Systematic Survey of Vibrio spp. and Salmonella spp. in Bivalve Shellfish in Apulia Region (Italy): Prevalence and Antimicrobial Resistance

Abstract

The emergence of antimicrobial resistance (AMR) is increasingly common across the globe and aquatic ecosystems could be considered a reservoir of antibiotic-resistant bacteria. This study aimed to determine prevalence and antibiotic susceptibility of the potential pathogenic bacteria Salmonella spp. and Vibrio spp. in bivalve molluscs intended for human consumption, collected over a period of 19 months along the northern coast of Apulia region. The AMR profile was also determined in non-pathogenic Vibrio species, common natural inhabitants of seawater and a useful indicator for the surveillance of AMR in the environment. The current study presents data on the AMR of 5 Salmonella and 126 Vibrio isolates by broth microdilution MIC. Multidrug resistance (MDR) was observed in one S. Typhimurium strain towards sulfamethoxazole, trimethoprim, tetracycline, gentamicin, and ampicillin and in 41.3% of the Vibrio strains, mostly towards sulphonamides, penicillin, and cephems. All Vibrio isolates were sensitive to azithromycin, chloramphenicol, tetracycline, amoxicillin/clavulanic acid, gentamicin, streptomycin, amikacin, and levofloxacin. The AMR phenomenon in the investigated area is not highly worrying but not entirely negligible; therefore, in-depth continuous monitoring is suggested. Results concerning the antibiotic agents without available specific clinical breakpoints could be useful to upgrade the MIC distribution for Vibrio spp. but, also, the establishment of interpretative criteria for environmental species is necessary to obtain a more complete view of this issue.

1. Introduction

The global increasingly spreading antimicrobial resistance (AMR) is such an alarming current issue that many scientific activities have been implemented worldwide to analyse the phenomenon and to conceive effective contrast measures. Several bacterial species are involved in this phenomenon and the zoonotic ones are clearly of greater concern. Resistant bacteria can affect humans directly through consumption of contaminated food or indirectly by transferring mobile genetic elements (e.g., plasmids, transposable elements, super-integron, and integrating conjugative elements genes) for antibiotic resistance to human pathogens [1,2]. Antibiotic-resistant bacteria (ARB) have emerged in the marine environment as a consequence of the excessive use of antibiotics in human, agriculture, and aquaculture systems during the past few decades [3]. Despite the use of antimicrobials in aquaculture being considered a leading cause of development of ARB [1], only five active substances are approved and registered in Italy: amoxicillin, flumequine, oxytetracycline, chlortetracycline, and sulfadiazine-trimethoprim.

Vibrio species could represent an important indicator of the presence of antibiotic resistance in the marine and estuarine ecosystems, as they are natural inhabitants of coastal waters. Approximately 12 Vibrio spp. can cause infections in humans via oral route by ingestion of contaminated water or raw or undercooked contaminated seafood (in particular, Vibrio parahaemolyticus and V. cholerae, but also V. vulnificus, V. fluvialis, V. mimicus, and V. hollisae) or through skin wound exposure to contaminated water, which could result in secondary septicaemia (especially V. vulnificus and, to a lesser extent, V. alginolyticus) [4]. Shellfish can easily harbour pathogenic microorganisms because of their filter-feeding behaviour and Vibrio have been proposed as the most common bacteria responsible for food poisoning after shellfish consumption [2]. V. cholerae is the causative agent of cholera, a severe worldwide disease affecting mainly children <5 years of age and with frequent cases of deaths. V. cholerae strains are classified into more than 200 serogroups on the basis of the chemical composition of the O antigen of lipopolysaccharide (LPS): strains belonging to serogroup O1 and O139 Bengal are responsible for the vast majority of cholera cases, while non-O1 and non-O139 strains can cause sporadic gastrointestinal and extraintestinal infections [4]. The detection of virulence-associated factors is useful to discriminate between pathogenic and non-pathogenic strains: the stn/sto genes encoding the heat-stable enterotoxin are one of the main factors associated with enteropathogenicity in V. cholerae [5]. V. parahaemolyticus is the most common pathogen causing seafood-borne illnesses in many countries by eating raw or undercooked shellfish, and the strains harbouring the tdh (encoding thermostable direct hemolysin) and trh (encoding tdh-related hemolysin) genes are pathogenic in humans [6,7]; moreover, it is able to form biofilm on mussel surface [8]. V. vulnificus is usually considered an opportunistic pathogen affecting mostly immunocompromised individuals, but it is a highly fatal human pathogen, responsible for >95% of seafood-related deaths in the United States. Vibrio alginolyticus causes mainly superficial wound and ear infections, generally cured by an appropriate antibiotic therapy, although rare cases of septicaemia may also occur [4]. However, this species is widespread in the marine environment; therefore, it may represent an essential indicator of environmental antibiotic resistance more than others. V. harveyi is one of the most common species infecting farmed aquatic animals [9,10]; however, sporadic cases of human infections, especially wound infections, have been reported in recent years, also, in the Mediterranean Sea [11,12]. Vibrio spp. is one of the zoonotic agents listed in Annex I to Directive 2003/99/EC [13] to be monitored according to the epidemiological situation. In Italy, vibriosis outbreaks occurred and occur nowadays, as recently reported in the European Union One Health Zoonoses Report 2019 [14], and, also, the Apulia region was historically involved in important cholera epidemics in the past years [15,16]. Therefore, research activities and monitoring of Vibrio spp. are of great interest in the Apulian territory.

