Enterococci are Gram-positive acid lactic bacteria ubiquitously found in nature in a wide range of habitats, including the gastrointestinal tract (GIT) of humans, vertebrate and invertebrate animals as part of its microbiome, as well as in the natural environment in plants, soil, fresh, and marine waters and in food (Murray, 1990; Byappanahalli et al., 2012; Lebreton et al., 2014; Cattoir, 2022). While the Enterococcus genus comprises >50 species, its distribution in nature is highly variable and associated with host species factors such as age and diet or environmental stress conditions (Lebreton et al., 2014; Cattoir, 2022). Enterococcus faecalis, Enterococcus faecium, Enterococcus hirae, and Enterococcus durans are the species most commonly found in the gut of mammals, with E. faecium and E. faecalis representing up to 1% of human adult microbiota (Torres et al., 2018; Cattoir, 2022). Some of these Enterococcus species, namely E. faecalis and E. faecium, evolved as human opportunistic pathogens being responsible for a broad range of infections from the urinary tract, intra-abdominal infections or wounds, to more severe life-threatening conditions such as bacteremia and endocarditis usually associated with high antibiotic resistance and virulence (Cetinkaya et al., 2000; Arias and Murray, 2012; Zhong et al., 2019; Cattoir, 2022; Wei et al., 2024). Although E. faecalis accounts for the majority of enterococcal infections in humans, E. faecium has been increasingly reported in hospitals usually associated with severe conditions (Gao et al., 2017; ECDC, 2023; Wei et al., 2024). In fact, a major hallmark of enterococci that allowed the success of nosocomial pathogens is its capacity to acquire antimicrobial resistance and adaptation to hostile environments. Enterococci are intrinsically resistant to a variety of antimicrobials and have remarkable genome plasticity that enables the accumulation of mutations and acquisition of external antimicrobial resistance determinants (Cetinkaya et al., 2000; Arias and Murray, 2012; Guzman Prieto et al., 2016).

Of major concern was the emergence of vancomycin-resistant isolates, which through the acquisition of the van operon were able to produce altered cell wall peptidoglycan precursors not recognized by glycopeptides (Uttley et al., 1988; Cetinkaya et al., 2000). Vancomycin-resistant E. faecium (VRE) become endemic in US hospitals as a consequence of the increasing usage of vancomycin in difficult-to-treat infections, while in European hospitals, VRE infections have been increasing since the COVID-19 pandemics, reaching a mean rate of 17.6% in 2022 (ECDC, 2023). Antibiotic-resistant E. faecium is among the top ten bacterial pathogens with the highest antimicrobial resistance-associated mortality, with treatment often restricted to last-resort antibiotics such as linezolid, tigecycline, and daptomycin. However, sporadic-resistant isolates to these antibiotics have been already reported (Bender et al., 2018a), leading VRE to be considered as a high-priority pathogen for drug development by WHO (2024).

Colonized individuals or animals are potential reservoirs of multidrug-resistant Enterococcus spp., representing a significant source for pathogen transmission (Garcia-Solache and Rice, 2019; Cattoir, 2022; Wei et al., 2024). As VRE have a high capacity to persist in environmental surfaces and be maintained as commensals among humans and animals in the gut until they achieve optimal conditions to develop infection (Chavers et al., 2003; Zaheer et al., 2020; Wada et al., 2021), the screening of community and environment reservoirs for multidrug-resistant E. faecium emerged as an urgent tool to support nosocomial infection control strategies.

In Portugal, VRE infections have been increasingly detected in hospitals paralleling the European scenario (ECDC, 2023). Moreover, healthy individuals without hospital and/or antibiotic exposure were identified as asymptomatic carriers of antibiotic-resistant Enterococci (Novais et al., 2006; Correia et al., 2014; Freitas et al., 2018). While VRE was sporadically detected in food-producing animals and pets (Zaheer et al., 2020; Cattoir, 2022; Gião et al., 2022), a comprehensive screening of environmental reservoirs in the country remains scarce.

