1. Virus-Extracellular Vesicles (EVs) Modulate Immune Responses
The researchers discussed the strategies adopted by viruses to hijack EV biosynthesis machinery to aid in their replication, egress and spread. The researchers explore the relationship between EVs, viruses and the immune system and summarize the antiviral and host immunomodulatory signaling that is mediated by EVs in the context of viral infections.
1.1. Virus-EVs Stimulate Antiviral Immune Responses
Virus-EVs containing either host or viral proteins or miRNAs can enhance all stages of the antiviral innate and adaptive immune response. Virus-EVs transport host miRNAs, molecules and proteins resulting in the activation of innate antiviral immune responses in uninfected recipient cells. For instance, in the late stages of Influenza virus (IAV) infection, when cells are undergoing apoptosis, the host Y5 non-coding RNA is degraded into miRNAs that are transported via EVs to uninfected recipient cells, where they induce IFN antiviral responses
[1]. During HIV-1 infection, the second messenger cGAMP is produced in response to virus sensing inside the cytoplasm. cGAMP is then transported inside HIV-1-EVs with HIV-1 viral particles to recipient cells, where cGAMP activates the stimulator of interferon genes (STING) pathway
[2]. Furthermore, virus-EVs also transport IFN signaling proteins, such as interferon-induced proteins with tetratricopeptide repeats (IFIT) and interferon-induced transmembrane protein (IFITM). IFITM is known for blocking enveloped virus entry
[3] and activating the adaptive immune response
[4], while IFIT inhibits viral replication
[5]. For instance, DV-EVs containing IFITM3 prevent DV entry into recipient cells
[6], while IAV-EVs contain both IFIT and IFITM activate the secretion of IL-6, MCP-1 and TNF by recipient cells
[7]. Additionally, IAV-EVs coated with α-2,3 and 2,6 sialic acids on their surface can competitively block viral entry by binding IAV particles before they can interact with targeted cells
[7].
Secondly, virus-EVs stimulate the secretion of chemokines and pro-inflammatory cytokines by recipient cells, which induce activation and recruitment of immune cells to the infection site. As mentioned above, the IAV-EVs transporting IFIT and IFITM activate the secretion of IL-6, MCP-1 and TNF
[7]. Respiratory syncytial virus (RSV)-EVs contain both viral RNA and proteins and host RNAs and either induce the secretion of MCP-1, IP-10 and CCL5 by recipient monocytes or of CCL5, IP-10, TNF-α by airway epithelial cells, without leading to an active infection of these cells
[8]. DV-EVs secreted by infected macrophages contain miRNAs and the viral NS3 protein and induce the secretion of MCP-1, IP-10, IL-10, TNF-α and CCL5 by endothelial cells, allowing the activation of the primary antiviral immune barrier
[9]. DV-EVs activation of endothelial cells is associated with an increased expression of cell adhesion proteins (ICAM and VE-cadherin) and modification of the trans-endothelial electrical resistance, supporting a change in the endothelial barrier permeability
[9]. Interestingly, infection of monocyte-derived DCs induces both activation of DCs and secretion of EVs containing a variety of host miRNAs and mRNAs that depends on the nature of the DV serotype. According to their functions, these miRNAs and mRNAs can induce an inflammatory response and apoptosis in cells integrating these DV-EVs
[10].
Lastly, virus-EVs can stimulate robust adaptive immune responses. As mentioned before, IAV-EVs and DV-EVs transporting IFITM could mediate this activation
[4][6][7]. Moreover, IAV-EVs containing MHC-I and -II proteins, and numerous viral proteins can act as a source of antigens used by DCs to initiate the adaptive immune response. However, these IAV-EVs do not activate T cells directly
[7].
