The impact of environmental contaminants on extracellular vesicles and their key molecular regulators: A literature and database-driven review

Abstract

Exposure to environmental chemicals is now well recognized as a significant factor contributing to the global burden of disease; however, there remain critical gaps in understanding the types of biological mechanisms that link environmental chemicals to adverse health outcomes. One type of mechanism that remains understudied involves extracellular vesicles (EVs), representing small cell-derived particles capable of carrying molecular signals such as RNAs, miRNAs, proteins, lipids, and chemicals through biological fluids and imparting beneficial, neutral, or negative effects on target cells. In fact, evidence is just now starting to grow that supports the role of EVs in various disease etiologies. This review aims to (1) Provide a landscape of the current understanding of the functional relationship between EVs and environmental chemicals; (2) Summarize current knowledge of EV regulatory processes including production, packaging, and release; and (3) Conduct a database-driven analysis of known chemical–gene interactions to predict and prioritize environmentally relevant chemicals that may impact EV regulatory genes and thus EV regulatory processes. This approach to predicting environmentally relevant chemicals that may alter EVs provides a novel method for evidence-based hypothesis generation for future studies evaluating the link between environmental exposures and EVs.

1 |. INTRODUCTION

1.1 |. Introduction to extracellular vesicles

Extracellular vesicles (EVs) are small, lipid membrane bound, microparticles that can originate from nearly all cell types and are categorized by physical properties such as size, biochemical composition, or cell of origin (Thery et al., 2018). Several proposed EV subtypes exist including exosomes, microvesicles, and apoptotic bodies among others, though official nomenclature has yet to be agreed upon given commonly overlapping parameters (Carberry, Keshava, et al., 2022; Thery et al., 2018). In general, exosomes are frequently characterized as small EVs (<200 nm diameter) generated within cellular endosomal compartments and released upon fusion with the cell membrane. Microvesicles are commonly defined as medium-to-large-sized EVs with approximate size ranging from 100 to 2000 nm in diameter. These vesicles are formed via direct budding from the cell membrane. Finally, apoptotic bodies are larger vesicles ranging from 1000 to 5000 nm in diameter and are released exclusively from cell death processes. While these specifications have been commonly reported, they are not mutually exclusive. Reporting of the physical properties of extracellular vesicles isolated in experiments serves to contextualize the biological relevance of findings. However, given the challenges of reaching a non-overlapping definition and/or mechanism for differentiating between exosomes, microvesicles, apoptotic bodies, or other EV categories, there is a growing preference for using the term “EVs” over other nomenclature (Carberry, Keshava, et al., 2022; Thery et al., 2018).

Biological relevance may be further characterized by evaluation of EV molecular content. EVs carry RNA fragments, microRNAs (miRNAs), proteins, and chemicals among other small molecules (Hartjes et al., 2019). While EVs are recognized in part for their role in cellular waste management, it is also understood that EV content may reflect cell of origin or be uniquely packaged in response to external factors such as cellular stress, exogenous insults, and other physiological or pathological changes (Abels & Breakefield, 2016; Couch et al., 2021). Once released from cells into the extracellular space, EVs can remain locally or travel through biological fluids to reach nearby and distal target cells. EVs function by transferring their biological cargo to recipient cells, thereby imparting neutral, beneficial, or adverse effects in target cells upon uptake (Kwok et al., 2021). Through this mechanism, EVs play a key role in normal cellular homeostasis as well as disease physiology. Numerous papers have proposed EV involvement in cellular health and regulation (e.g., cell death, cell stress mechanisms, immune responses, etc.) (Murao et al., 2021; Sanwlani & Gangoda, 2021; Zhou et al., 2022) as well as various disease pathways including cardiovascular disease, diabetes, aging, and cancer among others (Akhmerov & Parimon, 2022; Suire & Hade, 2022; Thery et al., 2018). As such, these molecules have emerged as important biomarkers and mediators of exposure and disease in clinical and epidemiological research (Carberry, Keshava, et al., 2022). One field that will benefit from expanded EV research is environmental toxicology where there is currently less discussion surrounding EVs compared to other fields such as biomedical fields.

1.2 |. Overview on human exposure and health effects research

Environmental chemicals are a major threat to global public health worldwide, with a growing incidence of associated adverse health outcomes ranging from irritation to acute and systemic disease, and mortality (Peters et al., 2021). Humans encounter potentially hundreds of chemicals in their daily environment via polluted air, contaminated water or food, or consumer products, which can then impact organ systems upon inhalation, ingestion, or dermal absorption. Current estimates report that there are between 80,000 and 150,000 chemicals in commerce, the majority of which having little to no toxicity information available (Bakand et al., 2005; Breen et al., 2021). Of these chemicals, several thousand are reported to have contaminated the environment (Landrigan et al., 2018).

Exposure to environmental pollution has been estimated to account for 16% of all deaths worldwide, disproportionately burdening lower income countries (Landrigan et al., 2018). Recent reports show that exposures including air pollution, water pollution, lead, and occupational hazards were responsible for nine million premature deaths in 2019 (Fuller et al., 2022). These estimates are likely greatly underestimated, given that the majority of chemicals in our environments remain under evaluated. Example chemical groups of particular concern include metals and metalloids, phthalates, flame retardants, pesticides and herbicides, polychlorinated biphenyls, per and polyfluoroalkyl substances (PFAS), and other manufactured chemicals that have established association with various adverse health outcomes: adverse reproductive health, endocrine disorders, neurological diseases, immune dysfunction, and cancers (Fuller et al., 2022; Grandjean & Bellanger, 2017; Rager et al., 2020).

Given the burden of disease estimated to be caused or impacted by environmental exposures, there is an increasing need to understand how our bodies respond and adapt to these insults. This includes understanding the underlying biological mechanisms linking exposure to disease. Understanding the biological mechanisms underlying environmental exposure-induced disease may help researchers develop treatments to mitigate or even reverse adverse effects. One understudied mechanism of environmental chemical toxicity is mediation by EVs (Figure 1). As previously mentioned, EVs represent a growing area of research within exposure science, toxicology, and public health research (Carberry, Keshava, et al., 2022).

FIGURE 1.

Overview of environmental exposures mediating EV regulation and function. This schematic depicts one potential mechanism through which environmental exposures may alter EV regulation and function. First, exposure to environmental chemicals (panel a) may alter gene and/or protein expression of known EV regulators (b). Alterations of these EV regulators result in differential release and molecular contents (c). These EVs are capable of remaining locally, or traveling through biological fluids to impart a wide range of effects on various target cell types (d).