Hence, in this study we applied a One Health approach to unveil the major VRE and antibiotic-resistant Enterococci reservoirs among healthy livestock and urban environment in Portugal. We will assess the significance of these reservoirs on the maintenance of hospital-associated multidrug-resistant Enterococcus, namely E. faecium, clonal lineages, and/or antibiotic resistance and virulence determinants in natural ecosystems, responding to the need for valuable epidemiological information to guide the design of targeted and effective infection control measures.

This study aimed at the identification of enterococci reservoirs outside hospitals, namely healthy livestock or environmental surfaces, providing valuable knowledge on the dissemination and resistance mechanisms of this major human pathogen. To the best of our knowledge, our study represents the first surveillance of non-clinical settings as the hands of passengers and outdoor urban surfaces near hospitals as possible reservoirs for VRE. Still, this is the first study in Portugal to explore livestock nares and udder regions as potential VRE colonization reservoirs.

In this study, E. faecalis and E. faecium were the unique species identified in all collections, although at different percentages depending on the sample source. Our results are in line with a recent One Health continuum surveillance of Enterococcus spp. that reveals E. faecium and E. faecalis as the predominant species related to human-associated environments (Zaheer et al., 2020). In fact, buses and hands of passengers were highly contaminated with E. faecalis and E. faecium, confirming that community settings could be reservoirs of major human pathogens, as previously shown for methicillin-resistant Staphylococcus aureus (Conceição et al., 2013). A similar study in Turkey supported also the role of public transports as reservoirs of Enterococcus spp.; however, isolates were not identified at the species level (Birteksöz Tan and Erdogdu, 2017). Moreover, we identified clinically relevant Enterococcus spp. in other urban environmental public surfaces highly touched by the hands of individuals, as previously identified in shared urban bicycles in China (Gu et al., 2020). As we screened urban environmental surfaces close to hospitals, and it is known that individuals with direct contact with the nosocomial environment have a high probability of having hands contaminated by VRE and transmitting these pathogens to the hospitals’ outside environment (Hayden et al., 2008), our study highlights the importance of hands’ hygiene measures to prevent human pathogens dissemination.

In our livestock samples, E. faecalis was the most prevalent species detected, as recently reported in pigs and to a lesser extent in cows cecal swabs collected at slaughterhouses in Portugal (Gião et al., 2022). Moreover, contrary to other studies that showed E. hirae as a prevalent species colonizing the gut of cattle and further detected in the food production chain (Zaheer et al., 2020; Gião et al., 2022), in our collection E. hirae was randomly detected comparing to the predominance of E. casseliflavus. We swabbed the nares and the udder region of cattle animals, which could justify this difference among the enterococcus species identified.

Pigs were highly nasally colonized by Enterococcus spp., namely E. faecalis has recently been reported in four pig farms in Spain (Abdullahi et al., 2023), which suggest that although the nares are not the common habitat for Enterococcus species, nasal sampling might be valuable for enterococci monitoring in animals.

Vancomycin resistance was not detected in our samples. Although isolates were able to initially grow in media supplemented with 6 μg/mL of vancomycin, none of them grew in VRE-selective media. The presence of other bacteria in our initial multibacterial samples might produce factors that protect enterococci from vancomycin or alter the local environment in a way that reduces the effective concentration of the antibiotic to concentrations that allow the growth of susceptible isolates, or the low-level resistance to vancomycin phenotype intrinsic of some enterococcus species could support this observation. VRE has already been identified in animals in Portugal (Freitas et al., 2009; Ramos et al., 2012; Braga et al., 2013; Novais et al., 2013), but there are no data regarding VRE in non-clinical environmental settings. Our study contributes to fulfilling this gap, identifying multidrug-resistant E. faecium isolates in buses, passengers, and urban environments, in addition to vancomycin resistance.

