The vagina is thought to be the main reservoir of GBS [21]. GBS colonization in the vagina of pregnant women is a major risk to the newborn. These virulence factors are associated with dissemination, immune evasion, and damage to tissue, and allow GBS survival in the hostile vaginal environment.

The first step for GBS colonization of the vagina is the adhesion to its epithelial cells via surface-associated adhesins. Several adhesion factors enable GBS to bind to components of the extracellular matrix (ECM), thereby enhancing its ability to penetrate the host mucosal barrier and spread to other host tissues.

Srr1/Srr2: Srr1/Srr2 are structurally characterized by an extended N-terminal signal sequence, two highly glycosylated serine-rich repeats (Srr), at least one non-repeat binding region (BR), and a C-terminal cell wall anchoring domain carrying the LPxTG (Leu-Pro-x-Thr-Gly) pattern [22]. Srr is glycosylated by glycosyltransferases (Gtfs), which includes a two-protein glycosyltransferase complex (GtfAB), Nss, and Gly [23]. The large loci encoding Srr1 and Srr2 are located at different chromosomal positions with similar genetic organization, and the genes mediating Srr glycosylation are also encoded by these loci [24]. Furthermore, Srr1 and Srr2 are structurally similar but show only 32% sequence identity at the amino acid level [25]. Therefore, only limited homology is shown (<20% concordance) [26]. Genome-wide association studies have shown that highly pathogenic strains of agalactiae could almost exclusively produce Srr2 and Gtfs [26,27].

FbsA, FbsB, FbsC: FbsA, FbsB, and FbsC are three members of the family of fibrinogen-binding proteins encoded by GBS. They adhere to human epithelial cells to promote vaginal colonization [34]. FbsA and FbsB were most frequently detected in infected pregnant women [35]. Meanwhile, not only is FbsB found in some of the major CCs in BJI strains, the CC10 and CC23, but it also promotes the epithelial invasion of GBS [29,36]. FbsC, which contains two bacterial immunoglobulin-like tandem repeat domains and a C-terminal cell-wall-anchoring motif (LPxTG), can also mediate biofilm formation, which can promote GBS colonization of the vagina. However, FbsC was not expressed in those clinical GBS isolates belonging to the highly pathogenic lineage ST17. In addition, it has been shown that FbsC can help GBS to colonize the brain with immunoprotective activity [34].

PbsP: The plasminogen binding surface protein (PbsP) is a surface protein with crucial functions in GBS pathophysiology. This adhesion protein has two streptococcal surface repeat structural domains, a methionine- and lysine-rich region, and an LPxTG cell wall anchoring pattern, mediating plasminogen binding to enhance the ability to colonize and invade host tissues [37]. A two-component system (TCS) homologous to the staphylococcal virulence regulator SaeRS is upregulated in vivo during GBS infection of the vagina, and PbsP is one of the targets of SaeRS. By employing a mouse model of vaginal carriage and using transcriptome sequencing (RNA-Seq), it was found that SaeRS promotes the expression of PbsP by using a component of vaginal irrigation fluid, which is still unknown as a signal to play an important regulatory role in PbsP [38]. A recent study discovered that the specific sequences required for this interaction are in the methionine and lysine-rich (MK-rich) domain of PbsP and the Kringle 4 lysine-binding sites (LBS) of plasminogen. They also found that the presence of lysine or other positively charged amino acid residues in the MK-enriched region is not necessary for the region to bind to plasminogen [39]. The finding of the MK-rich domain and Kringle 4 LBS may provide a new direction for vaccine development. In the pathogenesis of invasive infections, the binding of microbial pathogens to the host vitronectin (Vtn) is prevalent. In another study targeting Vtn, the pretreatment of cells with anti-Vtn antibodies or Vtn resulted in the inhibition and promotion of GBS adhesion and invasion of epithelial cells, respectively. Thus, Vtn was further found to act as a bridge between the streptococcal surface repeat (SSURE) domain of PbsP on the surface of GBS and the host integrin. In addition, inhibition of the interaction between PbsP and extracellular matrix components prevents GBS colonization [40].

Pili: The pili are GBS-encoded cell-wall-anchored appendages that protrude from the bacterial surface and each gene encodes three structural subunit proteins: capillary axis backbone protein (PilB), capillary tip (PilA, capillary-associated adhesion protein binding collagen type 1), and capillary base (PilC) [21], bearing a C-terminal LPxTG motif and two subfamily C sortases (SrtC) involved in covalent polymerization of the subunits [41,42,43]. GBS strains also possess the conserved “housekeeping” sorting enzyme A (SrtA), which is involved in the covalent assembly of pili on the cell wall. According to research, the deletion of SrtA does not affect the polymerization of a pilus but results in a reduction in pilus expression on the cell surface. In one model of GBS pilus assembly, the polymerization of pilus structures is handled by pilus island (PI) sortase, and SrtA utilizes the accessory pilus subunit GBS150 as an anchoring protein to covalently attach it to the cell wall [43].