Salmonella spp. is another important zoonotic pathogen frequently found in bivalve molluscs and for which the occurrence of multidrug resistance is widely reported also in the areas investigated in the present work [17]. Moreover, Salmonella spp. is one of the bacteria for which the monitoring and reporting of AMR is mandatory and specific guidelines on the antimicrobial susceptibility testing (AST) have been laid down [13,18,19].

In 2015, Member States adopted the Global Action Plan on AMR and Italy issued the National Action Plan on AMR 2022–2025 in September 2022. Following the principles announced in the Global Action Plan on AMR, it is necessary to have a consistent, standardised approach to collecting and reporting resistance data, so that trends and patterns of resistance evaluated at national, regional, and local level should guide targeted policy decisions to contrast the phenomenon [20].

Starting from these assumptions, the present study aimed primarily to investigate the presence of potential pathogenic bacteria belonging to the genera Salmonella and Vibrio in edible bivalve molluscs collected systematically along the northern Apulian coast for subsequent evaluation of the AMR profile of isolates, since antimicrobial resistance surveillance in the marine environment is a crucial aspect to implement effective local antibiotic reduction programs.

2. Materials and Methods

2.1. Sampling

Overall, 296 shellfish samples were collected from March 2021 to October 2022 along the northern Apulian coast (provinces of Foggia and Barletta-Andria-Trani). Most of them (263/296; 88.85%) derived from the implementation of the official classification programme of bivalve mollusc production and harvesting areas with regard to the Commission Implementing Regulation 2019/627 [21]. The geographical distribution of sampling points was established after a sanitary survey in order to choose the location at highest risk of faecal pollution and ensure that analytical results were representative of the area. At least 12 samples were taken from each sampling point over at least a 6-month period, as recommended in the community guide to the principles of good practice for the microbiological classification and monitoring of bivalve mollusc production and relaying areas with regard to Implementing Regulation 2019/627 [22]. The interval between two sampling occasions was approximatively 2 weeks, depending on the weather conditions and the availability of a commercial-size product. It is important to highlight that search for Salmonella is mandatory according to the European legislation [23]; otherwise, Vibrio detection was carried out only for research purposes.

The remaining samples (33/296; 11.15%) consisted of live bivalve molluscs originated from the same investigated areas and sampled for official control in accordance with the Integrated Regional Control Plan of Apulia region; they were collected in purification and dispatch centres or at retail and also tested for the aim of this study.

As regards the specimens collected for the official classification purpose, they consisted of oysters (Crassostrea gigas), mussels (Mytilus galloprovinciallis and Modiolus barbatus), clams (Venus gallina/Chamelea gallina), cockles (Acanthocardia tuberculata), and Japanese carpet shells (Ruditapes philippinarum). They were collected from fixed sampling stations, transported to the laboratory on the same day at temperatures between 4 and 10 °C and processed within 24 h of arrival. Environmental parameters, such as water temperature (°C) and pH, were measured on site during the sampling. The sampling stations are illustrated in Figure 1.

Detailed information about mollusc species and sampling points are reported in

Table 1. Number and type of samples for each sampling point.

Sampling Station	Area	Shellfish Species	n. Samples
1	Northern coast of Gargano	Mytilus galloprovincialis	12
2	Northern coast of Gargano	Mytilus galloprovincialis	13
3	Northern coast of Gargano	Crassostrea gigas	13
4	Northern coast of Gargano	Mytilus galloprovincialis	13
5	Northern coast of Gargano	Mytilus galloprovincialis	15
6	Varano lake	Ruditapes philippinarum	6
		Mytilus galloprovincialis	14
7	Varano lake	Crassostrea gigas	14
9	Varano lake	Mytilus galloprovincialis	13
10	Varano lake	Mytilus galloprovincialis	14
11	Southern coast of Gargano	Crassostrea gigas	6
		Mytilus galloprovincialis	11
12	Southern coast of Gargano	Mytilus galloprovincialis	14
13	Southern coast of Gargano	Modiolus barbatus	7
15	Southern coast of Gargano	Acanthocardia tuberculata	8
		Modiolus barbatus	1
16	Southern coast of Gargano	Acanthocardia tuberculata	8
17	Southern coast of Gargano	Acanthocardia tuberculata	8
18	Coastline of BAT 1 Province	Venus gallina/Chamelea gallina	17
19	Coastline of BAT 1 Province	Venus gallina/Chamelea gallina	15
20	Coastline of BAT 1 Province	Venus gallina/Chamelea gallina	17
21	Coastline of BAT 1 Province	Venus gallina/Chamelea gallina	12
22	Coastline of BAT 1 Province	Venus gallina/Chamelea gallina	12

¹ BAT: Barletta-Andria-Trani.

