Reversion of Ceftazidime Resistance in Pseudomonas aeruginosa under Clinical Setting

Pseudomonas aeruginosa is an important nosocomial pathogen which frequently becomes resistant to most antibiotics used in chemotherapy, resulting in treatment failure among infected individuals. Although the evolutionary trajectory and molecular mechanisms for becoming β-lactam resistant have been well established for P. aeruginosa, the molecular basis of reversion from β-lactam resistant to susceptible is largely unexplored. In this study, we investigated the molecular mechanisms by which a ceftazidime-resistant clinical strain is converted to a ceftazidime-susceptible isolate under the clinical setting. RNA sequencing and genomic DNA reference mapping were conducted to compare the transcriptional profiles and chromosomal mutations between these two isolates. Our results demonstrate that a gain-of-function mutation in ampD, via deletion of a 53 bp duplicated nucleotide sequence, is the contributory factor for the conversion. Furthermore, we show for the first time that AmpD is involved in intraspecies competitiveness in P. aeruginosa. We also found that AmpD is not responsible for phenotypic changes between R1 and S2, including growth rate, motilities, pyocyanin, rhamnolipid, and biofilm production. This finding provides novel insights into the alteration of β-lactam sensitivity in P. aeruginosa under the clinical setting.


Introduction
Pseudomonas aeruginosa is a clinically important opportunistic human pathogen and is responsible for both acute and chronic infections in vulnerable individuals. It causes a variety of infections, including respiratory tract and urinary tract infections, commonly in patients suffering from cystic fibrosis or in individuals with a compromised immune system [1]. Treatment for infections by P. aeruginosa is challenging due to its intrinsic resistance as well as acquired resistance mechanisms to a wide range of antibiotics, which often result in therapeutic failure [1].
β-lactams are first-line antibiotics used to treat DTR (difficult-to-treat) Gram-negative bacteria [2]. Due to their high efficacy and low toxicity, β-lactams are most commonly recommended for the treatment of serious infections by P. aeruginosa [3]. However, the development of resistance to β-lactams limits their effectiveness. Several important resistance mechanisms have been reported including: (i) overproduction of the chromosomally encoded β-lactamase (AmpC) [4]; (ii) loss/reduced production of the outer membrane porin OprD [5,6]; (iii) expression of multidrug efflux pumps, such as MexAB-OprM, MexCD-OprJ, and MexXY [4,6]; and (iv) acquisition of the extended-spectrum β-lactamase via horizontal gene transfer [7]. Of note, these resistance mechanisms have been found to contribute to the antibiotic resistance of P. aeruginosa in the clinical setting. Importantly, via characterization of clinical isolates from a single patient, evolutionary trajectories and molecular mechanisms for antibiotic resistance have been well established in P. aeruginosa [5,[8][9][10]. However,

Basic Characterization of the Pseudomonas aeruginosa Isolates
Bacterial strains and plasmids used in this study are shown in Table S1. All bacteria were grown at 37 • C in L-broth medium (10 g/L tryptone, 5 g/L yeast extract, and 5 g/L NaCl) or on L-agar plates (L-broth supplemented with 1.5% agar). Ceftazidime-resistant (R1) and ceftazidime-sensitive (S2) P. aeruginosa strains used in this study were sequentially isolated from sputum samples of the same patient with sequelae of craniocerebral trauma 48 days apart. The 16S rDNA was amplified by PCR (primers listed in Table S2) and analyzed by sequencing to identify the species of these two strains [11]. Random amplified polymorphic DNA (RAPD) typing was conducted using primers 272, 208, and 241 (Table  S2) following an early description [12]. Multilocus sequence typing (MLST) was carried out as described before with minor modifications [13] to determine the allelic profiles of these two isolates. Briefly, chromosomal DNA was extracted from overnight cultures of bacterial cells with a genomic DNA purification kit (Tiangen Biotec, Beijing, China). Using the genomic DNA as a template, the internal fragments of the aroE, acsA, mutL, guaA, ppsA, nuoD, and trpE genes were PCR-amplified and sequenced with specific primers (Table  S2) [13]. The sequence type (ST) was obtained for each isolate by submitting the gene sequences and analysis on the website http://pubmlst.org/paeruginosa/ (accessed on 5 June 2021). The minimum inhibitory concentration (MIC) of antibiotics was examined using the two-fold serial dilution method according to a previous description [14]. Susceptibility was determined based on the Clinical and Laboratory Standards Institute guidelines (CLSI 2011-2021).

