Genome Analysis of Pseudomonas aeruginosa Strains from Chronically Infected Patients with High Levels of Persister Formation

The appearance of persister cells with low metabolic rates are key factors leading to antibiotic treatment failure. Such persisters are multidrug tolerant and play a key role in the recalcitrance of biofilm-based chronic infections. Here, we present the genomic analyses of three distinct Pseudomonas aeruginosa Egyptian persister-isolates recovered from chronic human infections. To calculate the persister frequencies, viable counts were determined before and after treatment with levofloxacin. The susceptibilities of isolates to different antibiotics were determined using the agar-dilution method. To determine their recalcitrance, the levofloxacin persisters were further challenged with lethal concentrations of meropenem, tobramycin, or colistin. Furthermore, the biofilm formation of the persister strains was estimated phenotypically, and they were reported to be strong biofilm-forming strains. The genotypic characterization of the persisters was performed using whole genome sequencing (WGS) followed by phylogenetic analysis and resistome profiling. Interestingly, out of the thirty-eight clinical isolates, three isolates (8%) demonstrated a persister phenotype. The three levofloxacin-persister isolates were tested for their susceptibility to selected antibiotics; all of the tested isolates were multidrug resistant (MDR). Additionally, the P. aeruginosa persisters were capable of surviving over 24 h and were not eradicated after exposure to 100X-MIC of levofloxacin. WGS for the three persisters revealed a smaller genome size compared to PAO1-genome. Resistome profiling indicated the presence of a broad collection of antibiotic-resistance genes, including genes encoding for antibiotic-modifying enzymes and efflux pump. Phylogenetic analysis indicated that the persister isolates belong to a distinct clade rather than the deposited P. aeruginosa strains in the GenBank. Conclusively, the persister isolates in our study are MDR and form a highly strong biofilm. WGS revealed a smaller genome that belongs to a distinct clade.


Introduction
Pseudomonas aeruginosa is an opportunistic pathogen that poses a threat in clinical settings due to its intrinsic and acquired resistance to a wide range of antibiotics and biocides [1][2][3]. During chronic infection, P. aeruginosa can be evolved toward a highpersister state, producing more antibiotic-tolerant cells [4]. Persister-cell formation is a phenomenon that contributes to the tolerance of a bacterial sub-population to antimicrobial

Antimicrobial Agents
The antibiotics used in the study were pharmaceutical grade products. Meropenem was obtained from the ADWIA Company, Egypt. Cefoperazone and tobramycin were obtained from the Egyptian Pharmaceutical Industries Company (EPICO), 10th of Ramadan city, Egypt. Cefepime was obtained from RAMEDA, Egypt. Levofloxacin and amikacin were obtained from the AMOUN Company, Egypt. Colistin and chloramphenicol were the products of SIGMA-TEC, 6th October City, Egypt. The batch numbers of the used antimicrobials were listed in Supplementary Table S1.

Determination of Viable Bacterial Count
Ten-fold serial dilutions (final volume 1000 µL) of the bacterial suspensions, from an overnight culture in a nutrient broth (Oxoid, Hampshire, England), were performed in Eppendorf tubes. Triplicates of 100 µL of the appropriate dilutions were plated on nutrient agar plates (Oxoid, Hampshire, England). The plates were incubated for 24-48 h at 37 • C. The average numbers of colonies on the appropriate plate were counted and the colony forming units (CFU)/mL were calculated [27].

Selection and Quantification of Levofloxacin Persister Cells
Levofloxacin was added to a growing nutrient broth culture of the isolate at a concentration of 10-100 times its MIC. The suspensions were incubated in a shaker incubator (200 rpm) at 37 • C for 24 h. Falcone tubes containing bacterial cultures were centrifuged for 5 min at 8000 rpm and the supernatant was removed without disrupting the pellet. The cells were washed by re-suspending them in 1 mL phosphate-buffered saline (PBS) and transferred to an Eppendorf tube for centrifugation for 5 min at 14,000 rpm. The supernatant was gently discarded and the pellets were re-suspended in 1 mL of Mueller-Hinton broth (MHB). The viable count was made as described previously [28].