.

2.2. Vibrio Detection and Antimicrobial Susceptibility Testing

Vibrio spp. detection was performed according to the standard ISO 21872–1:2017 [24] by preparing two enrichment broths, which were incubated at 37 ± 1 °C and 41.5 ± 1 °C, respectively, to enhance the recovery of most Vibrio species. All the presumptive Vibrio spp. isolates were identified at species level based on the API^® ID 20E and 20NE systems (BioMérieux, Nürtingen, Germany), the halotolerance test with various concentrations of NaCl (0, 6 and 10%), and the MALDI-TOF MS (Bruker Daltonics, Bremen, Germany) procedure by the direct transfer method, as previously described [25]. Moreover, the isolates recognised as V. parahaemolyticus, V. cholerae, and V. vulnificus were confirmed by the conventional PCR method described in Annex C to ISO 21872–1:2017, which considers, also, the detection of V. parahaemolyticus virulence genes (tdh and trh).

The broth microdilution MIC method was carried out on colonies grown on nonselective medium saline nutrient agar with 1% NaCl (SNA) incubated at 36 ± 1 °C overnight, using Sensititre™ Gram Negative GN4F^® AST Plate and Sensititre™ NARMS^® Gram Negative CMV4AGNF AST Plate (Thermofisher Scientific, Paisley, UK). Sensititre™ Gram Negative GN4F^® AST Plate contained the following antimicrobials: amikacin (AMI 8–32 µg/mL), ampicillin (AMP 8–16 µg/mL), ampicillin/sulbactam (A/S2 4/2–16/8 µg/mL), aztreonam (AZT 1–16 µg/mL), cefazolin (FAZ 1–16 µg/mL), cefepime (FEP 4–32 µg/mL), ceftazidime (TAZ 1–16 µg/mL), ceftriaxone (AXO 0.5–32 µg/mL), ciprofloxacin (CIP 0.5–2 µg/mL), doripenem (DOR 0.5–4 µg/mL), ertapenem (ETP 0.25–8 µg/mL), gentamicin (GEN 2–8 µg/mL), imipenem (IMI 0.5–8 µg/mL), levofloxacin (LEVO 1–8 µg/mL), meropenem (MERO 0.5–8 µg/mL), minocycline (MIN 1–8 µg/mL), nitrofurantoin (NIT 32–64 µg/mL), piperacillin (PIP 16–64 µg/mL), piperacillin/tazobactam constant 4 (P/T4 8/4–128/4 µg/mL), tetracycline (TET 4–8 µg/mL), ticarcillin/clavulanic acid constant 2 (TIM2 8/2–64/2 µg/mL), tigecycline (TGC 1–8 µg/mL), tobramycin (TOB 2–8 µg/mL), and trimethoprim/sulfamethoxazole (SXT 2/38–4/76 µg/mL). Sensititre™ NARMS^® Gram Negative CMV4AGNF AST contained the following antimicrobials: amoxicillin/clavulanic acid 2:1 ratio (AUG2 1/0.5–32/16 µg/mL), ampicillin (AMP 1–32 µg/mL), azithromycin (AZI 0.25–32 µg/mL), cefoxitin (FOX 0.5–32 µg/mL), ceftriaxone (AXO 0.25–64 µg/mL), chloramphenicol (CHL 2–32 µg/mL), ciprofloxacin (CIP 0.015–4 µg/mL), gentamicin (GEN 0.25–16 µg/mL), meropenem (MERO 0.06–4 µg/mL), nalidixic acid (NAL 0.5–32 µg/mL), streptomycin (STR 2–64 µg/mL), sulfisoxazole (FIS 16–256 µg/mL), tetracycline (TET 4–32 µg/mL), and trimethoprim/sulfamethoxazole (SXT 0.12/2.38–4/76 µg/mL). The antibiotics used in the study were chosen in accordance with Clinical and Laboratory Standards Institute (CLSI) recommendations, including those for the treatment of Vibrio infections. It was decided to use these Sensititre™ plates in order to include most of the antibiotic compounds listed in CLSI M45 guidelines [26], even if some molecules were present in both antimicrobial plates but at different concentrations.