1.3 |. Overview of literature and database-driven review

This paper compiles evidence to explore whether environmental toxicants impact the genes involved in regulating EV production, packaging, and release. Here we implement a novel method for informing critical knowledge gaps involving the link between environmental exposure and EVs by leveraging both literature and databases. Specifically, current knowledge of genes and their associated proteins shown to regulate EV production, release, and packaging of molecular content is compiled using critical reviews. This gene list is further evaluated in conjunction with two publicly available databases, the Comparative Toxicogenomic Database (CTD) and the Environmental Protection Agency (EPA) CompTox Dashboard, to review and summarize previously published chemical-gene interactions between environmental contaminants and EV regulators. This approach to predicting environmentally relevant chemicals that may alter EVs sets a novel foundation for evidence-based hypothesis-generation for future evaluations connecting environmental contaminants to EVs.

2 |. METHODS

2.1 |. Reviewing existing literature evaluating environmental contaminants and changes in EVs

To evaluate the current state of literature supporting the hypothesis that environmental chemicals impact EV regulation, function, and downstream health effects, a search string was used to identify papers addressing this topic. Articles published before August 2022 were searched in PubMed using search terms: ((Extracellular Vesicle*) OR (Cell-Derived Microparticle*) OR (Exosome*) OR (Microvesicle*) OR (Ectosome*)) AND ((Environmental Pollutant*[MeSH Terms]) OR (Environmental Carcinogen*[MeSH Terms]) OR (Pesticide*[MeSH Terms]) OR (Environmental Pollution[MeSH Terms]) OR (Environmental Exposure[MeSH Terms])) NOT (Review[Publication Type]) Abstracts or full-text (when applicable) were reviewed to determine whether resulting studies met inclusion or exclusion criteria. To be included in this high-level review, papers must be an original research paper evaluating an environmentally relevant chemical or contaminant in relation to EV endpoints such as count, regulation, content, or function. Additionally, papers must be relevant to human exposure, toxicology, or epidemiology, spanning human cell culture, animal models routinely used to inform human health (e.g., mammalian models), and human subject study designs. Papers were excluded if they were review articles, were relevant only to ecotoxicity, or if exposures were exclusively medical treatments. Additional relevant literature not captured by the developed search string were manually queried and included.

2.2 |. Compiling key genes associated with EV production, packaging, and release

Several research groups have previously reviewed known mechanisms of EV regulation. Given the breadth of relevant reviews, five reviews were selected on this topic that have identified biological pathways and genes involved in EV production, packaging, and release (Abels & Breakefield, 2016; Gurunathan et al., 2021; Hessvik & Llorente, 2018; Kowal et al., 2014; Pegtel & Gould, 2019). Molecules identified as involved in EV regulation are reported using their associated official National Center for Biotechnology Information (NCBI) Entrez Gene names. The general regulatory function and specific EV category for each gene (and its associated protein) is also reported including production, packaging, or release. This list is not intended to be fully comprehensive, but to rather describe well-studied genes involved in EV regulation.

2.3 |. Predicting environmental chemicals with potential to alter mechanisms of EV regulation

To investigate potential chemical–gene interactions between environmental chemicals and EV regulator genes, a batch query was performed within CTD using official NCBI Entrez Gene names. EV regulator gene names were input using NCBI symbols and “any” available curated chemical–gene interaction data were downloaded. This included data from both vertebrates and invertebrates from published literature, as well as both direct and indirect associations. Direct interactions included chemicals that bind directly to proteins, whereas indirect interactions impacted proteins via intermediate events (i.e., increased phosphorylation). Data were then filtered for published chemical interactions reported specifically in humans. Output data were sorted by chemicals that impact the highest number of unique EV regulatory genes.

Chemicals previously demonstrated to impact EV regulatory genes were then filtered by environmental relevancy. Environmental relevance of the chemicals identified by CTD was determined by matching and filtering to include chemicals present in select lists within the EPA CompTox Chemical Dashboard. Environmentally relevant EPA CompTox lists were selected to include chemicals found in air, water, biosolids, and consumer products among other relevant categories. Specific lists queried and their respective justification of environmental relevance are detailed in Table 1. A batch search including all unique chemical names identified by CTD was performed using the CompTox Dashboard. Chemical names were filtered for presence in selected lists and matched with corresponding chemical-gene interaction data.

TABLE 1.

EPA CompTox dashboard chemical lists and environmental relevance

List name	Environmental relevance
BIOSOLIDS2021	Chemicals detected in biosolids from wastewater treatment plants
CPDAT	Chemicals in consumer products
FDAFOODSUBS	Substances added to food
FERTILIZERS	List of chemicals in fertilizers
FLAMERETARD	List of chemicals in flame retardants
INDOORCT16	List of chemicals in indoor dust
NATADB	List of US air toxics
SCDM	Superfund chemicals
USGSWATER	List of chemicals in water
WIKIHERBICIDES	List of herbicides
WIKIINSECTICIDES	List of insecticides

To be considered environmentally relevant, chemicals must be present in one or more of the selected lists. For chemicals either not present in selected lists, or not able to be identified within the CompTox Dashboard, environmental relevance was determined by manual query. Many unidentified chemicals included understudied chemicals, compounds or mixtures, or broad chemical categories that do not have specific chemical identifiers. Example environmentally relevant chemicals that were not identified using EPA chemical lists included “tobacco smoke pollution,” “particulate matter,” “vehicle emissions,” or “dust” among others. Unidentified chemicals were manually queried until the top 50 chemicals impacting EV regulatory genes were identified. Overall methods are further summarized in schematic Figure 2.

FIGURE 2.

Schematic of methods for predicting chemicals likely to alter EV regulation. The objective, method, and results for each step are described.

3 |. RESULTS AND DISCUSSION

3.1 |. Landscape of existing literature evaluating environmental contaminants and changes in EVs

EV research is a rapidly developing field, especially in cancer, pharmaceutical, and other medical related fields. Numerous studies have established that EVs may have beneficial, or adverse effects on disease progression and outcomes, and are sometimes thought to be neutral in consequence (Kwok et al., 2021). When evaluating EVs, it is of critical importance to understand the potential changes in EV release, content, and function that may occur in response to exogenous insults. Environmental toxicants have been shown to alter EV release, contents, and function that have associations with toxicant-mediated diseases (Benedikter et al., 2018). Despite the growing body of literature that links environmental exposures to the onset or progression of adverse health outcomes, the diversity of underlying mechanisms through which environmental exposures impact disease have yet to be fully accounted for and elucidated. A growing body of evidence supports that alterations in EVs are important mechanisms mediating exposure-associated toxicity and disease.

To survey the landscape of research linking environmental contaminants and EVs, a literature search was performed in PubMed, followed by a manual search. This search resulted in 152 papers of which 58 were determined to meet inclusion criteria as previously defined. Publication titles and relevant information for included papers are further detailed in Table S1. These papers may be generally categorized by publication year, primary evaluated EV endpoint, environmental exposure category, and experimental type as summarized in Figure 3. Note that the number of publications organized by year are shown to slightly dip from 2020 to 2022, though this trend is likely due to the lag in papers being fully annotated within the PubMed system.

FIGURE 3.

