As important mediators of bacterial virulence, O-BEVs are generated through distinct biogenesis pathways influence by bacterial cell envelope architecture^[22]. Gram-negative bacteria possess an outer membrane (OM), which serves as the direct source for vesicle budding, an inner membrane (IM), a periplasmic space (PS), and a thin peptidoglycan (PG) layer^[23]. In contrast, Gram-positive bacteria are characterized by a thick, multi-layered PG cell wall that encases the cytoplasmic membrane (CM) and presents a physical barrier to vesicle release^[12]. These structural distinctions give rise to several types of vesicles, including outer membrane vesicles (OMVs) from Gram-negative species, cytoplasmic membrane vesicles (CMVs) from Gram-positive species, as well as other less common or condition-dependent forms such as outer-inner membrane vesicles (OIMVs), explosive OMVs, and nanotubules^[24,25]. Among these, OMVs and CMVs represent the most abundant and best-characterized vesicle types among O-BEVs, and they are considered important mediators of bacterial virulence in periodontitis and associated systemic diseases^[26,27]. Therefore, the subsequent sections describe the biogenesis processes of OMVs and CMVs [Figure 1]^[28].

In Gram-negative oral bacteria such as Porphyromonas gingivalis (P. gingivalis), Tannerella forsythia (T. forsythia), Aggregatibacter actinomycetemcomitans (A. actinomycetemcomitans), Fusobacterium nucleatum (F. nucleatum), and Treponema denticola (T. denticola), OMVs are secreted with a typical diameter of 20-250 nm^[12]. These OMVs are enriched in surface-associated virulence factors, capable of inducing inflammatory responses and damaging host cells, thereby contributing to the onset or progression of oral diseases and systemic conditions^[29]. The biogenesis of OMVs is a multi-step process initiated by a dynamic imbalance between the OM and the PS, involving membrane remodeling, budding dynamics, and fission^[15]. Key factors promoting OM protrusion include the asymmetric distribution of OM lipids, the accumulation of hydrophobic molecules or protein complexes, and the remodeling of the PG layer, which may involve the dissociation of Braun’s lipoprotein^[30]. Subsequently, the OM undergoes reorganization, with phospholipids and LPSs rearranging to form spherical vesicles. Endolysins may further promote this process by locally degrading the PG layer, thereby facilitating OM budding^[31].

In contrast, CMVs derived from Gram-positive oral bacteria - including Filifactor alocis (F. alocis), Streptococcus gordonii (S. gordonii), Stomatobaculum longum (S. longum), Streptococcus mutans (S. mutans), Streptococcus sanguinis (S. sanguinis), and Streptococcus oralis (S. oralis) - must overcome the physical barrier of the thick cell wall during biogenesis^[15]. These CMVs are secreted with a typical diameter of 20-400 nm^[12]. This process is typically initiated by the localized degradation of the PG layer by endolysins, creating pores that expose the CM as budding sites. Subsequently, the membrane bulges outward through these pores, accompanied by the reorganization of lipids and proteins^[12]. Membrane-associated and curvature-stabilizing proteins interact with the lipid bilayer to facilitate the formation of sealed vesicular structures^[30]. During membrane lipid reorganization, specific lipids such as cardiolipin accumulate at budding sites, promoting vesicle formation and sealing by increasing membrane curvature and reducing fluidity. Meanwhile, cytoskeletal-like proteins and other curvature-associated proteins interact with lipids to regulate membrane shape and stability, facilitating CMVs biogenesis^[25]. For vesicle escape, CMVs traverse the cell wall through pores remaining after endolysin degradation or potentially through phage-like channel proteins^[15,25].

Emerging evidence has implicated periodontal pathogens, particularly P. gingivalis, in AD pathogenesis, with OMVs proposed as key mediators of the “oral-brain axis”^[71]. As detailed in Table 2, O-BEVs from P. gingivalis and A. actinomycetemcomitans have been reported to contribute to AD pathology through three interconnected mechanisms: (i) blood-brain barrier (BBB) disruption; (ii) neuroinflammation; and (iii) tau hyperphosphorylation, among which A. a-OMVs are involved only in BBB disruption and neuroinflammation.

