Enterococcus sp. are Gram-positive and facultative anaerobe lactic acid bacteria (LAB). Within this genus, 63 of the 87 identified species are validly published according to the German Collection of Microorganisms and Cell Culture GmbH (DSMZ; https://lpsn.dsmz.de/search?word=Enterococcus). Among them, Enterococcus faecium (E. faecium) represents one of the most ubiquitous species, as it is a commensal microorganism of the gastrointestinal tract of humans and animals, but it can also be found in environmental niches, such as soil, water, plants, or food (Delpech et al., 2012). In the latter, E. faecium is one of the most prevalent bacteria in traditionally manufactured cheeses (Serio et al., 2007), meat (Lauková et al., 2020), fermented vegetables, and fermented milk (Ben Belgacem et al., 2009), where it provides several beneficial effects, acting either as a probiotic (Foulquié Moreno et al., 2006; Jahansepas et al., 2019), or contributing to specific aroma, flavor, and taste (Delpech et al., 2012) during the fermentation process through the generation of flavor compounds (Abeijón et al., 2006). In several countries, the use of non-starter LAB is indispensable for producing traditional cheese and meat products (Delpech et al., 2012; Abeijón et al., 2006; Yerlikaya and Akbulut, 2019; Giraffa, 2003; Tsanasidou et al., 2021), suggesting that E. faecium endemic bacterial strains might have specific properties that make them a valuable resource for future food production (Daza Prieto et al., 2023; Woods et al., 2019; Ruppitsch et al., 2021). The UK Advisory Committee on Novel Foods and Processes (ACNFP) previously approved the use of E. faecium strain K77D as a starter culture in fermented dairy products (Hanchi et al., 2018). Additionally, E. faecium is a producer of enterocins, and numerous studies have aimed at their purification and characterization, as well as to determine their potential as a preservative in dairy products. For instance, E. faecium RZS C5 produces enterocins with activity against foodborne pathogens in milk and cheese (Hanchi et al., 2018; Huang et al., 2013; Leroy et al., 2003).

Numerous studies have demonstrated the benefits of enterococci as additives in food production (Santos et al., 2015; Fugaban et al., 2021; sim et al., 2018). However, other studies have shown that enterococci also play an important role as contaminants in foods, leading to spoilage (Giraffa, 2002). Additionally, their use has become controversial due to the increasing acquisition of multiple antimicrobial resistances, the most concerning the emergence of vancomycin-resistant E. faecium (VREfm) clones and their adaptation to the hospital environment (Lee et al., 2019; De Oliveira et al., 2020; Leclercq et al., 1988). The World Health Organization (WHO) classifies VREfm as a priority pathogen (Tacconelli et al., 2018) and this is one of the reasons why the genus Enterococcus has not been recommended for the Qualified Presumption of Safety (QPS) list of the European Food Safety Authority (EFSA; Koutsoumanis et al., 2023) neither has obtained the “generally recognized as safe” (GRAS) status of food additives that was first described in 1958 by the United States Food and Drug Administration (FDA) in the Food Additives Amendment of 1958 (1958).

The rapid evolution of NGS technologies, combined with the dual role of E. faecium in food microbiology and public health—as beneficial additives or starters in food production, as well as contaminants causing spoilage or carrying antimicrobial resistance genes (ARGs) or virulence determinants—has resulted in an increasing number of genomic studies that aimed to distinguish between food-grade and pathogenic strains (Martín-Platero et al., 2009; Montealegre et al., 2016).