The vanC genes are considered to be intrinsic to E. gallinarum (vanC1), E. casseliflavus (vanC2), and E. flavescens (vanC3) (Ahmed and Baptiste, 2018; Torres et al., 2018). However, vanC1 was detected in 36% of our E. faecalis isolates from cows, which is in line with the increasing reports of vanC1 in E. faecalis from animal-associated environment samples (Schwaiger et al., 2012; de Garnica et al., 2013; Nishiyama et al., 2015). On the other hand, none of the E. faecium isolated in our study harbor van genes, although vanC1 was already reported in E. faecium from human infection (Sun et al., 2014), evidencing that vanC1 could be spread to other clinically relevant Enterococcus species. Although a possible bias toward the selection of intrinsic vancomycin-resistant enterococcus could be suggested by the usage of 6 μg/mL of vancomycin in the first screening of samples, no other van genes or vancomycin non-susceptibility could be detected in the majority of the Enterococcus isolates in addition to naturally intrinsic resistant E. casseliflavus and E. gallinarum and 9% of the total E. faecalis isolates.

Given the clinical relevance of E. faecium in human infection, namely VRE that is considered a high-priority pathogen by the WHO (2024), a comprehensive characterization of enterococci isolated in the present study was focused on E. faecium isolates.

The rationale to further study the E. faecium isolates was supported by a significant increase of E. faecium in nosocomial infections worldwide, namely in Europe, associated not only with vancomycin resistance but also with other antibiotics used in human clinical practice and in veterinary medicine as ampicillin, aminoglycosides (high-level resistance), or last-resort antibiotics like linezolid and tigecycline. In the most recent Antimicrobial Surveillance Report 2022, the ECDC reported a significantly increasing trend of antimicrobial-resistant E. faecium in some individual countries in Europe and highlighted the urgent need for close monitoring to better understand the epidemiology, clonal diversity, and risk factors associated with vancomycin-resistant E. faecium infections (ECDC, 2023). In Portugal, the country where the study was carried out, VRE has been increasingly detected in hospitals compared to a decreasing trend observed in E. faecalis resistant to aminoglycosides since 2018, which heightened concern among hospitals about these resistant E. faecium infections and the need for monitoring reservoirs and the development of effective infection control guidelines (ECDC, 2023). Additionally, E. faecium reservoirs have been identified in a One Health continuum (Zaheer et al., 2020), suggesting that this relevant species for human health, may be prevalent in non-clinical environments and could potentially be transmitted to humans.

We identified high rates of antimicrobial resistance among the isolates independently of the source. Resistance of ampicillin and tetracycline was significantly associated with livestock samples, which could be a consequence of the usage of a β-lactam (amoxicillin) as prophylaxis in the feed regimen of pigs at the time of sampling, or with the global and widespread use of tetracycline for prophylactic purposes in livestock (Roberts and Schwarz, 2016). The gene tet(M) is considered the most frequent mechanism of tetracycline resistance in enterococci from humans, animals, food, and the environment (Cho et al., 2020), even though in this study tet(M) and tet(L) were equally detected. In addition, 32% of E. faecium from non-clinical settings presented phenotypic resistance to tigecycline, but only two isolates harbored tet(L) and tet(M), suggesting that other resistance mechanisms could be involved that deserve further investigation.

The ampicillin resistance rate of 56% among livestock in our study surpasses the 30% resistance reported in poultry E. faecium in Portugal (Poeta et al., 2006), but parallels the increasing rates of vancomycin-susceptible ampicillin-resistant E. faecium causing infections in European hospitals (Garcia-Solache and Rice, 2019). On the other hand, the significant use of aminoglycosides in veterinary medicine reflects the rates of high-level aminoglycosides resistance in E. faecium from livestock and human isolates (McGlinchey et al., 2008; Li et al., 2015; Diab et al., 2019). We detected approximately 24% HLAR in E. faecium isolates from non-clinical settings, associated with the presence of the ant(6)-la gene, as previously reported in animal E. faecium isolates (Han et al., 2011; El-Mahdy et al., 2018; Cagnoli et al., 2022) and suggesting a livestock reservoir for this acetyltransferase gene. In 2022, 34.9% of E. faecium isolated from invasive infections in Europe showed aminopenicillins and gentamicin high-level resistance (ECDC, 2023). We identified non-clinical environments and healthy livestock as reservoirs of ampicillin and high-level aminoglycosides-resistant E. faecium in Portugal. This finding is of major concern as the spread of these resistant isolates compromises the clinical utility of the commonly used synergistic protocol for aminoglycosides with a cell wall-active antibiotic such as ampicillin in the treatment of severe enterococcal infections (Miller et al., 2014; Echeverria-Esnal et al., 2023).

