Occurrence and temporal distribution of extended-spectrum β-lactamase-producing Escherichia coli in clams from the Central Adriatic, Italy

The spread of extended-spectrum β-lactamase (ESBL)-producing Escherichia coli is a major public health issue. Bivalves are filter-feeder animals capable of bioaccumulating the microorganisms present in water. This physiological characteristic makes them both good indicators of environmental contamination and possible carriers of pathogenic bacteria, including those resistant to antimicrobials. The aim of this study was to investigate the occurrence of ESBL-producing E. coli in clams (n = 308) collected from harvesting areas of the Central Adriatic Sea between 2018 and 2019. ESBL- /class C β-lactamase (AmpC)- producing E. coli and Escherichia spp. were isolated by streaking over the surface of MacConkey agar plates supplemented with cefotaxime enriched broths of the initial shellfish suspension. E. coli and Escherichia spp. resistant to cefotaxime were screened for ESBL production by using the double disk synergy test. Susceptibility to different antimicrobials and confirmation of ESBL-production were determined by the minimum inhibitory concentration (MIC) test. Isolates were further characterized by whole genome sequencing (WGS) and bioinformatic analysis of genomes with different tools. Overall, ESBL-producing E. coli were isolated from 3% of the samples. Of 13 ESBL- and ESBL−/AmpC-producing Escherichia spp. (n = 11 E. coli, n = 1 E. marmotae, n = 1 E. ruysiae) isolates, 13 were resistant to ampicillin and cefotaxime, 9 to sulfamethoxazole, 6 to tetracycline and nalidixic acid, 4 to trimethoprim, and 3 to ceftazidime, cefoxitin, ciprofloxacin, and chloramphenicol. Moreover, the majority (8/11) of the ESBL-producing E. coli isolates were multidrug-resistant. WGS showed that the isolates predominantly carried the blaCTX-M-15 gene (3/11) and blaCTX-M-14 and blaCTX-M-1 (2/11 each). The AmpC β-lactamase CMY-2 was found in two isolates. Phylogroup A was the most prevalent (5/11), followed by phylogroups D (4/11), F (1/11), and B2 (1/11). Ten different sequence types (STs) were identified. Occurrence at sampling sites ranged between 0 and 27%. To identify associations between the occurrence of ESBL-producing E. coli and E. coli levels, samples were divided into two groups, with E. coli at >230 MPN/100 g and E. coli at ≤230 MPN/100 g. ESBL-producing E. coli isolates were significantly more commonly recovered in samples with higher E. coli levels (14%) than in those with lower levels of E. coli (2%). Moreover, the majority (3/4) of the potentially pathogenic strains were isolated in samples with higher E. coli levels. These findings provided evidence for the bacterial indicator of fecal contamination, E. coli, as an index organism for ESBL-producing E. coli isolates in bivalves.


