Endemic High-Risk Clone ST277 Is Related to the Spread of SPM-1-Producing Pseudomonas aeruginosa during the COVID-19 Pandemic Period in Northern Brazil

Pseudomonas aeruginosa is a high-priority bacterial agent that causes healthcare-acquired infections (HAIs), which often leads to serious infections and poor prognosis in vulnerable patients. Its increasing resistance to antimicrobials, associated with SPM production, is a case of public health concern. Therefore, this study aims to determine the antimicrobial resistance, virulence, and genotyping features of P. aeruginosa strains producing SPM-1 in the Northern region of Brazil. To determine the presence of virulence and resistance genes, the PCR technique was used. For the susceptibility profile of antimicrobials, the Kirby–Bauer disk diffusion method was performed on Mueller–Hinton agar. The MLST technique was used to define the ST of the isolates. The exoS+/exoU− virulotype was standard for all strains, with the aprA, lasA, toxA, exoS, exoT, and exoY genes as the most prevalent. All the isolates showed an MDR or XDR profile against the six classes of antimicrobials tested. HRC ST277 played a major role in spreading the SPM-1-producing P. aeruginosa strains.


Introduction
Pseudomonas aeruginosa is a high-priority bacterial agent that causes healthcare-acquired infections (HAIs), which often lead to serious infections and poor prognosis in vulnerable patients, such as those who are in intensive care units (ICUs); those who have weakened immune systems, have undergone surgery, or have a history of inappropriate antibiotic use or severe burns; and those who have cystic fibrosis (CF), causing chronic lung colonization [1][2][3][4]. Globally, multi-drug and extensively resistant (MDR/XDR) strains of P. aeruginosa posing a difficult-to-treat resistance (DTR) phenotype have emerged in different clinical, hospital, and even environmental settings. These strains are of particular concern due to difficulties and limitations in treatment and their association with a high virulence potential, which can lead to severe and prolonged infections, and increased treatment costs, length of hospital stay, and patient mortality [5][6][7].
As a versatile opportunistic pathogen, P. aeruginosa is capable of causing both acute and chronic infections. Its pathogenic profile stems from the large and variable arsenal

Bacterial Isolates
This is a cross-sectional and descriptive study aiming to provide data on the SPM-1producing-P. aeruginosa isolates received at a reference center-the Special Pathogens Laboratory, Bacteriology and Mycology Evandro Chagas Institute (LabPate/SABMI/IEC)-for the routine surveillance of antimicrobial resistance. Since mid-2017, LabPate/SABMI/IEC has been acting in the antimicrobial resistance surveillance flow routine by confirming and detecting AMR mechanisms in bacterial isolates from public and private hospitals in the states of Pará (PA) and Acre (AC), northern Brazil. For the present study, 34 non-repeatable isolates of P. aeruginosa were obtained from various biological sample sources of patients admitted to healthcare services from 2018 to 2021, with suspected infection and/or colonization by MDR/XDR microorganisms and production of carbapenemases (resistance to carbapenems). All the isolates were identified using the Vitek-2 automated system at a routine hospital (BioMérieux). Subsequently, the isolates were sent to Evandro Chagas Institute for further analysis.