Moreover, in our study, almost all isolates presented phenotypic resistance to nitrofurantoin, a synthetic antibiotic used as the first-line therapy of uncomplicated lower urinary tract infection caused by HLAR Enterococcus and VRE (Mahdizade Ari et al., 2023). High nitrofurantoin resistance rates (56–84%) have been reported in clinical E. faecium isolates (Wang et al., 2013), but the identification of such a high prevalence of resistance in non-clinical isolates in our study is of major concern.

In our collections, E. faecium harbored an arsenal of virulence determinants associated with bacterial adherence and biofilm production such as efaA_fm, fnm, acm, and pili-associated proteins, which provide them an advantage on environmental persistence. Moreover, adhesion is usually the first step required for infection, suggesting that these environmental surfaces and healthy livestock represent reservoirs of potentially high pathogenic E. faecium isolates. This assumption is even supported by the high prevalence of virulence determinants with proven impact in biofilm and endocarditis infection as empA/empB/empC cluster, bepA or efaA_fm, widespread in E. faecium isolates, regardless of the isolation source. Putative virulence markers for E. faecium hospital-associated strains, such as insertion sequence 16 (IS16) or acm and scm adhesins (Werner et al., 2011), were also widely detected supporting the pathogenic potential of these environmental and livestock-associated E. faecium and showed a different scenario previously reported across a One Health continuum in Canada where livestock-associated Enterococcus spp. lacked virulence genes associated with clinical isolates (Zaheer et al., 2020). In addition, three genes (sgrA, fms15, and orf1481) exclusively associated with E. faecium hospital variants (Roer et al., 2024) were detected among E. faecium in this study, namely in bus isolates, suggesting a possible hospital-associated origin and dissemination to the community.

Bacteriocin production ensures niche competition of enterococci in the gut, namely against particular pathogens or adaptation to different environmental conditions (Tedim et al., 2023). In fact, in our collection, environmental isolates harbored a higher bacteriocin genetic determinants content, belonging exclusively to the heat-stable, non-lantibiotic class II bacteriocins. entA, a pediocin-like bacteriocin (class IIa) with strong antilisterial effects, and bacAS3 (subclass Iib) were the most prevalent bacteriocin determinants identified in our study (39 and 30%, respectively) as previously reported in a large and diverse collection of E. faecium isolates (Tedim et al., 2023). Enterocin gene entA was already identified not only in food-associated E. faecium but also in clinical outbreaks of VRE- and vancomycin-susceptible E. faecium in different European hospitals (Freitas et al., 2016). Moreover, 22% of our isolates, mainly isolates from the bus, contained >5 different bac genes, which is much higher than the previously bac mean content (1 to 5) among E. faecium (Tedim et al., 2023), suggesting a high adaptive potential of our environmental isolates.

Population analysis of the 23 E. faecium representative isolates showed a highly polyclonal bacterial collection, regardless of the isolation source. This variability in clonal population structure was not surprising given the non-clinical nature of these E. faecium isolates. Moreover, the identification of new STs, namely in cows (ST2203, ST2206, and ST2207), buses (ST2204, ST2210, and ST2211), and urban environmental surfaces (ST2208 and ST2209) suggests the acquisition of genetic variability to improve adaptation and maintenance in adverse environmental conditions or hosts.

