The distinctive biological mechanism of CRISPR systems, particularly their intrinsic ability to sequentially incorporate fragments of foreign DNA, thereby generating a hypervariable region within the host chromosome, places these loci as true biological records. This property renders them valuable for reconstructing historical events, such as host-virus interactions in natural environments, as well as for assessing genetic diversity (74, 75). The potential of CRISPR loci has already been explored for various applications, including tracking clinically relevant lineages and elucidating their origins and evolutionary paths (76, 77). In E. faecalis, the orphan CRISPR2 locus has been explored as an alternative target for genetic diversity analysis and phylogenetic studies, enabling clonality predictions and the identification of recombination events among STs (12–14).

Our original goal was to evaluate CRISPR2 array variability for genotyping and phylogenetic analysis of E. faecalis strains from diverse sources. However, initial analysis revealed an unexpected finding: a genetically related group of strains lacking CRISPR2. This observation challenges the prevailing assumption that CRISPR2 is ubiquitous, a component of the species’ core genome (8, 12). Given this intriguing observation and the fact that the orphan CRISPR2 locus, in the absence of associated cas genes, has no clearly defined biological role, we redirected our focus to explore the implications of CRISPR2 absence in the context of the species’ evolutionary trajectory. Due to the fact that none of the CRISPR2-negative strains were part of our local bacterial culture collection, our investigation was limited to genomic approaches.

Phylogenomic analyses revealed that the CRISPR2-negative strains form one of two E. faecalis major clades. Their most recent common ancestor (MRCA) diverged early in the evolutionary history of the species, leading to the independent evolution of these two lineages. These lineages can be classified into distinct subspecies based on dDDH thresholds for taxonomic delineation (referred to here as subspecies A and B, consisting exclusively of CRISPR2-positive and CRISPR2-negative genomes, respectively) (36). Objectively, this means that pairwise genomic comparisons between any representatives of each subspecies reveal hybridization values below 79%. Because no validly published E. faecalis child taxa were listed in the LPSN database at the time of writing, we believe this is the first study to describe a subspecies-level division for E. faecalis, based on the current criterion used for taxonomic classification of bacteria (33, 34, 78).

Our findings challenge the previous understanding of E. faecalis population structure and its generalist nature. By revealing well-supported and distantly related subspecies clades within E. faecalis, our results question the assumption that this species lacks a multiclade structure and is devoid of clearly divergent groups, as suggested by earlier studies (6, 15). Given that these prior studies did not include representative STs from subspecies B, it is plausible that their conclusions were influenced by a limited genomic sampling that lacked sufficient genetic diversity. In contrast, our study included a more comprehensive set of lineages, providing a broader perspective on the species’ population structure.

The internal topology of the subspecies A clade reflects its extensive genetic and ecological diversity. This clade includes STs predominantly found in hospital environments, with multidrug resistance profiles and a strong association with HAIs— such as STs 6, 103, 525, and 778, including all VRE isolates in our study (e.g., the well-characterized clinical isolate V583). Additionally, the clade encompasses generalist STs like STs 4, 16, 21, and 40, which are widespread across various ecological niches and display variable CRISPR-Cas content, alongside strains such as the commensal-like OG1RF (ST1), the species type strain ATCC 19433 (ST25), and the NCBI reference genome T5 (ST68) (14, 15, 79–81). The low bootstrap support observed among these lineages suggests complex evolutionary dynamics, likely driven by diverse selective pressures and rapid diversification, which complicate the reconstruction of a clear evolutionary pathway. While this finding is consistent with the aforementioned studies focused on subspecies A, it does not fully capture the population structure of E. faecalis, particularly regarding the phylogenetic positioning and evolutionary significance of subspecies B.