As a result, three different E. faecium “lineages” or “clades” have been described: clade A1, which includes hospital-associated strains; clade A2, which includes animal-associated strains; and clade B, which includes commensal/community-associated strains (Montealegre et al., 2016). The latter was recently reassigned to E. lactis in a genome-based study (Belloso Daza et al., 2021). Since then, several groups focused their studies on this re-classification of E. faecium clade B (Belloso Daza et al., 2022). The best example of E. faecium strains with different evolutionary trajectories is E. faecium clonal complex 17 (CC17; clade A1 hospital-associated E. faecium) and E. faecium clonal complex 94 (CC94; clade B community/commensal E. faecium; Peng et al., 2022). Additionally, several studies have reported that E. faecium clade A1 strains differ up to 5% from E. faecium clade B (now re-classified as E. lactis) strains (Galloway-Peña et al., 2012; Wei et al., 2024). CC17 E. faecium is the most dominant CC in hospital environments and is the main responsible for healthcare-associated infections worldwide due to its ability to acquire and disseminate ARGs (Lee et al., 2019; De Oliveira et al., 2020). On the other hand, community-associated E. faecium are commensal strains that are part of the human microbiome participating in the metabolism of nutrients and synthesis of vitamins (Krawczyk et al., 2021). Regarding antimicrobial resistance (AMR), E. faecium and E. lactis are generally intrinsically resistant to aminoglycosides, macrolides, and pleuromutilin antibiotics. As a result, these antibiotics are not commonly used to treat infections. Instead, vancomycin, linezolid, daptomycin, tigecycline, quinupristin/dalfopristin, and ampicillin are preferred. However, the usage of vancomycin and ampicillin has decreased, primarily due to the growing resistance to these antibiotics highlighting the need for new therapeutic options (Arias and Murray, 2012). Community-associated E. faecium and E. lactis strains are more often susceptible to the antibiotics typically used to treat E. faecium infections compared to clinical strains. However, some vancomycin-resistant strains have also been described among the community-associated isolates (D’Agata et al., 2001; Schwalbe et al., 1999). A crucial aspect to consider an E. faecium strain as “safe” is the absence of the ARGs vanA/vanB, the virulence genes (VGs) hylEfm and esp., and the insertion sequence IS16, as required by the European Food Safety Authority (EFSA; EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) et al., 2018). While clinically associated E. faecium strains typically carry the above-mentioned genetic markers, these markers are generally absent in community-associated E. faecium strains.

While VREfm is responsible for numerous outbreaks worldwide (Zhou et al., 2018; Gorrie et al., 2019), E. lactis is a commensal in the gastrointestinal tract of humans and animals and some strains of this species are used in the clinic for gastrointestinal disorders such as acute diarrhea (Holzapfel et al., 2018). The present study aimed to characterize E. faecium and E. lactis strains present in traditionally produced Montenegrin food products using whole-genome sequencing and to assess their safety status via comparison to hospital-associated strains and investigation of ARGs, VGs, bacteriocins, secondary metabolites, plasmids, and chromosomal point mutations.

Vancomycin-resistant E. faecium represents nowadays a significant public health challenge (Koutsoumanis et al., 2023). Conversely, Enterococcus species have been described in several countries as essential for the production of artisanal food products, underscoring the importance of correctly distinguishing between beneficial and pathogenic strains to ensure consumer safety. To achieve this, whole-genome sequencing (WGS) is today the method of choice to select safe and beneficial strains in food production (Ruppitsch et al., 2021; EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) et al., 2023; EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) et al., 2021; EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) et al., 2024a; EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) et al., 2024b; EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) et al., 2024c).

In our study, WGS revealed a high diversity among E. faecium isolates obtained from traditionally produced Montenegrin food, with one large cgMLST cluster (ST1453/CT2909) comprising isolates obtained from brine cheeses in the northern region of Montenegro. ST1453 E. faecium isolates were previously reported in traditional brine cheese from Montenegro (Ruppitsch et al., 2020). All ST1453 E. faecium isolates carried the intrinsic aac-(6′)-I and msrC ARGs, while only seven of the isolates possessed the intrinsic eatA ARGs. Additionally, five isolates were found to carry the acm virulence genes (VGs), and all isolates contained plasmids. To the best of our knowledge, no infections have been linked to ST1453 so far, which suggests that this clone seems to be adapted to artisanal cheese products in Montenegro. This observation is consistent with the review of Giraffa G. (Giraffa, 2003), who reported that enterococci might have a positive influence on cheese due to lipolytic activity, citrate utilization, and production of aromatic volatile compounds. The fact that all our ST1453-positive cheese samples were produced in the same municipality, probably with a similar supply of milk and infrastructures, may explain the colonization with the same E. faecium strain.