The bacterial inoculum was prepared by dissolving a fresh pure colony in 2.5% NaCl solution using a sterile cotton swab until it achieved the turbidity of the 0.5 McFarland standard. Then, 0.1 mL of this suspension was added to 9.9 mL of cation-adjusted Mueller–Hinton broth (Becton Dickinson, Milan, Italy), the antimicrobial plate wells were inoculated with 50 µL of this suspension, and the plate was incubated aerobically at 36 ± 1 °C for 24 h. Escherichia coli ATCC^® 25922 was used as quality control in each batch. The results were interpreted according to CLSI clinical breakpoints specific for Vibrio spp. No CLSI breakpoints were available for the following agents: ceftriaxone, nalidixic acid, streptomycin, tigecycline, ticarcillin/clavulanic acid, nitrofurantoin, doripenem, minocycline, ertapenem, tobramycin, and aztreonam. Hence, results of broth microdilution assays referring to the above-mentioned molecules are reported in the present study without any interpretative criteria. Only for streptomycin, the breakpoints described by other authors were used [27,28].

2.3. Salmonella Detection and Antimicrobial Susceptibility Testing

Salmonella spp. detection and serotyping were performed according to the ISO 6579–1:2017/Amd 1:2020 [29] and ISO/TR 6579–3:2014 [30], respectively. Antimicrobial susceptibility testing was performed on Salmonella strains by broth microdilution using Sensititre™ EUVSEC3^® (Termofisher Scientific, Paisley, UK), which contained the compounds specified in the Commission Implementing Decision (EU) 2020/1729: ampicillin (1–32 µg/mL), azithromycin (2–64 µg/mL), amikacin (4–128 µg/mL), gentamicin (0.5–16 µg/mL), tigecycline (0.25–8 µg/mL), ceftazidime (0.25–8 µg/mL), cefotaxime (0.25–4 µg/mL), colistin (1–16 µg/mL), nalidixic acid (4–64 µg/mL), tetracycline (2–32 µg/mL), trimethoprim (0.25–16 µg/mL), sulfamethoxazole (8–512 µg/mL), chloramphenicol (8–64 µg/mL), meropenem (0.03–16 µg/mL), and ciprofloxacin (0.015–8 µg/mL). The quality control of the batch was performed with E. coli ATCC^® 25922. The bacterial inoculum was prepared as described for Vibrio spp. but with 0.9% NaCl solution.

The epidemiological cut-off value (ECOFF) indicated in the Commission Implementing Decision (EU) 2020/1729 [18] was used as the interpretation threshold of AMR, except for colistin and tigecycline, for which the values defined for Enterobacteriales in the EUCAST clinical breakpoint table [31] were chosen, whereas the MIC breakpoints stated in CLSI document M100 [32] were used for azithromicyn and sulfamethoxazole (MIC values for sulphonamides). Multidrug resistance (MDR) was defined as nonsusceptibility to at least one agent in three or more antimicrobial categories [33].

Finally, the multiple antibiotic resistance (MAR) index was calculated as the ratio between the number of antibiotics to which a strain was resistant to and the total number of antibiotics used [34].

3. Results

3.1. Salmonella Isolates and Their Antimicrobial Resistance Profile

Among the 296 live bivalve molluscs analysed, only 1.7% of them were contaminated with Salmonella spp. Overall, five strains were isolated and tested for susceptibility to antimicrobial agents: three strains from clams (S. enterica subsp. enterica Kasenyi, S. enterica subsp. enterica Typhimurium, and S. bongori 48:z35:-) and two strains from mussels (S. enterica subsp. enterica Fischerhütte and S. enterica subsp. enterica Typhimurium). All isolates were sensitive to almost all antibiotics, except for S. Typhimurium isolated from Venus gallina, which showed resistance to sulfamethoxazole, trimethoprim, tetracycline, gentamicin, and ampicillin. Instead, S. Kasenyi and S. bongori 48:z35:- were resistant only to sulfamethoxazole.

3.2. Vibrio Isolates and Their Antimicrobial Resistance Profile

In total, 126 Vibrio strains (mostly V. alginolyticus and V. parahaemolyticus) were detected, with 38.2% (113/296) of shellfish samples contaminated with them. The API 20E system and the MALDI-TOF MS identification results were consistent, whereas the API 20NE displayed disputable and unreliable results. Furthermore, it is well known that the biochemical tests are inadequate for an accurate identification of V. harveyi [12,35] and that profiles for V. harveyi are not included in the bioMérieux database; thus, the API 20E and API 20NE systems do not allow this species to be properly identified via Apiweb™ and the identification of V. harveyi in this study was provided by the MALDI-TOF MS system. Vibrio spp. strains have been detected in shellfish collected from all sampling stations, with the exception of point 13. The temperature range for all isolates was of 9.4–29.9 °C, but isolation of potentially enteropathogenic species (V. cholerae, V. parahaemolyticus, and V. vulnificus) occurred when temperature ranged between 19.5 and 29.9 °C. The correlation between source and Vibrio species is shown in Figure 2. V. alginolyticus was the most prevalent species, found mostly in Mytilus galloprovincialis, followed by Venus gallina/Chamelea gallina.