1.2. EVs Secreted by Non-Infected Immune Cells Activate the Immune System
The antiviral immune response is also activated by EVs secreted by non-infected immune cells. EVs can protect the non-infected recipient cells from viral infection. For instance, during HBV infection, activated macrophages secrete IFN-α containing EVs, which are integrated by non-infected hepatocytes through TIM-1 (the host receptor for HAV), thus protecting cells against HBV infection
[11]. EVs also enhance the immune response against the virus-associated tumor cells. Activated Vδ2-T cells secrete EVs enhancing the immune response against EBV-associated tumors. These EVs contain Fas ligand, TRAIL, NKG2D, CD80/CD86 immunostimulatory ligands and MHC-I and –II. Fas ligand and TRAIL induce cell death in recipient cells, while transportation of NKG2D to recipient cells activates NK cells. The integration of CD80/CD86 and MHC complexes into the membrane of recipient cells also enhances tumoral and viral antigen presentation by EBV-infected tumor cells, resulting in activation of CD4 and CD8 specific T cell responses
[12].
1.3. Virus-EVs Inhibit Antiviral Immune Response
Similar to how viruses adapted the means to hijack host machinery and enhance their own replication, viruses have also found ways to suppress the antiviral immune responses by using EVs to transport host and viral immunomodulatory cargos from infected to uninfected cells. Some virus-EVs contain host miRNAs that inhibit the interferon (IFN) signaling in recipient cells, making them permissive to viral infection. EV71-EVs are enriched in host miR-146a, which represses the expression of signal transducer and activator of transcription 1 (STAT1), TNF receptor-associated factor 6 (TRAF6) and Interleukin 1 receptor-associated kinase 1 (IRAK1) and eventually suppresses the IFN-1 response in the recipient cells. Consequently, IFN-1-stimulated gene factors such as BST-2/tetherin are also inhibited. EV71-EVs integration by recipient cells not only enhances viral replication but also EV secretion by inhibiting the IFN antiviral immune response in the recipient cells since BST-2/tetherin is an inhibitor of Rab27a-dependent EV secretion
[13]. Similarly, Newcastle disease virus (NDV)-EVs contain host miR-1184, miR-1273f and miR-198 that inhibit IFN-β expression in the recipient cells and thus enhance NDV replication
[14]. Ebola virus-EVs transport the virus nucleocapsid, which impacts the IFN-1 response in the recipient cells
[15].
Viruses have also developed strategies to evade the innate immune responses through the modulation of the EV biosynthesis pathways. Viruses, such as the Tomato bushy stunt tombusvirus, hijack the ESCRT machinery to form replication centers protected from immune surveillance
[16][17]. Poliovirus, rhinovirus, coxsackievirus and DV have found ways to escape autophagolysosome degradation by hiding in autophagosome-derived EVs
[18][19]. Similarly, some viruses can escape phagocytic degradation. For instance, HIV-1 can enter MVB after being captured by immature DCs, and then is trafficked inside EVs, instead of being degraded. These EVs are enriched in HLA-DR1, CD63, CD9 and are associated with HIV-1 particles, which induce active infection of recipient CD4 T cells through the HIV-1 GP120 receptor
[20].
As noted earlier, virus-EVs are often enriched in PS, which not only improves uptake into recipient cells, but also plays a role in the modulation of the immune response
[21]. PS is a negatively charged glycerophospholipid that is asymmetrically distributed in the plasma membrane
[21]. While healthy cells present PS in the inner plasma membrane leaflet, PS translocation by Floppase to the outer leaflet acts as a cell signal. For instance, PS is externalized during cell stress and apoptosis, which leads to efferocytosis of the cell by macrophages, an alternative to phagocytosis that limits inflammation
[21]. These observations suggest that the transfer of PS via virus-EVs could induce efferocytosis of recipient cells while limiting the production of pro-inflammatory molecules. However, it remains to be established whether this effect benefits viral infection by limiting the antiviral immune response or benefits the host by limiting over-inflammation.
Virus-EVs also transport viral proteins that lower immune recognition through antigen presentation. For instance, HSV-1-EVs contain the viral envelope glycoprotein B (gB), which downregulates major histocompatibility complex class II (MHC-II) expression and promotes MHC-II uptake and secretion in EVs, thereby reducing the ability of recipient cells to present viral antigens and initiate immune responses
[22]. Similarly, HCMV-EVs also contain gB and gH, but also Fc-γ receptor homolog gp34. Gp34 binds the neutralizing IgG antibodies and prevents them from binding functional viral particles
[23].
