Characterization of Third Generation Cephalosporin- and Carbapenem-Resistant Aeromonas Isolates from Municipal and Hospital Wastewater

Antibiotic resistance (AR) remains one of the greatest threats to global health, and Aeromonas species have the potential to spread AR in the aquatic environment. The spread of resistance to antibiotics important to human health, such as third-generation cephalosporins (3GCs) and carbapenems, is of great concern. We isolated and identified 15 cefotaxime (3GC)- and 51 carbapenem-resistant Aeromonas spp. from untreated hospital and treated municipal wastewater in January 2020. The most common species were Aeromonas caviae (58%), A. hydrophila (17%), A. media (11%), and A. veronii (11%). Almost all isolates exhibited a multidrug-resistant phenotype and harboured a diverse plasmidome, with the plasmid replicons ColE, IncU, and IncR being the most frequently detected. The most prevalent carbapenemase gene was the plasmid-associated blaKPC-2 and, for the first time, the blaVIM-2, blaOXA-48, and blaIMP-13 genes were identified in Aeromonas spp. Among the 3GC-resistant isolates, the blaGES-5 and blaMOX genes were the most prevalent. Of the 10 isolates examined, three were capable of transferring carbapenem resistance to susceptible recipient E. coli. Our results suggest that conventionally treated municipal and untreated hospital wastewater is a reservoir for 3GC- and carbapenem-resistant, potentially harmful Aeromonas spp. that can be introduced into aquatic systems and pose a threat to both the environment and public health.


Introduction
Antibiotic resistance (AR) is undoubtedly one of the greatest threats to global health [1]. In addition to the issue of AR in the clinical setting, there has been increasing interest in recent years in the role that the environment, particularly wastewater, plays in the maintenance and spread of AR. Due to the multi-faceted nature of AR, a One Health approach that incorporates interactions between the human, animal, and environmental microbiome is needed to fully understand the spread of AR [1]. Continuous identification, assessment and monitoring of AR hotspots is incredibly important for controlling and preventing the spread of AR. Hospital wastewater assessment is critical because hospital sewage is a major contributor of antibiotic-resistant bacteria (ARB) and antibiotic-resistance genes (ARGs) to wastewater systems and, subsequently, to the environment [2,3]. Another well-known AR hotspot is wastewater treatment plants (WWTPs), where bacterial populations are reduced many times over, but which still discharge large amounts of ARB into water bodies on a daily basis [4]. One of the predominant genera in treated wastewater is Aeromonas [5,6], which are used as bacterial indicators of environmental AR [6][7][8].

