Biopsy of living tissue remains the gold standard for disease diagnosis. Yet, as an invasive technique, it is hindered by limitations such as restricted sample range, the risk of complications, and its static nature, which may not capture the disease’s dynamics [1]. Over the past decade, advancements in the study of circulating tumor cells [2], tumor-derived DNA [3], and exosomes [4, 5] have paved the way for liquid biopsy to emerge as a superior diagnostic approach. Saliva, in particular, offers a non-invasive and convenient alternative for biopsy, presenting significant promise for its integration into precision medicine.

Saliva is a vital component of human body fluids, acting as a repository for an array of biomarkers, including metabolic products, DNA, RNA, proteins, and microbial populations [6,7,8]. It mirrors the blood’s composition due to physiological exchange mechanisms between the salivary glands and the systemic circulation. Saliva’s role in clinical diagnostics is underscored by its non-invasive collection methods, which offer advantages over blood draws by eliminating the risk of cross-infection, reducing costs, and increasing patient compliance due to its minimally intrusive nature [9, 10]. Furthermore, saliva collection is not constrained by the location or time, facilitating immediate and flexible disease diagnostics, which is particularly advantageous over conventional tissue biopsies. These attributes make saliva an exceptionally promising medium for biomarker research and its translation into clinical practice.

Transitioning from the broader concept of saliva’s role in non-invasive diagnostics, we focus on the specific entities within saliva that could hold the key to unlocking a new era of disease detection and monitoring: salivary extracellular vesicles (EVs), including exosomes. Enclosed by a lipid bilayer, salivary EVs are crucial for cellular communication and carry biomolecules that mirror the health status of patients [11, 12]. Their analysis could enhance liquid biopsy techniques by providing a stable, non-invasive means to detect and monitor diseases, contributing to advances in precision medicine.

While a few review papers have focused on the clinical significance of salivary EVs, the novelty of our review should be noted for several reasons [13, 14]. Firstly, we have systematically summarized new methods for the isolation and enrichment of salivary EVs, offering methodological choices for future research. Secondly, there has been an abundance of recent studies utilizing salivary EVs for the diagnosis and monitoring of disease activity; our review provides a timely and objective synthesis of the latest findings. Thirdly, we discuss the potential and novel perspectives of engineering salivary EVs for therapeutic applications. Lastly, we highlight the emerging challenges and propose novel considerations for the clinical application of salivary EVs.

Overview of EV biology

EVs are primarily classified into three types: apoptotic bodies, microvesicles, and exosomes [15]. These subtypes of EVs are distinguished by their size and biogenesis. A challenge in this field is the absence of specific markers for each of these EV subpopulations. Apoptotic bodies are the largest, typically ranging from 500 to 5000 nm in diameter. They form during apoptosis and contain a variety of cellular components, playing a key role in the removal of dying cells and impacting immune responses [16]. Microvesicles, also known as ectosomes, are medium in size, ranging from approximately 100 to 1000 nm. They are produced by the outward budding and fission of the plasma membrane, differing from apoptotic bodies in size, formation process, content, and specific membrane antigens [17, 18]. Exosomes are the smallest EVs, measuring between 30 and 150 nm in diameter. They are generated within the endosomal network and released upon the fusion of multivesicular bodies (MVBs) with the plasma membrane [19]. MVBs selectively package cellular components through various sorting mechanisms [20]. Exosomes are particularly rich in membrane-associated proteins, such as CD63 and CD81 [21]. The secretion process relies on the RAB family of GTP-binding proteins for vesicle transport and the SNARE family for lipid bilayer fusion [22]. Not all MVBs are destined to release exosomes; some are directed to lysosomes for content degradation under certain conditions [23].