The transport of infectious material through EVs also provides an avenue to escape recognition by neutralizing antibodies. For example, HCV-EVs transporting viral RNA and proteins are only partially recognized by neutralizing IgG antibodies
[24]. Similarly, DV-infected cells secrete autophagosome-derived EVs containing viral RNA and proteins (envelope E, nonstructural NS1, prM, membrane M), which lead to active infection of recipient cells without being impaired by neutralizing antibodies
[19].
Finally, some viruses-EVs transporting viral proteins permit active infection or apoptosis of immune cells. As mentioned previously, HBV-EVs lead to an active infection of the recipient NK cells despite the tropism of HBV for hepatic cells, which suppresses the antiviral immune response and promotes chronic infection
[25]. Moreover, Ebola virus-infected cells secrete EVs containing the viral matrix protein VP40, which specifically induces apoptosis of recipient T cells and monocytes, potentially by inhibiting the expression of proteins of the miRNA machinery (Dicer, Drosha, Ago1)
[26][27]. Ultimately, virus-EVs enhance viral spreading by inhibiting a variety of innate and adaptive immune responses. These interactions between viruses and EVs not only disturb the anti-viral immune response but also lead to a wider impact on the tissue, such as virus-associated chronic inflammation, tumorigenesis and liver disease.
2. EV-Mediated Progression of Virus-Associated Disorders
2.1. Virus-EVs Induced Inflammatory Disease
The interactions between viruses, EVs and the immune system can sometimes result in overstimulation and lead to chronic inflammatory disease. Zika virus-EVs secreted by mosquito cells that contain viral RNA, and the E-protein are integrated by naïve human endothelial vascular cells resulting in a pro-inflammatory and pro-coagulant cellular state
[28]. These Zika virus-EVs induce the expression of protease-activated receptors, which activate MAPKs p38, ERK1/2 and NFκB inflammatory pathways and promote the expression of pro-inflammatory cytokines. Zika virus-Evs also increase the expression of the tissue factor receptor, which is a pro-coagulant factor. These changes support the systemic inflammation observed during Zika virus infection
[28]. Zika virus-Evs can also induce the differentiation of naïve human monocytes into a pro-inflammatory intermediate/non-classical phenotype and activation of TNF-α mRNA expression
[28].
Some virus-Evs transmit pathogen-associated molecular patterns (PAMPS), which activate PRRs in immune cells and promote the secretion of pro-inflammatory cytokines, eventually leading to chronic inflammation, as observed for HIV-1, Epstein–Barr virus (EBV), Ebola virus and HTLV-1. HIV-Evs containing TAR RNA and miRNA not only favor T cell infection, as mentioned above, but also support chronic inflammation
[29][30]. Bernard et al. showed that the serum of HIV-1 infected patients HIV-EVs contained two other viral miRNAs (vmiR88 and vmiR99)
[31], which are recognized by Toll-like receptors (TLR) -3 (TAR RNA
[30]), TLR-7 (TAR miRNA
[30]) or TLR-8 (TAR miRNA, vmiR88 and vmiR99
[31]) in the recipient macrophages. These interactions activate the NFκB pathway and lead to the secretion of pro-inflammatory cytokines (IL-6
[30], TNF-α
[30][31]) by the recipient macrophages
[30][31]. Similarly, EBV encodes 49 mature miRNAs that can all be secreted inside EVs
[32]. During the latent infection, EBV-EVs containing EBER1 miRNA and PS are recognized by the TIM-1 receptor of DCs. EBER1 uncapped 5′ triphosphate terminus is recognized by PRRs, leading to the activation of the antiviral immune responses and the consequential chronic inflammatory disease in individuals suffering from autoimmune disease
[33]. In addition, Ebola virus-EVs containing the viral glycoprotein can promote the secretion of pro-inflammatory cytokines by the recipient monocytes, macrophages and DCs through its interaction with TLR-4
[15]. HTLV-1 infected cells also secrete EVs containing viral tax protein and mRNA and host proinflammatory molecules (GM-CSF, IL-6). Upon reception of tax, recipient DCs secrete IL-10, IL-12, IL-17A, IFN-γ and G-CSF, which could activate Th1, Th17 and cytotoxic T cells. Altogether, these HTLV-1-EVs could either enhance the antiviral immune response or support HTLV-1-associated chronic inflammation and myelopathy/tropical spastic paraparesis neurological disorder
[34].