No pathogenic genes were found in V. cholerae and V. parahaemolyticus isolates, except for one V. parahaemolyticus strain harbouring trh gene. The highest MAR index values for each sampling point are reported in

Table 2. The highest values of multiple antibiotic resistance (MAR) index of Vibrio isolates.

Sampling Station	Matrix	Vibrio Species	Resistance Pattern	MAR Index
1	Mytilus galloprovincialis	V. alginolyticus	MERO, FIS, AMP, P/T4, PIP, FEP	0.193
2	Mytilus galloprovincialis	V. alginolyticus	FIS, SXT, AMP, PIP, FAZ	0.161
3	Crassostrea gigas	V. alginolyticus	FIS, AMP, P/T4, PIP, FAZ, TAZ, FEP	0.225
4	Mytilus galloprovincialis	V. alginolyticus	FIS, AMP, PIP, FAZ	0.129
		V. harveyi	FIS, AMP, PIP, FAZ	0.129
5	Mytilus galloprovincialis	V. alginolyticus	FIS, AMP, PIP, FAZ	0.129
		V. alginolyticus	FIS, AMP, PIP, FAZ	0.129
6	Mytilus galloprovincialis	V. alginolyticus	FIS, AMP, PIP, FAZ	0.129
		V. parahaemolyticus	FIS, AMP, PIP, FAZ	0.129
	Ruditapes philippinarum	V. alginolyticus	FIS, AMP, PIP, FAZ	0.129
		V. alginolyticus	FIS, AMP, PIP, FAZ	0.129
7	Crassostrea gigas	V. harveyi	AMP, PIP	0.064
9	Mytilus galloprovincialis	V. parahaemolyticus	FIS, AMP, PIP, FAZ	0.129
10	Mytilus galloprovincialis	V. alginolyticus	FIS, AMP, PIP, FAZ	0.129
		V. alginolyticus	FIS, AMP, PIP, FAZ	0.129
		V. parahaemolyticus	FIS, AMP, PIP, FAZ	0.129
11	Mytilus galloprovincialis	V. vulnificus	FIS, FAZ	0.064
12	Mytilus galloprovincialis	V. alginolyticus	FIS, AMP, FAZ	0.096
		V. alginolyticus	FIS, AMP, FAZ	0.096
		V. alginolyticus	FIS, AMP, FAZ	0.096
15	Acanthocardia tuberculata	V. alginolyticus	FIS, AMP, FAZ	0.096
16	Acanthocardia tuberculata	V. parahaemolyticus	FIS, AMP, FAZ	0.096
17	Acanthocardia tuberculata	V. alginolyticus	AMP, PIP, FAZ	0.096
18	Venus gallina/ Chamelea gallina	V. alginolyticus	FIS, AMP, PIP, FAZ	0.129
19	Venus gallina/ Chamelea gallina	V. alginolyticus	FIS, AMP, PIP, FAZ	0.129
20	Venus gallina/ Chamelea gallina	V. alginolyticus	FIS, AMP, PIP, FAZ	0.129
21	Venus gallina/ Chamelea gallina	V. alginolyticus	FIS, AMP, PIP, FAZ	0.129
22	Venus gallina/ Chamelea gallina	V. harveyi	FIS, AMP, PIP, FAZ	0.129
		V. parahaemolyticus	FIS, AMP, PIP, FAZ	0.129
		V. alginolyticus	FIS, AMP, PIP, FAZ	0.129

by specifying Vibrio species, resistance pattern, and source.

More specific details for each isolate about origin, resistance pattern, and MAR index are reported in Supplementary Table S1.

None of the 126 Vibrio isolates showed resistance to AZI, CHL, TET, AUG2, GEN, STR, AMI, or LEVO. Despite the interpretive criteria for Vibrio spp. other than V. cholerae provided in CLSI M45 guidelines [26] being uncertain for azithromycin, one V. harveyi (MIC = 4 µg/mL) could be surely stated as nonsusceptible and belonging to at least the intermediate category. Moreover, one V. parahaemolyticus displayed intermediate resistance to TET, one V. parahaemolyticus and five V. alginolyticus to AUG2, while intermediate resistance to STR was observed in all Vibrio species. The intermediate resistance percentages for each Vibrio species are detailed in

Table 3. Intermediate profile to single antimicrobial agents in percentage. The antimicrobial agents for which no intermediate resistance was found are not shown.