Identification, AR Profiles and ß-Lactamase Production in Aeromonas isolates
A total of 66 isolates of the genus Aeromonas were isolated from municipal wastewater (n = 59) and hospital wastewater (n = 7). Of these, fifteen isolates were selected on selective media with cefotaxime (representative of 3GCs) and fifty one isolates were selected on media with carbapenems. Fifty isolates were identified to the species level using matrixassisted laser desorption ionization-time-of-flight mass spectrometry (MALDI-TOF MS), and the other sixteen isolates, which could only be identified to the genus level using MALDI-TOF MS, were identified to the species level using their 16S rRNA gene sequences. The most prevalent species were A. caviae with 38 isolates (58%), followed by A. hydrophila with 11 isolates (17%), Aeromonas media and A. veronii with 7 isolates each (11%), and Aeromonas eucrenophila, Aeromonas rivipollensis, and A. salmonicida with only 1 isolate each ( Figure 1).
To compare the accuracy of the MALDI-TOF MS method in identifying Aeromonas at the species level, 16S rRNA sequencing of 35 Aeromonas spp. carrying carbapenemase genes was also performed (Table S3). Thirty-one isolates had the same species identification as MALDI-TOF MS, and four isolates showed a discrepancy between MALDI-TOF MS and 16S rRNA sequencing. Three isolates of A. caviae were identified as A. hydrophila by 16S sequencing, and A. salmonicida was identified as A. media by 16S sequencing. To compare the accuracy of the MALDI-TOF MS method in identifying Aeromonas at the species level, 16S rRNA sequencing of 35 Aeromonas spp. carrying carbapenemase genes was also performed (Table S3). Thirty-one isolates had the same species identification as MALDI-TOF MS, and four isolates showed a discrepancy between MALDI-TOF MS and 16S rRNA sequencing. Three isolates of A. caviae were identified as A. hydrophila by 16S sequencing, and A. salmonicida was identified as A. media by 16S sequencing.
All Aeromonas isolates were further subjected to antibiotic susceptibility testing against a range of clinically relevant antibiotics. More than 93% of the isolates were resistant to penicillins (amoxicillin (AML) and amoxicillin/clavulanic acid (AMC)) and first and second generation cephalosporins (cephalexin (CL) and cefuroxime (CXM)) ( Figure 2). In addition, more than 80% of the isolates were resistant to third and fourth generation cephalosporins (ceftazidime (CAZ) and cefepime (FEP)) and carbapenems (ertapenem (ETP), imipenem (IPM), and meropenem (MEM)), and more than 77% were resistant to ciprofloxacin (CIP). On the other hand, the isolates studied showed lower resistance to gentamicin (GM) (35%), sulfamethoxazole-trimethoprim (SXT) (38%), and colistin (COL) (20%) (Figure 2).  All Aeromonas isolates were further subjected to antibiotic susceptibility testing against a range of clinically relevant antibiotics. More than 93% of the isolates were resistant to penicillins (amoxicillin (AML) and amoxicillin/clavulanic acid (AMC)) and first and second generation cephalosporins (cephalexin (CL) and cefuroxime (CXM)) ( Figure 2). In addition, more than 80% of the isolates were resistant to third and fourth generation cephalosporins (ceftazidime (CAZ) and cefepime (FEP)) and carbapenems (ertapenem (ETP), imipenem (IPM), and meropenem (MEM)), and more than 77% were resistant to ciprofloxacin (CIP). On the other hand, the isolates studied showed lower resistance to gentamicin (GM) (35%), sulfamethoxazole-trimethoprim (SXT) (38%), and colistin (COL) (20%) (Figure 2). To compare the accuracy of the MALDI-TOF MS method in identifying Aeromonas at the species level, 16S rRNA sequencing of 35 Aeromonas spp. carrying carbapenemase genes was also performed (Table S3). Thirty-one isolates had the same species identification as MALDI-TOF MS, and four isolates showed a discrepancy between MALDI-TOF MS and 16S rRNA sequencing. Three isolates of A. caviae were identified as A. hydrophila by 16S sequencing, and A. salmonicida was identified as A. media by 16S sequencing.
All Aeromonas isolates were further subjected to antibiotic susceptibility testing against a range of clinically relevant antibiotics. More than 93% of the isolates were resistant to penicillins (amoxicillin (AML) and amoxicillin/clavulanic acid (AMC)) and first and second generation cephalosporins (cephalexin (CL) and cefuroxime (CXM)) ( Figure 2). In addition, more than 80% of the isolates were resistant to third and fourth generation cephalosporins (ceftazidime (CAZ) and cefepime (FEP)) and carbapenems (ertapenem (ETP), imipenem (IPM), and meropenem (MEM)), and more than 77% were resistant to ciprofloxacin (CIP). On the other hand, the isolates studied showed lower resistance to gentamicin (GM) (35%), sulfamethoxazole-trimethoprim (SXT) (38%), and colistin (COL) (20%) (Figure 2).  Different levels of resistance to carbapenem antibiotics were found in different Aeromonas spp. Almost all A. caviae, A. hydrophila, A. media, and A. veroni were resistant to ertapenem (Table S1). In addition, almost all A. media and A. veroni were also resistant to imipenem and meropenem. In contrast, the prevalence of resistance of A. caviae and A. hydrophila to imipenem and meropenem averaged 75% and 82%, respectively (Table S1). Colistin resistance was detected in single isolates of A. veronii and A. eucrenophila, three isolates of A. hydrophila and eight isolates of A. caviae (Table S1). With the exception of A. eucrenophila, all of these colistin-resistant isolates were co-resistant to at least one of the three carbapenem antibiotics.
Almost all Aeromonas isolates (65/66) were multidrug-resistant ( Figure 3), with 10 isolates showing co-resistance to 3-5 antibiotic classes and 55 isolates (majority A. caviae) showing co-resistance to 6-10 antibiotic classes ( Figure 3). methoxazole, CIP = ciprofloxacin, COL = colistin. Different levels of resistance to carbapenem antibiotics were found in different Aeromonas spp. Almost all A. caviae, A. hydrophila, A. media, and A. veroni were resistant to ertapenem (Table S1). In addition, almost all A. media and A. veroni were also resistant to imipenem and meropenem. In contrast, the prevalence of resistance of A. caviae and A. hydrophila to imipenem and meropenem averaged 75% and 82%, respectively (Table S1). Colistin resistance was detected in single isolates of A. veronii and A. eucrenophila, three isolates of A. hydrophila and eight isolates of A. caviae (Table S1). With the exception of A. eucrenophila, all of these colistin-resistant isolates were co-resistant to at least one of the three carbapenem antibiotics.