Introduction
Antimicrobial resistance (AMR) is one of the most significant public health threats (World Health Organization, 2021), responsible for hundreds of thousands of estimated deaths annually worldwide (O'Neill, 2016).
Third/fourth/fifth-generation cephalosporins are the highest priority critically important antimicrobials (HPCIAs) in human medicine (World Health Organization, 2019), and Enterobacteriaceae producers of extended-spectrum β-lactamases (ESBLs) are on the critical-priority WHO list of antibiotic-resistant bacteria for research and development of new antibiotics (Tacconelli et al., 2018).An 8-fold increase in the intestinal carriage rate of ESBL Escherichia coli in the community has occurred globally over the past two decades (Bezabih et al., 2021), and the global and regional human intestinal carriage of ESBL E. coli is increasing in both community and healthcare settings (Bezabih et al., 2022).The spread of ESBL poses a serious threat to public health; therefore, it is important to investigate sources and transmission routes and encourage studies contributing to the "One-Health" approach.
Human-and animal-gut bacteria, including those resistant to antimicrobials, can reach marine environments through various routes (e.g., runoff from land, sewage systems, and feces from birds and wild animals), with the potential to contaminate seafood products.Bivalves are invertebrate filter-feeder animals capable of bioaccumulating microorganisms present in the surrounding waters.Thus, they are good indicators of environmental contamination and may act as possible carriers of bacteria derived from fecal pollution (Lee and Silk, 2013), including those that are resistant to antimicrobials (Albini et al., 2022).
In the European Union, regulation concerning the sanitary safety of live bivalve mollusks (Anonymous, 2004(Anonymous, , 2019) ) stipulates that classified production areas shall be periodically monitored to check the microbiological quality of shellfish by using the bacteriological indicator of fecal contamination, E. coli.The latter is a genetically diverse species that comprises non-pathogenic gut commensals and strains responsible for intestinal and extra-intestinal disease.Enterotoxigenic E. coli (ETEC) strains are able to bind and colonize the intestinal epithelium and also produce various enterotoxins, of which heat-labile and heat-stable toxins and/or enteroaggregative heat-stable toxin 1 (EAST1) lead to diarrhea.Extra-intestinal pathogenic E. coli (ExPEC) are non-commensal E. coli isolates capable of causing extra-intestinal disease due to the possession of pathogenic virulence factors (Russo and Johnson, 2000).Escherichia coli isolates containing at least two genes coding virulence factors (papA and/or papC, sfa/foc, afa/draBC, kpsM II, and iutA) are defined as ExPEC (Peirano et al., 2013).
Antimicrobial resistance monitoring programs in the EU are focused on terrestrial animals (Aerts et al., 2019).Studies on the occurrence of ESBL-producing E. coli in bivalves are limited.The prevalence of ESBL-producing E. coli has been investigated in retail bivalves in studies conducted in European or North African countries (Boss et al., 2016;Vu et al., 2018;Sola et al., 2022).In European studies, ESBL-or AmpC-producing E. coli isolates were not recovered in retail oysters sampled in Switzerland (Boss et al., 2016), and ESBL-producing Enterobacteriaceae were isolated in 20% of bivalve samples collected at retail in Berlin (Vu et al., 2018).In another study conducted in Tunisia, bivalves were sampled in unrelated markets in different regions, and ESBL-producing Enterobacterales (mostly E. coli) were cultured from 1.6% of clam pools (Sola et al., 2022).Other studies investigated the occurrence of ESBL-producing E. coli isolates in bivalves from production areas (Rees et al., 2015;Bueris et al., 2022).Hence, there are relatively few studies assessing the occurrence of ESBL-producing E. coli in bivalves from production areas over different seasons, and none of them were carried out in Italy.Furthermore, to our knowledge, previous studies have not investigated the relationship between levels of the bacterial indicator of fecal contamination, E. coli, and ESBL-producing E. coli presence in bivalve mollusks.
Clam is a major commercial species in Italy, and among EU countries, Italy is the main producer, accounting for 77% of farmed clams in the EU, reaching 24,452 tons in 2020 (European Commission, 2022).As clams are grown in coastal waters, they can also represent a sentinel species in determining AMR in the marine environment.
The aim of this study was to investigate the occurrence of ESBLproducing E. coli in clams collected from harvesting areas of the Central Adriatic Sea between 2018 and 2019 and study the correlation between ESBL-producing E. coli and levels of E. coli, the bacterial indicator of fecal contamination of bivalve mollusks.Moreover, EBSL-and ESBL-/AmpCproducing E. coli isolates from bivalves were characterized phenotypically, for susceptibility to antimicrobials, and genotypically, by whole-genome sequencing (WGS), to assign them to a serotype, phylogroup, sequence type (ST), and identify the presence of resistance and virulence genes and mutations that confer antimicrobial resistance.

