Virulence Determinants of Enterococcus Faecium

Abstract

Enterococcus faecium, a member of the human gut microbiota, has emerged as a notable opportunistic pathogen, contributing to a diverse range of hospital-acquired infections. Its capacity to thrive in various anatomical sites and initiate infections is attributed to an elaborate suite of virulence determinants. Prominent among these are cell surface components and pili structures, which facilitate initial adhesion and subsequent biofilm formation. Additionally, temperature-regulated gene expression augments virulence by enhancing adherence and biofilm formation. E. faecium also employs sophisticated mechanisms to modulate host immune responses, including hindering leukocyte killing through membrane structures like lipoteichoic acids and capsular polysaccharides. Bacteriocins confer a competitive edge by inhibiting competing bacteria, while global regulators orchestrate biofilm formation and stress responses. The stringent response further enhances adaptation to stress conditions. Understanding these virulence factors is paramount for unraveling the intricacies of E. faecium infections and devising effective therapeutic strategies.

1. Introduction

Enterococci, typically considered benign members of the human gastrointestinal microbiota, have undergone a remarkable shift in their role, emerging as significant nosocomial pathogens worldwide [1]. Enterococcus faecalis and Enterococcus faecium have particularly risen in prominence due to their propensity for causing difficult-to-treat infections, facilitated by innate and acquired antibiotic resistance mechanisms [2]. Although E. faecalis historically dominated enterococcal infections, E. faecium has now become a significant part of the equation, being responsible for approximately 40% of enterococci-related cases [3, 4, 5, 6]. The emergence of vancomycin-resistant enterococci (VRE) has further complicated matters, revealing the ability of outbreak-related isolates to acquire unique genetic repertoires, enhancing their survival within healthcare environments.

Traditionally, research on enterococcal virulence factors has focused on E. faecalis, given its well-established pathogenic role. These factors, including secreted toxins, surface molecules, stress response proteins, and gene regulators, collectively contribute to enterococcal pathogenicity. However, our understanding of virulence determinants in E. faecium lags behind that of E. faecalis, despite its increasing clinical significance. As for E. faecalis, pathogenicity of E. faecium stems from a diverse array of virulence factors, enabling adhesion to host tissues, evasion of the immune system, and establishment of persistent infections [7]. Key among these factors are cell surface components and secreted proteins, which play pivotal roles in mediating interactions between the pathogen and its host and facilitate initial adhesion and biofilm formation, enhancing the pathogen’s ability to persist and cause chronic infections [8, 9, 10, 11, 12, 13, 14]. E. faecium also produces bacteriocins which confer a competitive advantage by inhibiting the growth of competing bacteria as well as temperature-regulated gene expression that enhances adherence and biofilm formation, thereby augmenting bacterial virulence [15, 16]. In addition, the intricate interplay between E. faecium and the host immune system involves sophisticated mechanisms aimed at modulating host immune responses [17, 18]. Understanding these multifaceted interactions is crucial for deciphering mechanisms driving E. faecium infections and for devising effective prevention and treatment strategies. Therefore, this chapter explores the diverse array of virulence factors employed by E. faecium to colonize host tissues, evade immune defenses, and instigate infections (also see Figure 1 and Table 1 for overview). By focusing exclusively on virulence factors experimentally validated using animal models, we aim to provide insights into potential targets for therapeutic intervention and vaccine development, thereby addressing the pressing clinical challenge posed by E. faecium infections.

Figure 1.

Overview of the virulence factors in E. faecium. They are depicted as families or pathways in which they have been described to play a role: Bacteriocins (blue), biofilm (orange), cell surface components (red), general regulators (pink), immune modulators (green), and secreted proteins (beige pink). The MSCRAMM class of proteins is highlighted with a yellow star, and the proteins containing an LPXTG motif are tagged with a blue square.

Category	Name	Product/designation	Gene ID (Aus0004)	Gene size	In vitro	In vivo	Putative role	Reference
Cell surface components in bacterial adherence	acm	Adhesin of collagen from E. faecium	EFAU004_02292	2166	X	X		[19]
	scm	Second collagen adhesin from E. faecium	EFAU004_02838	1983	X			[20]
	ecbA	Efm collagen-binding protein A	EFAU004_01867	3228	X			[12]
	empA	PilB subunit A	EFAU004_00930	3390	X	X		[10, 11]; [10, 11]; [10, 11]
	empB	PilB subunit B	EFAU004_00931	1422	X	X
	empC	PilB subunit C	EFAU004_00932	1878	X	X
	pilA	PilA subunit A	EFAU004_p1010	1974			X	[21]
	pilE	PilA subunit E	EFAU004_p1012	759			X	[21]
	pilF	PilA subunit F	EFAU004_p1013	2085			X	[21]
	prpA	Proline-rich protein A	EFAU004_00880	1290	X			[13]
	sgrA	Serine-glutamate repeat-containing protein A	EFAU004_01513	975	X			[12]
	esp	Enterococcal surface protein	EFAU004_02750	4938	X	X		[14, 22, 23]
	aggE	Aggregation promotting factor	—		X			[24]
	lwpa	Large WxL protein A	EFAU004_02118	2481	X	X		[25]
	swpa	Small WxL protein A	EFAU004_02121	681	X	X		[25]
	lwpB	Large WxL protein B	EFAU004_00588	3168			X	[25]
	swpB	Small WxL protein B	EFAU004_00591	759			X	[25]
	lwpC	Large WxL protein C	EFAU004_00154	2010			X	[25]
	swpC	Small WxL protein C	EFAU004_00153	729			X	[25]
Secreted proteins	hylEfm	Glycosyl hydrolase with β-N-acetylglucosaminidase activity	—		X			[26]
	sagA	Peptidoglycan hydrolase	EFAU004_02613	1539	X	X		[27]
Bacteriocinsa	entA	Enterocin	—		X			[28, 29]
	entP	Enterocin	—		X			[30, 31]
	bac43	Bacteriocin	—		X			[30, 32]
	RC714	Bacteriocin	—		X			[33]
	bac32	Bacteriocin	—				X	[30]
	entB	Enterocin	.		X			[34]
	bac51	Bacteriocin	.		X			[35]
	T8	Bacteriocin	.				X	[36]
Immune modulation	capD	Capsular polysaccharide protein D	EFAU004_01373	1008	X	X		[37]
	LTA	Amphiphilic glycoconjugate polymers	—		X	X		[38]
Biofilm regulators	ebrB	Enterococcal biofilm regulator B	EFAU004_02752	9112	X	X		[39]
	bepA	Biofilm and endocarditis-associated permease A	EFAU004_00700	1419	X	X		[40]
	atlAEfm	Autolysin	EFAU004_02584	2019	X			[41]
General regulators	asrR	Antibiotic stress response regulator	EFAU004_00882	441	X	X		[42]
	gls33-glsB gls20-glsB1	General stress proteins	EFAU004_01475	903	X	X		[43]
	ccpA	Catabolite control protein A	EFAU004_02017	1002	X	X		[44]
	ptsD	Sugar-specific membrane-associated subunit (enzyme IID) of PTS^clin	EFAU004_00680	822	X	X		[45]
	ern0160	Bacterial regulatory RNA	—	386		X		[46, 47, 48]
	relA	Alarmone	EFAU004_02500	2214	X			[49]