Determination of the MIC of Tested Antibiotics
The MICs of antibiotics were determined using the agar dilution method, according to the clinical laboratory standard institute (CLSI) guideline [29]. The used antibiotics were levofloxacin, chloramphenicol, amikacin, cefoperazone, meropenem, tobramycin, cefepime, and colistin. The MICs were reported and interpreted as sensitive, intermediate, or resistant, according to the CLSI breakpoints indicated in Supplementary Table S2.

The Killing Curve of Bactericidal Antibiotics against Levofloxacin Persisters
Time-killing curves were performed for selected persister-forming strains with antibiotics, as described previously [30]. Three water-soluble antimicrobial agents (colistin, meropenem, and tobramycin) were selected for time-killing curve assay.
Briefly, the overnight cultures of P. aeruginosa cells, treated with 100X MIC of levofloxacin (~2-5 × 10 3 CFU/mL), were pelleted (at 6000 rpm for 20 min), washed, and resuspended in 1 mL of a double-strength broth. The stock solutions of antimicrobial agents were prepared with concentrations equivalent to their 2x MICs and added to the suspension of persister cells. Tubes were incubated in a shaker incubator (200 rpm) at 37 • C for up to 24 h. Viable counts were made at zero time and after that at 1, 3, 5 and 24 h. The experiments were performed in triplicate with three independent cultures.

Biofilm Formation Capacity Test
The quantitative assay of biofilm formation was performed according to Stepanovic et al. [31]. A suspension of tryptone soy broth (TSB, Oxoid, Hampshire, UK) was prepared from the overnight culture of each isolate and adjusted to 10 6 CFU/mL in the freshly prepared culture broth of TSB. Aliquots of 200 µL of the bacterial suspension were transferred to the wells of a flat-bottom microtiter plate and incubated at 37 • C for 48 h. The TSB was removed gently and distilled water was used to wash the plates to remove any planktonic cells with subsequent air drying. After 20 min of treatment with 200 µL of 99% fixing methanol, the biofilm was then stained with 200 µL of 1% crystal violet (CV) solution for 15 min. After washing of the plate, 33% glacial acetic acid was used as a solvent for CV and the optical density (OD) of the solubilized dye was measured at 570 nm with a spectrofluorimeter (Biotek, Winooski, VT, USA). Wells containing only media were included in the study to serve as a negative control. The experiment was performed in triplicate.

Genomic DNA (gDNA) Extraction and Whole Genome Sequencing
The GeneJET™ Genomic DNA Purification Kit was used for gDNA extraction. The DNA purification protocol was performed according to the manufacturer's instructions. The purified gDNA was used immediately or stored at −20 • C. For the whole-genome sequencing, genomic, short-read sequencing was performed on an Illumina NextSeq 500 sequencer (Illumina, The Netherlands) using a library prepared with MiSeq v3 (2 × 75 bp) reagent kit.

Separation of P. aeruginosa Persister Cells and Determination of Their Antimicrobial Susceptibility
Thirty-eight P. aeruginosa isolates taken from chronically ill patients were studied for persister cells formation using 100X MIC of levofloxacin. From the thirty-eight P. aeruginosa isolates, only three isolates (8%) yielded persisters; these three isolates were isolated from respiratory tract infections and from patients with cystic fibrosis and were named P1, P2, and P3.
The resistance profile of these persisters showed increased resistance patterns to seven tested antimicrobials (Table 1). In comparison to the reference strain PAO1, all persister strains were resistant to meropenem, levofloxacin, chloramphenicol, and tobramycin (100%), while 90% of these strains were resistant to cefoperazone and colistin. Respective resistance was observed against cefepime (60%) and amikacin (70%). The three persister strains were found to be MDR.

Survival Curve of Persisters in the Presence of Levofloxacin, Tobramycin, Meropenem, and Colistin
After overnight growth in Mueller-Hinton (MH) broth, (Oxoid, Hampshire, England), cells were exposed to increasing concentrations of levofloxacin (250, 500, 1000 µg/mL). The analysis of the number of survivor cells after 5 and 24 h ( Figure 1) showed that the bulk of the population was effectively killed with 250 µg/mL levofloxacin. The killing curves were followed on the original strains.
A small fraction of cells (~10 −3 ) were tolerant for up to 1000 µg/mL levofloxacin (~100X MIC), The persisters after treatment with levofloxacin were exposed to lethal concentrations of colistin, meropenem, or tobramycin, and the survivor's numbers were followed for 24 h (Figure 2). A small fraction of cells (~10 −3 ) were tolerant for up to 1000 μg/mL levofloxacin (~100X MIC), The persisters after treatment with levofloxacin were exposed to lethal concentrations of colistin, meropenem, or tobramycin, and the survivor's numbers were followed for 24 h ( Figure 2).