On the other hand, the inclusion of subspecies B in the phylogenetic reconstructions of E. faecalis revealed significantly divergent lineages, as evidenced by their deep, well-supported branches. Unlike subspecies A, subspecies B displays a more homogeneous genetic and ecological profile, even though it is further divided into two well-resolved subclades. All strains of subspecies B are related by their MLST allelic profiles (STs 228, 624, 1468, and their respective loci variants), with the majority sharing common isolation sources, such as animal meat or meat processing facilities. From an epidemiological standpoint, the absence of association between subspecies B and human clinical sources, along with the uniform lack of traits commonly found in high-risk lineages (e.g., acquired multidrug resistance and virulence genes linked to treatment failure and increased pathogenicity in enterococcal infections; see reviews (71, 82) for comprehensive overviews) suggest that this subspecies may not pose a significant threat to human health, either as pathogens or as reservoirs of clinically relevant ARGs. However, it is also reasonable to consider that because subspecies B is not commonly found in humans and most molecular epidemiology studies on enterococci are centered around human clinical isolates, our understanding of its niche breadth and ecological roles within a one-health continuum is currently limited.

Drawing from the existing literature and available metadata on strains classified as subspecies B, including the STs represented in our study as well as their single and double-locus variants cataloged in the PubMLST database (https://pubmlst.org/), it appears that this subspecies is predominantly associated with animal-related sources. Notably, it has been isolated not only from meat and meat processing facilities but also directly from cloacal samples of chickens and wild birds such as common kingfishers (Alcedo atthis, a species within the Coraciiformes order). These reports are predominantly linked to ST228 (83–85).

The isolation of the ST624 strain from a mouse GIT, as documented in the NCBI under the identifier SAMEA13256522 (DSM111623 biosample), contributes to the understanding that subspecies B may also be associated with non-human mammals, in addition to livestock. Notably, we found no reports of subspecies B-associated STs being isolated from humans, with the sole exception of the ST228 SLV, ST247, found in a subgingival plaque sample from a patient with marginal periodontitis in Norway (86). This observation underscores the rarity of human association for subspecies B, reinforcing the notion of its distinct ecological profile.

The consistent absence of critical genetic determinants associated with E. faecalis persistence in hospital environments, enhanced pathogenicity, and treatment failures in HAIs aligns with the universal presence of the CRISPR3-Cas system across all subspecies B genomes examined. Johnson et al. (2021) (5) have reviewed how the evolution of enterococci, particularly E. faecalis and E. faecium, into MDR hospital-acquired pathogens is largely driven by the accumulation of MGEs, including plasmids, transposable elements, and phages, which serve as vehicles for the spread of ARGs and virulence factors. Given that these traits are predominantly acquired through HGT, and the CRISPR3-Cas systems are known to restrict HGT in E. faecalis (87), it seems that subspecies B genomes have been shaped by selective pressures markedly different from those maintaining the high-risk lineages of subspecies A.

An additional facet of this divergence can be seen in the variability of the ace gene between subspecies A and B. The Ace protein of E. faecalis mediates binding to host extracellular matrix proteins, such as collagen types I and IV and laminin (88, 89), playing a crucial role in the colonization and infection of various tissues (90, 91). Notably, the detection of ace genes in subspecies B genomes with significantly lower similarity to the VFDB reference sequence (NP_814829) compared to the ace sequences in subspecies A representatives may suggest tropism for different host environments, especially because this gene is known to be highly conserved among E. faecalis isolates (92).

Moreover, individual gut bacteria are thought to exhibit divergence patterns aligned with host phylogeny, especially during allopatric speciation, which limits bacterial dispersal between hosts and gradually leads to reproductive isolation (93). Given that enterococci are hypothesized to have been core members of the gut microbiome in the last common ancestor of mammals, birds, reptiles, and insects (94), it is plausible that the divergence between subspecies A and B was driven by host-specific adaptations and reduced gene flow over time. The ubiquitous presence of the CRISPR3-Cas system among subspecies B strains could further support this reduced genetic exchange, offering an alternative or complementary explanation for the reproductive restriction and cladogenesis that culminated in the divergence of E. faecalis strains at the subspecies level.

To further investigate the ecological adaptations that differentiate subspecies A and B, within the framework of reverse ecology, we conducted a pangenome analysis to identify genes enriched in each subspecies. This analysis revealed that both subspecies are enriched with genes distributed across the same 20 COG functional categories, suggesting that, in addition to the core genome, each subspecies has developed its own genetic repertoire supporting similar functional capacities. This finding is consistent with their independent evolutionary paths. Notably, subspecies A exhibited a significantly higher proportion of genes involved in carbohydrate transport and metabolism, supporting the hypothesis that subspecies A may be better adapted to environments with greater availability or diversity of carbohydrates, such as the human gut (95–97). In contrast, this suggests that subspecies B may be adapted to more restricted environments where carbohydrate variety is limited, or where alternative carbon sources are more prevalent, such as the guts of animals with more restrictive diets compared to humans, or ecosystems under distinct selective pressures (98). Carbohydrate availability in the host gut has been indicated as a major driver of enterococcal speciation since the emergence of the genus (99), which enhances the plausibility of our findings.