While intact P. gingivalis bacteria induce pronounced systemic inflammation, OMVs alone are sufficient to induce BBB disruption and promote neuroinflammation, suggesting a distinct pathogenic mechanism^[72]. The P. g-OMVs activate the NLRP3 inflammasome in microglia, leading to IL-1β release^[73]. This inflammatory cascade subsequently induces tau hyperphosphorylation at the Thr231 site in neurons and impairs spatial memory and learning in middle-aged mice^[74]. The LPS and gingipains carried by these OMVs have been implicated as key effector molecules driving these effects. For example, P. g-OMVs significantly increase the expression of IL-6, TNF-α, IL-8 and IL-1β in microglial cells, whereas a gingipain-deficient KDP163 strain fails to upregulate these genes^[75]. Moreover, LPS from P. g-OMVs can activate the NF-κB signaling pathway via TLR4, causing substantial release of pro-inflammatory cytokines and further aggravating neuroinflammation^[76]. In a mouse model, the P. g-OMVs significantly increased the expression of IL-1β and TNF-α in the hippocampus and cortex, activated astrocytes and microglia, and resulted in memory dysfunction^[77].

In contrast to the gingipain-centered mechanisms of P. g-OMVs, the A. a-OMVs have been reported to induce neuroinflammation primarily through their RNA cargo. These OMVs can alter action potentials in trigeminal ganglion (TG) neurons^[72]. In vitro and animal studies have suggested that these OMVs may cross the BBB and deliver extracellular RNA into brain monocytes and microglial cells, activating TLR4/MyD88 and TLR8-dependent NF-κB pathways and promoting TNF-α and IL-6 secretion^[78]. This finding suggests that different O-BEVs may converge on similar neuroinflammatory endpoints through distinct molecular routes - a nuance that would be missed by a simple enumeration of individual mechanisms.

Epidemiological and mechanistic studies have increasingly linked periodontitis to atherosclerosis, with O-BEVs emerging as potential mediators of this association^[79]. As detailed in Table 2, O-BEVs from P. gingivalis, F. nucleatum, and T. denticola have been implicated in atherosclerosis through three interconnected mechanisms: (i) endothelial dysfunction and barrier disruption; (ii) vascular calcification and smooth muscle cell dysfunction; and (iii) foam cell formation and lipid accumulation.

Studies have shown that P. g-OMVs promote vascular smooth muscle cell (VSMC) calcification in a concentration-dependent manner via the extracellular signal-regulated kinase 1/2-Runt-related transcription factor 2 (ERK1/2-RUNX2) pathway, a hallmark of atherosclerotic progression^[80]. Using in vitro and in vivo experiments, Farrugia et al. confirmed that P. g-OMVs increased vascular permeability by disrupting platelet endothelial cell adhesion molecule-1 (PECAM-1) in endothelial cells in a gingipain-dependent manner and enhanced vascular edema and mortality in a zebrafish model. Further evidence indicates that gingipain-containing OMVs impair intercellular contacts by cleaving junctional proteins such as CD31, thereby facilitating trans-endothelial migration of immune cells into the arterial intima^[81]. Moreover, P. g-OMVs have been reported to suppress the expression of endothelial nitric oxide synthase (eNOS) in human umbilical vein endothelial cells (HUVECs) and mouse aortic endothelium, potentially reducing NO secretion and impairing endothelial function^[82]. Once monocytes have infiltrated into the arterial wall, P. g-OMVs contribute to the aggregation and modification of low-density lipoprotein (LDL), promoting foam cell formation and accelerated lipid accumulation within the vessel wall. Fleetwood et al. found that stimulation of macrophages with P. g-OMVs triggered a metabolic shift from oxidative phosphorylation to glycolysis, accompanied by secretion of numerous inflammatory mediators. This metabolic reprogramming may activate the inflammasome and induces pyroptosis, resulting in the release of inflammatory cytokines and cytoplasmic components into the extracellular environment, thereby sustaining local inflammation and amplifying pyroptotic cell death^[83].