Biofilm Forming Capacity of Isolates
The biofilm-forming capacity of the three levofloxacin-persister isolates in addition to the PAO1 strain and a non-persister forming isolate (clinical isolate which had no survival cells after treatment with high concentration (100X MIC) of levofloxacin) was quantified using a spectrophotometric assay. All the tested isolates were strong biofilm producers, as the measured OD was greater than 4ODc (the ODc was 0.1). The persister-forming isolates showed the highest absorbance compared to PAO1, indicating their higher biofilm-forming capacity; the relative absorbance was calculated in comparison to the absorbance of ODc (Table 2 and Figure 3).  A small fraction of cells (~10 −3 ) were tolerant for up to 1000 μg/mL levofloxacin (~100X MIC), The persisters after treatment with levofloxacin were exposed to lethal concentrations of colistin, meropenem, or tobramycin, and the survivor's numbers were followed for 24 h ( Figure 2).

Figure 2.
Biphasic time killing curves of (a) tobramycin, (b) colistin, and (c) meropenem against levofloxacin persisters of P. aeruginosa. After treatment with 100X-MIC of levofloxacin, P. aeruginosa strains (P1, P2, P3) were exposed to bactericidal concentrations of antibiotics and viable counts were determined at different time points (0, 1, 3, 5, and 24 h). Data are expressed as the Mean ± SEM of three independent experiments.

Biofilm Forming Capacity of Isolates
The biofilm-forming capacity of the three levofloxacin-persister isolates in addition to the PAO1 strain and a non-persister forming isolate (clinical isolate which had no survival cells after treatment with high concentration (100X MIC) of levofloxacin) was quantified using a spectrophotometric assay. All the tested isolates were strong biofilm producers, as the measured OD was greater than 4ODc (the ODc was 0.1). The persister-forming isolates showed the highest absorbance compared to PAO1, indicating their higher biofilm-forming capacity; the relative absorbance was calculated in comparison to the absorbance of ODc (Table 2 and Figure 3).

Biofilm Forming Capacity of Isolates
The biofilm-forming capacity of the three levofloxacin-persister isolates in addition to the PAO1 strain and a non-persister forming isolate (clinical isolate which had no survival cells after treatment with high concentration (100X MIC) of levofloxacin) was quantified using a spectrophotometric assay. All the tested isolates were strong biofilm producers, as the measured OD was greater than 4ODc (the ODc was 0.1). The persister-forming isolates showed the highest absorbance compared to PAO1, indicating their higher biofilm-forming capacity; the relative absorbance was calculated in comparison to the absorbance of ODc (Table 2 and Figure 3).   Biofilm formation by persister isolates compared to PAO1 standard strain and a non-persister (NP) clinical isolate. The quantitative assay of biofilm formation was performed using a spectrophotometric assay. All isolates formed strong biofilm with the highest strength observed with persister isolates. Data are expressed as the Mean ± SEM of three independent experiments. Twotailed unpaired T-tests were employed to analyze significance; ns denotes no significance, *** p < 0.001.