Plasmid Construction and Gene Editing
To replace the ampD gene of the R1 strain with the ampD gene of the S2 strain, a 1038 bp DNA fragment containing ampD and partial immediately adjacent genes (ampE and PA4523) was PCR amplified using genomic DNA of S2 as a template (primers in Table S2), digested with HindIII and EcoRI, and then inserted into a pEX18Tc suicide vector, resulting in pEX18-ampD S2 . Gene replacement in R1 was carried out by conjugal transfer of pEX18-ampD S2 followed by selection for single crossover and then double crossover, as described previously [15]. The resulting strain R1ampD S2 was confirmed by PCR amplification and sequencing analysis (primers in Table S2). Strain S2ampD R1 was generated with similar manipulations.

Genomic DNA Reference Mapping
Genomic DNAs of R1 and S2 were extracted from overnight cultures of bacterial cells using DNA purification kit (Tiangen Biotec, Beijing, China). Genomic DNA reference mapping was carried out by GENEWIZ Life Sciences (Suzhou, China) as described in our previous studies [5,16]. The raw sequence data have been deposited in the NCBI with an accessible number PRJNA863030 (www.ncbi.nlm.nih.gov/sra/?term=PRJNA863030).

RNA Isolation, Reverse Transcription, and Real-Time qPCR
Overnight cultures of bacterial cells were diluted 50-fold into fresh L-broth medium and grown to an OD 600 of 1.0. Total RNA was extracted with RNAprep Pure Cell/Bacteria Kit (Tiangen Biotec, Beijing, China), and the RNA concentrations were measured using a NanoDrop spectrophotometer (Thermo Scientific, Waltham, MA, USA). cDNAs were then synthesized with random primers and PrimeScript Reverse Transcriptase (Takara, Dalian, China). The cDNAs were mixed with qPCR primers (Table S2), SYBR premix Ex Taq II (Takara, Dalian, China), and ddH 2 O for real-time qPCR. qPCR was carried out in a CFX Connect real-time machine (Bio-Rad, Hercules, CA, USA). A 30S ribosomal protein encoding gene, rpsL, was utilized as an internal control.

RNA Sequencing and Data Analysis
Total RNA was purified from the R1 or S2 strain as described above. Library construction, sequencing, and analysis services were conducted by GENEWIZ Life Sciences (Suzhou, China) as described in our early studies [5,16]. The raw RNA sequence data have been deposited in the NCBI under an accessible number PRJNA863024.

Pyocyanin, Rhamnolipid, and Motility Assays
Pyocyanin production was examined as previously described with minor modifications [17]. In brief, overnight bacterial culture was diluted 50-fold in fresh medium and cultivated for 24 h. 1 mL supernatant from each bacterial culture was extracted with 0.5 mL of chloroform. A 0.4 mL sample of the lower organic phase was then re-extracted into 0.3 mL of HCl (0.2 N), and the absorbance was measured at 520 nm with a Varioskan Flash (Thermo Fisher Scientific, Vantaa, Finland). Assays for rhamnolipid production by P. aeruginosa were conducted using a M8-based agar medium plate as previously described [16].

Quantification of Biofilm
Biofilm formation was examined as described previously with minor modifications [5]. Overnight cultures of bacterial cells were sub-cultured at 37 • C to an OD 600 of 1.0 in L-broth. The bacteria were then diluted 1:50 into fresh L-broth, and 150 µL of the suspension was aliquoted into each well of a 96-well plate, followed by cultivation at 37 • C for 48 h. After the bacterial culture was discarded, each well was gently washed three times with ddH 2 O, stained with 0.25% crystal violet for 15 min, washed three times with ddH 2 O, and then dried at 65 • C. 200 µL of eluate [methanol (4): acetic acid (1): ddH 2 O (5)] was added to each well and incubated for 15 min at room temperature. The crystal violet solution was examined at a wavelength of 590 nm with the Varioskan Flash (Thermo Fisher Scientific, Vantaa, Finland).

LL-37 Susceptibility Assay
Overnight cultures of P. aeruginosa strains were diluted 1:50 and grown at 37 • C to an OD 600 of 1.0. The bacterial cells were washed twice with PBS and resuspended in PBS. Then, the bacteria were treated with LL-37 for 4 h at a final concentration of 256 µg/mL. Viable bacterial cell numbers were determined by plating on L-agar plates. Bacterial survival rate% = (live bacterial number with LL-37 treatment) × 100/(initial inoculum bacterial number).