Lmb: The laminin binding protein (Lmb) mediates the attachment of GBS to human laminin, which is crucial for bacterial colonization and invasion [47]. According to recent research [48], GBS is less invasive in the lmb mutant strains, especially in human brain microvascular endothelial cells (HBMECs). During infections, neutrophils release large amounts of calprotectin, which binds zinc efficiently, thereby inhibiting bacterial growth [48]. Interestingly, a recent study has shown that the lmb gene in GBS promotes resistance to the reduction in zinc caused by calprotectin. As GBS alters its own zinc transport mechanisms, upregulating genes encode zinc-binding proteins, lmb, adcA, and adcAII, thus assisting GBS to bind zinc. In addition, mice infected with the Δlmb mutant not only had less GBS in their brains, but also had decreased mortality [49]. This further demonstrates the importance of the virulence of lmb to GBS.

ScpB: ScpB or C5a peptidase is a surface-associated serine protease. It not only prevents complement activation by cleaving C5a, a neutrophil chelator, but also helps bacteria bind fibronectin [50,51]. Through this fibronectin binding, it could aid in the GBS invasion of human epithelial cells [50]. In 90 isolates collected from invasive and non-invasive isolates from adults, serotype III accounted for 68.9% of all 90 isolates, and ScpB was detected in all 90 isolates [52]. In addition, analysis of 242 GBS strains, including 95 colonizers without pathogenicity and 68 pathogenic strains isolated from pregnant women, and 79 strains isolated from newborns with sepsis, revealed that the frequency of ScpB was significantly higher in neonatal strains. This shows that ScpB can be very well used to help GBS infect newborns with sepsis [53]. However, when ScpB is cross-linked to fibrin mediated by factor XIIIA (FXIIIA), GBS entrapment in the fibrin clot increases, allowing for reduced transmission of systemic infections [54]. This may provide new insights for future GBS treatments.

HylB: HylB denotes hyaluronidase or hyaluronate lyases, an exolytic enzyme released by the GBS. HylB can cleave the high-molecular-weight glycosaminoglycan polymer of hyaluronic acid, which serves as the epithelial extracellular matrix component. HylB cleavage of hyaluronic acid breaks the maternal–fetal barrier and enables the travel of GBS from the vagina to the fetus to cause fatal infection and damage [55]. The byproducts, such as disaccharide fragments produced from the cleavage of hyaluronic acid (HA), can bind to Toll-like receptors 2 and 4, blocking the pro-inflammatory cascade response induced by some GBS components [56]. The blocked TLR2- or TLR4-mediated response leads to immunosuppression, making the ascending infection possible. However, the non-hyaluronidase mutant GBS was found to be cleared by immune responses as they lack the HylB enzyme [55].

Hemolytic activity in GBS is due to the ornithine rhamnolipid pigment (hereafter called “hemolytic pigment” or “pigment”, also known as Granadaene) [18,59], which is produced by the genes of the cyl operon [60,61]. Surprisingly, it has been shown that Granadaene can also be produced by Lactobacillus heterologously expressing the GBS cyl operon, and that the Granadaene produced is functional [62]. Among the cyl operon, the cylE gene is necessary for pigment production [18,60,61], and transcription of cyl genes is negatively regulated by the CovR/S two-component system (also known as CovR/S TCS) [18]. As a major virulence factor in GBS, Granadaene not only has pigmentary and hemolytic activity but also effectively resists the elimination by mast cells, macrophages, and neutrophils [18,19]. Additionally, it can penetrate the human placenta [18], invading the amniotic cavity and causing severe fetal damage [63]. However, the hemolytic activity of GBS can be inhibited by photobleaching of granadaene and the sensitivity of GBS to reactive oxygen species, such as hydrogen peroxide, leading to increased membrane permeability, thus reducing its activity [64].

Additional research investigated the effect of hemolytic pigment on platelets [65]. By using the pigmented LUMC16 strain and its non-pigmented isogenic LUMC16ΔcylX-K mutant, it was observed that platelets responded to stimulation with the LUMC16 strain but not to the unpigmented LUMC16ΔcylX-K strain within 30 min. Thus, GBS pigment can induce initial platelet activation. Further observations revealed a gradual decrease in platelet viability with increasing time of infection in the LUMC16 strain and that platelets were killed when the pigment concentration increased to a certain level (1 and 2 μM). In contrast, the cells remained viable when infected with non-pigmented bacteria. That is, therefore, a good indication that pigments can cause platelet death [65]. Although some advancements have been made, the structural features responsible for hemolysis and cytotoxicity are still to be investigated [62].

The cervix controls and restricts the entry of microorganisms into the uterus by secreting mucus, cytokines, and antibacterial peptides. These secretions are not only effective in the prevention of infections but are also protective to the fetus developing in the uterus. If this barrier is breached, bacteria may enter the uterine cavity and cause premature birth [60].