Sampling
A total of 308 samples of clams (Venus gallina), collected from 28 sampling sites of harvesting areas located along the coast of the region of Marche (Supplementary Appendixes 1, 2), were analyzed for the presence of E. coli-producing ESBL and/or class C β-lactamase (AmpC).Of these, 127 were from bivalve mollusk harvesting areas classified as B, which requires a post-harvest treatment before being placed on the market to meet E. coli health parameters.The remaining 181 samples were from areas classified as A.
Monitoring for the presence of E. coli ESBL/AmpC was performed approximately each month between 2018 and 2019 for the majority (25) of the areas.
Bivalve mollusks were externally cleaned with running potable water; then, the flesh and liquor of the bivalve mollusks were aseptically collected, diluted, homogenized, and further diluted in a 0.1% sterile peptone water to achieve a final suspension of 1:10.Subsequent decimal dilutions were prepared in a 0.1% sterile peptone solution.E. coli enumeration on bivalve mollusks was performed by a most probable number (MPN) method according to ISO 16649-3 (Anonymous, 2015).
ESBL-/AmpC-producing E. coli were isolated by streaking over the surface of MacConkey (MC) agar plates supplemented with 1 μg/ mL of cefotaxime (Sigma Aldrich-Merck KGaA, Darmstadt, Germany) enriched broths of the initial shellfish suspension in double-strength mineral-modified glutamate (MMGB) broth from the E. coli enumeration method.Inoculated MC agar plates with cefotaxime were incubated at 37° C ± 1° C for 24 h ± 2 h in aerobic conditions.Two colonies showing typical characteristics of E. coli were randomly selected from each sample and isolated in trypticase soy agar (TSA, Biolife, Italy).

Escherichia coli identification
Presumptive colonies of E. coli were analyzed by PCR for the uidA gene with primers uidA-277F and uidA-934R 1 and by matrix-assisted laser desorption ionization-time-of-flight mass spectrometry (MALDI-TOF MS Biotyper, Bruker Daltonics) analysis.

Antimicrobial susceptibility testing and ESBL/AmpC phenotype
For the isolates identified as E. coli, disk diffusion susceptibility tests (EUCAST, 2017a) were conducted for nine antibiotics (Supplementary Appendix 2).Inhibition diameter sizes were interpreted by using the EUCAST breakpoint tables (EUCAST, 2018), except for nalidixic acid and tetracycline, for which CLSI breakpoint values were used (CLSI, 2019).
E. coli isolates resistant to cefotaxime were screened for ESBL production by using the double disk synergy test (DDST) (EUCAST, 2017b).
For each sample, according to antimicrobial resistance screening results, one E. coli isolate was selected and further investigated for the determination of the minimum inhibitory concentration (MIC) for different antimicrobial classes and for the confirmation of ESBL production.If differences were observed in the antimicrobial susceptibility profiles of E. coli isolates from the same sample, each isolate underwent MIC tests.MIC tests were performed with Sensititre EU Surveillance Salmonella/E.coli EUVSEC Plates and Sensititre EU Surveillance ESBL EUVSEC2 Plates (Thermo Fisher Scientific), according to the Thermo Scientific Sensititre Plate Guide for Antimicrobial Susceptibility Testing (Thermo Fisher Scientific).
Multidrug resistance (MDR) was considered when isolates were resistant to three or more antimicrobial classes (Magiorakos et al., 2012).

DNA extraction and whole-genome sequencing
Genomic DNAs were extracted from 1 mL of logarithmic phase broth cultures from pure E. coli cultures by using the QIAamp DNA Mini Kit (Qiagen Inc., Hilden, Germany) following the manufacturer's protocol for Gram-negative bacterial organisms.DNA was quantified with the Qubit fluorometer (QubitTM DNA HS Assay, Life Technologies, Thermo Fisher Scientific Inc.).DNA libraries were prepared by using the Nextera DNA Flex Library Prep Kit (Illumina Inc., San Diego, CA), according to the manufacturer's manual, loaded onto NextSeq 500/550 Mid Output Reagent Cartridge v2, 300 cycles kit (Illumina Inc., San Diego, CA) and then sequenced on an Illumina NextSeq 500 platform, to generate 150 bp paired-end reads.
The assembled genomes were analyzed by online tools available at the Center for Genomic Epidemiology (CGE), Technical University of Denmark. 2 Briefly, the FASTA files were analyzed using the following CGE databases: ResFinder (v.4.1) for antimicrobial resistance genes (ARGs) and chromosomal point mutations associated with resistance (Camacho et al., 2009;Zankari et al., 2017;Bortolaia et al., 2020), multilocus sequence typing (MLST v.2.0.9) for defining the ST (Lemee et al., 2004;Bartual et al., 2005;Wirth et al., 2006;Jaureguy et al., 2008;Camacho et al., 2009;Griffiths et al., 2010 (Camacho et al., 2009;Carattoli et al., 2014), VirulenceFinder (2.0.3) for virulence determinants (Camacho et al., 2009;Joensen et al., 2015;Malberg Tetzschner et al., 2020), and SeroTypeFinder (2.0) for serotyping (Joensen et al., 2015).Ribosomal multilocus sequence typing (rMLST, last updated 13 September 2022) at the Public databases for molecular typing and microbial genome diversity (PubMLST) was used for species identification (Jolley et al., 2012).The presence of chromosomal mutations was evaluated based on the criteria that one single chromosomal mutation in the gyrA gene confers low-level resistance to quinolones, and several mutations in DNA gyrase genes (gyrA and gyrB) and topoisomerase IV genes (parC and parE) are required to increase the level of quinolone resistance in Enterobacteriaceae (Correia et al., 2017).ARG or plasmid replicons were considered present if length coverage and identity to the reference sequence were 100% and ≥ 95%, respectively.Virulence genes were considered present if length coverage and identity to the reference sequence were 100% and ≥ 90%, respectively.