Growth Competition Assay
Overnight cultures of P. aeruginosa strains R1 and S2 were diluted 50-fold into fresh L-broth medium. The bacterial cells of R1 and S2 were mixed at a ratio of 3:1 and cultured for an additional 7 h with agitation at 200 rpm at 37 • C. For controls, the individual bacterial cells of R1 or S2 were cultured under the same conditions. Then, bacterial cells were serially diluted in physiological saline and plated onto L-agar plates with or without 10 µg/mL piperacillin to enumerate the CFUs of R1 and the total of R1 and S2. The S2/R1 competitive index represents the ratio of recovered bacterial CFU of S2 to that of R1. The competition between R1ampD S2 and R1 or S2 and S2ampD R1 was conducted using the same manipulation.

Statistical analysis
GraphPad Prism 7.0 software was used to conduct the statistical analyses. p values were calculated using the two-tailed unpaired Student's t test. Difference was considered statistically significant when p value was below 0.05.

Clinical Isolates R1 and S2 Are Clonally Related
Sputum samples from a patient with sequelae of craniocerebral trauma were collected at 48 days apart. Two P. aeruginosa strains were isolated, with the first one being resistant to ceftazidime (R1), while the later one being sensitive to ceftazidime (S2). Sequence analysis of the PCR-amplified 16S rDNAs revealed that they both shared 100% identity to that of P. aeruginosa. Multilocus sequence typing (MLST), based on sequence analysis of PCR amplicons and genomic DNA reference mapping, demonstrated that R1 and S2 belonged to the same sequence type (ST) of ST-3078 (allelic profiles for aroE, acsA, guaA, nuoD, mutL, ppsA, and trpE were 4, 18, 134, 1, 33, 6, and 4, respectively). RAPD typing for the R1 and S2 isolates also demonstrated that they were of the same clonal lineage ( Figure 1A-C). Further comparison of R1 and S2 strains via genomic DNA reference mapping revealed a total of 281 SNVs (single nucleotide variations) between the two isolates. In combination, these results suggested that the R1 and S2 isolates are clonally related [1].

Isolate R1 Is Resistant, While S2 Is Susceptible to Ceftazidime
Ceftazidime is a clinically effective β-lactam and is frequently used to treat P. aeruginosa infections [18]. Preliminary examination with a VITEK automatic microbe analysis instrument revealed a differential susceptibility against ceftazidime between the R1 and S2 strains. Therefore, we further examined their susceptibility to ceftazidime with the twofold serial dilution method [14]. As displayed in Table 1, according to the CLSI guidelines, the initial isolate (R1) was resistant to ceftazidime, with an MIC of 32 µg/mL, while the later isolate (S2) displayed susceptibility to ceftazidime, with an MIC of 4 µg/mL, representing an eight-fold decrease. Piperacillin, another β-lactam antibiotic, was also tested. Similar to the ceftazidime, R1 was resistant to piperacillin while S2 was susceptible, with a sixteen-fold difference in the MIC of piperacillin (Table 1).

Isolate R1 Is Resistant, While S2 Is Susceptible to Ceftazidime
Ceftazidime is a clinically effective β-lactam and is frequently used to treat P. aeruginosa infections [18]. Preliminary examination with a VITEK automatic microbe analysis instrument revealed a differential susceptibility against ceftazidime between the R1 and S2 strains. Therefore, we further examined their susceptibility to ceftazidime with the two-fold serial dilution method [14]. As displayed in Table 1, according to the CLSI guidelines, the initial isolate (R1) was resistant to ceftazidime, with an MIC of 32 µg/mL, while the later isolate (S2) displayed susceptibility to ceftazidime, with an MIC of 4 µg/mL, representing an eight-fold decrease. Piperacillin, another β-lactam antibiotic, was also tested. Similar to the ceftazidime, R1 was resistant to piperacillin while S2 was susceptible, with a sixteen-fold difference in the MIC of piperacillin (Table 1).