Regarding the E. faecium singletons, ST22, ST32, ST92, and ST296 were the most reported in the literature whereas other detected STs in our study have barely been reported. E. faecium ST32 strains have been previously isolated from chicken meat (Leinweber et al., 2018), having an association with vancomycin-resistant genes, which in contrast were absent in our case. E. faecium ST22 [which is a single locus variant of E. faecium ST32 (Freitas et al., 2020)] was reported by Cabal et al. (2022) with no acquired ARGs nor VGs, which is also in concordance with our findings. As for E. faecium ST92, Galloway-Peña et al. (2009) reported its association with vancomycin and ampicillin resistance in clinical isolates and outbreaks in the USA in 1980, which was the first detection of ampicillin resistance in E. faecium. In contrast, our E. faecium ST92 isolates did not carry any acquired ARGs. However, it carried several mutations in pbp5 associated with ampicillin resistance. Interestingly, most of the strains in our study, despite having specific mutations in pbp5 detailed in Supplementary Table 7, which are known to reduce the affinity for ß-lactam antibiotics and confer resistance, were susceptible. Specifically, the point mutations found are nucleotide substitutions that are responsible for increased resistance by decreasing the affinity for ß-lactam antibiotics, leading to ampicillin resistance. On the other hand, other authors such as Rice et al. (2004) reported that the decreased affinity of pbp5 is not the only factor involved in the expression of resistance to ß-lactams in E. faecium. In concordance with our findings, Belloso Daza et al. (2022) reported that some E. lactis strains harboring multiple mutations in pbp5 were still susceptible to ampicillin.

E. lactis ST697/CT6825 isolates were obtained from different beef sausages from the same producer in the northeast of Montenegro. This rare ST was reported from a patient isolated in Spain in 2010. Unfortunately, no genome nor a study was available from this isolate for comparison with our isolates. E. lactis ST697 strains, lacking VGs, were isolated from sausages from the same producer, therefore we assume that they might be also endemic within the company or producer’s infrastructure. E. lactis CC94/ST296/CT426, isolate CoE-451-22 was obtained from sausage and clustered with several isolates from different countries and sources. Due to its relatedness and the similarity of all the other strains within this cluster with E. lactis NCIMB 10415, an approved feed additive since 1999 (EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) et al., 2024c), we hypothesize that E. lactis ST296 could have been transferred from the feed to the animals and during slaughter and meat processing to the sausage. In concordance with previous studies, our strain E. lactis ST296 carried only intrinsic ARGs (Murray et al., 2020).

The absence of acquired ARGs, including van genes, in food-derived enterococci is an important parameter when assessing the safety of these food products for consumers. No vancomycin-resistance-associated ARGs were detected among our strains, which are mostly found in clinical strains, but not in food strains, as reported previously (Jahansepas et al., 2019). Most E. faecium and E. lactis isolates from our study carried the intrinsic ARGs aac (6′)-I, eatA, and msrC, which have been described in the literature (Diarra et al., 2010; Shridhar et al., 2022). The presence of intrinsic ARGs alone should not pose any health risk to consumers, as they are resistant to antibiotics that by default are not used in the treatment of clinical E. faecium infections. Additionally, some studies reported that macrolide resistance is not solely linked to the presence of the msrC gene but depends on specific mutations within the gene (Guido et al., 2001).

Acquired ARGs such as tetM/tetL, conferring resistance to tetracycline, were found in one E. faecium isolate (ST29) encoded in a mobilizable plasmid, obtained from dry sausage. Although these are acquired ARGs, this is not the first time they have been detected in E. faecium isolates obtained from food, specifically from fermented meat products, as reported by Jahan et al. (2013), who reported that the Enterococcus sp. isolates included in their study (including 13 E. faecium) had 65% resistance to tetracycline. Although E. faecium could serve as a reservoir for tetM/tetL genes for other bacteria, tetracycline is not a first-line treatment option for humans. None of our strains carried optrA ARGs conferring resistance to linezolid. Nevertheless, it is important to note that numerous studies warn about the danger of the presence of the optrA gene confirming resistance to linezolid (Xuan et al., 2023). Therefore, monitoring of optrA is recommended.