EVs are secreted by a diverse range of cell types, including immune cells, neurons, and tumor cells, and are present in various bodily fluids such as blood, urine, and saliva [23,24,25]. Their complex structure harbors an array of bioactive molecules, including nucleic acids, proteins, and metabolites, which reflect the state of their cells of origin [26,27,28]. Consequently, EVs present in biofluids offer an accurate representation of cellular and systemic health and are invaluable for the early detection and thorough monitoring of a variety of health conditions. Salivary EVs originate from a variety of cellular sources within and adjacent to the oral cavity. Primarily, they are secreted by the cells of salivary glands, including the parotid, submandibular, and sublingual glands. Additionally, a significant contribution comes from various cell types in the oral cavity, such as oral epithelial cells, immune cells, and neuronal cells, each adding unique biomolecular signatures to the EVs. The oral microbiome further enriches this profile by releasing its own set of EVs [29]. Beyond the local environment, systemic influences are evident, with blood-derived EVs entering the saliva through the salivary gland microvasculature, and in certain conditions, such as gastroesophageal reflux disease, exosomes from the esophagus and stomach may also be detected. This complex and integrated origin of salivary EVs underlines their potential as comprehensive indicators for both oral and systemic health assessments.

Advantages for salivary EVs compared to whole saliva

Salivary EVs offer significant advantages over whole saliva by enhancing biomarker detection, ensuring sample integrity, and allowing precise disease monitoring. Firstly, whole saliva samples are highly susceptible to contamination from external substances. During saliva formation and its expulsion from the oral cavity, contaminants such as food debris and oral bacteria can easily mix in, negatively impacting subsequent analyses [13]. Additionally, salivary glands secrete enzymes such as amylase into the saliva. The presence of these contaminants and enzymatic proteins, particularly, can severely interfere with and obscure the detection of low-abundance protein biomarkers in whole saliva. However, the process of extracting and concentrating salivary EVs is designed to separate them from contaminants, thereby reducing sample complexity. This allows for the detection of low-abundance proteins in whole saliva, significantly improving the sensitivity and specificity of the detection.

Secondly, although saliva contains components other than EVs, such as shed oral mucosal cells, DNA/RNA molecules, and cytokines, the proteins and RNA in saliva can degrade and denature quickly due to the presence of a high concentration of enzymes once they leave their optimal environment, leading to insufficient detection sensitivity and specificity, and reducing the accuracy and efficiency of the examination [30]. In contrast, salivary EVs have a lipid bilayer protective structure that prevents the degradation of their contents by exogenous enzymes, thereby better preserving the integrity and stability of the sample [31]. Moreover, salivary EVs act as a barrier to low-molecular-weight substances. The intact bilipid membrane of EVs selectively restricts the entry of hydrophilic molecules [32], potentially enhancing the specificity and accuracy of disease detection using these vesicles.

Finally, salivary EVs can enrich components from specific cellular origins, allowing for targeted detection of key changes in diseased tissues, greatly enhancing the quantitative detection accuracy for specific proteins or RNA biomarkers. For instance, tumor-specific proteins and RNA can be obtained from tumor cell-derived salivary EVs. Monitoring EVs derived from epithelial cells can help in the surveillance of oral mucosal lesions, among other things. Furthermore, salivary exosomes exhibit remarkable long-term stability, maintaining membrane integrity for up to 20 months at 4 °C [14, 33]. Key exosomal markers such as DPP IV and Alix remained intact, and the exosomes proved resilient to treatments like detergents and freeze-thaw cycles, highlighting their potential for clinical use [34] (Fig. 1).