Beyond P. gingivalis, T. d-OMVs can activate endothelial cells by inducing the expression of IL-8 and MCP-1, which may facilitate chemotaxis and aggregation of monocytes - an early step in atherogenesis^[84]. Additionally, the F. n-OMVs have been reported to upregulate the scavenger receptor CD36 via the TLR4/NF-κB pathway, enhancing the uptake of oxidized LDL (ox-LDL) and accelerating foam cell formation^[85].

In summary, OMVs derived from key periodontal pathogens do not act in isolation. Instead, they likely form a complex pathogenic network that collectively accelerates the initiation and progression of atherosclerosis through multiple interconnected mechanisms, including disruption of endothelial homeostasis, dysregulation of lipid metabolism, and perpetuation of chronic inflammation.

Epidemiological and mechanistic studies have long suggested a bidirectional relationship between periodontitis and diabetes mellitus (DM): periodontitis can impair glycemic control, insulin action, and diabetic outcomes, while DM can heighten periodontitis severity through delayed healing and enhanced infection risk^[86]. As detailed in Table 2, P. g-OMVs and F. n-OMVs have been implicated in the pathogenesis and complications of diabetes through three interconnected mechanisms: (i) direct impairment of hepatic insulin signaling; (ii) exacerbation of diabetic complications (e.g., retinopathy); and (iii) promotion of systemic and intestinal inflammation that may worsen insulin resistance.

The P. g-OMVs can deliver active gingipains to the liver. In HepG2 hepatocytes, these vesicles attenuate insulin-induced protein kinase B (Akt)/glycogen synthase kinase-3 beta (GSK-3β) signaling in a gingipain-dependent manner, thereby disrupting hepatic glucose metabolism and potentially promoting the onset and progression of DM^[87]. Furthermore, in mice, P. g-OMVs worsen diabetic retinopathy (DR), an effect that may involve mitochondrial-associated cell death and endothelial dysfunction triggered by PAR-2 signaling in human retinal microvascular endothelial cells^[88]. Emerging evidence also indicates that periodontal pathogens and their vesicles may promote insulin resistance in peripheral tissues. For instance, OMVs derived from other periodontopathic bacteria such as F. nucleatum have been implicated in impairing insulin signaling in adipocytes through TLR4-mediated inflammatory pathways^[89]. According to Engevik et al., F. n-OMVs activate TLR4 on intestinal epithelial cells. This may trigger a signaling cascade involving ERK, CREB, and NF-κB, which then stimulates the release of pro-inflammatory cytokines (e.g., IL-8, TNF-α) and promotes intestinal inflammation^[89]. Moreover, systemic inflammation triggered by OMVs-induced release of pro-inflammatory cytokines may further aggravate pancreatic β-cell dysfunction and insulin resistance^[90]. These mechanisms collectively underscore the role of O-BEVs as potential contributors to the pathogenesis and complications of diabetes, suggesting the potential benefits of oral microbiome management in glycemic control.

Autoantibodies produced in RA patients, such as rheumatoid factor (RF) and ACPAs, bind synovial antigens, activate the complement and trigger joint inflammation^[91]. Epidemiological and mechanistic studies have long suggested an association between periodontitis and RA, with O-BEVs emerging as potential mediators linking oral dysbiosis to joint autoimmunity^[92]. As detailed in Table 2, P. g-OMVs and F. n-OMVs have been linked to RA pathogenesis through two distinct but potentially complementary mechanisms: (i) PPAD-mediated protein citrullination and ACPA generation; and (ii) FadA-mediated activation of synovial macrophages.