Although these hypotheses align with our primary observations, it is important to note that nearly half of the genes enriched in subspecies B strains were not classified within any COG category. While this finding limits precise inferences regarding subspecies B’s ecological adaptations, it also provides compelling evidence of this lineage’s distinct evolutionary history and adaptive strategies. The significantly higher proportion of unclassified genes in subspecies B, compared to subspecies A, suggests that subspecies B may harbor novel genetic elements that have yet to be fully characterized. These unclassified genes could reflect adaptations to niche-specific pressures, possibly in environments that differ considerably from the well-studied human-associated habitats where subspecies A thrives. Moreover, the absence of functional annotation highlights the possibility that subspecies B is adapted to more specialized or less predictable environments, where distinct selective pressures —such as those found in non-human hosts or natural ecosystems— shape the evolution of unique genetic repertoires. Further research is required to elucidate the roles of these genes and their contributions to subspecies B’s fitness in specific conditions, which will enhance our understanding of E. faecalis niche breadth and its complex eco-evolutionary dynamics.

In line with these considerations, it is also relevant to address the observed variability in alleles of MLST housekeeping genes in subspecies B, which may provide additional insights into the ecological divergence between the subspecies. All subspecies B isolates share the same allelic variants of the gki and yqiL genes, which are key housekeeping loci in the E. faecalis MLST scheme (80). Interestingly, Fertner et al. (2011) demonstrated that ST228 — the most frequent ST within subspecies B — had the most phylogenetically distant relationship compared to other E. faecalis lineages based on a concatenated phylogeny of the seven MLST housekeeping genes (85). While this study did not directly link the phylogenetic divergence to distinct subspecies, our findings suggest that such genetic distance reflects an ancient and consistent divergence between subspecies A and B, particularly considering that housekeeping genes tend to be highly conserved, with slower rates of evolution compared to other genomic regions (17, 100, 101).

Building on the analysis of genes enriched in each subspecies, we applied a more stringent filter to identify genes present exclusively in one subspecies and absent in the other, across all representative genomes (i.e., clade/subspecies-specific genes, likely synapomorphic traits, and candidate niche-specifying genes) (17, 73). This approach aimed to pinpoint genes potentially associated with the primary niche of each subspecies—those likely acquired by the common ancestor and maintained through periodic selection due to their adaptive value (17, 73). Unlike the broader enrichment analysis, this method focused on genes that may have played a critical role in initiating ecological isolation and, thus, speciation, which is discussed here in light of the concept defined by Shapiro and Polz (2014) as “any stage of the dynamic process of ecological and genetic differentiation” (51).

The prediction of subspecies-specific genes and allelic variants provides valuable targets for understanding the genetic basis of subspecies B’s emancipation and unique ecology. Initially, certain proteins were classified as distinct orthologous groups, thought to be exclusive to each subspecies. However, following reannotation, these proteins were reclassified as products of subspecies-specific allelic variants within the same orthologous group, rather than distinct genes. This finding highlights the potential roles of both gene acquisition or loss events, and sequence variation in pre-existing core genes, in delineating the ecological niches of each subspecies, as supported by Cohan (2002) (102).

Although we cannot confirm whether such allelic variations correspond to actual adaptive mutations or to fixed neutral mutations solely based on these findings, the high degree of non-synonymous substitutions between subspecies A and B sequences suggests potential functional divergence. Even if these variants were neutral in the early stages of ecological speciation, they may have acquired adaptive significance over time, gradually shaped by distinct selective pressures and reflecting the niche specificity of each subspecies (103). It is important to note, however, that not all non-synonymous substitutions necessarily lead to functional changes (104). Therefore, further experimental validation is essential to determine the extent to which these variants contribute to bacterial adaptation (105). Investigating these potential functional changes could significantly enhance our understanding of subspecies-specific ecological adaptation in E. faecalis.