Virus-EV-induced activation of PRRs in immune cells can also be achieved indirectly. HIV-EVs containing the Nef protein downregulates the ATP-binding cassette transporter type A1 (ABCA1) in recipient macrophages, leading to a cascade of cellular modifications. Downregulation of ABCA1 reduces cholesterol efflux, resulting in the inactivation of Cdc42, decreased actin polymerization and increased abundance of lipid rafts. This influences TLR-4 concentration in lipid rafts, potentiating ERK1/2 signaling and activating NLRP3 inflammasome and interleukin-1β (IL-1β) responses
[35].
Finally, uninfected cells can also secrete EVs that promote chronic inflammation. During DV infection, IL-1β is secreted as part of the antiviral immune response, which promotes the secretion of EVs by non-infected platelets. These EVs containing various host proteins are detected by macrophages and neutrophils through CLEC5A and TLR-2, respectively, resulting in the secretion of inflammatory cytokines and activation of the neutrophile extracellular trap (NET), which lead to systemic inflammation, tissue damage, and vascular permeability
[36].
HBV and HCV-EVs secreted by infected hepatocytes are integrated by hepatic stellate cells (HSCs), which cannot usually be infected by HBV and HCV, thus exasperating the disease to other cell types. HBV-infected cells express the oncogenic viral protein HBx, which induces EV biogenesis by interacting with the cellular CD9, CD81 and neutral sphingomyelinase 2 (N-SMase). The HBx protein and mRNA are secreted in HBV-EVs, then integrated by HSCs in which HBx stimulates cell proliferation, supporting HBV-associated liver disease
[37]. Likewise, HCV-EVs contain the upregulated host miR-19a RNA, which activates the SOC3-STAT3-TGF-β pathway in HSCs, leading to fibrosis and worsened liver disease
[38].
2.2. Virus-EVs Promote Tumorigenesis
Such viruses as the human papillomavirus (HPV), EBV, HIV-1 and HTLV-1 are widely known for their ability to cause cancer. Not surprisingly, EVs produced from cells infected with these viruses can also have oncogenic properties. Virus-EVs can aid in the development of tumors by transferring viral oncogenes or RNAs. For instance, HPV type 16 (HPV-16) infection increases the expression of more than 50 host microRNA (miRNA) in infected cells, ten of which are upregulated and selectively packaged into EVs as a result of the E6/E7 oncogene. These miRNAs inhibit apoptosis and senescence while inducing proliferation in both the infected and recipient cells
[39][40]. In addition, keratinocytes transduced with E6 and E7 oncogenes secrete EVs containing E6 and E7 mRNA, resulting in the expression of these oncogenes in nearby cells
[41]. The HTLV-1 infected cells secrete EVs containing viral tax, HBZ and Env mRNAs and the viral oncogenic tax protein. These HTLV-1-EVs enhance the survival of recipient PBMCs by protecting them from FAS-mediated apoptosis due to tax-mediated up-regulation of pro-survival signaling molecules (AKT, Rb) and activation of the NFkB pathway. This supports HTLV-1-associated adult T cell leukemia/lymphoma
[34].
Virus-EVs can also cause cancer cells to adopt a more aggressive and invasive phenotype. For example, the EBV-infected cells express the viral latent membrane protein-1 (LMP-1), a constitutively active signaling protein mimicking CD40. LMP-1 can induce EV secretion through a CD63-dependent mechanism
[42]. In Burkitt’s lymphoma, the EBV-infected B cells secrete EVs containing LMP-1 that induce recipient B cell proliferation and differentiation into a plasma cell-like phenotype, causing a high production of IgG1 that promotes autoimmune disorders
[43]. In nasopharyngeal carcinoma, LMP-1 increases the packaging of hypoxia-inducible factor 1α (HIFα) into EVs
[44] and the secretion of EVs through syndecan-2 (SDC2) and synaptotagmin-like-4 (SYTL4)—NFκB pathway
[45]. In recipient nasopharyngeal carcinoma cells, HIFα serves as a transcription factor that modifies E- and N-cadherin expression, resulting in an epithelial-mesenchymal transition of cancer cells that promotes cancer invasion
[44]. The HIV-1-EVs secreted by T cells contain miR-155-5p that promotes the expression of proinflammatory factors (IL-6, IL-8, TGF-β) and migration molecules (collagen type I, matrix metallopeptidase 2) in recipient cervical cancer cells. In this tumoral context, IL-6 induces cancer cell proliferation and inhibits apoptosis through STAT3. Moreover, miR-155-5p targets AT-rich interactive domain (ARID2) DNA binding protein, which inhibits the migration of cervical cancer cells through the NFkB pathway. Altogether, these HIV-1-EVs promote cervical cancer proliferation and invasion
[46].