ampC Is Downregulated in the S2 Strain
To elucidate the molecular mechanism of increased susceptibility to ceftazidime in isolate S2, RNA sequencing was carried out, and the global transcriptional profiles were compared between strains R1 and S2. As shown in Table S3, the expression levels of 68 genes were significantly altered between the two strains. Among them, the β-lactamaseencoding ampC gene displayed a 265-fold decreased mRNA level in S2 compared to that in R1. To confirm the observation, we further examined and compared the relative mRNA levels of ampC between R1 and S2 by real-time qPCR. Agreeing with the RNA-seq data, the relative mRNA level of ampC showed a significant decrease in strain S2 compared to R1 ( Figure 1D). To further investigate whether the reduced ampC expression contributed to the increased susceptibility to ceftazidime in S2, the ampC gene was overexpressed on a plasmid (pUCP24) and introduced into the S2 strain background. Based on the MIC of ceftazidime ( Table 1), overexpression of the ampC gene reduced the susceptibility of S2 to ceftazidime, with a 16-fold increase in the MIC. These results indicated that decreased expression of the ampC gene indeed contributes to the increased susceptibility of S2 to ceftazidime.

A 53 bp Duplicated Nucleotide Deletion in ampD Contributes to the Decreased Expression of ampC and Increased Susceptibility to Ceftazidime in S2
To further explore the mechanism for the increased susceptibility to ceftazidime and the decreased expression of the ampC gene in S2, genomic DNA reference mapping was conducted to identify mutations in the genome of S2 relative to that of the R1 strain using the model strain PAO1 as a reference genome (www.pseudomonas.com, accessed on 8 July 2021). Between strains R1 and S2, there were three genes with frameshifts (insertions/deletions) or early stops, fifteen genes had non-synonymous SNVs, and four genes had both frameshifts and non-synonymous SNVs (Table S4). Among them, an ampD gene, encoding N-acetyl-anhydromuramyl-L-alanine amidase, which is involved in inducible ampC expression and β-lactam antibiotic resistance [19], showed a major difference. In the R1 strain, a 53 bp duplicated nucleotide sequence, corresponding to the 29th-81st nucleotide acids of the ampD gene in PAO1, was inserted between the 81st and 82nd base site, resulting in a frameshift after the 27th amino acid of AmpD. PCR amplification and sequencing analysis of the ampD gene further validated the genomic DNA reference mapping result. These data suggested that AmpD in R1 is likely inactive.
AmpD inactivation had been found to be the most frequent mechanism resulting in ampC hyperexpression and β-lactam resistance in clinical isolates of P. aeruginosa [20,21]. The AmpD protein in S2 is identical to that of the PAO1 model strain, while the 53 bp nucleotide frameshift insertion in R1 created a possible inactive AmpD. To assess whether the alteration of ampD contributed to the decreased expression of ampC, as well as the increased susceptibility to ceftazidime in the S2 strain relative to the R1 strain, the chromosomal ampD gene of the R1 and S2 strains was replaced by the ampD gene of S2 and R1 using homologous recombination, respectively. As shown in Figure 1D, replacement of ampD S2 with ampD R1 increased the mRNA level of ampC in the S2 strain, while replacement of ampD R1 with ampD S2 in R1 led to a significant decrease in ampC transcription ( Figure 1D). Consistent with the significantly altered expression of the ampC gene, the R1ampD S2 strain displayed the same MIC against ceftazidime as the S2 strain, while S2ampD R1 showed the same MIC of ceftazidime as the R1 strain (Table 1). These results demonstrated that AmpD of R1 is nonfunctional and thus unable to repress the expression of ampC. However, a gain-of-function mutation in ampD is responsible for the decreased expression of ampC and increased susceptibility to ceftazidime in the S2 strain.

The S2 Strain Has a Higher Growth Rate and Competitive Advantage over R1
It has been predicted that in the absence of antibiotic stress, the fitness cost of maintaining resistance would select for loss of the resistance allele, consequently resulting in re-sensitization [22]. Therefore, we examined the growth curves of the R1 and S2 strains.
As shown in Figure 2A, the S2 strain displayed an increased growth rate after 9 h compared to the R1 strain. However, R1ampD S2 and S2ampD R1 exhibited similar growth rates as their corresponding R1 and S2 parent strains, suggesting that the increased growth rate in S2 was not due to the alteration of ampD ( Figure S1). This result is consistent with previous studies which reported that inactivation of ampD had no effect on the fitness of P. aeruginosa due to the presence of two additional ampD homologs, ampDh2 and ampDh3 [23,24]. However, the absence of all three AmpD amidases severely compromised the growth rates of P. aeruginosa [24].
ampC. However, a gain-of-function mutation in ampD is responsible for the decreased expression of ampC and increased susceptibility to ceftazidime in the S2 strain.