Whole Genome Sequencing (WGS) Reveals Distinct Features of the Persisters
All the genome sequences were submitted to the public database under the project accession number PRJNA890046. Genomes of the P1, P2, and P3 isolates were deposited as 1 (SRR21888463), 2 (SRR21888462), and 4 (SRR21888461). Isolates were identified to be Pseudomonas aeruginosa, as all the three isolates showed more than 70% DDH to type strain. The genome sizes of the persister isolates were 4,471,320 bp, 4,951,902 bp, and 4,661,420 bp for the isolates P1, P2, and P3, respectively, compared to 6,264,404 bp the genome size of the PAO1_NC_002516 reference strain. The genomes were distinguished into 1547, 1688, and 1706 contigs, respectively. The numbers of protein-coding sequences within the genomes were 4606, 5130, and 4928 genes, for the persister isolates P1, P2, and P3, respectively, compared to the 5671 coding genes for the PAO1 reference strain. The GC contents of the genomes were 64.38, 63.91, and 64.14%, respectively, compared to 66.56% for the PAO1 genome. WGS confirmed species-level identification of all isolates with percent identity values of 98.5% for all assemblies to the reference strain PAO1.
The circular map representing the entire genome assembly is shown in Figure 4. The circular image depicts the comparison of the genome of isolates P1, P2, and P3 to the reference genome of the PAO1 (NC_002516) strain. From the inner to outer circle is depicted the nucleotide position, GC content, GC skew (of the PAO1), and genes of isolates P1, P2, Figure 3. Biofilm formation by persister isolates compared to PAO1 standard strain and a nonpersister (NP) clinical isolate. The quantitative assay of biofilm formation was performed using a spectrophotometric assay. All isolates formed strong biofilm with the highest strength observed with persister isolates. Data are expressed as the Mean ± SEM of three independent experiments. Two-tailed unpaired T-tests were employed to analyze significance; ns denotes no significance, *** p < 0.001.

Whole Genome Sequencing (WGS) Reveals Distinct Features of the Persisters
All the genome sequences were submitted to the public database under the project accession number PRJNA890046. Genomes of the P1, P2, and P3 isolates were deposited as 1 (SRR21888463), 2 (SRR21888462), and 4 (SRR21888461). Isolates were identified to be Pseudomonas aeruginosa, as all the three isolates showed more than 70% DDH to type strain. The genome sizes of the persister isolates were 4,471,320 bp, 4,951,902 bp, and 4,661,420 bp for the isolates P1, P2, and P3, respectively, compared to 6,264,404 bp the genome size of the PAO1_NC_002516 reference strain. The genomes were distinguished into 1547, 1688, and 1706 contigs, respectively. The numbers of protein-coding sequences within the genomes were 4606, 5130, and 4928 genes, for the persister isolates P1, P2, and P3, respectively, compared to the 5671 coding genes for the PAO1 reference strain. The GC contents of the genomes were 64.38, 63.91, and 64.14%, respectively, compared to 66.56% for the PAO1 genome. WGS confirmed species-level identification of all isolates with percent identity values of 98.5% for all assemblies to the reference strain PAO1.
The circular map representing the entire genome assembly is shown in Figure 4. The circular image depicts the comparison of the genome of isolates P1, P2, and P3 to the reference genome of the PAO1 (NC_002516) strain. From the inner to outer circle is depicted the nucleotide position, GC content, GC skew (of the PAO1), and genes of isolates P1, P2, and P3, respectively. In addition, a comparison of the PAO1 genome with persister and P3, respectively. In addition, a comparison of the PAO1 genome with persister genomes revealed the presence of a number of genes (2401) in the persister genomes that were absent in the PAO1 genome (Supplementary Table S3).

Antimicrobial Resistance Elements
Resistome profiling was carried out to detect the resistance elements contributing to the MDR-pattern of these persisters forming isolates against anti-pseudomonas antibiotics. The analysis of antibiotics and antiseptic resistance elements pinpointed 26 resistance genes distributed among the isolates, 21 of them only exist in these persister isolates and 4 are shared with the PAO1 genome, while the blaPDC-1 gene is present in only the standard strain of PAO1 ( Figure 5). These resistance elements represent a cocktail of resistance elements against most of the clinically used antibiotics, including the major aminoglycosides modifying enzymes acetyltransferases: aac(6′)-Ib and aac(6′)-Ib4, phosphotransferase: aph(3′)-IIb, nucleotidyltransferases: aadA1 and aadA11 beside the 16S-rRNA methylase rmtB4. Variants and sub-variants of different classes of β-lactamases were detected, including blaOXA-10, blaOXA-14, bla OXA-395, bla OXA-50, bla OXA-796, and bla OXA-864. Several pseudomonas-derived cephalosporinase (PDC) variants were also detected, i.e., bla PDC, blaPDC-1, blaPDC-11, and blaPDC-16 and finally the blaNDM-1 gene, which encodes for the New Delhi metallo-β-lactamase 1. Acquired resistance elements against chloramphenicol were also detected, including catB7, cmlA5, and floR2, while individual resistance factors against tetracycline (tetG), fosfomycin (fosA), sulphonamides (sul1), novel ciprofloxacin-modifying enzyme encoding gene crpP, quinolone resistance determinant (qnrVC1), and the antiseptic resistance gene (qacEΔ1) are mainly found in isolate number 3. The broad collection of antibiotic resistance genes found in the genomes are the reason for the MDR of these isolates.