Statistical analysis
Quantitative E. coli results were divided into two groups based on the level of fecal contamination (E. coli ≤ 230 MPN/100 g and E. coli > 230 MPN/100 g).Statistical analysis was performed with Fisher's test (Stata 16.1 ® ), and values of p < 0.05 were considered statistically significant.
To study the seasonality of ESBL E. coli in clams, samples were categorized as summer (21st of June to 22nd of September)-autumn (23rd of September to 20th of December), and winter (21st of December to 20th of March)-spring (21st of March to 20th of June), according to the season of collection.
Antimicrobial resistance phenotype of ESBL-and ESBL-/AmpC-producing Escherichia coli and other ESBL-producing Escherichia spp.
ESBLs were encoded in 3 out of 11 E. coli isolates from clams by the bla CTX-M-15 gene; other common CTX-M variants were bla CTX-M-14 and bla CTX-M-1 (2 out of 11 isolates each), whereas bla CTX-M27 and bla CTX- M55 were each present in 1 of the 11 isolates (Table 2).Other ESBLproducing genes found in the E. coli isolates from clams were the bla SHV12 gene (1 out of 11 isolates) and one bla TEM gene that had a sequence identity of 99.8% to bla TEM-106 (859/861 bp) and bla TEM-126 (859/861 bp), respectively .The combination of an ESBL (CTX-M-14 or a TEM enzyme with a gene sequence identity of 99.8% to bla TEM-106 and bla TEM-126 ) with a plasmidic AmpC β-lactamase (CMY-2) was found in two out of four of the isolates (Table 2).The remaining AmpCproducing E. coli isolates had mutations in the AmpC promoter (p.L9R, p.R8C, and p.R11Q; g.-28G > A, Table 2).The identified fluoroquinolone resistance genes were aac(6′)-Ib-cr and qnrS1 in three and two isolates, respectively.Moreover, seven isolates had at least one point mutation known to mediate quinolone resistances in the chromosomal gyrA, while three isolates also possessed at least one additional mutation in the genes parC/parE.Tetracycline resistance genes tetA, tetB, and tetM were found in 6, 1, and 1 of the 11 isolates, respectively.By the MOB-suite analysis of assembled genomes, 9 of the 11 contigs harboring ESBL-encoding genes and the two contigs harboring AmpC-encoding genes were classified as plasmid-derived sequences (Table 3; Supplementary Appendix 5).The presence of mobile elements and other antibiotic resistance genes in the same contigs harboring ESBL-/AmpC-encoding genes was also investigated and is reported in Table 3.