The alpha C protein (ACP) is the prototype of a family of Gram-positive bacterial surface proteins. It facilitates the entry of GBS into human cervical epithelial cells and traverses the cell layer by binding to glycosaminoglycans (GAG) on the surface of the host cell [66]. The prototype alpha C protein of GBS from strain Ia/C A909 includes a series of nine identical 246 bp tandem repeat units. It was found that deficiency in the tandem repeat region of the ACP occurs through a recA-independent recombination pathway and affects the protective potency and immunogenicity of the protein, so these deficiencies may be a mechanism of virulence in GBS [67].

GBS capsules consist mainly of carbohydrates with the capsule polysaccharide synthesis (CPS) operon driving its synthesis. The product of the cpsE gene in this operon is crucial for biofilm formation [68]. Recent studies demonstrated that, compared to the wild-type strain, the GBS cpsE mutant secretes fewer carbohydrates, resulting in weak capsules, and, therefore, a reduced growth of biofilms [68]. Based on the capsular antigen, GBS has been classified into ten serotypes (Ia, Ib, and II-IX) as of today. Type Ib (44.4%) was the most common type of capsule, and the next most common types were Type III (40.7%), Type II (11.1%), and Type Ia (3.7%) [69].

Streptococcal polysaccharide capsules defend bacterial cells from deposition of complement, opsonization, and phagocytosis [70,71]. Capsules contain the α2,3-linked sialic acid (Sia) residues, which are analogs to a human cell surface glycocomplex epitope [72]. Typically, platelets are degranulated by contacting microorganisms through chemotaxis and released kinins and small cationic platelet microbicidal proteins (PMPs). However, the direct contribution of platelets to the killing of GBS has not been described [73]. Studies have shown that CPS Sia can effectively inhibit the killing of GBS by human platelets and resist platelet-derived antimicrobial components. GBS without Sia expression have increased susceptibility to thrombin-activated platelet release [74,75]. Meanwhile, CPS Sia can also bind to Ig superfamily lectins (Siglecs) to inhibit the activation of neutrophils and macrophages [76]. Numerous studies have demonstrated that GBS can bind to the human platelet surface receptor Siglec-9 in a Sia-dependent manner [77,78], thereby inhibiting platelet activation. A recent study demonstrated that eliminating the inhibitory Siglec-E in a mouse model effectively abolished the inhibitory effect of GBS on platelet activity [68]. It follows that CPS Sia can not only inhibit platelet activation by interacting with inhibitory Siglecs but also directly impact intrinsic resistance to platelets.

The endometrium is the inner lining of the uterus containing crucial glands for menstruation and embryo implantation [79]. Although specific virulence factors that interact with the endometrium have not been identified yet, GBS invasion could be fatal to the fetus during gestation [2,4].

The feto–maternal interface includes decidual stromal cells (DSCs), cytotrophoblasts (CTBs), and macrophages (Mφs), which are necessary for the protection of the fetus. In particular, CTBs can secret factors to regulate immune cells at the feto–maternal interface such as factors inhibiting GBS-stimulated Mφ NFκB activity, tumor necrosis factor α (TNFα), and Matrix metallopeptidase 9 (MMP9) production [80]. However, GBS has developed stronger virulence factors to resist host defense and promote its survival.

Among the numerous causes of preterm birth, infection of the fetal membrane (amniotic and chorionic) with extracellular membrane vesicles (MVs) cannot be ignored. The GBS MVs contain nucleic acids, lipids, and virulence factors such as hyaluronate lyases, sialidases, and C5a peptidase [81,82]. Moreover, the production and composition of GBS MVs are strain-dependent, with specific lineage functions related to virulence [81]. Placental membranes enclose the fetus and play an important role in its protection. Recent studies have found that GBS MVs have a strong influence on choriodecidual membranes and that their secretion of various virulence factors can diminish the integrity of the membranes [82]. Surve et al. described that the protease activity present in MVs led to the degradation of collagen [82]. The fetal membranes are located on a collagen basement membrane, which is composed of collagen types II and IV. Beneath this membrane, there is a fibrous layer containing collagen types I, III, V, and VI. Therefore, the main structural strength of the membrane is provided by collagen, and its degradation can result in the loss of membrane integrity.

Additionally, GBS MVs can lead to infiltration of neutrophils and lymphocytes under the epithelium of the amnion, thus promoting extensive inflammation in the chorio-decidua [82]. Although the mechanism remains unclear, it can cause apoptosis in the chorio-decidual membrane [82].

Reactive oxygen species (ROS) produced by recruited neutrophils are essential for host clearance of GBS [63], and it has been demonstrated that it is likely that GBS destroys the ROS through the release of hemolytic pigments from MVs, thereby weakening the defense of hosts against GBS and allowing the bacteria to survive [83]. Meanwhile, hemolytic pigments in MVs can also inhibit GBS killing by H₂O₂, a major component of the oxidative burst. Furthermore, lung observations in mice treated with non-hemolytic (NH) and hyperhemolytic (HH) MVs also revealed that hemolytic MVs may exacerbate lung injury and/or bacteremia in neonates infected with NH GBS [83]. These, therefore, suggest that hemolytic GBS MVs could contribute to the pathogenesis of GBS and neonatal morbidity and mortality.