PPAD in P. g-OMVs serves as a key effector molecule mediating the pathogenic role of these vesicles in RA. The PPAD catalyzes the citrullination of host proteins such as fibrinogen and vimentin, generating ACPAs that are highly specific to RA and triggering inflammatory responses in the synovium^[93]. Notably, other virulence factors within P. g-OMVs may act synergistically with PPAD. For instance, gingipains cleave proteins at arginine residues, whereas PPAD preferentially citrullinates C-terminal arginine residues on polypeptide chains^[94]. Furthermore, elevated levels of TNF-α, IL-1β, and IL-6 in both serum and synovial fluid promote inflammatory responses and exacerbate joint damage^[95]. The O-BEVs derived DNA has been detected not only in serum^[96], but also in synovial fluid^[97], providing additional evidence for the systemic dissemination of oral bacterial components to the joint microenvironment.

A positive correlation exists between F. nucleatum abundance in RA patients and disease severity. F. n-OMVs carrying FadA can reach the joints and provoke local inflammation, and FadA targets synovial macrophages, resulting in activation of Rab5a GTPase (a vesicular trafficking regulator) and YB-1 (an inflammatory pathway modulator)^[98]. Moreover, certain O-BEVs-associated molecules can act as damage-associated molecular patterns (DAMPs), sustaining chronic activation of innate immune pathways within the joint microenvironment. Collectively, these mechanisms underscore how O-BEVs orchestrate a breach in immune tolerance and fuel autoimmunity in RA, providing a compelling mechanistic basis for the established association between periodontitis and RA.

Accumulating epidemiological data have revealed that adverse pregnancy outcomes (APOs) are strongly associated with periodontitis^[99,100]. As detailed in Table 2, P. g-OMVs have been implicated in APOs through three interconnected mechanisms: (i) direct targeting and dysfunction of trophoblast cells at the maternal-fetal interface; (ii) indirect damage mediated by host immune cell-derived EVs; and (iii) offspring transgenerational neurodevelopmental alterations.

The P. g-OMVs can directly target the maternal-fetal interface and disrupt normal pregnancy progression. For example, when pregnant mice were intraperitoneally injected with P. g-OMVs, fluorescent signals were subsequently detected at embryo implantation sites^[101]. Low doses of P. g-OMVs reduced fetal and placental weights, whereas high doses resulted in fetal loass^[102]. In vitro experiments from the same study revealed that P. g-OMVs decreased glucose uptake and glycolysis in trophoblast cells, impairing their migration and invasion capabilities. Proper migration of trophoblasts and endothelial cells is crucial for remodeling spiral arteries at the maternal-fetal interface^[101,102]. Under physiological conditions, trophoblasts contribute to placental homeostasis by inactivating neutrophils through the suppression of ROS release and neutrophil extracellular trap (NET) formation^[103]. Conditioned medium from P. g-OMV-treated trophoblast cells, however, promoted neutrophil chemotaxis and increased the production of ROS, IL-8, and TNF-α, potentially disrupting placental homeostasis and thereby inducing APOs^[104].

Beyond the direct effects of bacterial OMVs, the pathogenic impact of oral bacteria can be further amplified by host immune cells. EVs derived from P. gingivalis-infected macrophages (P. g-inf EVs) translocated to the fetoplacental unit and impaired fetal development, as evidenced by reduced fetal size and weight. Histological analysis of the placenta in the P. g-inf EV-injected group revealed disorganized vasculature, impaired angiogenesis, and compromised placental function, and proteomic analysis indicated a significant downregulation of VEGFR1 expression in the experimental placentas^[105].

More critically, maternal exposure to P. g-OMVs during pregnancy may exert intergenerational effects, directly disrupting the neurodevelopmental programming of offspring. Further evidence demonstrates that maternal exposure to P. g-OMVs induces marked neurodevelopmental alterations in the offspring. These include the suppression of key molecules (IL-6, Cux1 and SatB2) implicated in cortical development and neuronal differentiation, concurrent with an elevation in the Alzheimer’s disease-associated phospho-Tau (Thr²³¹) and changes in embryonic cortical neuron density^[106]. The ability of O-BEVs to translocate systemically and disrupt key physiological processes during pregnancy underscores their potential role as mediators of the “oral-systemic axis” in reproductive pathology. Further research is warranted to explore the translational potential of targeting O-BEVs or their cargo for the prevention and management of periodontitis-associated APOs.