The S2 Strain Has a Higher Growth Rate and Competitive Advantage over R1
It has been predicted that in the absence of antibiotic stress, the fitness cost of maintaining resistance would select for loss of the resistance allele, consequently resulting in re-sensitization [22]. Therefore, we examined the growth curves of the R1 and S2 strains. As shown in Figure 2A, the S2 strain displayed an increased growth rate after 9 h compared to the R1 strain. However, R1ampDS2 and S2ampDR1 exhibited similar growth rates as their corresponding R1 and S2 parent strains, suggesting that the increased growth rate in S2 was not due to the alteration of ampD ( Figure S1). This result is consistent with previous studies which reported that inactivation of ampD had no effect on the fitness of P. aeruginosa due to the presence of two additional ampD homologs, ampDh2 and ampDh3 [23,24]. However, the absence of all three AmpD amidases severely compromised the growth rates of P. aeruginosa [24]. Furthermore, we compared the competitiveness between the R1 and S2 strains. R1 was co-cultured with S2 with a 3:1 ratio of CFU in L-broth for 7 h, and the number of the population of R1 and S2 was determined by enumerating the CFU on L-agar medium with or without 10 µg/mL piperacillin. Piperacillin was used for selection due to the larger difference in MIC between the R1 and S2. We found that the S2 strain had a higher competitive advantage over the R1 strain, with a 16.3-fold higher competitive index ( Figure 2B). Since the individual culture of S2 under the same conditions showed a 2.0-fold higher CFU compared to that of R1 ( Figure S2A), we reasoned that the competitive advantage was not simply due to the higher growth rate of the S2 strain. Next, we performed competition assays for the R1 and R1ampDS2 strains, as well as the S2 and S2ampDR1 strains. As shown in Figure 2C, replacement of the ampD in R1 with ampDS2 rendered a higher competition advantage, with a 4.7-fold increased competitive index. Similarly, replacement of . Two bacterial strains, R1 and S2, R1 and R1ampD S2 , or S2ampD R1 and S2, were each mixed at a 3:1 ratio, co-incubated for 7 h at 37 • C, and plated on L-agar plates with or without piperacillin. The competitive indexes in (B-D) represent the ratio of recovered bacterial number (S2, R1ampD S2 , or S2) to the respective number of recovered R1, R1, or S2ampD R1 strains. Bars represent the means from three tests, and error bars indicate the standard deviation. ***, p < 0.001, ****, p < 0.0001, by Student's t test.
Furthermore, we compared the competitiveness between the R1 and S2 strains. R1 was co-cultured with S2 with a 3:1 ratio of CFU in L-broth for 7 h, and the number of the population of R1 and S2 was determined by enumerating the CFU on L-agar medium with or without 10 µg/mL piperacillin. Piperacillin was used for selection due to the larger difference in MIC between the R1 and S2. We found that the S2 strain had a higher competitive advantage over the R1 strain, with a 16.3-fold higher competitive index ( Figure 2B). Since the individual culture of S2 under the same conditions showed a 2.0-fold higher CFU compared to that of R1 ( Figure S2A), we reasoned that the competitive advantage was not simply due to the higher growth rate of the S2 strain. Next, we performed competition assays for the R1 and R1ampD S2 strains, as well as the S2 and S2ampD R1 strains. As shown in Figure 2C, replacement of the ampD in R1 with ampD S2 rendered a higher competition advantage, with a 4.7-fold increased competitive index. Similarly, replacement of ampD in S2 with ampD R1 resulted in a competitive disadvantage ( Figure 2D). No significant difference in individual growth rates between R1ampD S2 and R1 or S2 and S2ampD R1 was observed under the same culture conditions ( Figure S2B,C).

S2 Displays a Reduced Susceptibility to LL-37 Than R1
Antimicrobial peptide LL-37 is the only cathelicidin peptide produced by epithelial cells, macrophages, neutrophils, and lymphocytes of the human body [25,26]. LL-37 exhibits a broad spectrum of antimicrobial activity, functioning as an important first-line defense against pathogenic bacteria [27]. Since S2 might be evolved from R1 in vivo, we reasoned that S2 may have a higher tolerance to LL-37 than R1. To test this hypothesis, we first treated the R1 and S2 strains with 256 µg/mL LL-37 for 4 h in PBS and determined the percentage of viable bacteria by plating and enumerating the CFU on L-agar plates. We found that the S2 strain displayed a significantly higher survival rate in the presence of LL-37 ( Figure 3). Replacement of the ampD R1 with ampD S2 in the R1 strain resulted in a similar survival percentage in the presence of LL-37 as that of R1. Similarly, replacement of ampD S2 with ampD R1 in the S2 strain showed no impact on the survival rate in the presence of LL-37 ( Figure 3). These results suggested that the S2 strain has a reduced susceptibility to LL-37, which is not due to the AmpD alteration. Further studies are required to understand the mechanisms resulting in the increased tolerance to LL-37 of the S2 strain. ampD in S2 with ampDR1 resulted in a competitive disadvantage ( Figure 2D). No significant difference in individual growth rates between R1ampDS2 and R1 or S2 and S2ampDR1 was observed under the same culture conditions ( Figure S2B,C).