Landscape of literature investigating environmental exposure-associated EV alterations. (a) The number of publications per year, up until August 2022*. Note that number of paper in 2022 will likely increase as articles continue to be published and PubMed fully annotates existing literature. (b) Publication by primary EV endpoint evaluated is summarized. (c) Exposure categories that have been investigated in relation to EVs are summarized. (d) EV research by having been conducted in vitro, in vivo, or both is summarized.

There are still relatively few papers linking environmental exposures with EV alterations. However, there has been an apparent and steady increase in publications between 2012 and the present, indicating growth in this field. Within these studies, several endpoints have been evaluated including EV count, content, biological function, cell origin, regulation, uptake, and size among others. Of these endpoints, EV count, content and function have been most frequently evaluated. It should be noted that EV count and size are commonly measured for isolation validation, though may also serve as valuable endpoints. These endpoints have been evaluated using EVs isolated from in vivo samples (27 papers), in vitro samples (20), or a combination of both (8). While some overlap occurs, reviewed papers encompass various environmental exposures which may be broadly categorized by air pollution (33), occupational exposures (4), metals and metalloids (6), chemicals in household or consumer products (5), food/diet (4), pesticides (2), per-and poly fluoroalkyl substances (2), and additional environmentally relevant exposures (2). These papers are summarized below according to environmental exposure categories.

3.1.1 |. Air pollution

The majority of existing research evaluating environmental-exposure-associated alteration of EV production, release, packaging, or function focuses on air pollution including particulate matter, ozone, tobacco smoke, traffic pollution, wildfires, benzene, and polycyclic aromatic hydrocarbons as further detailed below. Given the well-established relationship between air pollution exposure and adverse cardiopulmonary health outcomes, many of these studies evaluated EVs as a potential mechanism or mediator of exposure-induced diseases such as cardiovascular disease (Du et al., 2022), asthma (Levanen et al., 2013), blood coagulation (Neri et al., 2016), venous thromboembolism (Emmerechts, Jacobs, et al., 2012), and high blood pressure (Rodosthenous et al., 2018). Several different endpoints have been evaluated including EV release and origin, EV miRNA and protein content, and EV biological function as further described.

Several papers evaluated the effect of air pollutants on the release of EVs in vivo and in vitro, with the majority demonstrating that PM, ozone, and cigarette smoke increase the release of EVs as reported by EV count and measured by nanoparticle tracking analysis or flow cytometry. More specifically, at least five studies found that exposure to PM increases EV release in a time and/or dose-dependent manner. In two of these studies, human exposure to PM was characterized by regional air monitors and EVs were isolated from human plasma samples. These studies both showed that EVs were released in a greater concentration after short-term exposure to high-level PM compared to low-level PM after 24 h (Bonzini et al., 2017; Rota et al., 2020). Both studies also identified overweight individuals as a particularly susceptible population. Another study evaluated short-term and long-term PM exposure among individuals with type 1 or type 2 diabetes (Emmerechts, Jacobs, et al., 2012). Here, EV release was significantly and positively associated with long-term, but not short-term exposure to PM. Similar findings have also been reported among in vitro studies where multiple primary and cancer cells lines treated with PM showed a time and concentration-dependent increase in release of EVs (Martin et al., 2019; Neri et al., 2016). Other air pollutants have also been shown to increase the release of EVs including ozone (Smith et al., 2021), cigarette smoke (Saxena et al., 2021), traffic pollution (Emmerechts, De Vooght, et al., 2012; Gao et al., 2017), benzene (Malovichko et al., 2021), and polycyclic aromatic hydrocarbons (Le Goff et al., 2019). Overall, these findings are significant given that clinical studies have linked increased EV release with increased risk of disease, including cardiovascular disease (Ueba et al., 2010). Biologically, it has been hypothesized that increased EV release is one potential mechanism through which increased molecular signals, such as miRNA, can be transferred to nearby or distal target cells and mediate disease outcomes (Rodosthenous et al., 2016).

To further understand the potential biological consequences of altered EV regulation, many studies have characterized EV content, including miRNAs, proteins, and chemicals. These findings provide a basis for another mechanism relating environmental exposures and EV-mediated disease outcomes. For example, at least three studies have reported that EVs released from human plasma or cells exposed to PM carry functionally active tissue factor on their surface, identifying a mechanism through which they may impact blood coagulation and thrombotic related diseases (Emmerechts, De Vooght, et al., 2012; Emmerechts, Jacobs, et al., 2012; Martin et al., 2019). Various types of air pollution including PM, ozone, cigarette smoke, and traffic pollution have been shown to significantly alter the expression of EV encapsulated miRNAs. Across several studies, PM and traffic pollution has been associated with altered levels of EV miRNA expression involved in cardiovascular disease pathways (Du et al., 2022; Pergoli et al., 2017; Rodosthenous et al., 2016; Rodosthenous et al., 2018). Notably, these relationships were dose and time-dependent among in vivo and in vitro exposures (Bollati et al., 2015; Rodosthenous et al., 2016; Rodosthenous et al., 2018). While the relationship between altered miRNA profiles and disease outcomes requires further characterization, studies have provided strong evidence that EV miRNA profiles can modify disease such as blood pressure or developmental delays (Rodosthenous et al., 2018; Wang et al., 2022). For example, EV miRNAs miR-223–3p and miR-199a/b positively modified the relationship between PM exposure and blood pressure in a cohort of elderly men (Rodosthenous et al., 2018). In addition to miRNAs, altered EV proteins have also been associated with air pollution exposures. Cigarette smoke has been associated with increased levels of high mobility group box 1 (HMGB1) and Suppressor of cytokine signaling 3 (SOCS3) proteins within EVs, potentially mediating disease pathways such as inflammation (Chen et al., 2016; Haggadone et al., 2020). EVs may also transport chemical components resulting from environmental exposures on their surface or within the vesicle. For example, cigarette smoke has been demonstrated to increase spermine on the surface and within EVs and consequently mediate pathways related to hypertension (Zhu et al., 2019).

Biological consequences of EVs have been further characterized using in vitro studies. EVs may be isolated from biological samples or cell culture media and used to investigate their biological effects on separate cells. In these studies, both adverse and protective effects have been observed. For example, EVs isolated from human plasma with varying PM exposure levels were used to treat primary endothelial cells (Rota et al., 2020). EVs isolated from human plasma with high PM exposure increased endothelial activation, while EVs isolated from human plasma with low PM exposure decreased endothelial activation (Rota et al., 2020). A similar relationship exists between with cytotoxic effects. EVs isolated from PM treated cells have been linked to increased cytotoxicity and increased pro-inflammatory cytokines (Martin et al., 2019). In contrast, EVs isolated from healthy individuals have been shown to increase cell survival in cells treated with PM (Zhou et al., 2021). Cell treatment with EVs isolated from healthy donors reduced cytotoxicity by reversing apoptosis through AKT phosphorylation. Protective effects have also been associated with EVs produced by endothelial cells exposed to cigarette smoke condensate (Saxena et al., 2021). Specifically, isolated EVs from untreated cells promoted cell survival in a dose-dependent manner among endothelial cells exposed to cigarette smoke condensate. Furthermore, biological consequences may be further characterized by evidence of cross-tissue talk mediated by EVs. Two studies have investigated EV mediated cross-talk between the lung and heart (Carberry, Koval, et al., 2022; Wang et al., 2022). For example, one recent found that wildfire exposure in vivo was associated with altered EV miRNAs in plasma. These EV miRNAs were then associated with overlapping altered mRNA targets in both the lung and heart, providing evidence of cross-tissue EV mediation of disease. These example studies demonstrate the robust biological and pathological relevance of air pollution-induced alterations of EV release and content.