Cancer escape from immune surveillance can also be facilitated by the transportation of viral regulatory factors in virus-EVs. HPV-16-infected keratinocytes expressing E7 oncogene secrete plasma-membrane-derived EVs that inhibit the adaptive immune response. These HPV-16-EVs inhibit CD40 expression and IL-12 p40 subunit secretion by recipient Langerhans DCs in the epidermis, which reduces antigen presentation and activation of anti-HPV cytotoxic T cells
[47]. The Lymphoma EBV-infected cells secrete EVs containing miR-BARTs and PS, which are integrated by monocytes. miR-BARTs enhance the upregulation of IL-10, TNF-α and Arginase 1 in active M1 monocytes, which support their phenotype switch into a regulatory M2-like phenotype and promote tumor growth and lymphoma severity
[48]. Similarly, gastric carcinoma EBV-infected cells secrete EVs targeting DCs and suppress their maturation, resulting in a worse prognosis for patients
[49].
3. EVs and Viruses as Therapeutic Tools
Ultimately, the relationships between EVs and viruses play an integral role in the spread and progression of multiple pathologies. A key contributor to these pathologies is the immune system. Virus-EVs act as potent immunomodulatory molecules that can serve to either activate or suppress the immune system. This can consequently lead to an onset of chronic inflammatory diseases or permit immune evasion and escape of other pathologies, such as cancer. However, the alternative can also be true. Virus-EV interactions can be a powerful immunotherapy strategy to target a variety of disorders.
Both viruses and EVs have a wide range of therapeutic applications. EVs have reached clinical trials for a wide variety of different illnesses—from cancer
[50][51][52][53] to cardiovascular disease, type 1 diabetes, neurodegenerative diseases (Huntington’s disease, Parkinson’s disease)
[54], autoimmune and inflammatory diseases
[55], and even for acute respiratory distress syndrome (ARDS) from SARS-CoV-2 infection
[56]. Similarly, viral vectors have been used for vaccine development for decades and are now being readily explored for gene therapies, vaccines and oncolytic viral therapies
[57][58][59][60]. However, EV and virus therapies both have inherent strengths and weaknesses. EVs are safe, stable, have low immunogenicity and can intrinsically cross tissue and cellular barriers but are challenging to produce and load efficiently with drugs
[50][54]. In contrast, viruses are easy to manufacture and can be engineered to overexpress therapeutic payloads but have an increased risk of toxicity and are more readily cleared by the immune system
[61].
In this respect, combinational virus-EV therapeutics have many promising characteristics that overcome the weaknesses of their individual counterparts. As discussed previously, EVs can promote the cell-to-cell spread of viral particles and genomes, permit infection of cells with low expression of the viral receptor and provide protection against immune responses. Hence, the viral therapies that spread via EVs, either naturally or by design, have the potential of reaching more target cells without clearance from the immune system. Viruses can also help overcome the low loading of therapeutic payloads in EVs by propagating in recipient cells and delivering prolonged expression of therapeutic transgenes. Virus infection is also known to promote EV production and secretion in many cells, so there is added potential for using engineered viruses with selectivity for different tissues to generate EVs with therapeutic payloads.