S2 Displays a Reduced Susceptibility to LL-37 Than R1
Antimicrobial peptide LL-37 is the only cathelicidin peptide produced by epithelial cells, macrophages, neutrophils, and lymphocytes of the human body [25,26]. LL-37 exhibits a broad spectrum of antimicrobial activity, functioning as an important first-line defense against pathogenic bacteria [27]. Since S2 might be evolved from R1 in vivo, we reasoned that S2 may have a higher tolerance to LL-37 than R1. To test this hypothesis, we first treated the R1 and S2 strains with 256 µg/mL LL-37 for 4 h in PBS and determined the percentage of viable bacteria by plating and enumerating the CFU on L-agar plates. We found that the S2 strain displayed a significantly higher survival rate in the presence of LL-37 ( Figure 3). Replacement of the ampDR1 with ampDS2 in the R1 strain resulted in a similar survival percentage in the presence of LL-37 as that of R1. Similarly, replacement of ampDS2 with ampDR1 in the S2 strain showed no impact on the survival rate in the presence of LL-37 ( Figure 3). These results suggested that the S2 strain has a reduced susceptibility to LL-37, which is not due to the AmpD alteration. Further studies are required to understand the mechanisms resulting in the increased tolerance to LL-37 of the S2 strain. The viable bacterial numbers were determined by plating on an L-agar plate. The bacterial survival rate represents the ratio of recovered bacterial CFU following treatment with LL-37 to that of the initial inoculum. The data represent the means from three samples ± standard deviation. ns, not significant; **, p < 0.01, by Student's t test.

S2 shows Decreased Motilities and Pyocyanin Production, but Increased Biofilm and Rhamnolipid Production Compared to R1
It has been demonstrated that P. aeruginosa commonly shows decreased motilities, enhanced biofilm formation, and altered virulence production to adapt to the environment in the airway of patients with cystic fibrosis [28][29][30][31]. To determine whether the S2 strain acquired phenotypic adaptations relative to the R1 strain, we examined the motilities, production of pyocyanin and rhamnolipid, and biofilm forming abilities of the R1 and S2 strains. As shown in Figure 4, the later isolate S2 displayed impaired swimming, swarming, and twitching motility, whereas the early isolate R1 was proficient in all these motilities. The S2 strain also exhibited significantly decreased production of pyocyanin compared to the R1 strain ( Figure 4). Interestingly, the S2 strain was a better biofilm former and rhamnolipid producer than the R1 strain ( Figure 5). The above data, in combination with the fact that S2 has a higher growth rate and competitive advantage over R1, as well as the increased tolerance to LL-37, suggests that S2 was likely evolved from strain R1. However, by testing the phenotypes of strains R1ampDS2 and S2ampDR1, we found that The viable bacterial numbers were determined by plating on an L-agar plate. The bacterial survival rate represents the ratio of recovered bacterial CFU following treatment with LL-37 to that of the initial inoculum. The data represent the means from three samples ± standard deviation. ns, not significant; **, p < 0.01, by Student's t test.