3.1.2 |. Occupational exposures

Occupationally relevant exposures such as metal-rich PM, dust, toluene diisocyanate, silica, and psychological stress have been associated with EV alterations. Increased concentrations of metallic PM exposure was associated with altered EV miRNA profiles related to inflammation and coagulation among participants working in a steel production plant (Pavanello et al., 2016). Similarly, dust-exposed workers with pneumoconiosis showed altered EV miRNA expression implicated in lung cancer pathways (Zhang et al., 2018). Other notable findings include the increased production and release of EVs associated with both toluene diisocyanate and silica exposure (Brostrom et al., 2015; Wang et al., 2020). In vitro studies have also evaluated occupationally relevant exposure such as silica, where macrophage-derived EVs contained significantly altered miRNA profiles related to inflammation, fibrosis, and cell migration (Zhang et al., 2018). These effects were further supported by EV induced differentiation in separate fibroblast cells.

3.1.3 |. Metals and metalloids

Metals and metalloids are another environmental exposure category of public health interest, especially those that are known carcinogens. Four reviewed studies investigated arsenic- or arsenite-associated EVs as mechanisms of carcinogenesis (Chen et al., 2017; Dai et al., 2018; Ngalame et al., 2018; Xu et al., 2015). Each study utilized arsenic- or arsenite-transformed cells to isolate EVs and expose normal or healthy cells. Several biological mechanisms of EV-mediated disease have been elucidated following these experiments. In one experiment, EVs from arsenic-transformed bronchiolar epithelial cells induced cell proliferation in normal bronchiolar epithelial cells (Xu et al., 2015). More specifically, EV miR-21 was increased among arsenic transformed cells and was demonstrated play a significant role in the mechanism through exosome depletion and miRNA knockdown. Similarly, another EV miRNA, miR-155, has been implicated in carcinogenesis pathways (Chen et al., 2017). This study showed upregulation of miR-155 in EVs released from arsenic-transformed liver cells. Exposing normal liver cells to these EVs resulted in a pro-inflammatory response which was prevented by exosome depletion, further supporting EV’s role in disease mechanisms. Other studies have shown EV content such as circular RNAs, cytokines, enzymes, and oncogenes as significantly increased in EVs released from arsenite transformed cells (Dai et al., 2018; Ngalame et al., 2018). These studies further supported the role of EVs in mechanisms of inflammation and proliferation, as well as induction of cancer morphology. Though limited, the current literature provides evidence that inhibition of EV release or EV packaging is a potential strategy for preventing adverse effects in vitro (Chen et al., 2017; Dai et al., 2018; Ngalame et al., 2018; Xu et al., 2015).

Two additional studies have evaluated the relationships between metals exposures and EV-encapsulated miRNAs related to disease (Harischandra et al., 2018; Howe et al., 2022). In a cell culture model of Parkinson’s disease, manganese was shown to significantly increase the number of EVs released and alter the expression of numerous EV-encapsulated miRNAs (Harischandra et al., 2018). The altered EV miRNAs identified in this study have previously been associated with biological pathways including mitochondrial dysfunction and protein aggregation among others, providing evidence for future studies to expand upon in the investigation of mechanisms contributing to Parkinson’s disease. In an epidemiological study, nine different metals including antimony, barium, cadmium, cobalt, mercury, molybdenum, nickel, thallium, and tin were measured in urinary samples collected during pregnancy and evaluated for associations with over 100 EV-miRNA isolated from maternal serum (Howe et al., 2022). Here, 35 unique miRNAs were significantly associated with cobalt, barium, thallium, and/or mercury, eight of which were associated with three different metals. Importantly, these altered miRNAs have known roles in placental development and function. Collectively, these studies support the need for further investigation of EVs as potential mediators of metal exposure associated diseases.

3.1.4 |. Household and consumer products

Chemicals within household and consumer products are of increasing public health concern, especially those with endocrine disrupting potential. Phthalates, which are found in cosmetics, food packaging, and personal care products, have been associated with altered EV miRNA profiles in at least two studies (Barnett-Itzhaki et al., 2021; Zhong et al., 2019). In two studies, follicular fluid samples were collected from women undergoing in vitro fertilization and analyzed for phenol and/or phthalate metabolite concentrations associated with EV miRNA content (Barnett-Itzhaki et al., 2021; Martinez et al., 2019). Biomarkers of these endocrine disrupting chemicals were detected among the majority of samples and were significantly associated with differentially loaded EV miRNAs. Example predicted pathways targeted by these miRNAs include some relevant to reproduction such as oocyte development and maturation (Barnett-Itzhaki et al., 2021). Similarly, another study found preliminary evidence that maternal exposure to phthalates is potentially associated with altered placenta-derived EV miRNAs (Zhong et al., 2019). Another endocrine disrupting chemical, bisphenol A (BPA), as well as flame retardants polybrominated diphenyl ether (PBDE), tetrabromobisphenol A (TBBPA), and 2,4,6-tribromophenol (TBP) have also been associated with altered EV content (Sheller-Miller et al., 2020). Placenta explants were treated with either BPA, PBDE, TBBPA, or TBP and resulted in the release of EVs with altered protein profiles. Example common pathways altered by these proteins include cell death, inflammation, and cell proliferation. EV miRNA related to cell proliferation also been altered in response to hydroquinone, a chemical found in skin-lightening creams (Jiang et al., 2020). Together these results provide mechanistic evidence relating exposure to common household or consumer product chemicals and EV mediated disease pathways.

3.1.5 |. Pesticides and herbicides

At least two studies have evaluated EVs alterations associated with herbicide or pesticide exposure (Faria Waziry et al., 2020; Vujic et al., 2021). Paraquat, a widely used herbicide, was used to expose human brain microvascular endothelial and evaluate resulting changes in EV release and protein content (Vujic et al., 2021). EV release was significantly decreased in response to Paraquat exposure, and several EV proteins were identified with differential expression. EV protein alterations mimicked protein alterations measured in parent cells, and were predicted to be associated with pathways such as ubiquinone metabolism and HIF1a transcription. Another study investigated the insecticide pyriproxyfen, and its relationship with EV regulation in human cells, and their effect on viral transmission (Faria Waziry et al., 2020). Interestingly, cell viability and EV release were not altered with exposure to pyriproxyfen alone. However, the combination of pyriproxyfen and viral introduction induced increased concentrations of EVs compared to control groups with just virus or pyriproxyfen alone. The amount of virus also was significantly increased in the treatment group, suggesting EVs may be able to mediate viral replication/transmission. These studies further support the communicative and mechanistic roles of EVs in environmental exposure induced/mediated diseases.