Fig. 1

EV biogenesis and the advantages of salivary EVs for disease detection and monitoring. EV biogenesis and the advantages of salivary EVs for disease detection and monitoring. (A) Salivary EVs, secreted by various types of cells, can be detected in bodily fluids. (B) Compared to whole saliva, salivary EVs offer notable advantages, including effective removal of contaminants, maintenance of integrity, and enrichment for targets. These comprehensive properties establish salivary EVs as a key component in biomarker detection

The conventional methodologies for isolating salivary EVsFig. 2

The conventional isolation method for salivary EVs. Conventional methodologies for isolating salivary EVs encompass a range of techniques, including ultracentrifugation, ultrafiltration, chemical precipitation, size exclusion chromatography, immunoaffinity capture, and microfluidics

Ultracentrifugation

Ultracentrifugation, a key technique for isolating EVs from saliva, includes differential and density gradient methods. Ultracentrifugation-isolated salivary EVs showed a higher expression of specific markers, were monodispersed, and teacup-shaped, while polyethylene glycol (PEG) resulted in irregular shapes and a broader size range. Proteomic analysis revealed ultracentrifugation as superior for isolating EV-related proteins [35]. Importantly, density gradient ultracentrifugation yielded higher quantities and purity of salivary EVs compared to size-exclusion chromatography, ultracentrifugation, and ultracentrifugation plus filtration [34]. Interestingly, using equilibrium density-gradient centrifugation with extended 96-hour ultracentrifugation, saliva-derived EVs from healthy individuals showed distinct buoyancy drifts, identifying subclasses with classical exosomal markers and those linked to membrane remodeling and vesicle trafficking [36].

Ultrafiltration

Ultrafiltration, a common size-based separation technique, uses membranes with specific pore sizes to concentrate EVs from large volumes, like cell culture media. This method allows smaller molecules or EVs to pass through, retaining larger particles. Known for processing multiple samples simultaneously, ultrafiltration saves time but faces challenges such as membrane clogging and protein contamination. To address these, it’s often combined with other methodologies. For instance, ultrafiltration and size-exclusion chromatography have been used to isolate human salivary exosomes, revealing diverse RNA compositions through next-generation sequencing. Similarly, salivary EVs can be isolated effectively using ultrafiltration with chemical precipitation [37].

Chemical precipitation method

Chemical precipitation methods, particularly PEG precipitation, are employed to isolate EVs based on their interactions with hydrophilic polymers. PEG interacts with hydrophobic protein and lipid molecules of EVs, leading to co-precipitation [38]. Although PEG precipitation can be costly and has challenges in avoiding protein aggregate contamination, its simplicity and effectiveness have led to its widespread use in commercial kits like ExoQuick and Total Exosome Isolation. When applied to saliva samples, these kits can efficiently isolate intact EVs, yielding high amounts of EVs suitable for downstream analyses. Notably, Total Exosome Isolation Solution resulted in the highest protein concentrations in salivary EVs when compared to several other techniques, including ultracentrifugation, ExoGAG, and ExoQuick [38].

Size exclusion chromatography (SEC)

SEC effectively isolates salivary EVs, separating molecules by size as smaller ones penetrate the porous beads in the column, while larger EVs elute earlier. This gentle method preserves EV integrity and minimizes protein contamination, making it suitable for functional studies and biomarker discovery. However, a drawback of this method is that it can dilute the sample, which may require additional steps to concentrate the sample afterward. Despite these limitations, SEC yielded a higher quantity of salivary EVs than ultracentrifugation, with no notable differences in dimensions between the two methods. Moreover, Izon’s qEVOriginal-70 nm columns yielded the highest purity in salivary EV isolation when compared to various other methods, indicating superior efficacy in separating EVs from other similarly sized particles [38].

Immunoaffinity capture

Immunocapture techniques for EV isolation utilize antibodies specific to EV membrane proteins. This method involves capturing EVs with magnetic beads coated with targeted antibodies, followed by magnetic separation. It enables the isolation of intact, high-purity EVs while preserving their biological activity, making them suitable for experimental applications like cell co-culture and in vivo studies. Additionally, immunocapture can be customized to isolate EVs from certain cell types or disease states, especially cancer, by targeting tumor-specific biomarkers. This specificity renders immunocapture a rapid and clinically relevant method for isolating EVs from saliva samples.