VGs are crucial for the adaptation, pathogenicity, and colonization of bacterial infection (Kim et al., 2016; Fu et al., 2022). None of our isolates carried hylefm and esp VGs, whose absence is a key aspect according to EFSA guidelines to consider a strain of E. faecium as “safe” (EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP), 2012). The VG acm and efaAfm were present in all E. faecium isolates and all E. lactis isolates in our study, which is in concordance with other food and probiotics Enterococcus studies (Strateva et al., 2016; Yuksekdag et al., 2021). For instance, acm has been reported in a fermented dry sausage E. faecium isolate from Italy, in the probiotic JDM1 E. lactis strain, and in the probiotic SF68 E. faecium strain (or E. faecium NCIMB 10415, now re-classified as E. lactis NCIMB 10415; Belloso Daza et al., 2022; Holzapfel et al., 2018; Fu et al., 2022). The VG EfaAfm, one of the most prevalent in E. faecium food isolates (Wang et al., 2024), has been obtained from white cheese (Yuksekdag et al., 2021), roe deer and boar meat (Guerrero-Ramos et al., 2016), and in the probiotic E. faecium OV3-6 strain (Choeisoongnern et al., 2021). Much less prevalent, sgrA was found in one E. lactis isolate, and ecbA was found in two different E. faecium isolates from beef dry sausages (sudzuk). Previous studies reported that the presence of only one of these virulence genes does not indicate pathogenicity; however, the expression and/or combination of these genes could pose a potential risk (Soheili et al., 2014).

All E. faecium isolates and E. lactis isolates from our study carried plasmids, being the most prevalent rep1 and repUS15. VREfm strains were reported harboring vanA gene cluster encoded on a rep1 replicon (Yu et al., 2012). While another VREfm strain was reported to carry repUS15 plasmid without ARGs (Katsuyama and Ohnishi, 2012). The strain with the acquired tetM/tetL was encoded on a mobilizable plasmid repUS43, which is in accordance with the results from a recent study (Yu et al., 2012). The presence of plasmids in E. faecium strains is considered as a potential risk (Schwalbe et al., 1999; EFSA, 2024), due to the horizontal transfer of ARGs. Therefore, monitoring is essential to mitigate the potential health risks associated with their spread in food.

E. faecium isolates obtained from food are reported to produce enterocins, proteins with antimicrobial activity against foodborne pathogens such as Listeria monocytogenes (Gontijo et al., 2020; Schittler et al., 2019; Tsigkrimani et al., 2022). Therefore, the application of enterocin-producing E. faecium has been described to show potential as a biopreservative against food-borne pathogens (Giraffa, 2003). Our strains carried Enterolysin A, Enterocin A, Enterocin P, Duracin Q, Enterocin B, Bacteriocin 31, Enterocin EJ97, Sactipeptides, and Enterocin SEK4. Enterocin A and enterocin B have been reported from E. faecium on several occasions (Martín-Platero et al., 2009) and some authors also informed about the presence of enterocin A, enterocin B, and enterocin P in one E. lactis strain (Ben Braïek et al., 2018). The most common secondary metabolites in our isolates were T3PKS (polyketides family), cyclic lactone autoinducer, and RiPP-like. The polyketides have been isolated from plants, bacteria, and fungi (Lv et al., 2014) and have been described to have anticancer, anti-cholesterol, and antimicrobial properties (Yu et al., 2012). Specifically, T3PKS in bacteria have significant biological functions such as biosynthesis of some lipidic, natural compounds, and various secondary metabolites, ranging from signaling molecules to bioactive natural products (Katsuyama and Ohnishi, 2012). Furthermore, PKS and RiPP-like cluster genes were detected in two E. lactis strains isolated from the rumen of a healthy calf (Korzhenkov et al., 2021), while lactones are used for adding flavors and fragrances to fermented and unfermented dairy products, which seem to have a beneficial use in the making of artisanal cheeses and dry sausages.

In conclusion, all but one investigated E. faecium and E. lactis isolates carried only intrinsic ARGs and some virulence genes, which alone are not associated with pathogenicity. Some of these isolates may be endemic and could play beneficial roles in traditional cheese and sausage production by enhancing organoleptic qualities and contributing to biopreservation through the production of enterocins, a possibility that warrants further research. However, since all isolates carry plasmids, which is a named risk factor for intentional use according to EFSA recommendations further investigation and ongoing genomic monitoring remain crucial to ensure consumer safety.