S2 shows Decreased Motilities and Pyocyanin Production, but Increased Biofilm and Rhamnolipid Production Compared to R1
It has been demonstrated that P. aeruginosa commonly shows decreased motilities, enhanced biofilm formation, and altered virulence production to adapt to the environment in the airway of patients with cystic fibrosis [28][29][30][31]. To determine whether the S2 strain acquired phenotypic adaptations relative to the R1 strain, we examined the motilities, production of pyocyanin and rhamnolipid, and biofilm forming abilities of the R1 and S2 strains. As shown in Figure 4, the later isolate S2 displayed impaired swimming, swarming, and twitching motility, whereas the early isolate R1 was proficient in all these motilities. The S2 strain also exhibited significantly decreased production of pyocyanin compared to the R1 strain ( Figure 4). Interestingly, the S2 strain was a better biofilm former and rhamnolipid producer than the R1 strain ( Figure 5). The above data, in combination with the fact that S2 has a higher growth rate and competitive advantage over R1, as well as the increased tolerance to LL-37, suggests that S2 was likely evolved from strain R1. However, by testing the phenotypes of strains R1ampD S2 and S2ampD R1 , we found that these phenotypic alterations were not due to the difference in ampD between the R1 and S2 strains (Figures 4 and 5).
Microorganisms 2022, 10, x FOR PEER REVIEW 9 of 14 these phenotypic alterations were not due to the difference in ampD between the R1 and S2 strains (Figures 4 and 5).   Microorganisms 2022, 10, x FOR PEER REVIEW 9 of 14 these phenotypic alterations were not due to the difference in ampD between the R1 and S2 strains (Figures 4 and 5).