Genomic analysis of the ESBL-producing
Escherichia spp.
Among the 13 sequenced Escherichia spp., two isolates of ESBL producers were identified by rMLST (100% with 53 exact matches each) as E. marmotae and E. ruysiae and assigned by phylogroup analysis to clades V and III, respectively.The E. marmotae strain harbored the astA gene for the heat-stable enterotoxin 1 and other virulence genes (Table 4), whereas the ESBL was encoded by the bla CTX-M-1 (Table 2).The latter was predicted to be located in a chromosome-derived sequence (Table 3, AN5).A new ST (14425) was identified in the E. marmotae strain (Table 4).The E. ruysiae strain was assigned to ST 3568 by MLST analysis.Several putative virulence genes were predicted from the genome sequence of the E. ruysiae strain, including the enterotoxin astA gene (Table 4).The ESBL was encoded by the bla CTX-M-15 gene (Table 2), which was located in a contig classified as a plasmid-derived sequence (Table 3, AN9).
Occurrence of ESBL-and ESBL-/ AmpC-producing Escherichia coli isolates and seasonality at sampling points Over the studied period, the prevalence of ESBL-(including ESBL-and AmpC-)-producing E. coli isolates (Figure 1) ranged from 0% (no isolation) at 21 (75%) of the 28 sampling sites to 27% (3 out of 11 samples) in an area that was not suitable for the direct human consumption of bivalve mollusks.
The occurrence of ESBL-/ESBL-and AmpC-producing E. coli STs at the sampling sites is reported in Table 6.Variability of STs and resistance to antimicrobials were found in strains isolated from clams sampled over time.One site (sampling point 2) had the greatest variability, with isolates of different STs harboring different ESBLproducing genes (Table 6).
The ETEC and two of the three ExPEC strains were isolated from samples with E. coli contamination levels >230 MPN/100 g.The majority (7 out of 11, 63%) of ESBL-and ESBL/AmpC-producing E. coli isolates were recovered from samples of areas classified as not suitable for direct human consumption.