Detection of ESBL, AmpC and Carbapenemase Genes
Polymerase chain reaction (PCR) of genomic DNA for the five clinically relevant carbapenemase genes (bla KPC , bla NDM , bla IMP , bla VIM , and bla OXA-48 ) and Sanger sequencing of the obtained amplicons were performed for all 66 Aeromonas isolates ( Table 1). The most frequently detected carbapenemase gene was the bla KPC-2 gene, which was found in 41 isolates (62%), 6 of which co-occurred with bla VIM-2 (in A. caviae), and 1 with bla IMP-13 (in A. media). Two A. caviae isolates were positive for the bla NDM-1 gene and one for the bla OXA-48 gene. Isolates identified as ESBL producers (n = 15) were screened for ESBL genes (bla CTX-M groups 1, 2, 9, bla TEM , bla SHV , bla PER , bla VEB , bla GES and bla SME ) by PCR and Sanger sequencing of the obtained amplicons (Table 1). Ten isolates were shown to carry at least one of the ESBL genes targeted. The most frequently detected ESBL gene was bla GES-5 , which was found in 8 of 10 isolates (in 5 A. caviae and 3 A. media). In two isolates (A. caviae and A. media), bla GES-5 co-occurred with bla TEM-1 and in one isolate with bla SHV-12 (in A. media). One A. caviae isolate carried bla GES-5+TEM-1+CTX-M-15 and one carried bla GES-5+TEM-1+OXA-48 . The gene bla VEB-9 was detected in one A. hydrophila isolate (Table 1). Of the 11 strains that produced pAmpC, 6 carried bla MOX (5 A. caviae, 1 A. hydrophila), 1 of them together with bla CIT (A. caviae), 1 with bla FOX (A. caviae), and 1 with both bla CIT and bla FOX (A. caviae). Two isolates carried only bla CIT , one carried only bla FOX , and two To determine whether the ESBL and carbapenemase genes were placed on plasmids, plasmid DNA was subjected to target PCR for the corresponding genes. Out of 15 ESBLproducing isolates, 2 A. media isolates carried bla GES-5 and 1 carried bla SHV-12 on plasmids. The bla CTX-M-15 gene was found on plasmid DNA from one A. caviae. All 5 isolates carrying bla TEM-1 were confirmed to carry it on plasmids. Regarding the carbapenemase genes, 40/41 isolates carried bla KPC-2 on plasmids. The gene bla VIM-2 was found on plasmids of 6/7 A. caviae isolates, and 1 A. caviae carried bla NDM-1 on the plasmid. The bla OXA-48 and bla IMP-13 genes were not detected on the plasmids.

Enterobacterial Repetitive Intergenic Consensus Polymerase Chain Reaction (ERIC-PCR) Fingerprinting
Genetic similarity of all 66 Aeromonas spp. was assessed by ERIC-PCR analysis ( Figure  S1A-D). The fingerprints of the isolates consisted of 1 to 13 amplification bands ranging in size from 100 bp to 10,000 bp. Overall, the examined isolates exhibited great intraspecies diversity as they were assigned to 50 different ERIC-PCR profiles (27/38 for A. caviae, 9/11 for A. hydrophila, 7/7 for A. media, and 7/7 for A. veronii).

Conjugal Transfer of Carbapenem Resistance and Genotypic Characterization of Captured Plasmids
Conjugation experiments were performed to determine the possible transfer of carbapenem resistance from Aeromonas spp. to other bacteria. These experiments were performed using 10 randomly selected carbapenem-resistant Aeromonas isolates carrying carbapenemase genes on a plasmid (4 A. caviae, 2 A. hydrophila, 1 A. media, 2 A. veronii, and 1 A. salmonicida) as donors and the susceptible E. coli CV601 as the recipient. On LB plates containing meropenem (0.12 mg/L), we selected transconjugants that matched the phenotype of the donors. Transfer of the meropenem resistance phenotype was successful in three of ten isolates tested, namely, one KPC-2-carrying A. caviae, one NDM-1-carrying A. caviae, and one KPC-2-carrying A. salmonicida ( Table 1). The transfer frequency of meropenemresistant transconjugants was 6.41 × 10 −8 and 7.52 × 10 −5 per recipient in KPC-2-carrying A. caviae and A. salmonicida, respectively, and 8.59 × 10 −7 per recipient in NDM-1-carrying A. caviae.
Plasmid DNA was extracted from all three transconjugants and analysed by PCR for the presence of carbapenemase genes and plasmid replicon types ( Table 1). The carbapenemase gene bla KPC-2 was detected in transconjugants from bla KPC-2 -PCR positive A. caviae and A. salmonicida, whereas bla NDM-1 was not detected in transconjugant from bla NDM-1 -PCR positive A. caviae. In transconjugants from bla NDM-1 -PCR positive A. caviae and bla KPC-2 -PCR positive A. salmonicida, a plasmid was detected that was assigned to the ColE replicon type. In contrast, a combination of the replicon types of ColE and IncN plasmids was detected in transconjugant from bla KPC-2 -PCR positive A. caviae (Table 1).