Phenotypic and Molecular Assays Associated with Antimicrobial Susceptibility and Genetic
Variant Definition of bla  Antimicrobial susceptibility testing (ATS) was performed by applying the Kirby-Bauer disk diffusion method on Mueller-Hinton Agar (MHA) for 12 antimicrobials belonging to six (06) different classes: piperacillin, piperacillin + tazobactam, and ticarcillin/clavulanic acid (penicillin + β-lactamase inhibitor class); ceftazidime and cefepime (cephalosporins class); aztreonam (monobactams class); imipenem (carbapenems class); gentamicin, tobramycin, and amikacin (aminoglycosides class); and ciprofloxacin and ofloxacin (fluoroquinolone class). The results were interpreted according to the criteria and breakpoints of Clinical and Laboratory Standards Institute, where isolates were classified as susceptible (S), intermediate (I), and resistant (R) [33,34]. Additionally, P. aeruginosa isolates were phenotypically classified based on their propensity to be MDR when they were resistant to ≥1 drug in ≥3 antimicrobial classes; XDR when they were not susceptible to 1 agent in all antimicrobial classes tested, except ≤2, according to the criteria described by Magiorakos et al. [35] and Mulet et al. [36]; and DTR based on the susceptibility results with ceftazidime, cefepime, imipenem, ciprofloxacin, and ofloxacin, as described by Kadri et al. [7].
Bacterial genomic DNA was obtained from a single overnight grown colony of P. aeruginosa cultures via the boil-and-freeze method and using the commercial Pure-Link™ Genomic DNA Mini Kit (Thermo Fisher Scientific, São Paulo, Brazil), following the manufacturer's recommendations. The genomic DNA obtained was quantified using a Picodrop PICO100 spectrophotometer (Picodrop Limited, Hinxton, UK) and concentrations set between 25-50 ng/µL were used for all molecular assays. The detection of AMR genes encoding carbapenemase bla SPM , bla IMP , bla VIM , bla NDM , bla KPC , and bla OXA-48 was performed via PCR in a Veriti thermal cycler (Applied Biosystem, Foster City, CA, USA) as described [37]. Visualization of PCR products was performed via electrophoresis in a 1.5% agarose gel at 110 V for 45 min in TAE 1× buffer (89 nM Tris-borate and 2 mM EDTA pH 8.0). DNA ladder 1 Kb (Invitrogen™, Carlsbad, CA, USA)) was used as molecular weight marker, gel stained with SyberSafe (Invitrogen™, Carlsbad, CA, USA)), and differentiation of bands visualized under ultraviolet light.
For determination of the bla SPM variant, the PCR products were direct sequenced bidirectionally using the Big Dye Terminator v3.1 kit on the ABI Prism 3100 or 3500XL Genetic Analyzer platform (Applied Biosystems, Foster City, CA, USA), and the sequences obtained were compared with those available in the BLAST database (https://blast.ncbi. nlm.nih.gov/Blast.cgi (accessed on 6 June 2023)).

Molecular and Phenotypic Detection of Virulence-Related Factors
The detection of invasion-related genes belonging to the T1SS, T2SS and T3SS was performed via PCR in a Veriti thermal cycler (Applied Biosystem, Foster City, CA, USA) according to the protocol described by Rodrigues et al. [32]. Visualization of PCR products was performed via 1.5% agarose gel electrophoresis at 110 V for 45 min in TAE 1× buffer (89 nM Tris-borate and 2 mM EDTA pH 8.0). As molecular weight marker, 1 Kb DNA ladder (Invitrogen™) was used, gel stained with SyberSafe (Invitrogen™, Carlsbad, CA, USA)) and differentiation of bands visualized under ultraviolet light. In addition, the pigment production and mucoid phenotype of P. aeruginosa isolates were verified by observing bacterial growth on MHA agar plates and slants.

Molecular Typing by Multilocus Sequencing Typing-MLST
The MLST genotyping procedure followed the protocol outlined by Curran et al. [38], with modifications by using new design primers, except for aroE gene (Supplementary Table S1). In brief, the Veriti thermocycler (Applied Biosystems, Fos-ter City, CA, USA) was used to amplify via PCR the seven housekeeping genes constituting the scheme (acsA, aroE, guaA, mutL, nuoD, ppsA, and trpE). The resulting reaction products were sequenced bidirectionally using Big Dye Terminator v3.1 chemistry on the ABI Prism 3100 or 3500XL Genetic Analyzer platforms (Applied Biosystems, Foster City, CA, USA). The obtained results were compared and matched to the data available at the PubMLST database (http://pubmlst.org/paeruginosa (accessed on 6 June 2023)) to determine the allelic profiles and sequence types (STs).