Discussion
Gut bacteria from humans or terrestrial animals, including those resistant to antimicrobials, can enter aquatic environments through various routes.Thus, bivalves, which are filter-feeder animals, may act as possible carriers of bacteria derived from fecal pollution.Antimicrobial resistance monitoring programs in the EU focus on terrestrial animals, while studies on the presence of ESBL-producing bacteria in seafood products are limited.
We report here on the presence of ESBL-producing E. coli strains in bivalves collected in Italy between 2018 and 2019 from sampling points of production areas of the Central Adriatic Sea.To our knowledge, this is one of the few studies performed worldwide to investigate the occurrence of ESBL-producing E. coli strains in bivalves from production areas (Rees et al., 2015;Bueris et al., 2022).Moreover, this study investigated the relationship between levels of E. coli, the bacterial indicator of fecal contamination, and the presence of E. coli strains with an ESBL phenotype in bivalves.Overall, E. coli producers of ESBL were recovered from 3% of clam samples collected at several sampling sites in the studied period; of these, 2% and 1% harbored ESBL-and ESBL-/AmpC-producing E. coli isolates, respectively.Previous studies in bivalve production areas were conducted in other countries and were limited in the sample collection period (Rees et al., 2015;Bueris et al., 2022).In a study conducted in Canada (Rees et al., 2015), with a study period of 2 months, ESBL-producing E. coli isolates were not recovered from oysters harvested from sampling points of an open oyster fishery and a restricted zone for bivalves (Rees et al., 2015).In another study from Brazil, ceftriaxone-resistant E. coli isolates with an ESBL phenotype were cultured from edible bivalves (oysters and brown mussels) collected from three locations of a polluted area on the South American Atlantic coast (Bueris et al., 2022).Other studies reporting on the prevalence of ESBL-producing E. coli in bivalves were performed at retail in European countries (Boss et al., 2016;Vu et al., 2018).In the latter studies, E. coli producers of ESBL or AmpC were not recovered from retail sampled oysters (n = 10) in Switzerland (Boss et al., 2016), whereas in Germany, ESBLproducing Enterobacteriaceae were isolated in 20% of bivalve samples collected at retail level in Berlin, with the bivalves originating from several European countries, including Italy (Vu et al., 2018).Another study from North Africa reported a prevalence of 1.6% of ESBLproducing Enterobacterales (mostly E. coli) among 215 analyzed pools of 5 clams (Ruditapes decussatus), which were sampled in unrelated markets in four different regions of Tunisia (Sola et al., 2022).
All ESBL-or ESBL-/AmpC-producing E. coli isolated in this study from bivalves were resistant to cefotaxime, a third-generation cephalosporin, and, to a lesser extent, to ceftazidime (27%) and fourth-generation cephalosporin cefepime (18%).Moreover, resistance to fluoroquinolones, another antibiotic class recognized as the highest priority critically important antimicrobials (HPCIAs) in human medicine (World Health Organization, 2019), was recorded in 27% of the ESBL-or ESBL-/AmpC-producing E. coli isolates.
Other studies have reported the proportion of third-or fourthgeneration cephalosporin-resistant isolates of E. coli relative to the total number of isolates recovered from bivalve samples from production areas (Vignaroli et al., 2016;Grevskott et al., 2017;Miotto et al., 2019;Jeong et al., 2021).However, a screening method of a selective medium with an antibiotic to isolate third-generation cephalosporin-resistant E. coli strains was not applied in these studies; therefore, the prevalence of ESBL in samples collected at harvesting areas was not determined.In a systematic review and meta-analysis on antimicrobial resistance in marine bivalves from our group, resistance to third/fourth/fifth-generation cephalosporins and fluoroquinolones was recorded in approximately 10% of E. coli isolates, while resistance to carbapenems was not reported in E. coli strains from bivalves (Albini et al., 2022).Accordingly, resistance to carbapenems (ertapenem, imipenem, and meropenem) was not found in the ESBL-and ESBL-/AmpC-producing E. coli strains we isolated from bivalves.
CTX-M-type enzymes are the most common global ESBL in E. coli; among these, CTX-M-15 is the most frequent, followed by CTX-M-14 (Peirano and Pitout, 2019).Noteworthy, a recent study performed in Italy has reported CTX-M types as prevalent in both ExPEC human and animal isolates, and among these, the CTX-M-15 enzyme is largely predominant in human isolates and in a consistent percentage of the isolates from different animal species (Giufrè et al., 2021).According to the same study, the second most common CTX-M enzyme in Italian isolates was CTX-M-27 in humans and CTX-M-1 in animals.Previous studies performed on bivalve mollusks  reported that the ESBL phenotype in E. coli was predominantly due to the presence of bla CTX-M genes (Vu et al., 2018;Bueris et al., 2022;Sola et al., 2022), with the most frequent one detected in isolates from clams sampled at retail in Tunisia being bla CTX-M-15, followed by bla CTX- M-1 and bla CTX-M-14 (Sola et al., 2022) .In agreement with these studies, our analysis of ESBL genes revealed that the most prevalent gene found in E. coli isolates from bivalves of the Central Adriatic Sea in the studied period was the bla CTX-M-15 gene, followed by bla CTX-M-14 and bla CTX-M-1 , whereas bla CTX-M27 and bla CTX-M55 were less represented.Interestingly, other ESBL-producing genes were found in a minor percentage of the E. coli isolates from clams (bla SHV12 and bla TEM genes).
Additionally, 18% of ESBL-producing E. coli isolates from clams possessed the bla CMY-2 gene for the plasmidic class C β-lactamases.Of the latter, only the CMY-2 enzyme was identified in clams.Analysis of the chromosomal or plasmid location evidenced that the majority of ESBL-/AmpC-encoding genes were harbored in sequences classified as plasmid-derived by the MOB suite.The genomic location of ESBL-/ AmpC-encoding genes on plasmids in E. coli from the marine environment is worrying from a public health perspective as plasmids play an important role in the horizontal transfer of resistance genes.Moreover, mobile genetic elements such as insertion sequences (ISEcp1, synonym of ISEc9, and IS6100) and transposons (Tn2) were also found in the same contigs of the bla CTX-M (in five isolates) and the bla CMY-2 genes (in two isolates).Phylogenetic group analysis showed that phylogroup A was the most prevalent (45%) in ESBL-and ESBL/AmpC-producing E. coli isolates, followed by other phylogroups that include isolates associated with human extra-intestinal infections (D, F, and B2).Among the sequenced ESBL-and ESBL-/AmpC-producing E. coli isolates, a high genomic diversity (10 different STs in 11 isolates) was observed, yet some clinically important STs were identified (ST131, ST38, ST10, ST69, ST457, and ST398).Of these, E. coli ST131, ST69, and ST457 showed ExPEC status and carried bla CTX-M variants.The E. coli ST131 isolate from clams