Discussion
The present study aimed to characterize Aeromonas spp. from treated municipal and untreated hospital wastewater that exhibited resistance to the antibiotics 3GCs and carbapenems, which are important for human health, using culture-dependent and cultureindependent approaches. Aeromonas spp. are ubiquitous bacteria found in various aquatic habitats worldwide [9,11]. Considering this and the fact that they can easily acquire and exchange ARGs [30], research needs to focus more on the impact of these potentially harmful bacteria on human health.
In this study, 66 resistant Aeromonas strains were successfully isolated and identified to the species level by MALDI-TOF MS and/or 16S rRNA gene sequencing. These isolates could be assigned to 7 different species, with A. caviae (58%) dominating, followed by A. hydrophila, (17%), A. media (11%) and A. veronii (11%), while the prevalence of A. eucrenophila, A. rivipollensis and A. salmonicida was 2%. Of note, MALDI-TOF MS showed very good correlation with the results of 16S rRNA sequencing, with the exception of three cases of A. caviae identified as A. hydrophila and one case of A. salmonicida identified as A. media by 16S rRNA sequencing. In a previous study [31], MALDI-TOF MS was shown to be a more accurate method than 16S sequencing for distinguishing Aeromonas spp., particularly A. caviae and A. hydrophila and, therefore, we relied on the results from MALDI-TOF MS as the correct identification.
In other studies on aeromonads from wastewater environments, A. caviae was identified as the most abundant species in Brazil [32], while A. veronii was the predominant species in wastewater in Portugal [5]. Further assessment of genetic relatedness within these species by ERIC-PCR revealed high intraspecies diversity, as 7 species were assigned to 50 different ERIC-PCR profiles. This is in agreement with other studies that also showed high heterogeneity in ERIC sequences of Aeromonas strains [33][34][35][36].
Although Aeromonas spp. are environmental bacteria found primarily in aquatic environments, some members of this genus are pathogenic [13]. These include A. caviae, A. dhakensis, A. hydrophila, and A. veronii, which have been associated with gastrointestinal and respiratory tract infections, soft tissue and skin infections, etc. [11][12][13]. These species have also been isolated from untreated and treated wastewater [5,32]. Therefore, the predominant isolation of these species from hospital and municipal wastewater in this study suggests that these sources serve as reservoirs for these pathogens and as potential pathways for their transmission to humans and animals.
Characterizing the AR profile of Aeromonas isolates from wastewater is critical to understanding AR from the perspective of the One Health approach. In our study, all but one isolate was multidrug-resistant (i.e., resistant to ≥3 antibiotic classes). As expected, of all antibiotics tested, resistance to cephalosporins (first-fourth generation) and carbapenems (80%) was most common, followed by fluoroquinolones (CIP) (77%). Interestingly, 18% of our Aeromonas isolates (mainly A. caviae and A. hydrophila) were resistant to both carbapenems and colistin, limiting treatment options for Aeromonas infections, as carbapenems and colistin play an important role as "last resort" in the treatment of infections caused by gram-negative MDR bacteria. Therefore, the discharge of untreated hospital wastewater and treated municipal wastewater could increase the prevalence of these opportunistic pathogens with clinically relevant AR phenotypes in environmental waters and pose a threat to human health.
The positive AmpC β-lactamase result in some Aeromonas isolates was not unexpected as Aeromonas spp. are known to harbor chromosomal AmpC β-lactamase genes [11]. Among these genes, bla MOX was the most prevalent. Previously, bla MOX had been described in A. caviae [27] and in Aeromonas sp. from treated wastewater [14]. In addition, Aeromonas sanarellii, A. caviae, and A. media have been identified as carriers of three different variants of bla MOX [37]. Many previous studies have shown that Aeromonas isolates from the environment were susceptible to the 3GCs [5,38,39]. However, in this study, 15 Aeromonas isolates were isolated on selective agar with cefotaxime (3GC), all of which were resistant to ceftazidime (3GC) and produced ESBLs. Four different ESBL genes (bla GES-5 , bla SHV-12 , bla CTX-M-15, and bla VEB-9 ) and one narrow spectrum β-lactamase gene (bla TEM-1 ) were found in these isolates, with bla GES-5 being the predominant gene and found mainly in A. caviae and A. media isolates. This is consistent with a previous study in which the bla GES-5 gene was also found in A. caviae isolated from hospital wastewater in Brazil [29]. Although this gene is mainly of chromosomal origin in A. caviae and A. media isolates, it was also detected in an A. media isolate on the plasmid replicon ColE which, to our knowledge, had not previously been associated with the carriage of the bla GES-5 gene. The other ESBL genes detected have also been found in environmental Aeromonas spp. in previous studies [10,28,40,41].