3.1.6 |. Diet

Few studies have evaluated the link between diet and EVs, though there is evidence that diet may impact disease directly or mediate the relationship between environmental exposures and disease. Diet-induced EVs have been shown to have either positive or negative effects on disease outcomes (Mudd et al., 2020; Nordgren et al., 2019; Rompala et al., 2020; Zhu et al., 2021). Ochratoxin, a dietary contaminate, may alter cell cycle and oxidation via EVs (Zhu et al., 2021). In vitro, cells were treated with ochratoxin released EVs that increased cytotoxicity, induced reactive oxygen species, and affected cellular gene expression related to metabolism and cell cycle in target cells. EVs within bovine milk have also been demonstrated to impact disease endpoints (Nordgren et al., 2019). Mice fed diets with bovine milk containing EVs showed increased inflammation in response to agricultural dust exposure, whereas mice fed diets with EV depleted bovine showed anti-inflammatory responses. Adverse effects related to dietary-altered EVs have also been shown to have transgenerational effects in vivo (Rompala et al., 2020). Epididymal EVs from ethanol-exposed mice co-incubated with sperm resulted in ethanol drinking and anxiety-like phenotypes among offspring, indicating intergenerational effects of EVs. Dietary-induced EV exposures are also capable of inducing positive effects on disease phenotypes. Berry derived anthocyanidins loaded into EVs were used to treat colon cancer cells, resulting in anti-proliferative effects in vitro (Mudd et al., 2020). These effects were more pronounced when anthocyanidins were loaded into EVs vs direct exposure to anthocyanidins, indicating potential preferential uptake.

3.1.7 |. Per and poly-fluoroalkyl substances

PFAS, a high-priority emerging class of contaminants, has been investigated among two in vivo studies for its association with EVs. One study provides preliminary evidence that increased EV concentrations enhance the association between perfluorooctanesulfonic acid (PFOS) and carotid intima-media thickness, a cardiovascular disease endpoint (Lin et al., 2016). Furthermore, another recent epidemiological study provides evidence that PFAS including perfluorooctanoic acid (PFOA), perfluorodecanoic acid (PFDA), perfluorononanoic acid (PFNA), perfluorohexanesulphonic acid (PFHxS), and PFOS (along with several other marine pollutants including pesticide metabolites, polychlorinated biphenyls, and mercury) may also affect EV miRNA profiles (Kupsco et al., 2022). Though not statistically significant in vivo, these preliminary findings support the need for future research to expand upon PFAS research in vitro (e.g., further dosings, timepoints, and/or environmentally relevant mixtures-based exposures) to fully elucidate relationships between these ubiquitous environmental exposures and EV mediated mechanisms of disease. As literature involving both PFAS and EV research continues to expand, more conclusive evidence may be found linking these two research areas.

3.1.8 |. Other environmentally-relevant exposures

Other environmentally relevant exposures such as asbestos and space radiation have been associated with altered EVs. More specifically, asbestos and radiation have been associated with altered EV content including protein and miRNAs, respectively (Gaines & Nestorova, 2022; Munson et al., 2018). Asbestos exposed cells released EVs with altered protein content related to cancer pathways, which subsequently caused gene expression changes in target cells. Cancer related pathways were also associated with EV miRNA induced by radiation exposure in vitro.

3.1.9 |. Landscape of EV literature summary

Overall, currently published literature strongly supports the relationship between environmental exposure, EVs, and mechanisms of disease. Further research is required to fully elucidate the mechanisms through which EVs impact disease outcomes in response to environmental exposures. Furthermore, these findings highlight the need for exposure prioritization and subsequent evaluation of potential influences on EV production, packaging, and release, as well as downstream effects of content on other cells.

3.2 |. Overview of mechanisms and key genes regulating EV production, packaging, and release

One primary mechanism through which environmental exposures may alter EV production, packaging, or release is through cellular alterations of EV regulator genes or proteins. Several reviews have previously identified key genes and their associated proteins involved in EV regulation (Abels & Breakefield, 2016; Gurunathan et al., 2021; Hessvik & Llorente, 2018; Kowal et al., 2014; Pegtel & Gould, 2019). Here, key pathways and molecules involved in EV production, packaging, and release are broadly summarized. Furthermore, specific genes are compiled and reported to provide a foundation for identifying chemicals that alter these genes and subsequently predicting those that may alter EV regulation.

3.2.1 |. Mechanisms involved in regulation of EV production

Mechanisms through which EVs are produced is one marker used to inform EV subcategories. Three primary EV subcategories, exosomes, microvesicles, and apoptotic bodies, are thought to have differences in their mechanisms of production, though there may be overlap (Gurunathan et al., 2021). Apoptotic bodies originate exclusively from dying cells, where larger-sized vesicles are formed upon cellular breakdown. Microvesicles are formed directly from the plasma membrane of cells, though the complete mechanisms of origin are currently not understood (Abels & Breakefield, 2016). In general, microvesicles originate from outward budding of the cell’s plasma membrane, forming a vesicle. Mechanisms related to exosomal production, however, have been extensively studied and reviewed and thus will be the focus below (Abels & Breakefield, 2016; Hessvik & Llorente, 2018; Pegtel & Gould, 2019).

In brief, exosomal biogenesis has been demonstrated to occur in the endosomal system via formation of intraluminal vesicles (ILVs) within late-endosomes/multivesicular bodies (Hessvik & Llorente, 2018). This formation occurs through two primary pathways: endosomal sorting complex required for transport (ESCRT) dependent or ESCRT independent (Gurunathan et al., 2021). In the ESCRT dependent pathway, exosome biogenesis begins when the endosomal membrane becomes enriched with tetraspanins, and ESCRTs are recruited. This is thought to initiate the formation of ILVs from endosomal membranes (Hessvik & Llorente, 2018). Several specific tetraspanins, ESCRT proteins, and other supportive proteins have been identified in these mechanisms, detailed below. Although the ESCRT-dependent pathway is the most well-established mechanism through which exosomes are formed, studies have shown that while reduced, ILVs are still formed within endosomes in the absence of ESCRT-related proteins(Kowal et al., 2014). These mechanisms still involve tetraspanins, but also include heat shock proteins or lipids as further detailed in the following text. Exosome formation via endosomal budding remains the most widely accepted and investigated mechanism of production, though recent research supports an additional pathway of exosome formation whereby exosomes bud directly from the plasma membrane (Pegtel & Gould, 2019).