Whole-Genome Sequencing (WGS) and Bioinformatics Analysis
Libraries were prepared from the previously extracted DNA using the Nextera XT kit (Illumina, San Diego, CA, USA) with the addition of I5 and i7 indexes, according to the manufacturers' protocol. The quality of the libraries was verified using the Bioanalyzer High Sensitivity DNA Analysis kit (Agilent™, Santa Clara, CA, USA) and quantified using the High Sensitivity Double Strand DNA Quibit kit (Invitrogen™, Carlsbad, CA, USA)). Subsequently, the libraries were added into a pool and sequenced with the 2 × 151 pairedend protocol on Illumina nextseq 550 using Mid Output (Illumina™) reagent cartridges and flow cells at the Arbovirology section of the Instituto Evandro Chagas.
The quality of the reads was checked using the fastqc v0.11.9 tool, treated using the fastp v0.23.2 tool to remove low quality reads and remove adapters. Subsequently, genome assembly was performed using the spades tool v3.15.3 based on the reference strains for P. aeruginosa CCBH4851 (NZ_CP021380.2), which belongs to clone ST277 reported as cause of endemic outbreak in Brazil in 2008 [39]. After assembly, the scaffolds were evaluated using the quast software (v 5.2.0) and submitted to the bactopia v2.2 pipeline for annotation using the prokka tool v1.14.6, resistance prediction using amrfinder v3.10.45. The modular tools of the bactopia pipeline were also used for downstream analysis: abricate for searching resistance genes, amrfinderplus for predicting resistance and proteins, MLST typing was predicted by searching for sequence in the PubMLST database, pasty for predicting P. aeruginosa serogroup, and plasmidfinder for predicting plasmid presence in sequencing. Finally, the annotated genomes produced by Bactopia were finally submitted type Strain Genome Server (TGYS) [40] web server for whole-genome similarity, clusterization and phylogenetic inference.

Ethical Considerations
The present study is in accordance with the principles of the Declaration of Helsinki and the terms of the CNS Resolution No. 466/2012 of the National Health Council. Since this is an experimental study, which used stored and provided samples by the institutions involved, without any contact and possibility of identifying the respective patients, the project did not need to be referred to the Ethics Committee on Research Involving Human Beings.

Genotyping by MLST Data
From the pool of P. aeruginosa isolates presenting MDR phenotypes, nine (9) randomly selected isolates were subjected to molecular typing via MLST, revealing that all nine MDR-P. aeruginosa isolates belonged to the high-risk clone (HRC) and endemic clone ST277 determined by the combination of the seven housekeeping genes used in the MLST scheme for P. aeruginosa (acsA 39, aroE 5, guaA 9, mutL 11, nuoD 27, ppsA 5, and trpE 2) ( Table 2).

Genotyping by MLST Data
From the pool of P. aeruginosa isolates presenting MDR phenotypes, nine (9) randomly selected isolates were subjected to molecular typing via MLST, revealing that all nine MDR-P. aeruginosa isolates belonged to the high-risk clone (HRC) and endemic clone ST277 determined by the combination of the seven housekeeping genes used in the MLST scheme for P. aeruginosa (acsA 39, aroE 5, guaA 9, mutL 11, nuoD 27, ppsA 5, and trpE 2) ( Table 2).

Discussion
Recently, the rapid emergence of CR-PA strains has become prominent in scientific interest and epidemiological surveillance, mainly due to the dissemination of MβLs that break down antibiotic compounds that are commonly used as a last-resort treatment to serious infections, rendering penicillin, cephalosporins, and carbapenems ineffective. This scenario is the result of several factors, including the overuse and misuse of antibiotics,