3.1. Engineering viruses to Target EVs
Adeno-associated virus (AAV) is a powerful tool for gene delivery that has been widely studied in clinical trials for therapies targeting the brain, spinal cord, liver and muscle
[62]. Despite having an excellent safety profile and high efficiency in tissue transduction, there are still some drawbacks to using AAV for gene therapy. Many patients have already acquired an immunity against AAV
[63], leading to the cleansing of viral particles by neutralizing antibodies and killing of the AAV-transduced cell by cytotoxic T cells
[64]. As therapeutic AAVs rarely integrate into the host genome except at specific sites, lack of integration can lead to loss of gene expression after cell division, which occurs in some tissues such as the liver. Off-target gene delivery may also occur, particularly in the liver, which can lead to other side effects
[64].
To overcome these issues, AAVs can be targeted to EVs to provide protection against neutralizing antibodies while keeping the ability to target specific tissues and limiting off-target infection. Maguire et al. showed that the cells transfected with DNA for producing AAV particles also secreted EVs containing the fully functional AAV (12,2% of EVs contained AAV1)
[65]. Transfection of the AAV producer cells with vesicular stomatitis virus glycoprotein G (VSV-G) further increased the production of AAV-EVs. These AAV-EVs also contained VSV-G at their surface, which led to higher transduction efficiency (AAV1 and 2)
[65].
AAV-EVs have a high potential for the treatment of neurological disorders. AAV-EVs (AAV1, 2, 9) avoid immune neutralization by host antibodies in vitro and in vivo
[66] and are more efficient at transducing targeted brain cells in mice
[66][67]. The selectivity of AAV can also be improved further by the creation of chimeric receptors that target neurons. Gyorgy et al. fused the rabies virus glycoprotein (RVG) to the transmembrane domain of platelet-derived growth factor (PDGF-TD) and transfected this construct along with AAV producer vectors. Due to this procedure, they generated AAV-EVs expressing RVG, which binds to cells expressing α-7 nicotinic acetylcholine receptor, and enhanced AAV-EVs uptake into the brain by neurons, astrocytes and endothelial cells
[66]. It has also been suggested that integrating AAV into EVs does not increase infection efficiency but rather enhances the transport of AAV from the vessels to the tissue while protecting the virus from neutralizing antibodies
[68]. This study was based on in vitro observations of the potential transduction of primary astrocytes and neuronal N2A cells after incubation with AAV-EVs, AAV or AAV in suspension with EVs
[68].
AAV-EVs also showed promising results against hemophilia B genetic disease. Meliani et al. transfected HEK293 cells with AAV8 plasmids encoding the human coagulation factor IX and later harvested AAV-EVs presenting TSG101 and CD9 specific markers. After intravenous injection, these AAV-EVs are protected from antibody recognition and successfully target and transduce liver cells, brain cells and skeletal muscle cells in mice. AAV-EVs could thus improve gene therapy safety and efficiency in liver genetic disease
[69]. Finally, AAV-EV strategies have also been studied for cancer therapy. Transfected HEK293 cells with AAV6 plasmids expressing luciferase have been tested for the effect of AAV-EVs on non-small cell lung cancer (carcinoma and adenocarcinoma) and small cell lung cancer cell lines. These AAV-EVs were more effective at transducing cancer cell lines in comparison to AAV alone in vitro. In vivo, AAV-EVs evade neutralizing antibodies and efficiently transduce lung cancer cells after an intratumoral injection
[70].
3.2. Natural Transport of OVs in EVs Enhances Oncolytic Viral Therapy
The intertwined relationship between EVs, viruses and activation of the immune response has a promising role in the development of oncolytic viral therapy. Oncolytic viruses (OVs) selectively infect tumor cells causing cell death, the release of tumor and viral antigens and activation of antiviral and anti-tumoral immune responses. Moreover, OVs can be genetically modified to express factors increasing the immune response, such as immune checkpoint inhibitors (anti-PD1, anti-PD-L1 and anti-CTLA-4), cytokines (IL-2, IL-12, etc.) or T cell engagers
[71][72]. Many OV therapies are currently in clinical trials, with various degrees of success. For example, talimogene laherparepvec (T-VEC) is a genetically modified HSV-1 encoding granulocyte-macrophage colony-stimulating factor (GM-CSF) that was approved by the U.S Food and Drug Administration and the European Medicine Agency in 2015 for the treatment of advanced melanoma
[71][73]. The triple mutant HSV-1 G47Δ OV (Delytact) has also received approval in Japan for the treatment of malignant glioma and in clinical trials for the treatment of prostate cancer, malignant pleural mesothelioma and recurrent olfactory neuroblastoma
[74][75][76].