Microfluidics-based techniques

Microfluidics-based techniques for isolating EVs represent a significant advancement in EV research. These methods utilize small-scale fluidic channels to manipulate and isolate EVs with high precision. Various microfluidic devices have been designed for EV isolation, offering the dual benefits of high-purity separation and the ability to detect specific EV subpopulations in small volumes of body fluid with high throughput. For instance, microfluidic devices that employ dielectrophoretic (DEP) forces have shown great promise in the isolation of EVs. These devices apply non-uniform electric fields to selectively separate exosomes based on their polarizability. This integration of DEP in microfluidics allows for precise, efficient, and less invasive EV isolation, enhancing purity and yield while reducing processing time [39]. For instance, a novel insulator-based DEP microfluidic device with a borosilicate micropipette array efficiently isolates exosomes from biofluids like plasma, serum, and saliva. Processing 200 µL samples in just 20 min at low voltage, this easily fabricated device provides a cost-effective, high-yield, and pure extraction method, enhancing biomarker research and diagnostics [40]. Deterministic lateral displacement (DLD) technology, a microfluidic method for size-based passive particle separation, excels in EV isolation with its ability to alter the flow paths of larger particles while sparing smaller ones. This feature renders DLD ideal for exosome separation and purification. The device, comprising an array of cylindrical structures tailored to specific sizes, allows for efficient EV isolation from larger cells, catering to high-throughput screening needs. DLD’s label-free process significantly reduces sample contamination, enhancing both separation efficiency and sample purity, vital for biomedical research and clinical applications. Despite its advantages, DLD faces challenges like low throughput, pillar clogging, and cumbersome setup, with current limitations in size-based separation impacting EV saturation and recovery [41]. Notably, viscoelastic flow-based microfluidic devices exploit the unique properties of viscoelastic fluids to separate EVs from other cellular components in a biological sample. As the sample flows through a microchannel filled with a viscoelastic medium, exosomes experience different flow-induced forces compared to larger particles, leading to their effective separation. This method is particularly advantageous as it enables gentle, label-free, and low-shear isolation of EVs, preserving their integrity while ensuring efficient separation [42]. Similarly, acoustic-wave-based microfluidic devices have emerged as a highly efficient method for EV isolation. Utilizing acoustic waves to generate pressure differences within microfluidic channels, these devices can separate EVs from other cellular components based on size and density. This technique is gentle, preserving the biological and structural integrity of the EVs. Its label-free, non-invasive nature allows for high-throughput and high-purity isolation [43]. For instance, the acoustofluidic platform, fusing acoustics and microfluidics, effectively isolates salivary exosomes, outperforming differential centrifugation by yielding 15 times more exosomal small RNA. This method shows promise for high-purity and high-yield extraction of salivary exosomes, optimizing HPV detection in liquid biopsy applications [44]. Moreover, immunoaffinity capture-based microfluidic devices utilize surface-bound antibodies that selectively bind to certain proteins present on the EVs’ surface. This method ensures highly specific and targeted isolation of EVs, particularly useful for isolating disease-specific or subtype-specific EVs. The integration of immunoaffinity principles into microfluidic systems facilitates a streamlined, efficient process, enhancing the purity and relevance of the isolated EVs (Fig. 2).

Novel methodologies for enriching, quantifying and identifying salivary EVs

The enrichment of salivary EVs is crucial for the early detection and monitoring of various medical conditions. By concentrating these vesicles from saliva, it becomes possible to enhance the accuracy and sensitivity of diagnostic assays, enabling the detection of disease markers that are often present in only trace amounts in the early stages of a condition. Improved enrichment techniques are key to unlocking their full potential in personalized medicine and clinical research. Recently, a novel strategy employs Fe3O4@SiO2-aptamer nanoparticles to capture and concentrate lung cancer exosomes, with duplex-specific nuclease serving as an amplification tool to enhance miRNA signal detection. This method shows promising potential for lung cancer diagnosis by enabling the sensitive and stable detection of exosomal miR-205 in both saliva and urine samples, with a sensitivity reaching as low as 7.76 pM [45].