The majority of A. caviae (37/38), A. hydrophila (7/8), and A. media (6/7) that were resistant to at least one of the carbapenem antibiotics tested produced carbapenemases and mainly possessed the bla KPC-2 gene. Previous studies have demonstrated the presence of bla KPC-2 in Aeromonas strains from natural waters in Brazil [42,43] and from wastewater in Brazil [3], the United States [30], China [20], and Japan [18]. Besides the present study, only one previous study detected bla KPC-2 in Aeromonas sp. from activated sludge in Europe (Poland) [14]. In addition, in the present study, bla VIM-2 , bla NDM-1, and bla OXA-48 were detected in A. caviae, while bla IMP-13 was found in A. media. To our knowledge, the bla VIM-2 , bla OXA-48 , and bla IMP-13 genes have never been detected in Aeromonas strains, and the bla NDM-1 gene was only recently discovered in two separate clinical cases of A. caviae in China [24,25]. Some of these genes (e.g., bla OXA-48 , bla IMP-13 ) in the Aeromonas isolates studied in this work were mainly of chromosomal origin. However, most of our Aeromonas isolates carried carbapenemase genes on plasmids, especially bla KPC-2 , bla VIM-2 , and bla NDM-1 , suggesting that they were acquired by HGT, probably under the selection pressure of antibiotics in hospital and municipal wastewater [2]. Interestingly, in six A. caviae isolates bla KPC-2 co-occurred with bla VIM-2 on a plasmid DNA, while in one A. media bla KPC-2 co-occurred with bla IMP-13 but was not present on plasmids. To our knowledge, this is the first report of the co-occurrence of both gene combinations (bla KPC-2 + bla VIM-2 and bla KPC-2 + bla IMP-13 ) in Aeromonas strains.
Plasmid-mediated multidrug resistance plays an important role in the worldwide spread of ARGs [44]. This study showed that Aeromonas isolates harbored various plasmid replicons, with ColE (n = 38), IncR (n = 15) and IncU (n = 11) being the most frequently detected. Our carbapenem-resistant Aeromonas isolates carrying these plasmids were also positive for bla KPC-2 , bla VIM-2 , or bla NDM-1 genes, suggesting possible mobilization of these genes by plasmid-mediated transfer. The plasmid replicons ColE and IncR have been associated with the carriage of bla KPC-2 in Enterobacterales isolates [45][46][47] while IncU (also known as IncP-6 [48]) has been associated with the transfer of bla KPC-2 from Aeromonas to E. coli [18]. In addition to bla KPC-2 , other carbapenamase genes such as bla NDM-1 and bla OXA-48 [49,50] and ESBL genes such as bla CTX-M-15 [51] have also been found on the ColE plasmid in enterobacteria. To our knowledge, this is one of the first reports of Aeromonas isolates carrying bla KPC-2 , bla VIM-2 or bla NDM-1 genes associated with conjugative ColE plasmids.
In the present study, transferable carbapenem (meropenem)-resistant plasmids were also captured from two A. caviae and one A. salmonicida strains to carbapenem-susceptible E. coli CV601. One ColE plasmid was captured from A. caviae (bla NDM-1 positive) and A. salmonicida (bla KPC-2 positive), and another one was captured from A. caviae (bla KPC-2 positive) in combination with replicon type IncN. The presence of bla KPC-2 in the ColE plasmid in transconjugant from A. salmonicida suggests that these plasmids may play an important role in the spread of Aeromonas carrying plasmid-localized bla KPC-2 genes. Moreover, the observation made in this study that two plasmids (ColE and IncN) were co-transferred from A. caviae (bla KPC-2 positive) to E. coli is also interesting in light of the findings of Barry et al. (2019) [52]. They demonstrated co-transfer of small cryptic plasmids such as Col440I along with larger plasmids carrying bla KPC-2 from bla KPC -positive Citrobacter freundii to E. coli. They found that the larger bla KPC -plasmids were never transferred alone and, therefore, hypothesized that these small plasmids are often transferred with conjugative plasmids carrying ARGs because they might play a helper role. The same could be true for our small (ColE) and larger (IncN) plasmids detected in bla KPC -positive transconjugant. The conjugative plasmid replicon IncN has already been associated with the carriage of bla KPC-2 in enterobacteria [53]. However, we cannot exclude the possibility that bla KPC-2 is present on both plasmids or only on ColE. In addition, bla NDM-1 was not detected in the transconjugant from NDM-1-positive A. caviae, suggesting that it may be localized on other mobile genetic element that is integrated into the host chromosome.
Aeromonas spp. are considered susceptible to colistin, with the exception of Aeromonas jandaei and A. hydrophila [54,55]. In our study, 3/11 A. hydrophila were resistant to colistin, one A. veronii and one A. eucrenophila, and most A. caviae (8/38). Although mcr genes (mcr-1, mcr-3, and mcr-5) have been found previously in Aeromonas spp. [56][57][58], our A. caviae, A. eucrenophila and A. hydrophila isolates were negative for all mcr genes tested (mcr1-5). Nevertheless, it is possible that these isolates have a novel variant of the mcr gene or another resistance mechanism, such as the addition of L-Ara-4N to the lipopolysaccharide layer [59]. Therefore, whole genome sequencing of these isolates would be required to fully elucidate the mechanism of colistin resistance.