Table 1.

Virulence genes in E. faecium.

^a

EntA, EntP, Bac43, Bac32, EntB, and Bac51 are composed of two genes encoding for the bacteriocin precursor and the immunity protein. T8 is comprised of 4 genes encoding for the bacteriocin, the immunity protein, the mobilization protein, and the relaxase nuclease.

Name, locus tag, function and link with virulence are presented. Of note that some bacteriocins/enterocins are not found in E. faecium Aus0004 reference strain.

2. Cell surface components

Bacterial adherence marks the initial step in the infectious process, facilitating the colonization of various niches within the human body and fostering the development of biofilms. Strategies for cell adhesion exhibit a wide spectrum, encompassing both individual proteins and complex multimeric macromolecules, each endowed with intricate functions [50]. In Gram-positive bacteria, numerous proteins anchor the bacterial cell wall via an LPXTG motif. The motif undergoes cleavage by a transpeptidase named sortase, ensuring accurate localization and effective binding to the cell wall [51].

A subset of these proteins forms the microbial surface components recognizing adhesive matrix molecules (MSCRAMMs), engaging in ligand binding through mechanisms such as “dock and lock” or “collagen hug”, resulting in robust interactions with ligands. MSCRAMMs are ubiquitous among Gram-positive bacteria and are associated with diverse substrates [8]. In E. faecium, notable members include Acm, Scm, EcbA, and PilB, although additional homologous proteins have been identified [20].

The cell wall anchored collagen adhesin Acm (adhesin of collagen from E. faecium) is a well-known virulence factor of E. faecium, exhibiting strong binding affinity to type I collagen and, to a lesser extent, to type IV collagen. Interestingly, Acm shares a higher protein homology with Cna of Staphylococcus aureus (62% overall similarity) than with Ace of E. faecalis (47% overall similarity), despite the similar functions of these proteins [52]. In vivo endocarditis models have demonstrated that a Δacm mutant exhibits impaired initial adherence crucial for valve colonization and vegetation establishment [53]. Studies have reported a higher prevalence of the acm gene in infectious strains compared to non-infectious strains (i.e. food or stool origins), particularly within clade A1 (formerly CC17) [19, 54]. However, Freitas et al., conducted a large retrospective study identifying a similar rate of acm genes in both infectious and non-infectious strains, albeit with a higher rate of complete genes, with an entire sequence, observed in clinical setting [55]. Notably, IS-mediated disruptions of the acm gene are frequently observed in isolates of non-infectious origins [19]. The second collagen adhesion, Scm, confers binding to type V collagen and, to a less extent, to type I collagen and fibrinogen. Another collagen adhesin, EcbA, facilitates binding to type V collagen and fibrinogen. Despite limited data on these adhesins, Scm appears to be particularly relevant in clinical settings, with a higher prevalence of scm in clinical isolates compared to community-derived isolates. Conversely, no notable difference in the prevalence of ecbA was observed according to their origin [12, 55, 56].

Pili, large multimeric proteins located on the cell surface, form adhesive hair-like structures. In E. faecium, these are several pilus gene clusters located on pathogenic genomic islands, in contrast to E. faecalis where a single pili gene cluster has been identified. Pili exhibit high protein sequence homology (>70%) among enterococci and are highly conserved within each species [9]. In E. faecium, two types of pili, PilA and PilB, have been identified. PilB, encoded by empABC genes (formerly known as the endocarditis and biofilm-associated pilus operon ebpABC_fm), comprises three different subunits including the major sub-unit EmpC and the putative accessory subunit EmpA and EmpB.

The ΔempABC mutant exhibits impaired initial adherence and biofilm formation. Furthermore, PilB promotes infection of kidneys and bladder in a urinary tract infection (UTI) mice model [10]. Adherence and biofilm defects were also observed in ΔempA and ΔempB mutants albeit not in the ΔempC mutant. While EmpC is dispensable for biofilm formation, it constitutes the pili backbone. Nonetheless, all three genes are essential for the full virulence of E. faecium in a UTI mouse model, with the EmpA subunit at the top being the primary determinant for biofilm formation thereby facilitating the development of vegetation in an endocarditis rat model [11]. The second type of pili, PilA, is also encoded within a gene cluster comprising pilA, pilE, and pilF genes. PilA forms the backbone while PilF is located at the top. Both subunits assemble into a high-mass pili structure, although the role of PilE remains undetermined [21]. Notably, PilA and PilB polymerization at the cell surface occurs exclusively at 37°C with pili being undetectable in transmission electron microscopy in cells grown at 21°C. This suggests that pili are not expressed in the environment but may be upregulated in vivo to enhance bacterial colonization and infection [21]. Interestingly, pilA and empABC genes are more frequently encountered in clinical isolates [54, 55].