Vibrio Species	n	Antimicrobial Agents
		FOX	AZI	TET	AUG2	STR	P/T4	PIP	FAZ	TAZ	A/S2
V. alginolyticus	90	4.4			5.6	3.3	1.1	26.7	30	1.1	6.7
V. parahaemolyticus	17	5.9		5.9	5.9	5.9		11.8
V. harveyi	17		5.9			17.6		29.4	11.8
V. vulnificus	1	100				100
V. cholerae	1	100				100

n: number of isolates.

.

High resistance percentages to FIS (57.1%; 72/126) (MIC > 256 µg/mL), AMP (85.7%; 108/126), and FAZ (56.3%; 71/126) were found among all Vibrio species, except for V. vulnificus and V. cholerae, which were sensitive to AMP. Moreover, 38% and 24.6% of isolates were resistant (MIC > 64 µg/mL) and intermediate resistant (MIC = 32–64 µg/mL) to PIP, respectively. Resistance to FOX was expressed only by one V. alginolyticus isolate, although four V. alginolyticus, one V. vulnificus, one V. cholerae, and one V. parahaemolyticus were intermediate resistant. Resistance to MERO, SXT, PIP, P/T4, IMI, TAZ, A/S2, and FEP was observed only in V. alginolyticus strains with low resistance percentages, as results show in

Table 4. Resistance percentages to single antimicrobial agents. The antimicrobial agents for which no resistance was found are not shown.

Vibrio Species	n	Antimicrobial Agents
		FOX	MERO	FIS	SXT	AMP	P/T4	IMI	PIP	FAZ	TAZ	A/S2	FEP
V. alginolyticus	90	1.1	1.1	61.1	5.6	93.3	2.2	1.1	42.2	55.6	1.1	1.1	3.3
V. parahaemolyticus	17			70.6		76.5			35.3	94.1
V. harveyi	17			17.6		64.7			23.5	17.6
V. vulnificus	1			100						100
V. cholerae	1			100						100

n: number of isolates.

.

Among V. parahaemolyticus strains, 94.1% (16/17) were resistant to FAZ, 76.5% (13/17) to AMP, 70.6% (12/17) to FIS, and 35.3% (6/17) to PIP. Susceptibility to all tested antimicrobials, except for FAZ, was found in the V. parahaemolyticus trh+ strain.

Overall, 41.3% (52/126) of strains displayed MDR, mostly towards sulphonamides, penicillin, and cephems. The most prevalent MDR profile was FIS-AMP-PIP-FAZ, followed by FIS-AMP-FAZ, both of them found primarily in V. alginolyticus and secondarily in V. parahaemolyticus.

The MIC values are shown in Supplementary Table S2, including those relating to the antibiotic agents without available specific CLSI breakpoints.

4. Discussion

Overall, few shellfish samples (5/296) have been found to be contaminated with Salmonella spp. in the current study and only one S. Typhimurium strain showed multidrug resistance. Despite the low prevalence of Salmonella, the detection of pathogenic serovars raises concerns about consumer health, especially given the recent finding of S. Typhimurium strains of human origin, collected from the same region and resistant mainly to ampicillin, tetracycline, azithromycin, and sulfamethoxazole [17], which are the same antimicrobials depicted in our multidrug-resistant strain, except for azithromycin replaced with gentamicin. Giacometti et al. [36] carried out a similar investigation on 102 Salmonella isolates from bivalve molluscs and water samples collected during the official monitoring programme in the area of the province of Ferrara (northwestern area of the Adriatic Sea) between 2001 and 2017. The most common resistances observed by the authors were to streptomycin (58.8%), ampicillin (52%), and tetracycline (45.1%), whereas 44.12% of isolates were MDR. Thus, it seems to suggest that resistance to these antimicrobial classes (penicillin, aminoglycoside, and tetracycline) is widely spread in the Adriatic Sea, in accordance with the AMR profile exhibited by our S. Typhimurium strain. However, any comparison with our results is difficult due to the long period of sampling (7 years), the type of samples (molluscs and water), and the method used to evaluate the AMR (agar disk diffusion). It is easy to understand that the AMR patterns vary according to the geographical origin of the samples; indeed, results concerning 27 Salmonella strains isolated by Lozano-Leon et al. [37] from mussel samples harvested in Galicia are completely different: all isolates showed MDR and were resistant to cefuroxime and cefuroxime/axetil; the majority of them expressed resistance to cefoxitin and gentamicin and some resistance was observed towards ampicillin, amikacin, cephalothin, and tobramycin, although the antimicrobial susceptibility was performed only on some strains.