3.2.2 |. Mechanisms involved in regulation of EV packaging

During the process of EV formation molecular contents such as nucleic acids, proteins, and lipids may be packaged into the vesicles and transferred to target cells (Hessvik & Llorente, 2018). These contents may reflect their parent cell and be used in determination of cell-origin; however, it has been shown that EV contents may be significantly altered in response to external stimuli such as environmental exposures as previously reviewed. The underlying mechanisms that influence EV packaging are not fully known, though several papers have identified potential pathways involved in miRNA, mRNA, and protein packaging which have been reviewed (Abels & Breakefield, 2016; Hessvik & Llorente, 2018; Pegtel & Gould, 2019).

Currently, miRNA is one of the most evaluated EV contents. One identified mechanism through which miRNA can be selectively loaded in EVs is via a sequence motif, or a short recurring nucleotide sequence “GGAG” (Abels & Breakefield, 2016; Hessvik & Llorente, 2018; Pegtel & Gould, 2019). This sequence is enriched among EV miRNAs and preferentially loaded into EVs via protein heterogeneous nuclear ribonucleoprotein A2B1 (Abels & Breakefield, 2016; Hessvik & Llorente, 2018). Another molecule, synatotagmin binding cytoplasmic RNA interacting protein, has also been associated with EV miRNA motifs. Other molecules associated with miRNA loading into EVs include RISC component argonaute 2 (AGO2), sphingomyelin phosphodiesterase 3 (SMPD3) (also known as nSMASE), and Y-box binding protein 1 (Pegtel & Gould, 2019). Increased gene expression of AGO2 or nSMASE and their associated proteins have been associated with increased EV miRNAs, while knockdown of these genes has been associated with decreased EV miRNAs (Abels & Breakefield, 2016). Other pathways such as post-transcriptional modifications may also be involved (Abels & Breakefield, 2016).

In addition to miRNA loading, mechanisms of mRNA and protein packaging have been elucidated. Fragments of mRNA within EVs have been shown to be enriched for sequences closer to the 3′ untranslated region (Batagov & Kurochkin, 2013; Hessvik & Llorente, 2018). This enrichment may play a role in mRNA sorting into EVs, though further research is required to understand underlying mechanisms. Protein packaging has also been evaluated with protein ubiquination and lipids being involved (Hessvik & Llorente, 2018). EV proteins have previously been shown to be enriched for ubiquinated proteins and thus ubiquination is thought to play a potential role in protein selection and sorting. Several lipids have also been implicated in mechanisms sorting proteins into EVs (Hessvik & Llorente, 2018). For example, lipids such as cholesterol, sphingomyelin and glycosphingolipids have been shown to be elevated in EVs compared to their cell of origin (Skotland et al., 2017). Overall, findings elucidating mechanisms of EV packaging may provide insight as to how external exposures such as environmental chemicals may impact these processes and thus EV content.

3.2.3 |. Mechanisms involved in regulation of EV release

Similar to EV production and biogenesis, mechanisms of EV release help to distinguish EV subcategories. For microvesicles and apoptotic bodies, mechanisms of production and release are closely linked. Microvesicles are released upon direct budding and fission of the plasma membrane. Specific pathways have been described for microvesicle release including several genes and their associated proteins as well as external factors. For example, activation of the myosin light chain kinase by extracellular signal-regulated kinase causes the release of microvesicles (Abels & Breakefield, 2016). Alternatively, microvesicles have been shown to be released after recruitment of an endosomal sorting complex subunit, tumor susceptibility gene 101 (TSG101), and interaction with Arrestin 1 domain-containing protein 1 (ARRDC1). Finally, microvesicles may be released in response to external factors such as calcium influx or hypoxia (Abels & Breakefield, 2016).

Release of exosomes involves transportation of multivesicular bodies to the cellular membrane and release of contained exosomes via exocytosis into the extracellular space. This process involves two primary protein families: members of the RAS oncogene family (RABs) and members of the SNAP Receptor (SNARE) family (Gurunathan et al., 2021; Hessvik & Llorente, 2018; Kowal et al., 2014). Broadly, RAB genes and their associated proteins are responsible for intercellular vesicle transport, and SNAREs facilitate fusion with the membrane and subsequent release. Expression of specific RAB proteins including RAB11, RAB27, RAB2, RAB35, RAB5, RAB7, and RAB9 have been associated with EV release in several studies which have been extensively reviewed (Gurunathan et al., 2021; Hessvik & Llorente, 2018; Kowal et al., 2014). SNARE family proteins such as vesicle-associated membrane protein (VAMP) and YKT6 v-SNARE homolog (YKT6) have been implicated in mechanisms of release (Abels & Breakefield, 2016; Kowal et al., 2014).

3.2.4 |. Compiling the EV gene regulator list

Several research groups have previously reviewed known mechanisms of EV regulation; these have been broadly summarized above. Here, 76 genes and their associated proteins were identified as having a role in EV regulation and are summarized in Table 2. This list is not intended to be fully comprehensive, but to rather describe well-studied genes involved in EV regulation. This list notably contains many of the genes compiled into the Mouse Genome Informatics’ Extracellular Vesicle Biogenesis gene ontology (Mouse Genome Database, 2022), though the list generated in this study is more human-focused and is expanded in scope to include EV packaging, production, and release.

TABLE 2.

Genes associated with regulation of EV production, packaging, and release summarized by gene name, function, and EV category