Temperature-regulated gene expression appears pivotal in enterococcal colonization. Another protein upregulated at 37°C, identified through transcriptomic analysis, is the proline-rich protein PrpA, which binds to extracellular matrix proteins such as fibrinogen and fibronectin, mediating adherence to platelets. However, the precise role of PrpA in colonization or infection remains elusive [13]. Another surface adhesin, the serine-glutamate repeat-containing protein A (SgrA), binds to nidogen 1 and nidogen 2, two components of the basal lamina, as well as fibrinogen. However, SgrA does not facilitate binding to biotic surfaces such as human intestinal epithelial cells or kidney cells. Instead, it mediates biofilm formation on abiotic surface and may be involved in material-related infections, such as intravascular catheter infections. Consequently, a higher prevalence of the sgrA gene has been noted in clinical isolates, particularly those belonging to clade A1 [12].

The enterococcal surface protein (Esp), a large cell surface protein (>200 kDa) and composed of two distinct tandem repeat units, exhibits structural homology with Bap, a biofilm-associated protein of S. aureus [57]. Esp is present not only in E. faecium but also in other enterococci, sharing high protein sequence homology (>90%). Recent elucidation of the molecular mechanism of adhesion revealed that in E. faecalis, the N-terminal domain of Esp forms aggregates with amyloid-like properties, acting as an adhesin and contributing to fibrillary structure formation involved in biofilm formation [58]. The Δesp mutant exhibits deficient initial adherence associated with impaired biofilm formation on abiotic surfaces compared to esp-positive isolates but similar to esp-negative isolates [59, 60]. These observations were corroborated in a rat endocarditis model [22]. A serological response has been observed in humans during enterococcal endocarditis, confirming Esp expression in during infection [22]. Esp is also associated with UTI in mice, binding to bladder and kidney epithelial cells, and contributing to kidney inflammation, renal failure, and bacteremia [14]. However, Esp is dispensable for gastrointestinal colonization and adherence to intestinal cells in mice [61]. The esp gene is located on an enterococcal pathogenic island, frequently associated with clinical isolates, particularly those resistant to ampicillin [62, 63]. The association between esp and antibiotic resistance is significant, as esp-positive isolates demonstrate a higher capacity to acquire antibiotic resistance genes through conjugation compared to esp-negative isolates, notably the vanA gene cluster [23]. Interestingly esp gene expression is temperature- and environment-dependent, with higher expression at 37°C and in anaerobic conditions, facilitating early stages of the infectious process [60].

WxL proteins, widespread among Gram-positive bacteria, constitute another class of cell surface proteins. In E. faecium, three gene clusters (wxlA, wxlB, and wxlC loci) encode three distinct proteins each: a small WxL protein (SwpA-C), a large WxL protein (LwpA-C), and a domain of unknown function family protein (DufA-C). While SwpA and DufA of the locus A display binding to collagen (type I and IV) and fibrinogen, LwpA does not mediate binding to these molecules. A ΔwxlABC mutant exhibits impaired establishment of vegetation in a rat endocarditis model, although no differences were observed in biofilm formation, UTI or peritonitis models compared to the parental strain. Additionally, Wxl proteins play a role in response to bile stress as demonstrated by enhanced survival of the ΔwxlABC mutant in the presence of bile salts, suggesting their potential involvement in bile uptake [25].

In E. faecalis, aggregation substance ([AS], Asp1, Asa1 and Asc10) encoded by sex pheromone plasmids, enhances bacterial aggregates formation and increases binding to extracellular matrix proteins such as fibronectin, thombospondin, vitronectin, and type I collagen [64]. AS also facilitates cell adhesion and internalization, notably through Arg-Gly-Asp (RGD) domains recognized by integrins localized on the surface of eukaryotic cells [65, 66]. In an ex-vivo endocarditis model, AS promotes valve colonization and biofilm formation [67]. While AS is absent in E. faecium, AggE, another aggregation protein, enhances aggregation, binding to extracellular matrix proteins (collagen, fibronectin, and mucin), and biofilm formation. Structurally distinct from AS by the lack of RGD domain, AggE is a high-molecular-weight protein (>170 kDa) and remains to be characterized for its clinical impact [24].

3. Secreted proteins

Secreted proteins, also known as exoenzymes, are proteins synthesized within bacterial cells but subsequently released into the extracellular environment, where they perform diverse roles in various physiological processes such as nutrient acquisition, pathogenesis, and environmental adaptation, thereby contributing to the bacterium’s survival and virulence.

The enterococcal cytolysin Cyl, alongside two proteases—gelatinase GelE and serine protease SprE—play significant roles in enterococcal pathogenesis [68]. Their functions have been extensively examined in E. faecalis. Cytolysin aids bacteria in evading the host immune response by destroying cells [69]. The gelE gene, encoding a gelatinase, and the sprE gene, encoding a serine protease, are located on the same transcript. Both proteases are governed by a quorum sensing signal transduction system termed Fsr, and their crucial roles in infections have been demonstrated through various in vivo models, including Caenorhabditis elegans nematode killing [70], rabbit endophthalmitis [71], and murine peritonitis [72]. Clinical isolates of E. faecalis often exhibit an enrichment of the proteolytic trait compared to community fecal isolates [73]. However, unlike E. faecalis, E. faecium lacks both proteolytic (GelE and SprE) and cytolytic (Cyl) activities, despite the identification of sequence homologies of these three proteins in clinical isolates [74, 75].