Several proteins encoded by subspecies-specific gene variants are potentially linked to stress response and host colonization. One example is the EF1546-encoded protein, which contains a LysM domain, a well-characterized motif in autolysins responsible for binding bacterial cell wall peptidoglycan and facilitating precise catalytic cleavage during daughter cell separation (106). Controlled septum cleavage is critical for defining cell size and chain length in E. faecalis, factors directly impacting bacterial fitness (106). Mutations within the LysM domain have previously been associated with elongated cell chains, impaired immune evasion, and reduced virulence in E. faecalis (106–109). Given that such phenotypic alterations can profoundly affect bacterial morphology and its interaction with host defenses, the divergence in LysM-containing proteins between subspecies A and B may underlie their differential success in host colonization.

Subspecies-specific sequence divergence was also observed in key transcriptional regulators, which may influence the adaptive responses of E. faecalis to varying environmental pressures. The EF1369 gene, encoding a Cro/CI family transcriptional regulator, has been linked to stress tolerance, including survival under high-salt conditions (8% NaCl), acidic environments (pH 2.8), and within mouse peritoneal macrophages (110, 111). Similarly, the EF0876 gene, encoding a Mga helix-turn-helix protein, plays a crucial role in virulence regulation, particularly in host tissue adhesion in response to carbon metabolism (112, 113). Notably, inactivation of EF0876 resulted in a 100-fold reduction in gut colonization in mice in a previous study (113), highlighting its importance for E. faecalis fitness in the GIT. Mga family regulators, including key E. faecalis regulons like EbpR (114), are activated by elevated CO₂ levels (115, 116), typically found inside mammalian hosts (∼5-6%), as opposed to atmospheric levels in natural settings (∼0.036%) (117, 118). Given the significant role of EF1369 and EF0876 in mediating E. faecalis colonization and survival under mammalian host conditions, and considering that these findings are based on subspecies A (i.e., strains derived from the V583 clinical isolate) (110, 113), subspecies-specific sequence divergence in these regulators could impact the fitness of subspecies B under similar selective pressures. This is consistent with the lineage’s unusual isolation from human hosts, and might reflect the genetic basis supporting ecological niche distinctions between the subspecies.

While the adaptive significance of subspecies-specific allelic variants of core genes is less predictable, the acquisition of entirely novel genes encoding new cellular functions is more likely to enable a strain to exploit previously inaccessible resources and establish itself within a new ecological niche (i.e., niche-specifying genes) (17, 51, 102). From this perspective, we highlight our main inferences and speculations regarding the gains and losses of key genetic determinants and the corresponding selective pressures driving the ecological separation of E. faecalis subspecies, as discussed below.

Interestingly, the E. faecalis V583 EF2063 homologue, identified in the present study as subspecies A-specific, is one of the flanking regions of the CRISPR2 locus in E. faecalis genomes (28), whose absence we propose as a genomic marker for subspecies B. The absence of this region in subspecies B suggests that it may have been lost in a common ancestor through a single recombination event, akin to the mechanism of CRISPR1-Cas and CRISPR3-Cas variability described previously (28). While the precise functions of both the EF2063 gene (predicted to encode an AraC-family transcriptional regulator) and the orphan CRISPR2 locus remain unclear, we speculate that their combined loss in subspecies B could reflect a selective pressure favoring the genotype lacking this region, potentially dispensable or deleterious in its primary niche. This hypothesis is supported by the recent description of a wild-type E. faecalis strain harboring all three canonical CRISPR loci (CRISPR1-Cas, CRISPR2, and CRISPR3-Cas), leading the authors to conclude that CRISPR genotype variation in the species is likely due to CRISPR loss rather than locus acquisition by HGT (119, 120). However, fitness trade-offs associated with CRISPR variability in E. faecalis have been primarily attributed to the presence or absence of functional CRISPR systems, not the orphan CRISPR2 (5). Thus, any further exploration of how CRISPR2 variation contributed to the ecological separation of the subspecies must also consider the role of EF2063 in this process.