There are many OVs, such as VSV
[77], HSV-1
[78] and NDV
[14], which induce the secretion of EVs to improve their spread or sensitize cancer cells to infection. In addition, tumor cells secrete high levels of EVs to support tumorigenesis
[79]. These EVs create an immunosuppressive environment for the tumor to grow or transfer oncogenes to healthy cells to promote tumor invasion. EVs can also promote cancer metastasis by enhancing cell migration, modulating the extracellular matrix and modifying stromal cells to prepare a pre-metastatic niche
[79]. Because of these intense EVs exchanges within the tumor and the capacity of OVs to induce EV secretion to either enhance viral spreading or increase the antiviral immune response, it is interesting to study whether EVs enhance virotherapy efficiency.
Several studies have shown that EVs increase the spreading of oncolytic adenoviruses, enhance the immune response, and help target metastatic niches. Indeed, Ad5/3-D24-GMCSF leads to the secretion of EVs containing viral proteins and genomes that actively infect the recipient tumor cells, thus enhancing viral spreading
[80]. Moreover, melanoma cells secrete EVs during LOAd-CD40L or LOAd-4-1BBL Ad infection that activate DCs
[81]. EVs secreted by the OBP-301 Ad-infected tumor cells can reach metastatic niches. OBP-301 is an oncolytic telomerase-specific Ad studied in a phase I clinical trial. OBP-301 administration in HCT116 colorectal primary tumors in a mouse model leads to the secretion of EVs having a strong tropism toward tumor cells, which were then transported to metastatic tumors where they induce the recruitment of immune cells, just like the OV
[82]. Altogether, these observations support the potential of EVs in enhancing OV therapy efficiency. Many studies are thus turning toward EVs as a vector for OVs safe and specific transport to the tumor site.
3.3. Engineering of OV-EV Targeted Therapies
A major challenge with OV therapy is the presence of pre-existing anti-viral immunity. OVs are also often delivered via intratumoral injection, which limits their ability to target metastatic tumors. Whereas, intravenous injection of OVs faces issues with viral clearance in the vascular system and peripheral tissues by immune cells, neutralizing antibodies and other molecules
[83]. However, OVs can be genetically modified to escape host antibodies
[84] and phagocytosis
[85]. In addition, there are delivery strategies for OVs, such as the use of liposomes to cloak OVs
[86], the use of tumor-infiltrating T cells to which oncolytic Ad binds
[87], or even the use of NK cells
[88] or mesenchymal stem cells
[89] that get infected by the OV and then migrate to the tumor site. In addition, EVs also have the ability to naturally uptake some OVs (OV-EVs) and improve delivery to target tissues.
Cancer cells can be used as OV-EVs producers, which will have a natural tropism for cancer cells
[90]. The OV-EVs can also be loaded with chemotherapeutic agents via co-incubation, which helps increase their delivery to tumors
[90]. For example, Garofalo et al. created oncolytic Ad (Ad5D24-CpG) encapsulated in EVs with the chemotherapeutic agent paclitaxel (Ad-EV-Chemo) from A549 lung carcinoma epithelial cells
[90]. They found that Ad-EV-Chemo increased apoptosis of tumor cells in vitro while simultaneously raising virus replication and transduction efficiency. In nude mice harboring A549 tumors, Ad-EV-Chemo improved survival and decreased tumor growth while completely modifying the transcriptome of xenograft tumor cells
[90]. They then demonstrated the efficiency of the systemic injection of the combined therapy in an immunocompetent mouse model. EVs were produced by the LLC1 murine Lewis lung carcinoma cell line. Ad-EV-Chemo induced T cells activation and infiltration inside the tumor. Interestingly, the treatment leads to a localized inflammation in the peritumoral environment, suggesting that EVs both protect OVs from alerting the immune system while ensuring the specific delivery to the tumor, making systemic injection safer
[91]. In another study, Garofalo et al. produced the Ad-EVs from LLC1 cells and showed in the C57BL/6 immunocompetent mouse model that intravenous but not an intraperitoneal injection of Ad-EVs leads to the successful targeting of the treatment to the tumor and infection of tumor cells
[92].