Wastewater Sampling
Treated municipal wastewater (secondary effluent; 24-h composite samples) and untreated hospital wastewater (grab samples) from two hospitals (H1 and H2) in Zagreb, Croatia were collected on three consecutive days in winter 2020 (January). Municipal wastewater underwent stepwise treatment (primary and secondary-aerobic biodegradation) prior to discharge into the Sava River. Hospital wastewater was not treated in the hospital prior to discharge into the municipal wastewater system, as is common practice in all hospitals in Croatia. Samples were collected in sterile glass bottles (2.5 L), transported in cool boxes with ice blocks, and processed in the laboratory within 2 h.

Isolation of 3GC-and Carbapenem-Resistant Aeromonas
A series of dilutions of municipal and hospital wastewater samples were prepared in 0.85% NaCl (tenfold dilution up to 1:10,000). The dilutions were filtered in triplicate under vacuum through sterile mixed cellulose ester membrane filters (47 mm diameter, 0.22 µm pore size, Whatman, GE Healthcare, Life Science, Chicago, IL, USA) and the filters were placed on Rapid' E. coli 2 (BioRad, Hercules, CA, USA) agar plates supplemented with 4 mg/L cefotaxime (CTX) and CHROMagar mSuperCARBA (CHROMagar, Paris, France) agar plates. The plates were incubated at 37 • C for 24 h. Different colonies of Rapid' E. coli 2 + CTX and CHROMagar mSuperCARBA were selected for isolation of presumptive 3GCor carbapenem-resistant Aeromonas spp. A total of 200 presumptive isolates were purified on the same medium and stored in a 20% glycerol stock at −80 • C.

Identification of Isolates
Isolates were identified to species level by MALDI-TOF MS. Isolates were streaked on Mueller-Hinton plates (Oxoid, Basingstoke, UK), incubated overnight for 18-24 h at 37 • C, and sent to the Laboratory of Mass Spectrometry and Functional Proteomics at the Ruder Bošković Institute. Pure cultures were transferred directly to the spots of the MALDI-TOF MS target using a toothpick. A score value greater than 2.00 for the species level and a score value between 1.70 and 1.99 was indicated for successful identification. Isolates that could not be identified to species level by MALDI-TOF MS and those that carried carbapenemase gene(s) were additionally analyzed by sequencing of the 16S rRNA gene. The 1465 pb fragment was amplified with the universal primers 27F (5 -AGAGTTTGATCCTGGCTCAG-3 ) and 1492R (5 -GGTTACCTTGTTACGACTT-3 ) [60]. Thermocycling conditions were as follows: initial denaturation at 98 • C for 5 min, followed by 35 amplification cycles of denaturation for 10 s at 98 • C, annealing for 30 s at 60 • C and extension for 1:30 min at 72 • C, followed by a final extension at 72 • C for 5 min. PCR products were sent to Macrogen Europe (Amsterdam, The Netherlands) for purification and Sanger sequencing. Partial nucleotide sequences of the 16S rRNA genes were compared with the homologous sequences of the different Aeromonas species available in the GenBank database using BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi, accessed on 20 February 2023). All sequences were identified to species level (≥99% sequence identity).