E. faecium yet possesses another virulence factor, Hyl_Efm. Initially misidentified based on homology as a hyaluronidase—an enzyme known for its ability to degrade hyaluronic acid and disrupt connective tissue—Hyl_Efm is associated with virulence in other gram-positive pathogens [76, 77]. However, despite initial assumptions, no hyaluronidase activity has been detected in hyl_Efm-carrying E. faecium isolates. Further investigation revealed Hyl_Efm to be a glycosyl hydrolase with β-N-acetylglucosaminidase activity [26], which does not directly contribute to E. faecium colonization [78]. Interestingly, Hyl_Efm-encoding genes can be found either on the chromosome or on plasmids. Notably, isolates carrying hyl_Efm on plasmids have demonstrated higher bacterial loads during colonization [79]. This observation suggests that virulence factors, which are often co-located on the same plasmid, could potentially explain the increased pathogenesis of the strains carrying the hyl_Efm-carrying plasmid, which is why HylEfm is considered one of the virulence markers of E. faecium [26, 80, 81].

SagA, a 75-kDa peptidoglycan hydrolase, is a major secreted antigen crucial for growth and cell wall metabolism [82, 83]. In E. faecium, SagA from clades A1 and B have different repeated sequences that link the N and C terminal domains. These distinct sequences are believed to influence the protein structure, thereby affecting the functionality and role of SagA in biofilm formation [84]. Hence, it is speculated that this sequence variation in SagA, dependent on the clade of E. faecium, plays a pivotal role in biofilm formation, among other functions. Additionally, anti-SagA antibodies have been noted for their opsonic-killing activities against E. faecium, suggesting a potential of SagA as a vaccine candidate [85]. Moreover, a study demonstrated the ability of SagA to provide protection against enteric infections in both C. elegans and mouse models [27].

4. Bacteriocins

Bacteriocins, primarily produced by the Firmicutes phylum, especially lactic acid bacteria like Enterococcus spp., are potent antimicrobial peptides (AMPs) with a wide range of actions [15]. They exhibit remarkable diversity in structures and are often encoded by complex gene clusters, subject to rapid evolution and horizontal transfer [86]. Termed enterocins in enterococci, these ribosomally-synthesized extracellular proteins exert mostly bactericidal effects by targeting the cytoplasmic membranes of neighboring bacteria, disrupting transmembrane potential and pH gradients to induce leakage of intracellular content [87, 88]. While most of these AMPs are active against closely related bacteria, some display broader activity. Enterocins confer a significant competitive advantage by inhibiting the growth of other bacteria without compromising the producer’s fitness. This advantage is reinforced by co-existing genes in the enterocin cluster, which provides protection through self-immunity proteins and efflux transporters [86]. Enterocin production can thus clear bacteriocin-susceptible bacteria from a community or shield against colonization by bacteriocin-susceptible invaders. Although comprehensive reviews cover aspects such as classification, synthesis, and application of enterocins [15, 16, 89, 90], this chapter focuses on clinically derived enterocins produced by E. faecium and their role in pathogenesis. Recently, an excellent review has noted E. faecalis and E. faecium as leading producers of enterocins with 24 and 19 enterocins described, respectively [16], potentially linked to their plasmid pool. Notably, some species harbor multiple enterocin-related genes, providing a competitive edge against other strains [16].

Among clinical isolates of E. faecium, eight enterocins have been identified (enterocins EntA, EntB, and EntP; bacteriocins Bac32, Bac43, Bac51, RC714 and T8) located independently in the chromosome or plasmids, each with unique activity spectra. While this chapter emphasizes clinically-derived enterocins, it is worth mentioning the significant role of enterocins from other sources, such as strains isolated from food products. For instance, the CRL35 toxin from an E. faecium strain found in Argentinean cheese exhibits antibacterial and antiviral activities against Listeria monocytogenes and Herpes virus, respectively [91, 92, 93].

The pediocin-like bacteriocins family counts the most and best-documented enterocins produced by clinical isolates of E. faecium as Enterocins A (EntA) and P (EntP) as well as bacteriocins 43 and RC714 fall into this category. This group of enterocins is particularly potent against Listeria species and harbor a recognizable conserved sequence (YGNGVXC) in their N-terminal region, thought to allow them to act at low concentrations against their neighborhoods [15, 94].

EntA is one of the commonest enterocin from this group and is characterized by its potent antimicrobial activity primarily against L. monocytogenes, attributed to the presence of two disulfide bridges in its structure [28, 29]. It is produced by various strains of E. faecium sourced from different environments, including food products [95] and clinical settings [30, 31]. Studies have reported a high incidence of EntA in clinical isolates of E. faecium, especially in vancomycin-susceptible E. faecium (VSEfm), suggesting its importance in hospital-acquired infections [30, 31]. EntA is described as a part of the core genome of E. faecium, located chromosomally, possibly providing a competitive advantage to this species [96]. Additionally, E. faecium strains producing EntA often carry genes encoding other bacteriocins, suggesting a potential synergistic effect in antimicrobial activity, with EntB for example [97].

EntP has been identified in both chromosomes and plasmids across various E. faecium strains isolated from food sources and clinical settings [98, 99, 100]. While its antimicrobial spectrum is wide, encompassing foodborne pathogens and spoilage bacteria, its presence in clinical isolates, albeit at low rates, suggests a potential role in enterococcal virulence or colonization within human hosts [30, 31]. Studies have shown a higher occurrence of EntP and other bacteriocins among clinical vancomycin-resistant E. faecium (VREfm)/VSEfm strains compared to non-clinical isolates, indicating their relevance in clinical settings [95, 98]. In addition, Farias et al. tested the antimicrobial activity of EntP against 14 vanA-positive E. faecium and E. faecalis strains as a result of all the strains tested being susceptible to the bacteriocin, therefore highlighting their potential as therapeutic agents against antibiotic-resistant pathogens [101].