Gene symbol	Official gene name	Regulatory function	EV category	Reference*
AGO2	Argonaute RISC catalytic component 2	Packaging	Exosome	2, 3
ARF6	ADP ribosylation factor	Production	Exosome	1, 2, 3, 5
ARRDC1	Arrestin domain containing 1	Production	Exosome	3, 2
ATG5	Autophagy related 5	Release	Exosome	3, 5
BST2	Bone marrow stromal cell antigen 2	Release	Exosome	1
CD63	CD63 molecule	Production	Exosome	1, 2, 3, 4, 5
CD82	CD82 molecule	Production	Exosome	1
CD9	CD9 molecule	Production	Exosome	1, 2, 3, 5
CHMP2A	Charged multivesicular body protein 2A	Release	Exosome	3, 5
CHMP4C	Charged multivesicular body protein 4C	Production	Exosome	1, 4, 5
CHMP4A	Charged multivesicular body protein 4A	Production	Exosome	2
CHMP4B	Charged multivesicular body protein 4B	Production	Exosome	4
CIT	Citron rho-interacting serine/threonine kinase	Release	Exosome	1
CTTN	Cortactin	Release	Exosome	1
DGKA	Diacylglycerol kinase alpha	Production/Release	Exosome	1, 2
EXPH5	Exophilin 5	Production/Release	Exosome	3, 5
FLOT2	Flotillin 2	Production	Exosome	4
HGS	Hepatocyte growth factor regulated tyrosine kinase substrate	Production	Exosome	1, 2, 4, 5
HNRNPA2B1	Heterogeneous nuclear ribonucleoprotein A2/B1	Packaging	Exosome	3, 2
HPSE	Heparanase	Release	Exosome	3
HSPA8	Heat shock protein family A (Hsp70) member 8	Production	Exosome	4
HSPG2	Heparan sulfate proteoglycan 2	Production	Exosome	2
ISG15	ISG15 ubiquitin like modifier	Release	Exosome	1
IZUMO1R	IZUMO1 receptor, JUNO	Production	Exosome	3
LITAF	Lipopolysaccharide induced TNF factor	Production	Exosome	1
LMP-1	Latent membrane protein	Production	Exosome	1
MAPK3	Mitogen-activated protein kinase 3	Release	Exosome	5
MRGPRX3	MAS related GPR family member X3	Production	Exosome	3
PDCD6IP	Programmed cell death 6-interacting protein	Production	Exosome	1, 2, 3, 4, 5
PIKFYVE	Phosphoinositide kinase, FYVE-type zinc finger containing	Release	Exosome	1
PKM	Pyruvate kinase M1/2	Release	Exosome	1
PLD2	Phospholipase D2	Production	Exosome	1, 2, 2, 3, 4
RAB11A	RAB11A, member RAS oncogene family	Release	Exosome	1, 2, 3, 4, 5
RAB27A	RAB27A, member RAS oncogene family	Release	Exosome	1, 2, 3, 4, 5
RAB27B	RAB27B, member RAS oncogene family	Release	Exosome	1, 2, 3, 4, 5
RAB2B	RAB2B, member RAS oncogene family	Release	Exosome	1, 4, 5
RAB35	RAB35, member RAS oncogene family	Release	Exosome	1, 2, 3, 4, 5
RAB5A	RAB5A, member RAS oncogene family	Release	Exosome	1, 4, 5
RAB7A	RAB7A, member RAS oncogene family	Release	Exosome	1, 2, 4, 5
RAB9A	RAB9A, member RAS oncogene family	Release	Exosome	1, 4, 5
RALA	RAS like proto-oncogene A	Release	Exosome	1, 3, 5
RALB	RAS like proto-oncogene B	Release	Exosome	1
SDC1	Syndecan 1	Production	Exosome	1, 2, 5
SDCBP	Syndecan binding protein	Production	Exosome	1, 2, 3, 5
SMPD2/3	Sphingomyelin phosphodiesterase 2/3	Production/Packaging	Exosome	2, 1, 2, 5
SNAP23	Synaptosome associated protein 23	Release	Exosome	1, 3, 4
STAM	Signal transducing adapter molecule	Production	Exosome	1, 4, 5
STX1A	Syntaxin 1A	Release	Exosome	1
STX4	Syntaxin 4	Production	Exosome	3
SYNCRIP	synaptotagmin binding cytoplasmic RNA interacting protein	Packaging	Exosome	3
SYT7	Synaptotagmin 7	Release	Exosome	1
SYTL4	Synaptotagmin like 4	Production/Release	Exosome	3, 5
TFRC	Transferrin receptor	Production	Exosome	4
TSG101	Tumor susceptibility 101	Production	Exosome	1, 2, 3, 4, 5
TSPAN8	Tetraspanin 8	Production	Exosome	1, 4
UNC13D	Unc-13 homolog D	Production/Release	Exosome	3, 5
VAMP7	Vesicle associated membrane protein 7	Release	Exosome	1, 2, 4, 5
VAMP8	Vesicle associated membrane protein 8	Release	Exosome	4
VPS4A	Vacuolar protein sorting 4 homolog A	Production	Exosome	2, 4
VPS4B	Vacuolar protein sorting 4 homolog B	Production	Exosome	1, 2, 3, 4, 5
VTA1	Vesicle trafficking 1	Production	Exosome	1
YBX1	Y-box binding protein 1	Packaging	Exosome	3
YKT6	YKT6 v-SNARE homolog	Release	Exosome	1, 4, 5
ZHX2	Zinc fingers and homeoboxes 2	Release	Exosome	5
ARF6	ADP ribosylation factor 6	Production/Release	Microvesicle	2
PLD2	Phospholipase D2	Production/Release	Microvesicle	2
MAPK1	Mitogen-activated protein kinase 1	Production/Release	Microvesicle	2
MYLK	Myosin light chain kinase	Production/Release	Microvesicle	2
TSG101	Tumor susceptibility 101	Production/Release	Microvesicle	2
ARRDC1	Arrestin domain containing 1	Production/Release	Microvesicle	2
ITGB1	Integrin subunit beta 1	Production/Release	Microvesicle	2
VAMP3	Vesicle associated membrane protein 3	Production/Release	Microvesicle	2
MMP14	Matrix metallopeptidase 14	Production/Release	Microvesicle	2
VPS4B	Vacuolar protein sorting 4 homolog B	Production/Release	Microvesicle	2
WWP2	WW domain containing E3 ubiquitin protein ligase 2	Production/Release	Microvesicle	2
RAB22A	RAB22A, member RAS oncogene family	Production/Release	Microvesicle	2

^a1 Hessvik and Llorente (2018) 2 Abels and Breakefield (2016), 3 Pegtel and Gould (2019), 4 Kowal et al. (2014), 5 Gurunathan et al. (2021).

Alterations to these genes and/or their associated proteins at any step of EV regulation have the potential to alter EV production, content, or release (Hessvik & Llorente, 2018). Genes involved in EV regulatory pathways have been experimentally elucidated or confirmed using RNA interface screening, knockdown techniques, viral transfection, and other genomic techniques (Colombo et al., 2013; Hoshino et al., 2013; Hurwitz et al., 2016). Several additional factors may influence EV regulation including cell type and confluency, serum media conditions, growth factors, cellular stress, and external exposures among others (Gurunathan et al., 2021). In particular, one potential mechanisms through which external exposures such as environmental chemicals may impact EV production, packaging, or release is via alterations of these genes as depicted in Figure 1.

3.3 |. Predicting environmental contaminants that may disrupt EVs by acting on EV regulatory genes

Given the robust evidence that environmental exposures impact EV regulatory outcomes such as EV release, content, and function, it is important to establish methods to prioritize understudied environmental chemicals for future evaluation. Here, the Comparative Toxicogenomics Database (CTD) was used as tool to identify chemicals that have been shown to alter genes involved in EV regulation. The CTD is a publicly available database (http://ctdbase.org/) with curated data on chemical–gene interactions, chemical–disease relationships, and gene–disease relationships. This database aims to support hypothesis generation surrounding the relationship between environmental exposures and human health. Using the EV regulator gene list curated from previous review papers, CTD was leveraged as a tool to identify previously published evidence of chemicals targeting regulators of EV production, packaging, or release.