Discussion
Recently, the rapid emergence of CR-PA strains has become prominent in scientific interest and epidemiological surveillance, mainly due to the dissemination of MβLs that break down antibiotic compounds that are commonly used as a last-resort treatment to serious infections, rendering penicillin, cephalosporins, and carbapenems ineffective. This scenario is the result of several factors, including the overuse and misuse of antibiotics, and the poor infection control practices in healthcare settings. Certainly, the COVID-19 pandemic has also placed a tremendous pressure on healthcare systems worldwide, as critically ill patients were at increased risk for secondary bacterial infections associated with MDR/XDR/DTR strains, including MβL-producing-P. aeruginosa. In the present investigation, we report the spread of SPM-1-producing-P. aeruginosa strains mostly associated with the HRC ST277, and detected since the pre and early COVID-19 pandemic period in healthcare institutions in northern Brazilian.
Worrying rates of AMR associated with XDR/MDR/DTR P. aeruginosa isolates have been reported in the last decade, as demonstrated by Jean et al. [41] in Taiwan, where the AMR rate in 2015 was less than 18.0%, while in the following years (2016 and 2018), the rate increased to 19.7% and 27.5%, respectively. A study conducted in Spain reported that 17.0% of P. aeruginosa infections were caused by XDR strains, and high rates of over 30.0% of CR-PA were linked to hospital-acquired pneumonia (HAP) as reported in many European Union states since 2015 [42,43]. Additionally, DTR among P. aeruginosa were related to almost 8.0% of isolates causing BSIs [44]. Despite this, there is still scarce global information on the prevalence of MDR/XDR/DTR-P. aeruginosa [20]. Further, due to the similarity of symptoms between hospitalized patients with SARS-CoV-2 infection and those with hospital-acquired and ventilator-associated pneumonia, it is a common practice to administer broad-spectrum antibiotics as empirical treatments [45]. According to a review conducted by Fattorini et al. [29], 476 out of 539 patients (88.3%) diagnosed with COVID-19 received broad-spectrum antibiotics, such as expanded-spectrum cephalosporins (e.g., ceftriaxone, ceftazidime, and cefepime), fluoroquinolones, and carbapenems. Consequently, the use of antibiotics has significantly increased in many healthcare settings globally during this period [46].
As per national data by the Brazilian National Health Surveillance Agency (ANVISA), from 2018 to 2021 in adult ICUs, CR-PA was the third-most-detected bacterial pathogen related to BSIs and urinary-tract infections (UTIs), and demonstrated carbapenem resistance rates from 30.9% to 41.4%, and from 41.7% to 43.0%, respectively [47][48][49][50][51]. Worryingly, it is relevant to emphasize the staggering increase in the number of P. aeruginosa isolates causing BSIs in 2021 (pandemic-period), totaling 3,845 cases, a remarkable 168.1% surge compared to 2019 (pre-pandemic period), which recorded only 1432 cases. Surely, the resistance phenotypes of the CR-PA in this study, which included MDR/XDR/DTR isolates, reflect this worrisome scenario, further complicated by the presence of SPM-producing isolates. Finally, such findings also align with our research group's previous data, in which Rodrigues et al. [32] documented the early spread of MDR/XDR CR-PA within local ICUs in the state of PA from 2010 to 2013.
The monobactam antibiotic ATM has presented potential in the treatment of infections caused by MDR/XDR CR-PA [52,53]. In the present report, ATM has been indicated as an effective antimicrobial against CR-PA, with a resistance rate of only 37.1%. This sensitivity profile can be attributed to the fact that the antibiotic is not broken down by SPM. Studies conducted worldwide and in Brazil have reported similar findings, suggesting the strong efficacy of ATM against CR-PA [54,55]. However, it is noteworthy that resistance to ATM was observed in some isolates, pointing out the presence of other AMR mechanisms, such as mutations observed in mexAB-oprM efflux system [56]. Further investigations are needed to fully understand the role of ATM and its potential strategies in the management of CR-PA infections [57].