Bacteriocin 43 (Bac43) was initially discovered in VREfm strains isolated from hospitalized patients in the USA during the 1990s whereas it was also found once in a fecal sample from a healthy Japanese medical student, suggesting its potential dissemination beyond clinical settings [32]. Bac43 exhibits antimicrobial activity against a range of pathogens including L. monocytogenes and various Enterococcus spp. [30, 32]. Studies have revealed the presence of Bac43 on small mobilizable plasmids, particularly pDT1-like plasmids, which have shown spread among different clinical VREfm isolates, indicating potential horizontal transfer mechanisms contributing to its distribution [32]. The presence of Bac43, always found exclusively associated with clinical VREfm, has been consistently observed in clonally diverse outbreak VREfm strains, underscoring its potential role in enterococcal infections and highlighting its clinical relevance [95, 98, 102].

Bacteriocin RC714 was first described in 1996 in a clinical vanA-positive E. faecium RC714 strain. It shares 98% identity with Bac43 [32] and shows antimicrobial activity against vancomycin-susceptible, as well as VanA, E. faecalis, E. faecium, and E. hirae from human clinical/fecal and sewage samples but also against non-enterococcal species such as Listeria spp. [33].

Four other enterocins have been identified in clinical isolates but because of the absence of the specific domain in the N-terminal part, they are not considered part of the pediocin-like bacteriocins family.

Bac32, initially identified in clinical isolates of both VREfm and VSEfm from the USA and Japan, as well as in a non-clinical E. faecium strain from healthy feces in Japan, is associated with a 12.5-kb plasmid known as pTI1-type [103]. This plasmid, found to be highly transferable, harbors the bac32 genes, which are bacA encoding for the bacteriocin and bacB encoded for an immunity protein. Although Bac32 was detected at low rates in VREfm outbreaks across different regions, it often coincided with common clones within the same geographical area, suggesting potential clonal and/or regional expansion [30].

EntB was initially isolated from E. faecium T136 in Spanish dry sausage and later one detected in clinical isolates [30, 95, 96]. Not systematically found in clinical isolates, EntB has been mostly noticed in VREfm isolates originating from Latin America [30, 31]. This regional pattern suggests potential clonal or plasmid-mediated spread. EntB is often found alongside EntA and exhibits synergistic effects. It demonstrates activity against a broad range of Gram-positive bacteria, particularly foodborne pathogens such as Clostridium, Listeria or Staphylococcus spp. [34].

Bac51 was first identified in a clinical VanA-type VREfm from Japan. The bac51 genes, bacA encoding for the bacteriocin and bacB for the immunity protein are located on a small 6.0-kb mobilizable plasmid, pHY, and show antimicrobial activity against different enterococcal species [35].

Bacteriocin T8 was first isolated from vaginal secretions of children with human immunodeficiency virus. The four T8 genes, encoded for the bacteriocin, an immunity protein, a mobilization protein and a relaxase are located on a 7.0-kb plasmid and has a bactericidal activity, notably toward E. faecalis [36].

5. Immune modulators

Professional phagocytes, such as neutrophils, monocytes, and macrophages, which mount responses against pathogens, orchestrate host immunity. Pathogens, including Gram-positive bacteria like E. faecium, must possess a myriad of mechanisms to neutralize, evade, or limit this host response in order to adhere to host tissues or cells and establish infection. Gram-positive bacteria, including E. faecium, have evolved a diverse array of virulence factors that enhance their survival against host defenses. Membrane structures, such as lipoteichoic acids (LTAs), wall teichoic acids (WTAs), and capsular polysaccharides, are crucial immunogenic components of the bacterial cell wall surface [17, 18].

Polysaccharides located on bacterial surfaces interact with the human host and play significant roles in immune response and bacterial pathogenesis. In E. faecium, polysaccharides hinder leukocyte killing [104]. CapD (capsular polysaccharide protein D), a 336-amino acid protein, is involved in polysaccharide biosynthesis, particularly by catalyzing N-linked glycosylation. Compared to the wild-type, a strain lacking capD displayed reduced growth, less biofilm production, and increased cell surface hydrophobicity. Microscopic analysis revealed abnormal cell shapes in the ΔcapD mutant, where polysaccharide material was not properly exposed. This was further confirmed by analysis of the cell wall crude extract, demonstrating the absence of two WTAs and a different glycopolymer composition [105]. In in-cellulo experiments on various cell lines, such as uroepithelium or colon carcinoma, the ΔcapD mutant strain showed reduced adherence to host cells but enhanced opsonization in rabbit complement. Furthermore, virulence was assessed in vivo in a bacteremia mouse model, with surprisingly higher bacterial burden observed in different organs of mice infected with the mutant strain compared to the wild-type while no difference were observed in an virulence UTI model [37]. If clarification of the specific function of capD during pathogenesis is required, it underscores the importance of polysaccharide components such as capD in host-cell interaction and virulence in E. faecium [37].

Amphiphilic glycoconjugate polymers, LTAs are essential constituents of the cell wall of many Gram-positive bacteria. Immunotherapies targeting LTAs have been effective, particularly in E. faecalis, in modulating the host response. Studies on LTAs in E. faecalis have demonstrated its ability to induce opsonic antibodies, protecting against E. faecalis and E. faecium bacteremia [106]. A synthetic LTA fragment has been identified capable of inducing specific opsonic antibodies that mediate killing of clinical isolates. Additionally, in both sepsis and endocarditis animal models, antibodies raised against this synthetic LTA fragment significantly reduced bacterial colony counts, indicating its potential as a vaccine candidate against E. faecalis infection [107]. Furthermore, D-alanine esters of LTAs, generated by D-alanine ligase encoded by the dlt operon, have been implicated in antibiotic susceptibility and virulence in an invertebrate Galleria mellonella model. In both E. faecalis and E. faecium, inhibition of D-alanylation increased susceptibility to antibiotics and potentiated the effect of antibiotic combinations, suggesting a role in enhancing antibiotic efficacy against enterococcal infections [38].