Antimicrobial Resistance Elements
Resistome profiling was carried out to detect the resistance elements contributing to the MDR-pattern of these persisters forming isolates against anti-pseudomonas antibiotics. The analysis of antibiotics and antiseptic resistance elements pinpointed 26 resistance genes distributed among the isolates, 21 of them only exist in these persister isolates and 4 are shared with the PAO1 genome, while the blaPDC-1 gene is present in only the standard strain of PAO1 ( Figure 5). These resistance elements represent a cocktail of resistance elements against most of the clinically used antibiotics, including the major aminoglycosides modifying enzymes acetyltransferases: aac(6 )-Ib and aac(6 )-Ib4, phosphotransferase: aph(3 )-IIb, nucleotidyltransferases: aadA1 and aadA11 beside the 16S-rRNA methylase rmtB4. Variants and sub-variants of different classes of β-lactamases were detected, including bla OXA-10 , bla OXA-14 , bla OXA-395 , bla OXA-50 , bla OXA-796 , and bla OXA-864 . Several pseudomonas-derived cephalosporinase (PDC) variants were also detected, i.e., bla PDC , bla PDC-1 , bla PDC-11 , and bla PDC-16 and finally the bla NDM-1 gene, which encodes for the New Delhi metallo-β-lactamase 1. Acquired resistance elements against chloramphenicol were also detected, including catB7, cmlA5, and floR2, while individual resistance factors against tetracycline (tetG), fosfomycin (fosA), sulphonamides (sul1), novel ciprofloxacinmodifying enzyme encoding gene crpP, quinolone resistance determinant (qnrVC1), and the antiseptic resistance gene (qacE∆1) are mainly found in isolate number 3. The broad collection of antibiotic resistance genes found in the genomes are the reason for the MDR of these isolates. Aside from the antibiotic modifying enzymes, the persisters' genome harbored gene clusters that correspond to the efflux pump-mediated resistance. Genome analysis identified 24-30 genes shared among the three isolates. These genes code for membrane fusion proteins (MFP), resistance nodulation division proteins (RND), and outer membrane proteins (OMP) that constitute efflux pump systems (Table 3)   Aside from the antibiotic modifying enzymes, the persisters' genome harbored gene clusters that correspond to the efflux pump-mediated resistance. Genome analysis identified 24-30 genes shared among the three isolates. These genes code for membrane fusion proteins (MFP), resistance nodulation division proteins (RND), and outer membrane proteins (OMP) that constitute efflux pump systems (Table 3). Based on the identified gene clusters, nine multidrug resistance efflux pumps (MexAB-OprM, MexCD-OprJ, MexEF-OprN, MexGHI-OpmD, MexJK-OpmH, MexMN-OprM, MexPQ-OpmE, MexXY-OprM, and MexVW-OprM) were chromosomally encoded in the persister isolates. Moreover, mutagenic analysis of the MexAB-OprM efflux pump regulators recognized a few point mutations in the MexAB-OprM repressors (MexR, NalD, and NalC). Table 3. Genotypic characteristics of efflux pump systems in P. aeruginosa persister isolates.

Global Phylogenetic Analysis of the Persister Isolates
The phylogenomic location of the studied isolates was performed. Completely closed genomes of 445 P. aeruginosa isolates were downloaded from the public database and compared with these persister-forming isolates. When phylogenomically compared to previously known P. aeruginosa strains from different countries, the studied isolates were placed in a unified but separate clade ( Figure 6 and Supplementary Table S4).

Global Phylogenetic Analysis of the Persister Isolates
The phylogenomic location of the studied isolates was performed. Completely closed genomes of 445 P. aeruginosa isolates were downloaded from the public database and compared with these persister-forming isolates. When phylogenomically compared to previously known P. aeruginosa strains from different countries, the studied isolates were placed in a unified but separate clade ( Figure 6 and Supplementary Table S4). Figure 6. Phylogenetic tree analysis of persister isolates compared to the genome of 445 P. aeruginosa isolates downloaded from the NCBI database. The phylogenetic tree was constructed based on the MASH distance using Ondov et al. [35]. Our persister isolates were placed in a unified but separate clade (indicated by the red color). Figure 6. Phylogenetic tree analysis of persister isolates compared to the genome of 445 P. aeruginosa isolates downloaded from the NCBI database. The phylogenetic tree was constructed based on the MASH distance using Ondov et al. [35]. Our persister isolates were placed in a unified but separate clade (indicated by the red color).