Results obtained through the WGS analysis of the 10 XDR SPM-1-producing P. aeruginosa allowed further insights into the AMR mechanisms presented in such strains, in which the aac(6 )-Ib', aadA7, aph(3 )-IIb, bla OXA-56 , bla PDC-374 , bla SPM-1 , catB7, cmx, crpP, fosA-354827590 and rmtD1 markers were commonly found. For the bla OXA-494 gene, only one sample was negative; in contrast, for the bla OXA-50 gene, only one sample was positive. This bacterial resistome echoes the findings in the Brazilian study published by Galetti et al. (2018), where in genomic analysis of 13 different P. aeruginosa strains belonging to ST277 revealed a highly conserved resistome (bla SPM-1 , rmtD, aacA4, aadA7, bla OXA-56 , bla OXA-396 , bla PAO , aph(3 )-IIb, aac(6 )Ib-cr, crpP, catB7, cmx, and fosA), playing an important role in the persistence of this clone in infections occurring in Brazilian hospitals. The bla OXA gene variants are considered as naturally occurring in the P. aeruginosa genome, and its high prevalence indicates a potential horizontal transfer in which class D β-lactamases can be introduced by other co-habituating bacterial species [58,59]. According to Horcajada et al. [1] and Nicolau [60], the bla OXA-50 gene plays an important role in P. aeruginosa resistance, since classical β-lactamase inhibitors show weak activity against it. Indeed, kinetic analysis of β-Lactams hydrolysis by OXA-50 variants of P. aeruginosa demonstrated that chromosomally encoded AMR mechanisms mainly provided weak carbapenemase activity, but may act synergically [61]. Among the aminoglycoside-modifying enzymes presented, the aac(6 ) acetyltransferase is one of the most frequently described, conferring resistance to both tobramycin and amikacin, or tobramycin alone [1,62,63].
To fuel its pathogenicity, P. aeruginosa possesses an array of virulence factors that enable the colonization, invasion, and persistence within host tissues, often leading to acute and chronic challenging-to-treat infections. Gaining a comprehensive understanding of these virulence mechanisms is imperative for the development of effective strategies to manage P. aeruginosa infections [13]. In relation to presence of pigments like pyocyanin and pyoverdine, it has been implicated in exacerbating infections as these pigments sequester iron from host cells, serving the metabolic needs of the bacterium, and consequently intensifying the infection and pathogenesis [64]. A study conducted by Fothergill et al. [65] reported pyocyanin production in P. aeruginosa isolates ranging from 41.3% to 81.5%, findings consistent with the data obtained in the current investigation, where pyocyanin production among isolates was of 42.9%. With regard to pyoverdin, Prado et al. [66] observed pyoverdin production in over 74.0% of clinical strains, while Silva et al. [67] found that more than 90.0% of the isolates investigated exhibited pyoverdin production. Interestingly, the present study recorded a pyoverdine production rate of 42.9%, which contrasts with the previous findings. In this study, aprA, a gene belonging to T1SS, and lasA and lasB genes belonging to T2SS, showed high positive occurrence. Other studies with SPM-1-producing P. aeruginosa also reported a strong presence of these virulence genes, as in the studies by Adonizio et al. [68] and Silva et al. [67].
In addition, the translocation of up to four cytotoxic effector proteins by the T3SS is responsible for distinct tissue injury to the host, with exoU having a higher impact on bacterial virulence [11]. The distribution of the genes encoding these cytotoxins is not uniform among P. aeruginosa strains, and some of them, particularly exoS and exoU, are almost mutually exclusive [69]. In fact, a large, multicenter study conducted in Spain revealed that the exoU + /exoS − genotype was an independent risk factor for early mortality in P. aeruginosa BSIs, and was negatively linked to XDR profiles [14].Thus, the T3SS factor is an important differential factor that needs to be considered when analyzing virulence and clinical outcomes associated with HRC [70]. Results on the present study highlight the fact that all evaluated strains were related to the invasive virulotype (exoS + /exoU − ), genotypic virulence profile usually observed among MDR/XDR P. aeruginosa strains, as exoU carriage along with several AMR mechanisms may pose a fitness cost to bacterial cell [71][72][73]. Furthermore, all 10 samples analyzed belonged to serogroup O2. According to Stanislavsky [74], polysaccharide O (OPS), the most variable region of the lipopolysaccharide (LPS), is of major relevance in the virulence and is responsible for conferring serogroup specificity. According to Donta et al. [75], serogroup O2, along with serogroups O1, O3, O4, O5, O6, O7, O10, and O16, accounts for 90% of bacteriemic strains of P. aeruginosa. In a study by Nasrin et al. [17], serotype O2, along with serotypes O5, O16, O18, and O20, were among the most common, which corroborates with the findings of the present study.