Discussion
Antibiotic resistance poses a considerable challenge for the treatment of P. aeruginosa infections. Due to their high efficacy and low toxicity, β-lactams are most commonly recommended to treat serious infections by difficult-to-treat P. aeruginosa. Understanding the swift and molecular mechanisms of bacterial susceptibility to β-lactams in the clinical setting may provide clues for the development of new or improved therapeutic strategies effective against P. aeruginosa. In this study, we identified a gain-of-function mutation in ampD, via deletion of a 53 bp duplicated nucleotide sequence, as the contributory factor for the conversion from resistant to susceptible against ceftazidime in a clinical P. aeruginosa strain. Of note, our study also indicated that recovery of susceptibility to β-lactam antibiotics in P. aeruginosa may occur during the P. aeruginosa infections. Hence, antibiotic susceptibility testing was required for repeated isolates at some intervals for optimal antibiotic therapy.
AmpC cephalosporinase is a chromosomally encoded resistance mechanism in P. aeruginosa, whose overproduction confers resistance to most β-lactam antibiotics [4]. ampD mutations resulting in ampC de-repression and decreased susceptibility to β-lactam antibiotics have been reported in P. aeruginosa clinical isolates [19][20][21]32,33]. Mutations in the ampD gene with amino acid substitutions [32,33], frameshift deletions [21,32], in-frame deletions [32], and IS element insertions [20] have been demonstrated to be the molecular mechanisms of AmpC-hyperproduction-mediated β-lactam resistance in clinical isolates of P. aeruginosa. Of note, by investigating a collection of isogenic ceftazidime-susceptible and -resistant pairs of strains, each pair sequentially isolated from a patient treated with β-lactams, Juan et al., found that inactivating mutations of ampD, including frameshift mutations, premature stop, and gene deletions, were responsible for AmpC hyperproduction and ceftazidime resistance [21]. In the current study, we sequentially recovered a pair of ceftazidime-resistant and -susceptible P. aeruginosa isolates from a single patient and found that a gain-of-function in AmpD, via deletion of a 53 bp duplicated nucleotide, was the molecular mechanism for the decreased ampC expression and increased ceftazidime susceptibility. Since the 53 bp inserted nucleotide acid between 81st and 82nd base sites in the R1 strain was a duplication of the 29th-81st nucleotide sequence of the ampD gene, we speculate that this fragment in R1 was gained during genome duplication. Similarly, during genome duplication, it might have been lost in the S2 strain. To the best of our knowledge, this is the first time reporting this nucleotide fragment duplication in the ampD gene in P. aeruginosa clinical strains. Interestingly, this is not the first report of a gain-of-function mutation in P. aeruginosa clinical strains. Gain-of-function mutations in the LysR family transcriptional regulator MexT result in constitutive overproduction of the efflux pump MexEF-OprN and increased resistance of P. aeruginosa clinical isolates to fluoroquinolones, chloramphenicol, and trimethoprim [34,35]. Gain-of-function mutations in the two-component regulatory system sensor kinase PmrB have been observed to promote polymyxin resistance of P. aeruginosa isolates from cystic fibrosis patients [36]. In addition, gain-of-function mutations in the LysR family transcriptional regulator have already been found to constitutively activate the transcription of a complex regulon for aromatic compound degradation in Acinetobacter baylyi and the transcription of genes of the cysteine regulon in Salmonella enterica serovar Typhimurium [37,38].
Our study reveals the function of AmpD in intraspecies competitiveness in P. aeruginosa. To the best of our knowledge, this is the first study to link the amidase AmpD with the intraspecies competition of P. aeruginosa. However, whether the AmpD alteration is the only contributor to the competitiveness advantage for S2 and the underlying mechanisms of AmpD-mediated competitiveness in P. aeruginosa remain elusive and warrant further studies.
Our study demonstrates that S2 displays decreased motilities and pyocyanin production, but increased biofilm and rhamnolipid production compared to the R1, while these phenotypic changes were not caused by the ampD mutation. Notably, simultaneous mutation of ampD and ampDh3 had wild-type swimming, twitching, and swarming, while simultaneous inactivation of the three amidases (AmpD, AmpDh2, and AmpDh3) had a substantial effect on twitching, swimming, and swarming in the P. aeruginosa PAO1 and PA14 background [24]. By analyzing the RNA-seq results, no difference in the expression levels of genes related to motility, pyocyanin production, biofilm formation, and rhamnolipid production was observed between the R1 and S2 strains (Table S3). Interestingly, genomic DNA reference mapping revealed mutations in the genes flgK, pilB, pilY1, and pslA of S2 compared to R1, which were reported to be involved in motility and biofilm formation [39][40][41][42]. Efforts are underway to determine the underlying mechanisms resulting in phenotypic changes in motility and biofilm formation in S2 in comparison to the R1 strain.
Pyocyanin and motility phenotypes (swimming, twitching, and swarming) are considered to be important virulence factors for acute infections, while biofilm formation is important for chronic infections [29]. P. aeruginosa undergoes evolutionary adaptations during the chronic infection process, including the loss of motility, reduced production of virulence factors, transition to a biofilm lifestyle, and evolution to a high-level antibiotic resistance [28,29]. Our later isolate S2 exhibited decreased pyocyanin production, impaired motilities (swimming, swarming, and twitching), and increased biofilm production. However, the S2 strain displayed an increased susceptibility to β-lactam antibiotics. Although we did not know the exact medical record of the patient before R1 isolation, the patient did not receive β-lactams in the 48-day period between R1 and S2 isolation. Therefore, we speculate that the R1 bacterial cells encountered a β-lactam-antibiotic-free environment in the host. A similar loss of resistance has been previously observed for polymyxin [36]. Although the molecular mechanisms have been unclear, some highly polymyxin-resistant clinical strains have been observed to gradually lose resistance when they are passaged repeatedly under non-antibiotic selection pressures [36].
In summary, using the sequentially recovered ceftazidime-resistant and ceftazidimesusceptible isolates, we identified that a functional gain of AmpD due to loss of a 53 bp duplicated nucleotide resulted in an increased susceptibility to ceftazidime in the later P. aeruginosa isolate S2. We also found that the altered AmpD contributed to the bacterial competition between R1 and S2, but AmpD was not responsible for phenotypic changes between R1 and S2, including growth rate, motilities, pyocyanin, rhamnolipid, and biofilm production.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/microorganisms10122395/s1, Figure S1. Growth curves of strains R1 and R1ampD S2 (A) and S2 and S2ampD R1 (B) in L-broth medium; Figure S2. Growth of individual strains R1, S2 (A), R1, R1ampD S2 (B), and S2, S2ampD R1 (C) in L-broth medium under otherwise identical competition conditions. ns, not significant; **, p < 0.01, by Student's t test; Table S1. Bacterial strains and plasmids used in this study; Table S2. Primers used in this study; Table S3. Transcriptome analysis: differentially expressed genes by RNA sequencing; Table S4. SNVs identified by genomic DNA reference mapping with elimination of SNVs with heterozygous mutations and the same mutation between R1 and S2. References [43][44][45][46] are cited in the supplementary materials. Funding: This work was funded by the National Science Foundation of China (31970680, 31870130, 32170177, 32170199, and 31970179) and the National Key Research and Development Project of China (2021YFE0201300, 2021YFE0101700, and 82061148018). The funders had no role in the study design, data collection and interpretation, or the decision to submit the work for publication.
Institutional Review Board Statement: Ethical review and approval were waived for this study as the clinical strains used in the study come from the routine procedures of the clinical laboratory rather than the clinical trials.

Informed Consent Statement:
Written informed consent was waived as the clinical strains used in the study come from the routine procedures of the clinical laboratory rather than the clinical trials.

Data Availability Statement:
The raw data of genomic DNA reference mapping have been deposited in the NCBI with an accessible number PRJNA863030. The RNA sequence raw data have been deposited in the NCBI under an accessible number PRJNA863024.