As regards Vibrio spp., three different identification techniques were performed simultaneously in the present work to obtain a more reliable result, given that differentiation of closely related Vibrio species can be difficult. Although MALDI TOF MS is considered a valid tool, its discriminative power depends on the fullness of the reference library, which mainly contains clinically relevant species; thus, the reference library could be increased by inserting environmental bacterial isolates, especially marine bacteria, to make the MALDI TOF MS result more consistent. The API 20E and MALDI TOF MS systems gave the same results, except for the V. harveyi species because of the lack in the bioMérieux database. Therefore, the MALDI TOF identification of this Vibrio species was considered reliable and might perhaps be supported by biomolecular analysis, such as whole genome sequencing, which offers a great identification accuracy. Vibrio detection occurred during the entire investigation period but it is worth noting that potentially enteropathogenic species were isolated when water temperature was above 19.5 °C. It is well known that occurrence and densities of V. parahaemolyticus in molluscs are positively correlated to water temperatures [38] and outbreaks occur mainly during the warmer months in temperate zones [39]. However, the increase in surface temperature in coastal European seas in recent years has been linked to outbreaks caused by Vibrio cholerae nonO1-nonO139, V. parahaemolyticus, and V. vulnificus in several European countries [40]. Among potentially enteropathogenic species, V. parahaemolyticus was the most frequent one, isolated mainly from Mytilus galloprovincialis, but the only strain harbouring trh gene, hence proving to be harmful to humans, was isolated from Venus gallina o Chamelea gallina. V. cholerae isolation from Acanthocardia tuberculata cockles is a rare finding and certainly noteworthy given the scarcity of studies concerning this shellfish, mostly carried out in Morocco. For example, Boutaib et al. [41] found Salmonella spp. and non-pathogenic Vibrio parahaemolyticus in seven and five Acanthocardia tuberculata samples, respectively, but no results concerned V. cholerae, since this species was not included in the study.

In the current survey, the majority of Vibrio strains displayed resistance to ampicillin, sulfisoxazole, and cefazolin. Resistance to ampicillin was prevalent, as in previous studies [1,2,42,43].

If we compare our results with similar studies previously conducted in Italy, data seem to be quite varied. For instance, in the study carried out by Ottaviani et al. in 2001 [44] on several Vibrio species isolated from seafood, all strains exhibited susceptibility to imipenem, meropenem, chloramphenicol, and tetracycline, except 58% of V. alginolyticus, which were resistant to tetracycline; more than 90% of isolates showed susceptibility to oxolinic acid, cefotaxime, flumequine, doxycycline, and trimethoprim–sulphamethoxazole and more than 80% to nalidixic acid and ciprofloxacin. Moreover, resistance to streptomycin and lincomycin was found in more than 90% of isolates and many strains of V. alginolyticus, V. harveyi, V. vulnificus, and V. parahaemolyticus were resistant to penicillin, carbenicillin, ampicillin, cephalothin, and kanamycin, while V. alginolyticus and V. parahaemolyticus strains were resistant to rifampicin. A later investigation of antimicrobial resistance in V. parahaemolyticus from indigenous bivalves collected from harvesting areas along Italian coasts of the south Adriatic Sea, the central Tyrrhenian Sea, and the central Adriatic Sea reported that all isolates were resistant to ampicillin and amoxicillin but no resistances were observed to chloramphenicol, tetracycline, oxytetracycline, doxycycline, and trimethoprim/sulfamethoxazole [45]. Some strains exhibited resistance to cefotaxime (24.1%), cefalothin (43.7%), cefalexin (67.8%), colistin sulphate (13.8%), erythromycin (20.7%), and streptomycin (32.2%), whereas very low resistance percentages were found towards polymyxin B, nalidixic acid, oxolinic acid, nitrofurantoin, ciprofloxacin, kanamycin, and neomycin. Lopatek et al. [28] investigated the antimicrobial susceptibility of V. parahaemolyticus strains isolated from different species of raw shellfish and marine fish originated from various countries. They found that the majority of the strains were resistant to ampicillin and streptomycin and were recovered mainly from Italian samples; resistance to gentamicin was found in 12.5% of the strains, isolated from Italian, Dutch, and Norwegian samples, whereas only one strain was resistant to ciprofloxacin, isolated from Italian clams.

It should be noted that the majority of the studies on AMR in Vibrio spp. performed the agar disk diffusion method; thus, few works can be properly compared with our results.

In the current study, the V. vulnificus strain exhibited a poor resistance pattern, showing resistance only to cefazolin and an intermediate resistance to cefoxitin. However, Baker-Austin et al. [46] found numerous coastal and, also, septicaemia isolates in the USA resistant to antibiotics routinely prescribed for V. vulnificus infections, such as doxycycline, tetracycline, aminoglycosides, and cephalosporins, thus suggesting the importance of continued monitoring. In contrast, Bier et al. [27] reported that most antimicrobial agents recommended for treatment of V. vulnificus and V. cholerae non-O1/non-O139 infections were effective in vitro; likewise, in our study, both V. vulnificus and V. cholerae isolates did not display any worrisome resistance.