Discussion
Chronic infections represent a terrible therapeutic challenge and have a puzzling property. These infections are difficult and even impossible to eradicate. There is growing evidence that bacterial persisters are involved in the relapse of bacterial infections [42]. Th characterization of persister cells is important, not only in regarding the understanding of population dynamics but also concerning antibiotic tolerance in chronic-infections therapy [43]. Few studies have assayed persister formation in clinical isolates from complex host environments [44,45]. In the current study, the characterization of P. aeruginosa persisters isolated from chronic respiratory tract infections was performed using phenotypic and genotypic methods.
Initially, we isolated P. aeruginosa multidrug-tolerant persister forming strains. Although 100-fold MIC is unlikely to be achieved in the therapy of clinical infections, the application of high doses of antimicrobials would select for high persister mutants. The treatment with 100-fold MIC is a method of choice to identify highly drug-tolerant persister cells of P. aeruginosa [46]. Accordingly, in our study, three different P. aeruginosa isolates were identified as persister-forming strains. These isolates were recovered from 38 P. aeruginosa clinical isolates (8% persister-forming capacity). A low persister level, approaching 1% of the total population, was reported by Mulcahy et al. [44]. While a higher persister forming percentage (19%) was reported recently [45], their isolates were only obtained from CF infections, which is a deadly disease accompanied with a chronic respiratory infection.
Two of the three persister isolates (P2 and P3) formed both large colonies and small colonies on agar plates. Similarly, it was reported that S. aureus persisters induce a phenotypic change into small colony variants due to genomic inversion [47]. Furthermore, a recent study indicated that P. aeruginosa can persist for a long time in an acidic environment, through phenotypic alteration to a slow-growing colony phenotype [13].
The frequency of persisters was calculated for each isolate, and it ranged from 0.0017 to 0.0027 (approximately 10 −3 , meaning 1 out of 10 3 cells can form persister). Our result is quite consistent with the results of Salcedo-Sora and Kell [48], who reported a persister level within the range of 0.0012-0.0016. However, our results reveal much higher frequencies than the 10 −5 and the 10 -6 reported previously [24,49].
P. aeruginosa strains are intrinsically resistant to a variety of antimicrobials, and can further develop resistance during therapy [3]. In this study, the MIC results revealed that the persister-forming isolates were highly resistant to most of the tested antibiotics. Similarly, a recent study of Pseudomonas CF isolates revealed that more than 75% of the CF isolates were resistant to at least half of the antibiotics. Additionally, 25% of their isolates were found to be resistant to all the tested antibiotics, and only 36% of CF-isolates were resistant to tobramycin [50]. Furthermore, it was reported that persister (Hip) isolates were resistant to ciprofloxacin by 72.4% [45]. In this investigation, the time-killing curves using colistin, meropenem, and tobramycin revealed that the curves substantially differentiate not only between clinical isolates of P. aeruginosa but also between the different antibiotics. For example, isolate P1 showed a higher number of survival after exposure to tobramycin and meropenem, while isolate P2 had more survivor cells upon exposure to colistin.
The threat of P. aeruginosa is complicated due to its ability to form biofilms that provide a protected environment to bacterial cells which help in tolerating various stresses, including antimicrobials [51,52]. It was reported that the formation of P. aeruginosa persisters and biofilms are interrelated [53]. The biofilm formation was evaluated in the three persisterforming isolates, in addition to the PAO1 standard strain and a non-persister clinical isolate. All the tested isolates were strong biofilm formers; however, the persister isolates had stronger biofilms (as their OD was twice that of the PAO1 strain). In contrast, a previous study reported that the biofilm of PAO1 and their persister strain had the same strength [26].