Similarly, building on the understanding that functionally related genes are often arranged contiguously in the genome, reflecting potential co-selection under specific environmental pressures (121), we observed that many of the subspecies A-specific genes exhibited synteny, (i.e., based on the gene arrangement of corresponding homologs in the V583 genome sequence) (57). This suggests that they may have been selected together in response to a specific ecological pressure, either gained by the most recent common ancestor of subspecies A or lost in subspecies B.

For instance, the proteins encoded by EF1919 and EF1920 were annotated as a GNAT family acetyltransferase and a C4-dicarboxylate anaerobic carrier DcuC, respectively. While GNAT members are involved in various cellular processes, and this one has not been linked to any specific pathway so far, DcuC carriers are mainly responsible for succinate efflux produced during glucose fermentation (122). Interestingly, most bacteria containing dcuC homologs are pathogenic and all known DcuC family members are from enteric bacteria (122). This pattern supports the hypothesis that this locus may be integral to a key carbon and energy metabolic pathway in subspecies A strains, particularly under anaerobic conditions in the gut, the primary habitat of E. faecalis where it occurs as both a commensal and opportunistic pathogen (1, 123). In contrast, the absence of these genes in subspecies B could indicate a competitive disadvantage for gut colonization, compared to subspecies A.

Interestingly, many other subspecies A-specific genes seem to encode ecologically relevant determinants involved in E. faecalis ability to either colonize, persist or cause infection in a host environment. The EF1672 gene, encoding an ABC superfamily protein permease, was also exclusively and universally detected within subspecies A genomes. Notably, transcriptomic data have supported this gene’s crucial involvement in E. faecalis physiological adaptation during the course of infection in a mammalian host model, being hypothesized to be one of the core genes required for the species to thrive within a host, and thus a potential target for antimicrobial agents and vaccines to treat and prevent enterococcal infections (124).

Another example implies that subspecies B strains might lack or have a reduced ability to utilize sucrose as a carbon and energy source. This is supported by the exclusive detection of determinants of two operons encoding a sucrose PTS (phosphoenolpyruvate: sugar phosphotransferase system) transporter (EF1602) and sucrose metabolism (EF1603, also named scrB-1) within subspecies A strains. Previous reports have shown that both loci likely contribute to virulence since they were up-regulated in E. faecalis strains grown in human urine (125), where sucrose levels can be increased in high-sugar diets (126). Knock-out mutants of EF1603-04 showed reduced virulence in a Caenorhabditis elegans infection model (127), which corroborates the differential role of these loci in E. faecalis pathogenesis associated with sucrose’s contribution to growth in infection. Additionally, studies have demonstrated that sucrose utilization enhances the expression of E. faecalis virulence-associated determinants and the production of biofilm matrix components (e.g., eDNA and EPS) in biofilms (128). Altogether, these findings support the genetic basis of subspecies A pathogenic potential in contrast to subspecies B by highlighting the contribution of sucrose uptake and metabolism in infections caused by E. faecalis, which seems consistent with this species (specifically, subspecies A) being a notorious agent of UTIs and aligned with the relevance of biofilm formation in the context of HAIs (82).

Additionally, the subspecies A-specific EF1138 gene encodes an oxidoreductase of the aldo/keto reductase family, strongly related to E. faecalis V583 stress response induced by bovine bile exposure, being potentially implicated in bile salt modification (129, 130). Bile tolerance mechanisms are critical for bacterial survival and colonization of the GIT, where bile’s emulsifying action poses a challenge (131). The absence of EF1138 in subspecies B may suggest that bile is not a strong selective pressure in its core ecological niche, or that subspecies B has evolved alternative mechanisms to tolerate bile exposure or to cope with different bile salt compositions, which includes the possibility of adaptation to a different host range (132–134).

Considering that most studies on E. faecalis-host interactions have focused on its pathogenic potential in human infections and often rely on conventional animal models, primarily mammals (e.g., mice and rabbits), it is important to acknowledge that conclusions from these studies are inherently biased by the unique physiology of each host species (135). While the absence of subspecies A-specific genes in subspecies B suggests reduced fitness in humans and other mammalian hosts, it remains uncertain whether these genes are essential for colonization and survival across a broader host range, particularly in non-mammalian vertebrates (97, 136). This knowledge gap limits our understanding of subspecies B’s ecological niche and adaptability.