Attached to N-acetylmuramic acid residues of peptidoglycan, WTAs play diverse roles in bacterial physiology, including cell division, autolysin activity, surface protein scaffolding, attachment to host cells and abiotic surfaces, and cation homeostasis [108]. While no studies have investigated WTAs in E. faecium, research on E. faecalis has shown that disruptions in WTA biosynthesis render the bacteria highly susceptible to opsonophagocytosis, modulating host complement recognition and immune response [105].

Enterococcal polysaccharide antigen (Epa), described as an immunoreactive rhamnose-containing polysaccharide cell wall component, plays a significant role in enterococcal pathogenesis [109]. Epa has been found involved in biofilm formation and tissue invasion [110, 111, 112, 113, 114]. In E. faecalis, the Δepa mutant strain exhibits reduced translocation toward colonic epithelial cells and is associated with attenuation and deficiencies in translocation across a polarized monolayer of colonic epithelial cells [110, 111, 113]. The epa locus has also been identified in E. faecium, appearing to be also part of the core genome as well and therefore being found in clinical isolates [115].

6. Biofilm regulators

Biofilm formation is a fundamental aspect of E. faecium adaptability to diverse environments. Several virulence factors within this bacterium have been identified as regulators of biofilm formation.

EbrB (for enterococcal biofilm regulator B) characterized by its helix-turn-helix motif, belongs to the AraC family of transcriptional regulators [39]. Positioned upstream of the esp gene—an essential component in E. faecium biofilm formation —EbrB operates independently of the esp operon [39, 59]. Unlike its E. faecaliscounterpart, PerA, which exhibits broader regulatory influence but lacks control over esp expression, EbrB is tailored specifically for the esp-containing operon in E. faecium [39, 116, 117]. The pivotal role of EbrB in maintaining the basal level of Esp cell-surface expression, essential for biofilm integrity, has been elucidated [39]. A reduction in Esp surface exposure was observed in the ΔebrB mutant strain, akin to the observation in the Δesp mutant strain, with phenotypic restoration evident in the ebrB-complemented strain. Intriguingly, unlike esp expression, which fluctuated with cell density, ebrB expression appeared unaffected by growth phase. Top et al. confirmed the indispensability of EbrB for biofilm formation using a semi-static biofilm model and a flow cell system [39]. Biofilm formation was significantly more impaired in the ΔebrB mutant strain compared to the Δesp mutant strain, suggesting that EbrB regulates the expression of three genes located within the same pathogenicity island as esp. In vivo implication of EbrB was demonstrated in a mouse model of intestinal colonization, showing a reduced bacterial load in stools, small intestine, and colon with the ΔebrB mutant strain, suggesting the regulator involvement in intestinal colonization.

Carbohydrate phosphotransferase systems (PTS) were initially described in E. faecium in 2013 [45]. BepA (for biofilm and endocarditis-associated permease A), identified in E. faecium a few years later, is a carbohydrate PTS permease [40]. Enriched in E. faecium hospital outbreak isolates, bepA belongs to the accessory genome of E. faecium, and is absent in commensal strains [40, 45]. BepA, composed of eight transmembrane domains, is implicated in sugar transport, sharing homologies with FruA identified in Streptococcus pyogenes and Deinococcus radiodurans, involved in fructose uptake [118, 119]. Using human serum, deletion mutant of bepA and complemented mutant strains were tested to elucidate the role of BepA in biofilm formation. The ΔbepA mutant strain exhibited significantly reduced biofilm formation (lower biomass and thickness) in these conditions [40]. Confirmation of these results came from a competition assay between the ΔbepA mutant and the wild-type strains in a rat endocarditis model, revealing a significant fitness disadvantage for E. faecium in the absence of bepA. Thus, BepA appears to be crucial in the pathogenesis of E. faecium endocarditis. As BepA shares similarities with FruA, efforts were made to highlight its impact on fructose transport and general carbohydrate metabolism. Both the wild-type and mutant strains were able to grow in various media containing different carbohydrates, except for β-methyl-D-glucoside, which the mutant appeared unable to metabolize, though the mechanism of action remains unclear [40].

Peptidoglycan hydrolases, also known as autolysins, play a vital role in biofilm formation and stability. Autolysis in many bacterial species results in the release of extracellular DNA (eDNA), an integral component of the biofilm matrix [120, 121, 122]. AtlA, a major autolysin found in both Gram-positive and Gram-negative bacteria, serves various functions including cell division and cellular autolysis [123]. In 2013, a study identified six putative autolysins in E. faecium, confirming one as the major autolysin, named AtlA_Efm [41]. Structurally similar to the autolysin found in E. faecalis, AtlA_Efm comprises a C-terminus with six LysM domains and a lysozyme-like superfamily domain, involved in hydrolyzing β-1,4-linked polysaccharides [41, 124, 125]. To confirm AtlA_Efm role, experiments with a mutant strain where atlA_Efm was inactivated showed reduced initial adhesion and peptidoglycan lysis, confirming its autolysin function. In a polystyrene biofilm model, the atlA_Efm-deficient mutant produced significantly less biofilm than the wild-type strain. Observation of overnight cultures revealed long chains in the ΔatlA_Efm mutant strain, indicating impaired peptidoglycan hydrolysis and cell separation, results confirmed by confocal microscopy also in planktonic cells. The complemented strain reinstated wild-type phenotypes, confirming AtlA_Efm role in biofilm formation and cell lysis/separation. Additionally, experiments exploring AtlA_Efm function in eDNA release showed its involvement in cell-cell interactions and facilitation of eDNA release [41, 126]_. Similar to other species, AtlA_Efm was found to localize on the cell surface, particularly at the septum of cells in the log phase and in the biofilm matrix. The protein also contributed to the expression of Acm, a collagen-binding adhesin described above [41]. However, recent observation by Wei et al. noted that E. faecium strains were able of producing biofilms without atlA_Efm expression, suggesting that AtlA_Efm may not be the sole determinant of biofilm formation [127].