Banerjee and Farber [42] characterised 1021 Vibrio strains isolated from molluscs harvested in Canada between 2006 and 2012 and found that only 4.9% of them were sensitive to all tested drugs, while the antibiotics contributing the most to AMR were ampicillin, cephalothin, erythromycin, kanamycin, and streptomycin, although a declining trend in the frequency of MDR/AMR Vibrio spp. was registered until 2012.

Recently, Chahouri et al. [47] conducted a similar microbiological investigation in the Agadir Bay (Morocco), performing the search and AMR characterisation of Vibrio and Salmonella strains from mussels, sediment, and water samples. In accordance with our results, they isolated Vibrio strains at a high frequency, while a low percentage was noted for Salmonella. The eight Salmonella isolates showed resistance to ampicillin (100%), chloramphenicol (87.5%), and amoxicillin/clavulanic acid (62.5%) but were sensitive to all the other antibiotics used. Vibrio strains were mainly resistant to ampicillin (57.7%), cephalothin (62%), amikacin (60.6%), and, to a lesser extent, to ciprofloxacin (26.8%). Moreover, the authors perfectly agree with us when they highlight the importance of environmental survey to properly assess the microbiological quality in aquatic ecosystems.

In our study, almost all Vibrio strains exhibited MAR index values ranging from 0.064 to 0.193, below the arbitrary value of 0.2, indicating low-risk contamination sites [34], with the only exception of one V. alginolyticus isolate, which was resistant to seven antibiotics (MAR = 0.225). Since indices between 0.20 and 0.25 are in a range of ambiguity [34] and no Vibrio strains displayed MAR indices above 0.25, it could be stated that isolates originated from a low-risk environment where antibiotics are not regularly used perhaps. Moreover, resistance was observed mostly towards unusual antimicrobial compounds, such as ampicillin, piperacillin, sulfisoxazole, and the first-generation cephalosporin cefazolin, while few strains exhibited resistance to trimethoprim/sulfamethoxazole, piperacillin/tazobactam, and ampicillin/sulbactam (five, two, and one V. alginolyticus, respectively), proving that the antimicrobial combination therapy is still quite effective. In conclusion, the level of AMR in the tested Salmonella and above all Vibrio strains, which represent a consistent bacterial population in the marine environment, seems to indicate a poor diffusion of this phenomenon in the investigated area, suggesting that the actual condition is not highly worrying but perhaps promising, especially in view of the strategies currently implemented for the prudent use of antimicrobials in both human medicine and the zootechnical sector. However, our findings are not negligible, as some V. alginolyticus isolates exhibited resistance towards the critically important antimicrobials for human medicine. The epidemiological value of our study is of great relevance given the methodical approach used for monitoring: the survey was conducted over a period of 19 months and in a rather large area particularly devoted to shellfish farming and shellfish harvesting. Thus, the large amount of systematic data collected give an overview of the current scenario in one of the most representative regions for the Italian shellfish production sector [48].

The present study, like other previous ones, suggests the opportunity to implement a national/European programme to monitor the prevalence and distribution of antimicrobial resistance pattern in several not only pathogenic, but also environmental bacteria. Special consideration should be given to bacteria isolated from seafood which are generally eaten raw or undercooked and could represent a great threat to human health by transferring mobile genetic elements for antibiotic resistance to human pathogens. With regard to the latter aspect, it is worth mentioning, for example, the isolation of a V. parahaemolyticus strain carrying the bla_NDM-1 gene from seafood and displaying in vitro carbapenemase activity but not phenotypical resistance [49]; this is a certainly highly worrisome finding, which could lead to therapeutic failure; thus, further research should be encouraged. Investigation of AMR in microorganisms within a specific area is necessary to formulate effective antibiotic reduction programmes. Indeed, although antibiotic resistance is an ancient and natural phenomenon, it is generally recognised that it occurs mainly in bacteria inhabiting the gastrointestinal tract of subjects receiving antibiotics and that distribution of antibiotic resistance genes (ARGs) in the aquatic ecosystem reflects mostly the faecal contamination by ARB. Furthermore, antibiotic pollution contributes to promoting the emergence and maintenance of ARGs and ARB in a delimited area [50]. Hence, insightful information on trends of AMR distribution in site-specific microbial populations is fundamental to better understand the local use and abuse of antibiotics and realise appropriate corrective measures. In this perspective, it would be appropriate to fix MIC breakpoints, also, for environmental Vibrio species and our results could be useful to upgrade the MIC distribution for Vibrio spp. relating to the antibiotic agents without available specific CLSI breakpoints.