The isolation sources of subspecies B, which include wild birds, poultry, and meat processing environments, suggest a potential adaptation to hosts or environments beyond humans (83–85). Although the specific ecological role of birds in the lifecycle of subspecies B remains uncertain, previous studies support the hypothesis that wild migratory birds act as reservoirs for genetically diverse E. faecalis strains. Some of these strains are prevalent in poultry, indicating possible transmission routes from wild to domestic birds (16, 137). These findings reinforce the idea that subspecies B may have evolved to thrive in non-mammalian hosts, particularly avian species, whose gut microbiota composition is consistently distinct from that of mammals (138).

Furthermore, according to conclusions of a study by Chaillou and colleagues on meat microbiota (139), subspecies B’s common isolation from meat products and processing environments likely points to two primary sources of contamination: (1) animal-derived microbiota, particularly from the skin or gut, which may include Enterococcus, and (2) environmental psychrotrophic bacteria, largely represented by Firmicutes associated with water reservoirs, plants, or plant-derived animal feed (139). These hypotheses are not mutually exclusive in the context of E. faecalis, but rather suggest multiple potential dissemination pathways. This also implies that subspecies B may have undergone selective adaptations to thrive in such environments, where particular pressures could have driven and maintained its ecological and genetic differentiation over time (51).

Our efforts to obtain further insights into the unique ecological niche of subspecies B focused on the functional annotation of its specific genes, a reverse ecology approach (17, 73). While many of these genes encode proteins with uncharacterized functions, functional predictions were made based on putative homologs and conserved domains. Despite the significant homology between subspecies B sequences and their closest matches in the UniProtKB database, strongly supported by low E-values in most cases (54), notable differences in sequence identity were observed, making functional predictions less definitive. Nevertheless, the detection of conserved protein signatures provided valuable insights into potential ecological roles. Interestingly, most of the predicted homologs were derived from taxa commonly associated with meat microbiota and the contamination routes proposed by Chaillou et al., such as Vagococcus, Carnobacterium, Leuconostoc, various enterococcal species, and others (139). This correlation suggests that these genetic determinants may confer adaptations to shared ecological niches, potentially aiding subspecies B in coping with the selective pressures of these environments.

One of our most compelling findings was the detection of sequences potentially encoding the toxin-antitoxin (TA) protein components HicA and HicB, along with a ClpP/crotonase-like domain-containing protein. These elements may play a coordinated role in stress responses linked to environmental conditions such as nutrient deprivation and heat shock (140, 141). It has also been suggested that TA systems, including HicAB, are typically abundant in free-living bacteria but tend to be lost in host-associated prokaryotes (142, 143). This raises the possibility that subspecies B’s recurrent isolation from meat products could be influenced by environmental contamination routes, potentially involving water, plants, or soil.

Additionally, the ubiquitous detection of the CRISPR3-cas system within subspecies B further supports the contribution of environmental pressures, as functional CRISPR-Cas systems are known to provide defense against lytic phages, which are particularly abundant in water and soil environments (144–146). These environments could impose strong selection, contributing to the maintenance of CRISPR-mediated defense mechanisms. Although the CRISPR3-cas system was not exclusive to subspecies B, we observed that its Cas proteins (Cas1, Cas2, and Csn2) clustered distinctly from those of subspecies A, showing greater similarity to those found in Vagococcus elongatus, a species originally isolated from a swine-manure storage pit (147), which aligns with the transmission routes involving either animal or environmental reservoirs to meat microbiota, as proposed by Chaillou and colleagues (139). This divergence could reflect adaptations to different ecological pressures, though further investigation is needed to clarify the specific ecological role of these systems.

We also identified subspecies B-specific traits that may be involved in the uptake and metabolism of alternative carbohydrates when glucose is scarce, which could suggest an adaptive advantage in diverse environments. Notably, we detected protein clusters resembling the IIABC component (BglF-like permease) of a PTS beta-glucoside-specific system and a PRD (PTS regulation domains)-containing protein (LicT-like transcriptional antiterminator) (148, 149). These elements likely participate in the transport and positive regulation of genes linked to beta-glucoside metabolism, responding to substrates like salicin and arbutin found primarily in plant material but rarely in mammals (150).