Genomics has been a key element in the study of P. aeruginosa evolution. It has been reported that P. aeruginosa strains had a variable genome size of approximately 5-7 Mbp [54]. The omnipresence of P. aeruginosa may be related to its high genome plasticity. The genome of the P. aeruginosa PAO1 standard strain was sequenced approximately two decades ago [54]. However, the genetic mediators of persister formation for P. aeruginosa are poorly understood [4]. In the current study, genome sequencing of persister-forming isolates revealed that they had smaller genomes (4.66-4.9 Mbp), compared to PAO1-genome (6.3 Mbp) sequenced previously [55], and the genome of Pseudomonas CF-isolates (6.36 Mbp) recently sequenced [50]. The smaller genome size for our isolates could be due to the shortreads-based sequencing and assembly used in our study. Furthermore, it is documented that P. aeruginosa can customize its genome as individual strains have the ability to acquire or discard genomic segments to fit their needs for survival in virtually any environment [56].
Resistome profiling indicated the presence of a broad collection of antibiotic-resistance genes in our persisters' genome, with 25 resistance genes distributed in persister-isolates, most of which confer resistance to β-lactams (11 genes), aminoglycoside resistance (6 genes), and chloramphenicol resistance (3 genes), while resistance to other antibiotic classes is mediated by one gene each. CARD analysis revealed that the CF-isolates genome contains a high number of resistance genes, including beta-lactam resistance genes (n = 6), efflux pumps (n = 37), antibiotic inactivation enzymes (n = 8), and only one aminoglycoside resistance gene [50].
Multidrug efflux pumps constitute a group of antibiotic resistance elements that are a core part of bacterial genomes [57]. The P. aeruginosa PAO1 genome contains several drug efflux systems, predominantly of the RND and MFS families [56]. In the current study, several multidrug efflux tripartite pumps that encoded in operon were recognized at the genome level in our persister isolates (6-8 classes of the RND multidrug efflux pumpoperon in each genome), which explains their profound resistance against tested antibiotics. Consistent with our results, MexA/B pumps were reported to be closely associated with antimicrobial resistance and biofilm formation in clinical P. aeruginosa isolates recovered from respiratory infections [58].
It was reported previously that mutations in mexR, nalC, and nalD regulatory genes yield a MDR phenotype in P. aeruginosa [59,60]. Surprisingly, the identified point mutations within the MexAB-OprM efflux pump regulatory genes (MexR, NalC) have been previously described in clinical MDR-P. aeruginosa isolates [61,62]. In addition, an inactivating mutation in mexZ, encoding a repressor of the MexXY-OprM pump, was reported in P. aeruginosa persister isolates from CF patients [44]. All of these mutations favored the overexpression of these pump systems and caused increased resistance to different antimicrobials. Future work will focus on studying the transcriptional changes associated with persister formation to identify potential targets for the anti-persister drug.
Our results in resistome profiling of persister isolates allow the identification of genes that are likely to be involved in the antibiotic resistance of P. aeruginosa persister isolates. Knowledge of the complete genome sequence and encoded processes provides a wealth of information for the discovery of new antibiotic targets and for the development of more effective strategies to treat the life-threatening opportunistic infections caused by P. aeruginosa in humans.
Furthermore, phylogenetic analysis indicated that persister isolates belong to a unified but separate clade rather than the deposited P. aeruginosa strains in the public repositories. Further investigation is required to explore the validity and features of this clade. Similarly, the phylogenetic analysis of P. aeruginosa CF isolates performed by Datar et al. [50] revealed their unique clustering and indicate their location in different clades. To the best of our knowledge, our study is the first to study the genome sequence of persister isolates from local hospitals in Egypt.

Conclusions
The present study characterized, for the first time in Egypt, P. aeruginosa persister isolates. The persister isolates were strong biofilm producers and have a highly resistant phenotype when compared to the PAO1 standard strain. These results were confirmed by genome sequencing revealing a large number of resistance genes. Furthermore, WGS revealed a smaller genome that belong to a unified but separate clade compared to the genome of other P. aeruginosa strains deposited in the GenBank.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/pathogens12030426/s1, Table S1: The batch number of antimicrobial agents used in the current study; Table S2: CLSI breakpoints of antimicrobial agents used in the current study; Table S3: The list of genes that found in persister strains genome and absent from PAO1 genome; Table S4: List of the 445 P. aeruginosa isolates downloaded from the NCBI database and used for Phylogenetic tree analysis