Similarly, plasma membrane-derived EVs can also carry chemotherapeutic drugs (cisplatin)
[93] and oncolytic Ad (Ad5)
[94] to the tumor site. Ad5-EVs secreted by A549 cancer cells are protected from antibody neutralization and deliver Ad5 inside tumor cells in a viral receptor-independent manner. Moreover, this therapy also successfully targets stem-like tumor-repopulating cells, thus preventing cancer relapse
[94]. Altogether, these studies support the potential of Ad-EVs and Ad-EVs- Chemo in cancer therapy, which can target tumor cells after intravenous injection due to their protection from neutralizing antibodies, eventually inducing T cell infiltration inside the tumor and decreasing tumor growth.
OV-EVs can also be engineered using EV-mimetic nanovesicle drug loading technology, which allows the production of a much higher quantity of EVs containing the drug of interest. Producer cells are first transduced to express a drug and then passed step-by-step through smaller and smaller nanosized filters, eventually forcing the formation of EV-mimetic nanovesicles containing the drug
[95]. Zhang et al. used this method to produce a higher amount of EVs loaded with an oncolytic Ad5-P expressing the extracellular domain of the programmed cell death protein 1 (PD-1)
[96]. The team genetically modified HEK293T cells to express VSV-G transmembrane protein and then infected these cells with Ad5-P OV. Using EV-mimetic nanovesicle drug loading technology, they successfully produced a high quantity of EVs loaded with Ad5-P and presented VSV-G at their surface. They tested these Ad5-P-EVs in various cancer cell lines and in an ascitic tumor model produced by intraperitoneal injection of hepatocellular carcinoma cell line H22 in mice. Ad5-P-EVs efficiently entered the recipient cells through the interaction of VSV-G with the low-density protein (LDL) receptor. They showed that this method of transport of Ad5-P through EVs protects the virus from neutralizing antibodies, enhances virus infection and ultimately increases the production of soluble PD-1 by infected tumor cells while prolonging mice survival
[96].
3.4. Packaging of Therapeutic miRNAs into EVs via Engineered OVs
In the past, the researchers' group has shown that it is possible to design and encode artificial miRNAs (amiRNAs) into oncolytic VSVΔ51 that get packaged into EVs and transmitted to surrounding uninfected cells
[77]. This strategy allows for the use of oncolytic viruses to deliver amiRNAs that target any cellular gene product throughout the TME. In the study, the researchers screened a library of amiRNAs to identify candidates that sensitize cancer cells to virus infection. One such amiRNA that the researchers found targets transcripts encoded by the cellular gene ARID1A, which enhanced infection in tumors but not normal cells. ARID1A is a member of the SWI/SNF gene family encoding helicases and ATPases that regulates gene transcription by altering chromatin structure
[97][98][99][100][101]. ARID1A-knockout cells display greater susceptibility to several OV platforms, including oncolytic VSVΔ51, vaccinia virus, herpes simplex virus-1 and reovirus
[77]. Interestingly, two coincident discoveries by Shen et al.
[102] and Pan et al.
[103] showed that disruption of the SWI/SNF complex in tumor cells also leads to enhanced immunotherapy but through the use of immune checkpoint inhibitors. In addition, ARID1A has a synthetic lethal pathway with EZH2, and thus cells that lack ARID1A can be killed by the drug GSK126, a specific EZH2 inhibitor
[104].
EVs produced by cancer cells upon infection with VSVΔ51 encoding an amiRNA targeting ARID1A, but not a non-targeting amiRNA, can sensitize uninfected cells to GSK126
[77]. Through enhanced OV replication and synthetic lethality, the researchers showed increased survival in mice bearing aggressive pancreatic, ovarian and melanoma tumors
[77]. The researchers also demonstrated that the researchers can engineer an amiRNA to target PD-L1 and combine the two amiRNAs into a single virus entity for cancer therapy
[77]. Taken together, these findings support the development of virally encoded, EV-delivered amiRNAs as a strategy to promote virus spread within tumors and modify the TME.