7. General regulators

Bacterial stress responses are crucial for virulence and adaptation to diverse host environments. Certain proteins serve as global regulators, influencing intestinal colonization, stress response, and various stages of the infectious process.

AsrR (for antibiotic and stress response regulator), encoded by the asrR gene, is an oxidative-sensing regulator first described in 2012 [42]. It belongs to the extensive family of MarR regulators, well-known for their involvement in various virulence and antibiotic resistance mechanisms across both Gram-positive and -negative bacteria. Notable examples include MgrA and OspR, identified in S. aureus and Pseudomonas aeruginosa, respectively, both functioning as oxidative-sensing regulators with pleiotropic effects [128, 129]. In E. faecium, AsrR shares a structural similarity with other MarR family regulators, featuring a winged-helix DNA binding motif and two cysteine residues essential for oxidative-stress sensing, akin to MgrA and OspR proteins. AsrR acts as a global repressor, influencing oxidative stress response, antibiotic, cationic antimicrobial peptide (CAMP) resistance, autolysis adhesion, biofilm formation, pathogenesis, and mutagenesis [42]. Lebreton et al. demonstrated that AsrR activity is inhibited by hydrogen peroxide, confirming its role in oxidative stress response through in vitro and in vivo survival analyses. Using a ΔasrR mutant strain, they also showed that AsrR modulates susceptibility to penicillin and CAMPs, likely by repressing expression of the pbp5 gene and the dlt operon, respectively. Additionally, they observed increased cell adhesion, biofilm formation, and enhanced colonization in a G. mellonella model, suggesting AsrR down-regulatory effect on adhesion expression genes such as acm and ecbA. In a murine model of systemic infection, higher bacterial loads in organs with the deleted strain further confirmed the involvement AsrR in pathogenicity. Down-regulation of uvrA and mutS2, two genes involved in DNA transfer and genome adaptation, was also noted in the absence of AsrR [130]. Lebreton et al. also highlighted the conservation of the asrR sequence among E. faecium strains, although it is absent in E. faecalis. Some putative homologs have also been described in E. gallinarum and E. casseliflavus [42].

Gls24, encoded by the gls24 gene, is a general stress protein identified in E. faecalis. It is part of an operon comprising six genes and is induced by various stressors such as glucose starvation or bile salt exposure. Gls24 enables E. faecalis to survive bile salts stress and has been implicated in virulence [131, 132, 133]. Adjacent to gls24 within the same operon is glsB, an open reading frame cotranscribed with gls24 [132]. Two close homologs of gls24 in E. faecium, named gls33 and gls20, due to their predictive molecular mass of 33 and 20 kDa, respectively, have been identified [43]. These genes are part of nine-gene and four-gene operons, respectively. Additionally, two upstream ORFs, sharing 80% amino acid homology with GlsB, were identified and named GlsB and GlsB1, with their loci confirmed through transcriptional analysis. These gene clusters (orf1,2,3,4,5,6,7-gls33-glsB and orfA,B-gls20-glsB1) were found to be conserved and closely located in eight E. faecium clinical isolates [43]. To confirm the role of gls33-glsB and gls20-glsB1 genes in E. faecium virulence, single and double-deletion mutant strains were constructed. In vitro experiments revealed no significant differences in growth rates compared to wild-type strains, suggesting that these genes are not essential for E. faecium growth in vitro. However, single-deleted strains showed significantly reduced resistance to physiologically relevant bile salt concentration compared to the wild-type strain. The double deletion exhibited even lower resistance compared to the two single mutants, confirming the necessity of both gls loci for full tolerance to physiological bile salt concentrations. Finally, the involvement of gls33-glsB and gls20-glsB1 genes in virulence was assessed in a mouse peritonitis model. Deletion of both gene pairs resulted in significantly prolonged survival compared to wild-type strains, while single deletion had no impact on mouse mortality. These findings suggest that the gls loci are crucial for the pathophysiology of E. faecium infections [43].

CcpA (for catabolite control protein A) is a global regulator belonging to the LacI-GalR family of transcriptional regulators, implicated in carbon catabolite repression (CCR) [44]. CCR is a process that represses the synthesis of enzymes required for the transport and metabolism of non-preferred sugars when preferred carbon sources such as glucose or fructose are present in the growth environment [134]. This protein consists of 339 amino acids, with an N terminus containing a helix-turn-helix DNA-binding motif and a C terminus composed of a ligand-binding domain [135]. Studies have shown that CcpA is crucial for E. faecium growth fitness [44]. Deletion of ccpA resulted in reduced growth of the strain, with smaller colony compared to the wild-type strain. This growth defect was reversed by the complementation of ccpA in the deleted strain. Additionally, an E. faecium strain with a growth defect was identified due to a unique nonsense mutation in the ccpA gene, converting a glutamine to a stop codon [44]. The role of CcpA in bacterial growth has been established in various Gram-positive bacteria such as E. faecalis or S. aureus [136, 137]. Interestingly, the importance of CcpA in the pathogenesis of E. faecium infective endocarditis was demonstrated in a rat experimental model. In this model, rats were infected with a mix of both the wild-type and the ΔccpA mutant strains [44]. A reduction in vegetation size was observed with the deleted strain, and there was a significant domination of the wild-type strain over the deleted strain in vegetation. These findings were further supported by similar experiments using the ΔccpA mutant and complemented strains, with the deleted strain identified in significantly lower proportions. In vitro additional data suggested that the attenuated phenotype observed for the ΔccpA mutant strain in the endocarditis model might be due to a growth defect or a reduced capacity to form biofilm [44]. CcpA exhibits pleiotropic effects, including the regulation of growth fitness and biofilm formation, and more broadly contributes to E. faecium virulence. However, the precise mechanism of action of this transcriptional regulator remains unclear.