Moreover, we observed the exclusive presence of putative carbohydrate-binding proteins in subspecies B, such as the GlcNAc-binding protein A (GbpA) and a WxL domain-containing protein. The GbpA homolog contains domains frequently associated with cellulose- and chitin-binding capabilities, polysaccharides predominantly found in plants and arthropods, or fungi, respectively (151, 152). In Vibrio cholerae, GbpA facilitates adhesion to zooplankton chitinous surfaces in aquatic habitats and to GlcNAc moieties on human intestinal epithelium (153). This dual functionality strongly suggests an exaptive process where a protein initially evolved for environmental persistence (e.g., attachment to chitin) later adapted to facilitate host colonization (154). By analogy, a similar exaptation in E. faecalis subspecies B could enable these bacteria to persist in environmental reservoirs (potentially associating with arthropods or adhering to plant cell walls) and also to colonize animal gut surfaces via GlcNAc binding, supporting its ecological adaptability across diverse environments (153, 154).

The WxL domain-containing proteins, commonly associated with adhesion to cellulose and xylan in other Gram-positive bacteria, such as Enterococcus faecium (155), reinforce this notion. These polysaccharides, abundant in plant cell walls, are often present in animal diets as dietary fibers. Interestingly, the ability of gut bacteria to adhere to these fibers confers a competitive advantage, especially in environments where plant-based feeds are supplemented with xylanase, facilitating the growth of beneficial microbes like lactobacilli (155, 156). Thus, the exclusive presence of a WxL domain-containing protein in subspecies B suggests that these bacteria might have adapted to adhere to dietary fibers in the gastrointestinal tract of animals, potentially contributing to their isolation from meat products.

Collectively, the subspecies B-specific GlcNAc-binding and WxL domain-containing proteins exemplify how exaptation might facilitate bacterial adaptation to shifting habitats, supporting the two potential transmission routes proposed by Chaillou et al. (2015). These traits could either enable direct environmental persistence, leading to contamination of meat products, or facilitate colonization of animal hosts, with subsequent transmission through the food production chain. However, it is important to note that identifying these proteins as exaptive traits remains speculative and requires further investigation for confirmation. Nevertheless, considering such possibilities underscores the adaptive potential of subspecies B and highlights the value of genomic surveillance across diverse settings to understand the evolutionary mechanisms driving E. faecalis adaptation, especially from a One Health perspective.

Our findings also revealed specific subspecies B traits that suggest the existence of an unique repertoire of surface proteins, including several containing the LPxTG cell wall anchor motif, a feature typically associated with cell surface adhesion in Gram-positive bacteria (157). This repertoire appears to include proteins that may mediate adhesion to various substrates, as inferred from other conserved domains identified in their sequences. For example, one protein exhibits the fimbrial isopeptide formation D2 domain, which may facilitate pili subunit cross-linking and mediate surface adhesion to host structures (158, 159). Another predicted protein contains a T-Q ester bond domain, an unusual bond type thought to stabilize pilin subunits and commonly found in cell surface adhesion proteins of Gram-positive bacteria (160), similar to a known collagen adhesin from Bacillus cereus.

Moreover, we identified three proteins with MucBP domains, matching putative adhesins encoded by other enterococcal species. Notably, two of these proteins showed the highest similarity to a sequence from Enterococcus phoeniculicola ATCC BAA-412, a strain isolated from the uropygial gland of wild Red-billed Woodhoopoes (Phoeniculus purpureus) (161). This finding supports our earlier suggestion that subspecies B may be adapted to avian hosts, as the presence of proteins homologous to those in a bird-associated E. phoeniculicola highlights the potential role of birds as hosts or reservoirs for subspecies B. MucBP domain-containing proteins are generally linked to adherence in the mucosal environment of the GIT, providing a potential mechanism for subspecies B to colonize this niche. Notably, previous studies have identified these proteins as common in other commensal and environmental Gram-positive species, such as Lactobacillus plantarum (162–164). This suggests that subspecies B might exhibit similar ecological flexibility, capable of thriving in a range of environments, including the GIT of animals.