A second database, CompTox Chemicals Dashboard, was used to identify which of these predicted chemicals have environmental relevance including chemicals found in water, air, soil, biosolids, and consumer products among other relevant categories. The CompTox Chemicals Dashboard is a publicly available database through the United States Environmental Protection Agency (EPA) compiling chemical information such as chemistry, exposure information, and toxicity data for over 900,000 chemicals. Within this database over 300 chemical lists are available categorizing chemical by class, exposure source, or regulatory status among others. Chemicals within environmentally relevant lists may be used to support future hypothesis generation, by highlighting environmental contaminants that may alter EV production, packaging, and release mechanisms, based on previously identified chemical-gene interactions.

3.3.1 |. Environmental chemicals predicted to alter EV regulatory genes

Using the CTD as a tool to review established interactions between EV regulatory genes and environmental chemicals, 1418 unique chemicals were identified with at least one interaction reported with any of the 76 EV regulatory genes listed in Table 2. Upon evaluation of environmental relevance, this list was further narrowed down to 858 chemicals present in at least one environmental list from Table 2. Additionally, 494 chemicals were either not present in a list or not able to be identified. The top 50 environmentally relevant chemicals were identified after manual query of the remaining unidentified chemicals. These top 50 chemicals represent environmentally relevant chemicals previously shown to target or alter EV regulatory genes. A summary of the top 50 chemicals based on most unique gene targets is represented in Figure 4.

FIGURE 4.

Top 50 chemicals demonstrated to alter genes involved in EV regulation. For each chemical, environmental relevance is visualized by presence or absence in CompTox chemical lists with white being absent and increasing blue intensity indicating presence in a list. The corresponding number of unique gene targets are visualized on the bar graph from greatest to least number of gene targets. *Indicates chemicals manually queried for environmental relevance.

Of the top 50 environmental chemicals previously shown to impact genes involved in EV regulation, tobacco smoke pollution was the top chemical impacting 60 unique genes. Interestingly, several other chemicals or mixtures related to air pollution were also identified among the top 50 chemicals. For example, PM is associated with interactions with 25 unique EV regulatory genes, air pollutants with 21, and ozone with 20. This finding is significant given that most existing EV research evaluating environmental-exposure induced alterations of EV regulation focuses on air pollution, as previously described. Furthermore, other chemicals previously discussed such as bisphenol A, benzo[A]pyrene, arsenic, and arsenite also appear in this list of top 50 chemicals predicted to impact EV regulation. While these results provide proof of concept, it should be noted that these chemicals are highly studied and thus more data exists regarding their gene targets, Other notable chemicals identified outside of the top 50 that overlap with those previously discussed in association with EVs include asbestos, benzene, ethanol alcohol, ochratoxin A, paraquat, PCBs, and PFAS. Though not currently ranking among the top 50, increasing data reported for high-priority emerging contaminants such as PFAS and other persistent chemicals will likely support continued research in this area. Chemicals and their associated gene targets and environmental relevance are further detailed in Table S2. This overlap represents a proof-of-concept and provides confidence in these methods for predicting and prioritizing chemicals that yet to be studied in relation to EV regulation.

3.3.2 |. Chemicals predicted to alter EVs that remain understudied in terms of EV research

In addition to chemicals previously discussed, this analysis identified numerous chemicals that have not yet been studied but have predicted potential to alter EV regulatory genes and thus deserve attention. Several environmentally relevant categories were queried including biosolids, consumer products, foods, flame retardants, indoor and outdoor air, superfund chemicals, water, and herbicides/pesticides. Several chemicals identified within the top 50 impacting the most unique genes were present across multiple environmentally relevant lists, and thus multiple potential exposure sources. Notably, acrolein and formaldehyde were found in four or more environmentally relevant lists. Acrolein is an unsaturated aldehyde released into the air via cigarette smoke and burning of fossil fuels, or formed as a cooking byproduct when animal or vegetable fats are heated (Jiang et al., 2022; Maiyo et al., 2022). This chemical is highly toxic and as such as been associated with adverse health outcomes such as cardiovascular disease, diabetes, alcoholic liver disease, and other chronic diseases (Jiang et al., 2022). Similarly, formaldehyde is another simple aldehyde and is a product of fires, cigarette smoke, and is manufactured for building materials and household products. Formaldehyde’s toxicity has been studied extensively and is a known carcinogen (Swenberg et al., 2013). Though much is already understood about mechanisms of toxicity for these chemicals, evaluation of EV’s potential role in mediating these mechanisms may be valuable for filling in remaining gaps in knowledge.

Public health implications of exposure to herbicides and pesticides are abundant given their ubiquity and pervasiveness in the environment and food supply (El-Nahhal & El-Nahhal, 2021). Two pesticides, dicrotophos and ivermectin, were identified as chemicals that may impact EV regulation. More specifically, dicrotophos is an organophosphate typically used to prevent garden insects but, similar to other organophosphates, can cause cholinesterase inhibition in humans and at very high exposures lead to respiratory paralysis and death (Lorke & Petroianu, 2019). Notably, dicrotophos is found across multiple environmental exposure sources including biosolids, household products, and water. Ivermectin is an FDA-approved anti-parasite treatment, though several adverse side effects including nausea, vomiting, low blood pressure and facial swelling can occur. Similarly, copper sulfate, an herbicide, was identified as a chemical likely to impact EV regulation across multiple exposure sources including household chemicals and food substances. Copper sulfate has been established as toxic to humans with adverse effects ranging from irritation to nausea, vomiting, and damage to organs (Gamakaranage et al., 2011; Pankit & Bhave, 2002). Given the ubiquity of pesticide and herbicide exposure in the environment, it is important to understand how these chemicals induce toxicity, and what options may be available for potential treatments.

In this analysis, several plant derivatives or toxins were also identified as potential disruptors of EV regulation. Alflatoxin, alpha-pineine, and abrine are a few examples of identified toxins that humans may be exposed to through agriculture and food, inhalation of contaminated air, and household or consumer products. Additionally, plant extracts were found to be associated with alterations of several EV regulatory genes. As previously discussed, plant derivatives may promote the release of EVs carrying signals that impart beneficial effects on target cells (Mudd et al., 2020) while others may further exacerbate or promote adverse effects (Zhu et al., 2021). Continued evaluation of this relationship between natural toxins and EVs may serve to elucidate unique mechanisms of toxicity or opportunity for therapeutics.

4 |. CONCLUSIONS AND FUTURE DIRECTIONS

EVs have recently been established as an important biomarker of exposure and disease, though much work is needed to understand how environmental chemicals alter EV regulation, content and function. While numerous environmental chemicals have been directly linked to altered EVs, the majority of chemicals in our environment have yet to be studied. Given robust evidence in literature that EVs may contribute to mechanisms of exposure-related disease, it is important to develop methods to prioritize which chemicals should be evaluated. We propose leveraging databases to identify known chemical–gene interactions in conjunction with environmentally relevant chemical lists to predict and prioritize chemicals that alter EV regulatory processes. Implementing this method validated several chemicals with direct links to EV alterations, as well as highlighted numerous unstudied chemicals for future evaluation. Overall, this review serves as a basis for evidence-based hypothesis testing as this research field linking EVs and the environment continues to grow.