was MDR and genomic features of clade C of the pandemic E. coli ST131 lineage (Denamur et al., 2021), which is the most prevalent ExPEC clonal group isolated in extra-intestinal infections in humans (Nicolas-Chanoine et al., 2014).The ST131 clone has previously been reported in water environments (Colomer-Lluch et al., 2013;Nicolas-Chanoine et al., 2014;Jørgensen et al., 2017), influent (Nicolas-Chanoine et al., 2014;Jørgensen et al., 2017), effluent treated wastewaters of water treatment plants (Zhi et al., 2020;Sekizuka et al., 2022), and bivalves (Vignaroli et al., 2016;Bueris et al., 2022;Sola et al., 2022).The ST457 is a broad host range, globally disseminated diverse E. coli lineage that can cause human extra-intestinal disease (Nesporova et al., 2021).A study focusing on ST457 evidenced that Australian human clinical and silver gull strains were closely related, suggesting that ST457 was an emerging ESBL lineage with reservoirs in wildlife and food-producing animals (Nesporova et al., 2021).ST38, ST10, and ST69 found in clams in a study conducted in Italy were reported to be more frequently detected in both human and animal isolates (Giufrè et al., 2021).Moreover, ST10 was one of the most represented STs among isolates from cattle and pigs, while ST 69 was also largely represented in isolates from pigs (Giufrè et al., 2021).
In this study, other ESBL-producing Escherichia spp., E. marmotae (Liu et al., 2015) and E. ruysiae (Van der Putten et al., 2021), were identified.Phylogroup analysis had previously assigned these species to Escherichia cryptic clades V and III, respectively.Previous studies on Escherichia cryptic clades have speculated that these may represent environmentally adapted Escherichia lineages that may be more abundant outside the gastrointestinal tract of the host (Walk et al., 2009;Ingle et al., 2011).Cryptic lineages of Escherichia were unlikely to be detected in human fecal samples and were more abundant in animal feces, ranging from 3% to 8% in non-human mammals to 8-28% in birds (Clermont et al., 2011).E. marmotae isolated from the feces of wild rodents (Marmota himalayana) has been reported as a potential invasive pathogen (Liu et al., 2019).In addition, humaninvasive infections caused by E. marmotae have recently been described, and this Escherichia species has only recently been identified as a new common pathogen because it can be easily misidentified as E. coli in routine diagnostic laboratories (Sivertsen et al., 2022).Like animal isolates from wild rodents (Liu et al., 2019) and clinical isolates from human-invasive infections (Sivertsen et al., 2022), the ESBL-producing isolate of E. marmotae from clams harbored the enterotoxin-encoding gene astA.Moreover, the E. coli in silico serotyper evidenced that it contained the fliC-H56 flagellar antigen gene, as did the human clinical isolates (Sivertsen et al., 2022).Available data on E. marmotae isolates from different sources showed infrequent occurrences of antimicrobial resistance (Sivertsen et al., 2022); isolates from reported human-invasive infections were phenotypically susceptible to tested antimicrobials, and resistance genes were not identified in their genomes (Sivertsen et al., 2022).Differently from these findings, the E. marmotae strain from clams was an ESBL-producing isolate, phenotypically resistant to cefotaxime, which possessed the bla CTX-M-1 gene.
Escherichia ruysiae sp.nov.was proposed by van der Putten et al. ( 2021) as a novel species, encompassing Escherichia cryptic clades III and IV.The strain of E. ruysiae was isolated from the fecal material of an international traveler, harbored the bla CTX-M-14 gene, and was of cryptic clade IV (van der Putten et al., 2021).Differently from the human isolate, the isolate of ESBL-producing E. ruysiae from clams in this study harbored the bla CTX-M-15 gene, the enterotoxin-encoding gene astA, and belonged to cryptic clade III.
In the EU, the microbiological safety of bivalve mollusks is based on the classification and monitoring of production areas.Assessing the sources and types of fecal contamination in the vicinity of the areas and how these affect mollusk production areas, combined with the quantitative monitoring of the fecal indicator organism E. coli, is critical to providing an estimate of the risk of contamination of an area by microbial pathogens.
To the best of our knowledge, this study was the first to investigate the correlation between the presence of ESBL-producing E. coli and the bacterial indicator of fecal contamination of E. coli in bivalve mollusks.Over the studied period, ESBL (including ESBL-/AmpC-)producing E. coli were not isolated in clams from most (75%) of the studied sampling sites (n = 28), while the highest frequency of isolation (27%) was observed in an area requiring post-harvest treatment to reduce microbiological contamination before human consumption.A significant correlation was found between the indicator of fecal contamination by E. coli above 230 MPN/100 g and the presence of ESBL-producing E. coli.Thus, this study provided evidence of E. coli in molluscan shellfish as an index of the potential presence of ESBLproducing E. coli isolates, which are bacteria resistant to a critically important class of highest-priority antimicrobials.
Considering seasonality, ESBL-producing E. coli were not isolated in clam samples collected in summer but in other seasons (8% in spring, 5% in winter, and 4% in autumn), although these differences 10.3389/fmicb.2023.1219008Frontiers in Microbiology 11 frontiersin.orgin prevalence were not found to be significant.One area had the greatest variability in STs and ESBL-producing genes, whereas the bla CTX-M-1 variant was present in one specific area.These findings can be explained by the variability in the presence and type of pollution sources affecting the microbiological contamination of bivalve mollusk areas and by environmental effects (e.g., rainfalls, winds, and tidal currents) on pollution sources.A large number of uncertainties exist with respect to the sources and transmission routes of antimicrobial-resistant bacteria and antimicrobial-resistance genes in food-producing environments.
In future studies, we will investigate the presence in bivalves of ESBL-producing E. coli, other antimicrobial-resistant bacteria, and antimicrobial-resistance genes of importance in human health in relation to the sources and types of pollution, seasonal variations, and climatic factors that influence production areas.
conclusion, this study presents novel observations on the prevalence, seasonality, genomic, and phenotypic characteristics of ESBL-producing E. coli isolates in bivalves from production areas.ESBL-producing E. coli isolates were significantly more likely to be present among clam samples with higher levels of E. coli contamination (> 230 MPN/100 g) than among those with lower levels (≤ 230 MPN/100 g).Furthermore, potentially pathogenic ESBLproducing E. coli strains (ETEC and ExPEC) were isolated mainly in samples with E. coli contamination levels above 230 MPN/100 g.These findings provided evidence in support of E. coli as an index organism for the presence of ESBL-producing E. coli.

