in AmpG, AmpD, AmpC, and penicillin-binding protein genes, which meant D135G substitution of AmpR may be the main reason for P. aeruginosa HS110 resistance to aztreonam. Structural changes in AmpR may affect the efficiency of UDP-NAM-P5 to bind to it, thereby deinhibiting AmpC, which needs further research to confirm. The choice of antimicrobial agents for treating infections caused by MBL-producing P. aeruginosa is limited because MBL-producing P. aeruginosa often plays resistance to a variety of antimicrobial agents (including carbapenems and ceftazidime-avibactam). It is gratifying that several current antimicrobial agents showed good in vitro activity against MBL-producing P. aeruginosa, such as cefiderocol and cefepime-zidebactam (23,24). In this study, the susceptibility of cefiderocol and cefepime-zidebactam were not affected by the expression of PDC-16. These new antimicrobial agents may be potential options for treating MBL-producing P. aeruginosa.
Conclusions. P. aeruginosa may develop resistance during prolonged therapy with all antimicrobial agents. This study found that the D135G substitution of AmpR resulted in the loss of inhibition and overexpression of its downstream PDC-16, ultimately making the bacteria resistant to aztreonam. This finding re-emphasizes the importance of timely detection of antimicrobial susceptibility of P. aeruginosa during its infections. In addition, since the emergence of MBL-producing P. aeruginosa, the appropriate therapy to treat its infections is unclear, which is a cause for public concern.

MATERIALS AND METHODS
Clinical strains. P. aeruginosa HS90 and P. aeruginosa HS110, isolated from a male outpatient who suffered from an arm infection after an electrical injury, were collected at Huashan hospital (Shanghai, China) on May 24, 2021, and June 4, 2021, respectively. The first isolate P. aeruginosa HS90 displayed resistance to meropenem, ceftazidime, cefepime, piperacillin-tazobactam, ceftazidime-avibactam, and amikacin but was susceptible to aztreonam, imipenem, cefepime-zidebactam, and polymyxin B. In addition, the fosfomycin MIC was 32 mg/liter. After obtaining antimicrobial susceptibility patterns of P. aeruginosa HS90, anti-infective therapy was administered using a regimen of aztreonam (2 g every 8 h) combined with fosfomycin (12 g every day). During that period, fosfomycin was discontinued because of the shortage of fosfomycin. Unfortunately, rapid mutation occurred after 10 days, making the subsequent isolate P. aeruginosa HS110 resistant to aztreonam. In antimicrobial susceptibility testing, P. aeruginosa ATCC 27853 was included for quality control assessment.
According to the manufacturer's instructions, the NG test Carba 5 assay (NG Biotech, France) was used to detect carbapenemases carried by clinic strains. Briefly, a 1-mL loopful of bacteria was mixed with five drops of Carba-5 extraction buffer. A total of 100 mL of the mixture was transferred into the Carba-5 cassette after vortexing, and the results were evaluated after incubation at 23 6 2°C for 15 min (26). 3-Aminophenyl boronic acid (APB) is known to be a potent inhibitor of class C b-lactamases and has been used successfully in detecting the production of plasmid-mediated class C b-lactamases in Enterobacterales (27)(28)(29). Aztreonam combined-disc tests alone and with 300 mg of APB were performed to compare inhibitory zone diameters.
Analysis of gene expression. The relative mRNA levels of b-lactams resistant genes (bla PDC-16 , bla IMP-45 , bla OXA-1 , and bla OXA-395 ) and efflux pump genes (including MexA, MexB, MexR, and OprM) were determined by real-time-qPCR, according to the referenced method, with some modifications (30). Total RNA was obtained with the TaKaRa MiniBEST Universal RNA extraction kit (TaKaRa, Dalian, China) according to the manufacturer's instructions. RT-PCR was performed using SYBR Premix Ex Taq (TaKaRa, Dalian, China) on an ABI ViiA 7 realtime PCR system (Thermo Fisher Scientific, USA). The reaction parameters were as follows: 50°C for 2 min and 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 34 s; finally 95°C for 15 s, 60°C for 1 min, and 95°C for 15 s. The rpsL gene was performed as the endogenous reference gene. All experiments were repeated in triplicate independently.
Efflux pump inhibitory assays. The CCCP and PAbN and inhibitory tests were used to detect the effect of efflux pumps on b-lactam resistance of P. aeruginosa HS110, especially aztreonam. Briefly, both CCCP (MedChemExpress, Shanghai, China) and PAbN (MedChemExpress, Shanghai, China) were prepared at 25 mg/liter. The MICs of b-lactams were performed using the broth microdilution method. Bacterial growth in CAMHB containing b-lactams with and without CCCP or PAbN was evaluated in parallel.
Data availability. The data used during the current study are available from the corresponding author upon reasonable request. The bacterial genome sequences have been uploaded to NCBI with the following BioProject ID: PRJNA780409. The study protocol was approved by the Institutional Review Board of Huashan Hospital, Fudan University (approval 2018-408).