Antibiotic Susceptibility Testing and Phenotypic Identification of ESBLs, pAmpC and Carbapenemases
Antibiotic susceptibility testing was performed by the disk diffusion method using commercially available disks according to EUCAST guidelines [61]. . Antibiotic disks AML, CL, and CAZ were purchased from OXOID (Basingstoke, UK) and the others from BD BBL (Franklin Lakes, NJ, USA). In addition, for isolates resistant to any of the carbapenems, the minimum inhibitory concentration (MIC) was determined by broth microdilution according to EUCAST guidelines. In addition, the susceptibility to COL was tested for all isolates using the MIC. Briefly, the initial concentration of 64 mg/L (COL, IPM, and MEM) or 16 mg/L (ETP) was serially double diluted in a sterile 96-well plate to a final concentration of 1 mg/L (COL, IPM, MEM) or 0.25 mg/L (ETP). Wells contained 90 µL Mueller-Hinton broth (Merck, Darmstadt, Germany) or, in the case of colistin, cation-adjusted Mueller-Hinton broth 2 (Sigma-Aldrich, Steinheim am Albuch, Germany) and serially diluted antibiotics. Each well was inoculated with 10 µL of an overnight bacterial culture diluted to a concentration of 5 × 10 5 CFU/mL. The plates were incubated overnight at 37 • C, and the lowest concentration at which no visible growth occurred was determined as the MIC of the isolates. Escherichia coli ATCC 25,922 and Escherichia coli NCTC 13,846 were used for quality control.
For phenotypic determination of ESBL and pAmpC production, 3GC-resistant isolates were subjected to DDST and pAmpC tests, respectively. For ESBL production, overnight cultures were diluted in 0.85% NaCl to a concentration of 0.5 McFarland and plated onto Mueller-Hinton agar using a sterile cotton swab. The CAZ (30 µg) and CTX (30 µg) discs were placed 20 mm and 30 mm (centre to centre) away from the amoxicillin-clavulanate (AMC, 20 + 10 µg) disc, respectively. If, after overnight incubation at 37 • C, synergy with clavulanate occurred with one of the 3GCs (enlargement of the zone of inhibition), this was considered a positive result for ESBL production (EUCAST guidelines).
The pAmpC production was determined using a cefoxitin disk (30 µg) alone and in combination with phenylboronic acid (300 µg) [62] applied to the inoculated Muller-Hinton plates. If there was an increase in the zone of inhibition of ≥5 mm after overnight incubation at 37 • C, the isolate was classified as a pAmpC producer.
All isolates that were resistant to carbapenems (both 3GC-and carbapenem-resistant) were subjected to an in-house CarbaNP assay to determine carbapenemase production [63]. Briefly, bacterial suspensions in Tris-HCL lysis buffer were mixed with 100 µL phenol red solution containing ZnSO 4 × 7H 2 O (0.1 mM) and imipenem-cilastatin (12 mg/mL). After incubation at 37 • C for a maximum of 2 h, the bacterial strains that changed the color of the suspension from red to orange or yellow were classified as carbapenemase producers.

Genomic and Plasmid DNA Extraction
Genomic DNA was extracted from overnight cultures of all Aeromonas isolates and transconjugants using the Quick-DNA TM Miniprep Plus Kit (Zymo, Irvine, CA, USA) according to the manufacturer's instructions. Plasmid DNA was extracted from overnight cultures of ESBL-, pAmpC-, and carbapenemase-producing Aeromonas isolates and three carbapenem-resistant transconjugants using the ExtractNow Plasmid Mini Kit (Minerva Biolabs, Berlin, Germany) according to the manufacturer's protocol. The extracted genomic and plasmid DNA was stored at −20 • C until further analysis.

Detection of Target ARGs by PCR and Sanger Sequencing of the Amplicons
Target ARGs in genomic and plasmid DNA extracted from Aeromonas isolates and transconjugants were detected by regular PCR with specific primers and conditions (Table 2), and the gene variant was identified by Sanger sequencing of the amplicons obtained. 3GC-resistant isolates (n = 15) were screened for ESBL genes by singleplex PCR (bla CTX-M groups 1, 2, and 9) and multiplex PCR (bla TEM , bla SHV , bla PER , bla VEB , bla GES, and bla SME ) and for pAmpC genes (bla FOX , bla EBC , bla CIT , bla ACC , bla MOX ) by multiplex PCR ( Table 2). All Aeromonas isolates (n = 66) and three transconjugants were screened for the presence of carbapenemase genes (bla KPC , bla NDM , bla OXA-48 -like, bla IMP , and bla VIM ). The colistinresistant strains were screened for the mcr-1, mcr-2, mcr-3, mcr-4, and mcr-5 genes ( Table 2). All PCR products were separated by 1.5% gel electrophoresis at 100 V for 60 min, stained with ethidium bromide, and visualized in the UV transilluminator. All positive PCR products were sent to Macrogen (Amsterdam, The Netherlands) for purification and sequencing in forward direction. The resulting sequences were compared with the reference in the NCBI database using BLASTX search.