Accurately quantifying salivary EVs is of great importance for advancing both diagnostic and therapeutic applications. This precise quantification is not just a technical achievement; it represents a fundamental key to unlocking a deeper understanding of the role of EVs in various physiological and pathological states. By accurately measuring EV concentrations in saliva, we can establish more definitive correlations between EV profiles and specific health conditions. This enhances diagnostic precision in the early detection of diseases, in monitoring disease progression, and in evaluating responses to treatments. Recently, a novel fluorescent biosensor was developed for the one-step, sensitive quantification of salivary exosomes using magnetic and fluorescent bio-probes. This biosensor combines DNA concatamers, quantum dots, and aptamers on magnetic microspheres for efficient exosome capture, enabling a “one exosome-numerous QDs” amplification effect. It offers rapid (0.5 h) and highly sensitive quantification, characterized by a low limit of detection (500 particles/µL), a wide detection range spanning three orders of magnitude, and robust performance in complex samples, evidenced by its impressive quantitative capacity with an R-squared value of 0.998 [46]. Interestingly, a new hybrid capture bioassay for measuring microvesicle tissue factor (MVTF) activity in human body fluids has been introduced, bypassing traditional high-speed centrifugation methods. This assay integrates specific immunocapture of MVs using anti-CD29 and anti-CD59 coated magnetic beads with accurate TF activity measurement. Its application across diverse body fluids like plasma, pleural fluid, and saliva has demonstrated improved reproducibility and maintained sensitivity [47]. This development could have profound implications for the future of MV research, potentially leading to more precise diagnostic and therapeutic strategies in clinical settings.

Differentiating EVs from other nanocarriers is crucial for avoiding misinterpretation of results due to the presence of similar-sized particles, ensuring more accurate biomarker identification and paving the way for more targeted and effective treatments. A new continuous isoelectric fractionation technique has been developed for efficient separation of EVs, lipoproteins, and ribonucleoproteins from biofluids including saliva. This high-throughput method, utilizing a linear pH profile and machine learning for recalibration, achieves not only high purity and yield across various biofluids but also reaches a resolution of 0.3 ΔpI. This level of resolution is sufficient to separate different nanocarriers and their subclasses effectively [48]. Importantly, NanoFCM demonstrated superior performance in analyzing EVs ranging from 40 to 200 nm, effectively quantifying individual EV particles, identifying uncommon EV marker subsets, and facilitating the simultaneous localization of multiple surface markers. In contrast, Aurora was more effective in analyzing EVs larger than 200 nm and excelled in identifying EVs stained with various surface markers [49]. Similarly, He et al. presents a new electrochemical method for detecting salivary exosomes, using a red blood cell membrane engineered with CD63 aptamer on a gold electrode for targeted capture. This technique provides sensitive detection of target salivary exosomes across a broad linear range from 5 × 10² to 1 × 102 particles per mL, coupled with a low detection limit of 2.07 × 10² particles per mL, showing significant potential for the clinical diagnosis of oral diseases using salivary exosomes [50]. The summary of novel methodologies for enriching, quantifying and identifying salivary EVs in the review is provided in Table 1.

Table 1 Novel methodologies for enriching, quantifying and identifying merged as key tools for salivary EVsSalivary EVs as biomarkers for oral disease diagnosis and monitoring

Salivary EVs hold significant potential for the detection and management of oral diseases, given their direct interaction with the oral environment. The anatomical proximity of salivary glands to the oral cavity ensures that salivary EVs are an abundant and rich source of biological markers that reflect local pathophysiological conditions. These vesicles encapsulate and transport molecular signatures of oral health status, including responses to infections, inflammation, and neoplastic transformations. Consequently, analyzing salivary EVs offers a non-invasive, real-time snapshot of oral tissue integrity, making them highly relevant for early diagnosis, monitoring disease progression, and evaluating treatment responses in oral health care.