TABLE 1
Distribution of MIC (minimum inhibitory concentration) values among the 13 ESBL-or ESBL-/AmpC-producing Escherichia coli (11 isolates) and ESBL-producing E. marmotae (1 isolate) and E. ruysiae (1 isolate) from clams.Percentages are shown in brackets.The shaded areas show the range of values tested for each antibiotic.

TABLE 2
Phenotypic antimicrobial resistance (AMR) profile, resistance genes, and gene mutations for AMR of ESBL-and ESBL-/AmpC-producing Escherichia spp.isolates from clams.

bla CTX-M-15 bla TEM-35 gyrA
Predicted genomic location (plasmid or chromosome) of ESBL-and AmpC-encoding genes using the MOB-Suite and presence of mobile genetic elements (MGEs) and other antimicrobial resistance genes in the same contig.

cvaC eilA etsC fyuA gad hlyF iroN irp2 iss iutA kpsE lpfA terC kpsMII_K5 mchF
Plasmid replicons were considered present if length coverage and identity to the reference sequence were 100% and ≥ 95%, respectively.Virulence genes were reported if length coverage and identity to the reference sequence were 100% and ≥ 90%, respectively.Genes indicated in black were identified with 100% coverage length and identity ≥ 95% for serotype.Genes indicated in black and bold were identified with 100% length coverage and 100% sequence identity.ExPEC, Extra-intestinal pathogenic Escherichia coli; ETEC, Enterotoxigenic E. coli.

TABLE 5
Phylogroups of 11 isolates of E. coli producers of ESBL or ESBL/AmpC from clam samples.

TABLE 6
ESBL-(ESBL-/ESBL-and AmpC-) producing E. coli recovered from 308 samples of clams collected from harvesting areas of the Central Adriatic between July 2018 and November 2019.