Plasmid Replicon Typing
PBRT was used to identify the replicon type of plasmids in Aeromonas isolates and transconjugants. This was achieved by PCR amplification with plasmid DNA of the strains and transconjugants using the specific primer sets for 22 replicons and conditions listed in Table S2 [64][65][66]. Reactions were performed using the Hot Start Core Kit Ab+ (Jena Bioscience, Jena, Germany). PCR products were separated by electrophoresis (100 V, 60 min) on a 1.5% agarose gel and stained with ethidium bromide.

ERIC-PCR
All Aeromonas were fingerprinted by ERIC-PCR using the primers ERIC-1R (5 -ATGTA AGCTCCTGGGGATTCAC-3 ) and ERIC-2 (5 -AAGTAAGTGACTGGGGTGAGCG-3 ) [73]. The temperature profiles for amplification were as follows: initial denaturation at 95 • C for 7 min, followed by 30 cycles of amplification with denaturation for 30 s at 90 • C, annealing for 1 min at 52 • C, and extension for 8 min at 65 • C, followed by a final extension at 65 • C for 16 min [33]. The amplification products were separated by 1.5% gel electrophoresis for 90 min at 100 V. The size of the amplified products was determined by comparison with a 1-kb DNA ladder (Promega, Fitchburg, WI, USA).
The pattern of DNA fingerprint bands generated by ERIC-PCR was analyzed, and dendrograms were generated with GelJ software v2.0 [74], using the Dice coefficient to calculate similarity between fingerprints and UPGMA (the unweighted pair-group method with average linkages) method for cluster analysis.

Conjugation Assay
To investigate the potential plasmid transfer of carbapenem resistance, in vitro conjugation experiments were performed using 10 different carbapenemase-producing Aeromonas strains (4 A. caviae, 2 A. hydrophila, 2 A. veronii, 1 A. media, 1 A. salmonicida) as donors and the rifampicin-and kanamycin-resistant E. coli strain CV601 as plasmid recipient. The filter conjugation assay was performed as previously described [75]. Briefly, donor and recipient strains were grown overnight at 28 • C in LB broth. A total of 500 µL of the overnight cultures of each donor and recipient strain were mixed, centrifuged, and the pellets were resuspended in 200 µL of physiological saline and then spotted onto a filter for mating. After overnight incubation at 28 • C, the filters were washed in physiological saline, and 100 µL of the conjugation mixture was spread on LB agar plates containing kanamycin (50 mg/L), rifampicin (50 mg/L), and MEM (0.12 mg/L) and incubated for 48 to 72 h at 28 • C. Transconjugants were determined by the green fluorescence of the green fluorescent protein. Putative transconjugants were verified by BOX-PCR as previously described [76] and tested for carbapenemase genes and plasmid types as described above. Transfer frequencies were calculated as the total number of transconjugants divided by the total number of recipients.

Conclusions
This study showed that hospital and municipal wastewater contains multidrugresistant Aeromonas species, some of which are opportunistic pathogens of clinical importance. These Aeromonas isolates exhibited a diverse plasmidome and harbored ESBL and/or carbapenemase genes, which are frequently localized on plasmids. Successful plasmid-mediated transfer of the carbapenem resistance phenotype from A. caviae and A. salmonicida strains to susceptible E. coli recipients was also demonstrated. All these data suggest that Aeromonas spp. in the environment may serve not only as a reservoir for clinically important carbapenemase genes but also as a source for their transfer to other bacteria, including pathogens.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/antibiotics12030513/s1, Figure S1. Dendrogram showing genetic relatedness of (A) 38 Aeromonas caviae, (B) 11 Aeromonas hydrophila, (C) 7 Aeromonas media and (D) 7 Aeromonas veronii strains determined by analysis of ERIC-PCR fingerprints using the Dice similarity coefficient and UPGMA clustering method; Table S1: Antimicrobial susceptibility patterns of Aeromonas isolates; Table S2. Specific primers and conditions used in plasmid replicon typing; Table S3. Identification of Aeromonas strains by MALDI-TOF MS and/or 16S rRNA gene sequencing. Red letters indicate discrepancies between MALDI-TOF MS and 16S identification. ID represents the isolate designation; TWW and H refer to the source of bacterial isolation, treated municipal wastewater and hospital wastewater, respectively. ND, not determined.