Periodontitis

Periodontitis is an inflammatory condition precipitated by dental plaque and influenced by multiple factors that elicit an immune response in the host. This condition not only results in the deterioration of periodontal support structures but may also have systemic health implications [51]. Distinct microRNAs (miRNAs), proteins, and molecular patterns in salivary EVs effectively differentiate periodontitis patients from healthy individuals, underscoring their diagnostic potential. In a pilot study with 29 participants, three miRNAs, namely miR-140-5p, miR-146a-5p, and miR-628-5p, were markedly elevated in the salivary EVs of periodontitis patients [52]. Another study found that levels of miR-223-3p in salivary exosomes were lower in periodontitis patients compared to healthy individuals. This particular miRNA plays a crucial role in regulating inflammation in periodontitis by targeting the NLRP3 protein [53]. Moreover, in comparison to healthy saliva, exosomal miRNAs from chronic periodontitis samples were predominantly down-regulated, although miR-125a-3p was notably upregulated. Owing to its strong correlation with periodontal pocket depth, miR-125a-3p emerges as a potential key biomarker for chronic periodontitis [54]. Beyond salivary exosomal miRNAs, other components have also been reported to be dysregulated in the saliva samples from periodontitis patients. For instance, in young adults with severe periodontitis, salivary exosomes contained 26 proteins not detected in the healthy cohort. Conversely, the healthy group exhibited 58 proteins absent in the periodontitis samples. Importantly, exosomes from the periodontitis group were enriched with immune-related proteins, such as complement components and chemokine ligand 28, suggesting their active involvement in the immune response characteristic of the disease [55]. Interestingly, although EV size and morphology remained consistent across participants, the CD9 + subpopulation was more prevalent in those with periodontitis. A decrease in osterix mRNA and an increase in tumor necrosis factor-alpha (TNFα) were also observed, indicating their potential as diagnostic markers for periodontitis [56]. Furthermore, in the salivary EVs from periodontitis patients, higher levels of LPS + outer membrane vesicles, m5C methylation, and four periodontal pathogens were identified compared to healthy individuals. Notably, m5C hypermethylation in salivary EVs effectively distinguished periodontitis patients from both healthy and gingivitis groups [57].

Beyond their diagnostic utility, salivary EV profiles also serve as robust biomarkers for disease severity and responsiveness to therapy in periodontitis. In patients with periodontitis, salivary samples demonstrated decreased levels of exosomal CD9 and CD81, when compared to healthy controls. Notably, these reduced levels correlated with clinical measures of periodontal disease severity [58]. Similarly, increased levels of PD-L1 mRNA were observed in salivary exosomes from periodontitis patients compared to controls, with variations in these levels indicating differences in disease severity. This suggests the potential utility of salivary PD-L1 mRNA as a diagnostic and prognostic marker [59]. Importantly, following initial periodontal therapy (IPT) for periodontitis, significant changes were observed in the composition of salivary exosomes. In particular, patients exhibiting increased periodontal inflammation post-IPT demonstrated elevated levels of C6 protein and miRNAs, including miR-142 and miR-144. Conversely, decreased or stable expressions of CD81 and TSG101 were associated with clinical improvements in periodontitis, such as reduced probing depth. Changes in HSP70 expression were consistent with persistent levels of periodontal inflammation [60] (Fig. 3).

Fig. 3

Salivary EVs in the diagnosis and monitoring of periodontitis. Notable differences in salivary exosomal miRNAs and proteins have been observed between healthy individuals and patients with periodontitis. Moreover, salivary EV profiles also provide robust biomarkers for assessing disease severity and responsiveness to therapy in periodontitis. This highlights the significant role of salivary EVs as valuable biomarkers in the diagnosis and ongoing monitoring of periodontitis

Oral cancer

Oral cancer, defined as a malignant neoplasm found in the oral cavity, greatly benefits from early detection and treatment, which can significantly improve patient survival rates [61,62,