Although cgMLST analysis could not trace a direct transmission link between E. faecium from invasive infections in Portuguese hospitals and our non-clinical E. faecium isolates, the genetic background of several isolates from pigs (ST32-CC17; ST624; ST127; ST296), cows (ST30; ST2206-CC94), buses (ST27, ST2201; ST800-CC94), and urban environment (ST1205-CC94) were already identified in human clinical infections and carriage (see text footnote 3) (Werner et al., 2012; Bender et al., 2018b; Zhou et al., 2019; Peng et al., 2022). In line with our observation, recent studies suggested that instead of an evident gene flow from livestock E. faecium to hospital E. faecium, it is supported by distinctive niche adaptations for the distribution of antimicrobial resistance genes, host adaptation, and virulence determinants among E. faecium population (Gouliouris et al., 2018; Arredondo-Alonso et al., 2020).

The identification in our study of E. faecium isolates belonging to genetic backgrounds prevalent in infection and associated with antimicrobial resistance to last-resort antibiotics is a relevant finding. Therefore, ST32 belonged to the CC17, the major hospital-associated clonal lineage responsible for human infections worldwide. While ST32 was detected in a pig nasal swab, it was already identified from carriage and infection in other livestock and humans as well as from environmental wastewaters (see text footnote 3) (Werner et al., 2012). Recent findings from Portugal reported linezolid-resistant ST32 E. faecium in a healthy human rectal swab (Freitas et al., 2020), highlighting its significant adaptive and pathogenic potential.

On the other hand, ST30 identified in our study in a cow sample was recently reported in German hospitals containing the optrA linezolid resistance gene (Bender et al., 2018b). In Portugal, ST30 E. faecium isolates were reported in a pig, dust, and food of pigs in the same farm (Novais et al., 2013). Our detection of ST30 in a cow sample underscores its capacity for host diversification.

CC94 E. faecium lineage is mostly associated with human community and commensal isolates that can cause sporadic infections in animals and humans (Peng et al., 2022) In our study, single isolates from livestock (a pigs ST296 isolate and a cow ST2206), a bus (ST800), and urban environment (ST1205) exhibited loci variations of the ST94, indicating their affiliation with the widespread CC94 lineage.

E. faecium isolated in this study and the E. faecium recovered from infection in Portuguese hospitals differed in the isolation period, which could be a limitation in the genomic comparison analysis. In fact, a more global approach as MLST identified isolates in this study sharing the same genetic background as infection isolates from hospitals, but the high discriminatory power of cgMLST could not identify a direct link between these isolates, which could be due to evolutionary pressure that allows the accumulation of small genetic events along time, promoting adaptation to selective pressure of each environment. An example to support this assumption was the identification in this study of an E. faecium bus isolate sampled in 2012 and characterized by ST27 that belonged to the same genetic background (sharing five MLST alleles) of an ST2020 isolate causing infection in hospitals in the same geographic area approximately 8 years later.

Another possible limitation of the study was the inclusion of livestock nasal swabs, which is not the primary niche of Enterococcus colonization in humans and animals. However, it has been shown that Enterococcus can be found in colonization of the anterior nares and co-habit with other pathogens such as S. aureus. Therefore, the previous identification of these samples as reservoirs of S. aureus and MRSA (Conceição et al., 2013, 2017a,b; Lopes et al., 2019) as well as the close living conditions of the animals (namely pigs, that live approximately 32 to 33 animals in 8–12 m² stockyards) made the samples to be considered potential reservoirs for multidrug-resistant VRE and therefore selected for this study.

Major achievements of this study included the unveiling of a global high prevalence of Enterococcus spp. in non-clinical settings such as livestock- and human-associated environments, including clinically relevant species such as E. faecium and E. faecalis. Although no vancomycin resistance could be detected, livestock and environmental surfaces showed to be reservoirs of E. faecium isolates with antimicrobial resistance to clinically relevant antibiotics, namely aminoglycosides (high-level resistance), ampicillin and tetracyclines including tigecycline. Moreover, these E. faecium isolates harbor a wide arsenal of virulence determinants associated with increased adaptation to hostile environments and different hosts, evasion of the immune system, and capacity of infection development, associated with genetic backgrounds already identified in infection isolates. Our results evidenced that surveillance of Enterococcus spp., with a special focus on E. faecium outside nosocomial settings is of paramount importance for effective infection control strategies and reduction of antimicrobial resistance spread, given its proven role as reservoirs of multidrug resistance and virulent strains.