In summary, O-BEVs establish a multiple pathogenic axis from local infection to systemic reproductive pathology and even intergenerational health effects, spanning from direct damage to placental trophoblasts, to secondary immune-mediated damage, and finally to interference with offspring neurodevelopment. Future research is essential to thoroughly explore the translational potential of targeting these vesicles or their pathogenic components, thereby paving the way for novel intervention strategies to prevent and manage periodontitis-associated APOs.

Osteoporosis and periodontitis exhibit a bidirectional relationship^[107]. Osteoporotic individuals face a twofold greater risk of developing periodontitis than healthy people^[108]. In both conditions, bone loss localized to alveolar bone in periodontitis or systemic in osteoporosis - is tied to heightened osteoclast differentiation and reduced osteogenesis. As detailed in Table 2, O-BEVs from multiple oral pathogens contribute to bone loss through two principal mechanisms: (i) promotion of osteoclast differentiation and (ii) suppressed osteogenesis.

Lipoproteins or LPS from O-BEVs from F. alocis, P. gingivalis, T. forsythia, and S. oralis drive osteoclast differentiation through activation of TLR2^[109]. Emerging evidence indicates that P. g-OMVs exacerbate osteoporosis by impairing mitochondrial dynamics, which downregulates the protein level of CPT2. This suppression of fatty acid oxidation (FAO) subsequently compromises ATP production in osteoblasts, ultimately contributing to bone loss^[110]. Furthermore, when added to osteogenic medium, F. a-CMVs suppressed bone formation in a dose-dependent manner, as indicated by reduced expression of osteogenic marker genes^[111]. Following intraperitoneal injection of DiO-labeled F. a-CMVs into mice, strong fluorescence was detected in the tibiae and femora, accompanied by decreased trabecular bone volume and elevated levels of the bone resorption marker CTX-1^[112]. These findings suggest that O-BEVs can travel to long bones and trigger bone loss, potentially via TLR2 signaling.

Beyond the systemic conditions discussed above, O-BEVs have also been implicated in several other human diseases, including respiratory diseases, hepatic steatosis, oral lichen planus (OLP), and HIV-1 infection, as summarized in Table 2.

Respiratory diseases: The P. g-OMVs trigger cell death in lung epithelial cells by disrupting the epithelial barrier, suggesting that these vesicles may be an important factor linking periodontitis to respiratory diseases^[113]. Notably, P. g-OMVs are enriched in core histones (e.g., H3) and translocate to the lungs, liver, and kidneys of mice. Both P. g-OMVs and recombinant H3 activated the NF-κB pathway, contributing to increased levels of pro-inflammatory cytokines in human lung epithelial A549 cells. The P. g-inf EVs induced lung injury, including edema, vascular congestion, inflammation, and collagen deposition, which was associated with alveolar damage^[114].

Hepatic steatosis: The F. a-CMVs have been associated with hepatic steatosis, particularly through increasing this condition in mice on a low-fat diet through mechanisms involving TLR-2 and PAI-1^[115].

OLP: A. a-OMVs and P. g-OMVs did not affect cell viability but potently increased the mRNA levels of TNF-α, IL-6, and IL-8. Moreover, these vesicles activated the STAT3 signaling pathway, as evidenced by increased phosphorylation and simultaneous upregulation of IL-1β mRNA, along with promoted NLRP3 protein accumulation. These changes suggest that the inflammasome complex may be activated^[116].

HIV-1 infection: In MT4 cells, P. g-OMVs promoted HIV-1 infection even when viral loads alone were too low to establish a productive infection. This suggests that OMVs may act as vectors for mucosal HIV transmission, thereby facilitating infection establishment and enhancing viral infectivity^[117].