Finally, two subspecies B-specific protein clusters were similar to predicted ABI family (Rce-like) proteins, both classified as type II CAAX prenyl endopeptidases. ABI proteins in enterococci have been previously associated with defense against phage infections (165), but there is also evidence suggesting their involvement in bacteriocin self-immunity mechanisms (166). This suggests that subspecies B might be able to produce additional bacteriocins. Unlike antibiotics, bacteriocins exhibit a narrow antimicrobial spectrum, targeting closely related bacterial strains (167). Such specificity implies that bacteriocin production might confer a competitive advantage by selectively inhibiting competitors within the same ecological niche. Additionally, bacteriocin production in Gram-positive bacteria is often linked to the transition from the logarithmic to stationary growth phase (168), highlighting their role as potent killing agents in resource-limited environments. Thus, the presence of subspecies B-specific proteins potentially involved in bacteriocin self-immunity hints at the possibility that this clade may possess a distinct competitive advantage in microbial communities, either in host-associated microbiomes or in environmental settings (169, 170).

Our findings suggest that the niche breadth of E. faecalis is likely underestimated, potentially encompassing a range of unexplored or insufficiently studied hosts and environments, which may explain the infrequent reporting of subspecies B-associated STs. By employing a reverse ecology approach, we have provided several insights into the hypothetical ecological properties and potential niches of subspecies B, aligning with current guidelines for the taxonomic description of prokaryotes from genome data (78).

Nonetheless, the limitations inherent to our methods must be acknowledged. Specifically, the focus on clade-specific genes to identify past adaptations conserved through periodic selection inevitably overlooks other genetic determinants that may have contributed to the lineage’s evolution and differentiation, as highlighted by Lassalle et al. (2015). Additionally, experimental validation is essential to confirm the speculated roles of clade-specific genes or alleles and corresponding selective pressures in driving the ecological differentiation between E. faecalis subspecies A and B, following similar approaches used in related studies (73, 99, 171, 172).

Despite these limitations, our study highlights how genome-based exploratory approaches, utilizing publicly available data, can provide novel insights into the eco-evolutionary dynamics of bacterial species while simultaneously revealing important knowledge gaps. In the present study, analyzing CRISPR system genetic patterns across diverse genomes (spanning a broad range of STs, isolation sources, etc.) within the operational definition of a single species enabled us to identify and partially characterize a distinct cluster of E. faecalis strains, corresponding to a phylogenetically, ecologically, and taxonomically cohesive unit —subspecies B, using approaches proposed by previous studies (17, 51, 73, 78).

Although previous works have depicted the phylogenetic prominence of subspecies B members, particularly ST228 strains, compared to other E. faecalis clusters (83, 85), neither the subspecies-level classification of this lineage, nor its positioning within the broader landscape of E. faecalis’s population structure have been acknowledged. Addressing this gap, our findings reinforce the need to reassess E. faecalis’s population structure in light of hidden genetic diversity in under-sampled environments and hosts (e.g., nonhuman isolates), as suggested by other studies (16, 123). Given the role of generalist species in microbial dispersal and their propensity to give rise to specialists confined to a narrower range of habitats (173, 174), it is plausible that the current view of E. faecalis as a species with little phylogenetic divergence and lacking a multiclade structure (15, 175) stems from an underrepresentation of its true biological diversity. This underrepresentation likely obscures the existence of host- or environment-specific clades in advanced stages of speciation, such as subspecies B.

Expanding intraspecific genomic surveillance to include isolates from nonhuman hosts and nonclinical environments will foster a more realistic understanding of the species’ population structure. Such efforts are critical from a One Health perspective, especially given previous evidence that E. faecalis’s adaptation to the hospital environment may be a byproduct of its evolution in a wider range of ecological niches (6, 99, 175). Investigating the existence of niche-specific E. faecalis clades can help predict and monitor potential adaptive novelties that may precede the emergence of high-risk lineages, allowing for more targeted prevention and control strategies aimed at curbing the spread of these genetic determinants and associated lineages (176, 177).

While opportunities for genetic exchange between subspecies A and B are likely provided in overlapping compartments of their multidimensional niche breadths (178), it remains to be elucidated whether and how gene flow between individuals of these clades is maintained.