Global epidemiology data further highlight a small geographic spread of bla SPM-1 strains when compared to its endemicity in Brazil, with rare reports of this variant in countries such as Iran [76,77], UK [78], Chile [79], Egypt [80], and the USA [81][82][83]. In Brazil, the clonal expansion of SPM-1-producing P. aeruginosa strains is related to the HRC ST277, with its detection in all Brazilian regions, including São Paulo [25], Rio de Janeiro [84], Paraná [85], Porto Alegre [86], Minas Gerais [87], and Pará [31,32], showing its dissemination potential, high adaptability, and establishment as an international clone. In the present report, the MLST genotyping revealed that 18 strains with MDR/XDR/DTR characteristics belonged to the ST277 lineage, and one related to the ST2711, which, to the best of our knowledge, is the first report of bla SPM-1 in another clone than the ST277. This finding also supports the limited genetic diversity of SPM-1-producing P. aeruginosa, and indicates the possible occurrence of an outbreak, with recent clonal expansion probably related to the high selective pressure on healthcare institutions in northern Brazil. The clonal expansion of such strain also raises concerns regarding the potential dissemination of AMR gene, and the limited effectiveness of conventional treatment options. Further, when comparing the genomic phylogenetic inference, we highlight the distance between sample 57508/ST2711 and the other samples, being the most distant sample when compared to CCBH4851; the remaining samples were clustered as a possible transmission chain due elevated similarity (above 99.9%). Further investigation is needed to understand the underlying mechanisms driving the persistence and spread of these particular STs in the clinical or environmental settings.
As the pandemic spread, hospitals globally observed an increase in patients infected with COVID-19, a situation requiring major adjustments in healthcare systems and infrastructure, especially in infection control and antimicrobial management programs [88]. In this regard, some reports indicate that the indiscriminate use of antibiotics determined by the therapeutic challenges in combating the pandemic has resulted in increased AMR rates, especially related to individuals infected with P. aeruginosa hospitalized in ICUs [89]. Unfortunately, less robust healthcare systems, such as those in the Latin American and Asian countries, where AMR rates are dangerously high and antimicrobial stewardship programs are just beginning to be implemented, are adjusting their response to the pandemic to varying degrees [90][91][92][93]. Regrettably, these circumstances create the so-called "perfect storm" for an accelerated evolution of AMR, especially in clinically important strains, such as P. aeruginosa [94]. The present study is one of first in Brazil to thoroughly report data on AMR after the COVID-19, pandemic reflecting a comparative perspective between studies conducted before the pandemic, where the peak detection of SPM-1-producing P. aeruginosa strains occurred between 2008 and 2015 [32,[95][96][97], and in the post-pandemic context, as with results in the current study, a re-emergence and the possibility of an outbreak of SPM-1-producing P. aeruginosa was observed.
The present study is not without its limitations. Firstly, a notable limitation was the loss of isolates during the culture process, which may have resulted in an incomplete dataset. Additionally, our laboratory faced the constraint of unavailability of certain essential testing disks for evaluating classical antipseudomonal drugs such as meropenem and colistin, and novel antibiotics including cefiderocol, ceftazidime-avibactam, and ceftolozane-tazobactam. Another limitation stems from the lack of comprehensive data regarding the origin and specific wards from which the P. aeruginosa isolates were recovered, limiting our ability to assess potential associations between strain characteristics and clinical settings, and outbreak investigation. Lastly, all included samples could not be genotyped via MLST and WGS due to a lack of necessary reagents, which could have provided valuable insights into genetic relatedness and transmission patterns. These limitations should be taken into account when interpreting the findings and highlight areas for further investigation and improvement in future studies.

Conclusions
The CR-PA isolates included in this study showed a high prevalence of virulence genes, where among them, aprA, lasA, toxA, exoS, exoT, and exoY were positive in all strains, suggesting a high pathogenicity capacity. The exoS + /exoU − virulotype was standard in all isolates, indicating an invasive characteristic. As for the phenotypic profile of resistance, all strains showed either MDR or XDR, in addition to a pool of DTR isolates, thus posing a challenge regarding the management and treatment of patients infected with P. aeruginosa producing SPM. Additionally, results obtained through MLST and WGS revealed the major role of the HRC ST277 in spreading SPM-1-producing strains, in addition to the novel report of the bla SPM-1 variant in the clone, ST2711, and a conserved resistome.