PTS^clin is a PTS system crucial for carbohydrate transport, initially identified in clinical isolates of E. faecium [45]. Genomic analysis revealed its specific presence in clinical E. faecium isolates while absent in commensal isolates, indicating its potential involvement in virulence. The ptsD gene encodes a sugar-specific membrane-associated subunit (enzyme IID) of PTS^clin. To investigate the involvement of PTS^clin and ptsD in E. faecium virulence, experiments were conducted, both in vitro and in vivo, using deletion mutant of the ptsD gene together with the E. faecium wild-type strain [45]. No significant difference in growth was observed between the two strains in medium containing 65 different carbohydrates, making it challenging to identify the preferred substrate for PTS^clin. The role of PTS^clin in intestinal colonization was examined through a competitive assay. In this chapter, the mutant and wild-type strains were orally administered (at a 1:1 ratio) to mice treated with antibiotics, and the bacterial load of each strain was determined in stool samples and intestines. In the absence of PTS^clin, a significantly lower level of colonization was observed, indicating the contribution of PTS^clin and ptsD to gut colonization of clinical isolates of E. faecium under antibiotic treatment. However, the exact role of ptsD in E. faecium infection pathogenesis remains unclear. More recently, a study evaluating the prevalence of virulence markers in over 300 E. faecium strains (clinical and non-clinical) confirmed the absence of ptsD in commensal isolates. Furthermore, it revealed a significantly higher prevalence of ptsD in hospital-acquired strains, with a strong positive association with ampicillin resistance [55].

Bacterial regulatory RNAs, known as small RNAs (sRNAs), play pivotal roles in diverse adaptive responses, including virulence and stress responses [138, 139]. In E. faecium, several putative sRNAs have been uncovered [47]. Among these, Ern0160 has emerged as a conserved element across all clinically studied isolates, hinting at its potential involvement in E. faecium environmental adaptation mechanisms. Remarkably, deletion of ern0160 yielded no discernible differences in growth in vitro or in intestinal colonization in a murine model compared to the wild-type strain. However, in a competition assay employing a co-colonization model in mice, the wild-type strain significantly outcompeted the ern0160-deleted strain after 3 days of colonization [46]. Subsequent experiments with the ern0160 trans-complemented strain, compared to the deleted strain complemented with an empty vector, yielded unexpected results. In the co-colonization model, the proportion of the complemented strain was significantly lower than that of the control strain, indicating a potential deleterious effect of ern0160 overexpression [46]. Additionally, this sRNA may contribute to antibiotic and biocide stress responses in E. faecium, as its expression is notably downregulated under sub-inhibitory concentrations of daptomycin and biocides [47, 48]. Further investigations are warranted to elucidate the targets of this sRNA and unravel its intricate role and regulatory networks.

The stringent response serves as a sophisticated stress response mechanism employed by bacteria to navigate nutrient scarcity. Central to this process are the alarmones guanosine tetraphosphate and pentaphosphate [(p)ppGpp], for which synthesis in proteobacteria is meticulously regulated by two enzymes: a synthetase and a bifunctional protein named RelA and SpoT, respectively [140]. In Gram-positive bacteria, RelA, also referred to as RSH (Rel SpoT Homolog), emerges as a pivotal regulator of (p)ppGpp synthesis [141, 142, 143]. Although the intricacies of this pathway in E. faecium remain incompletely understood, recent research has shed light on a mutation in the relA gene that instigated alterations in the stringent response and antibiotic tolerance [49]. In a compelling study, 22 strains of E. faecium, isolated from blood cultures during a single episode of persistent bacteremia in a 6-week-old child with acute myeloid leukemia, underwent sequencing and analysis [49]. Notably, a missense mutation was uncovered in the relA gene in eight strains, predominantly isolated after daptomycin was introduced into the antibiotic treatment regimen. Intriguingly, in vitro experiments unveiled that the ΔrelA mutant strains exhibited significantly elevated basal levels of (p)ppGpp compared to wild-type strains, indicative of an upsurge in the alarmone resting level. This adaptive response enabled these strains to swiftly acclimate to stress conditions such as antibiotic exposure. Additionally, the ΔrelA mutant strains showcased impaired biofilm production, despite displaying heightened tolerance when subjected to multiple antibiotic stresses [49].

8. Conclusions

While enterococci are typically members of the human gut microbiota, they have emerged as a leading cause of human infections, particularly with the worldwide spread of multidrug-resistant clinical isolates. Traditionally, E. faecalis has been recognized as the primary pathogenic species, with numerous virulence factors identified. Meanwhile, E. faecium has garnered attention as a multidrug-resistant species, accounting for the vast majority of VRE clinical isolates. However, extensive research has revealed diverse pathogenic traits in E. faecium, including cell surface components, secreted proteins, bacteriocins, immune evasion strategies, and different stress-response regulators. By combining virulence and resistance determinants, E. faecium is now a difficult-to-treat major opportunistic pathogen.

Conflict of interest

The authors declare no conflict of interest.

Acronyms and abbreviations

AS	aggregation substance
AsrR	antibiotic and stress response regulator
BepA	biofilm and endocarditis-associated permease A
CAMP	cationic antimicrobial peptide
CapD	capsular polysaccharide protein D
CcpA	catabolite control protein A
CCR	carbon catabolite repression
EbrB	enterococcal biofilm regulator B
eDNA	extracellular DNA
Epa	enterococcal polysaccharide antigen
Esp	enterococcal surface protein
LTA	lipoteichoic acid
MSCRAMM	microbial surface components recognizing adhesive matrix molecule
PTS	carbohydrate phosphotransferase system
SgrA	serine-glutamate repeat-containing protein A
sRNA	small RNA
VREfm	vancomycin-resistant E. faecium
VSEfm	vancomycin-susceptible E. faecium
WTA	wall teichoic acid