Several immunostimulating molecules are discharged when cells die including HMGB1, uric acid, ATP and heat shock proteins (HSP) (14, 16, 35–37). HMGB1 is the most frequently encountered DAMP in cells undergoing stress (14). This protein, initially found in the 1970s, is a nuclear protein and binds to chromatin (5, 38). HMGB1 is highly conserved and is involved in various cellular processes such as DNA repair, gene expression and replication (39–42). Upon inflammation or cellular stress, HMGB1 binds to immune cell receptors such as Toll-like receptor 2 (TLR2), Toll-like receptor 4 (TLR4), and Receptor for advanced glycation end products (RAGE) (14, 43, 44). Apart from being released by cells infected with pathogens, dying cells or cells undergoing necrosis are also able to secrete HMGB1 (45). The release of HMGB1 may induce inflammation upon binding to different immune receptors through the release of cytokines such as tumor necrosis factor alpha (TNF-α) and interferon gamma (IFN-γ) (45). In cancer, the role of HMGB1 needs further understanding as dual roles of HMGB1 have been reported (42, 46). Pro-tumorigenic roles of HMGB1 include initiation of inflammation (47), enhance tumor cell proliferation (48), promotes tumor invasion and metastasis (49), enforces angiogenesis (50), involves in chemoresistance (51) and promotes antitumor immunity (52). Nevertheless, there have been reports that HMGB1 may also play a role as a tumor suppressor (53, 54). The paradoxical roles of HMGB1 in cancer have been of interest to many researchers over the years. What is more interesting, HMGB1 has been reported to be present in EVs as well. A study by Deng et al. showed that hepatocytes release HMGB1 via vesicles after being stimulated by lipopolysaccharide (LPS) (55). A follow-up study by the authors showed that HMGB1 was indeed packaged in exosomes and released extracellularly (56). The authors showed that HMGB1 was released in exosomes via the TLR4 pathway (56). In a different study, it was discovered that large burn injuries (LBI) were able to secrete plasma microvesicles enriched with HMGB1 (57). The study found that the released HMGB1 formed complexes with pro-interleukin-1-beta (pro-IL-1β) in both human and mouse plasma, and this heterocomplexes were able to induce immune dysfunction in LBI (57).

According to vesiclepedia (58), a database for protein/mRNA enriched in extracellular vesicles, HMGB1 is present in EVs such as exosomes and microvesicles from various sources. For instance, HMGB1 protein was reported present in EVs derived from breast cancer cells (59), bronchial epithelial cells (60), chronic lymphocytic leukemia cells (61), colorectal cancer cells (62, 63), glioblastoma cells (64), among others. These studies show that while HMGB1 is secreted as free HMGB1, this protein can also be secreted and packaged in vesicles as well. However, the type of EVs that contain HMGB1 has not been reported exclusively for one type of EV. The abovementioned studies show that HMGB1 can be present in exosomes, general extracellular vesicles as well as microvesicles. The exact mechanism as to how HMGB1 is sorted into these vesicles is still lacking information. The method of isolation of different types of EVs may also influence the presence of HMGB1 in these EVs. HMGB1 that is present within EVs has been shown to affect other surrounding cells as well. Functionally, several reports have also shown that EV-derived HMGB1 can participate in the carcinogenesis process. A study by Li et al. suggested that exosomal HMGB1 derived from esophageal squamous cell carcinoma managed to differentiate monocytes into the pro-tumorigenic Programmed cell death-bearing-tumor-associated macrophages (PD1+ TAMs) phenotype (65). A different study by Ye et al. showed that exosome-derived HMGB1 in hepatocellular carcinoma can activate B cells (66). This subsequently leads to the enhanced proliferation of TIM-1+ regulatory B cells by the TLR2/4 and Mitogen-Activated Protein Kinase (MAPK) pathways (66). Additionally, it was also shown that exosomal HMGB1 play a role in platelet-driven cancer malignancy. It was reported that treatment with anti-platelet drug, dipyridamole and aspirin inhibited tumor progression in Lewis lung carcinoma (LLC) cell lines and reduced the exosomal HMGB1 content. Similar finding was displayed in a tumor-bearing mouse model where combined treatment of dipyridamole and exosome-release inhibitor, GW4869 significantly mitigated tumor growth (67). Exosomal HMGB1 was also found to be involved in angiogenesis. A recent study by Gao et al. showed that hypoxic bone marrow mesenchymal cells were able to release exosomal HMGB1 that further enhanced angiogenesis via c-Jun N-terminal Kinase JNK/Hypoxia-inducible factor (JNK/HIF) pathway (68). It is interesting to note that the role of HMGB1 may differ depending on the form it is released. For instance, a study by Ma et al. showed that extracellular HMGB1 had opposing effects towards the expression of SAM and SH3 domain containing protein 1 (SASH1) as compared to exosomal HMGB1 (69). Although it is well known that extracellular HMGB1 is able to activate the inflammatory pathway via the TLR/RAGE receptors, information on EV-derived HMGB1 is still lacking and this calls for the need of further research. Since HMGB1 is a nuclear protein, it can be assumed that HMGB1 is packaged within EVs and may not be present on the surface, but further verification is needed. Therefore, the mechanism by which HMGB1 is able to stimulate immune response upon internalization of EVs still needs to be determined.

Besides HMGB1, HSPs are commonly categorized as DAMPs as well (70). HSPs act as chaperones to ensure the proper folding of proteins (70, 71). These proteins are typically released when cells are under stress and are usually overexpressed in tumor cells due to the demand for cellular energy and the unstable environment (72). It was shown that certain HSPs trigger a pro-inflammatory response in mouse macrophage and human monocytes (73). Upon encountering HSPs, T regulatory cells (Tregs), T cytotoxic cells, natural killer (NK) cells, macrophages and DCs are activated (74). Nevertheless, the roles of HSP as DAMPs are still debatable. However, for the purpose of this review, we will consider HSPs as DAMPs and discuss the presence of HSPs in EVs. The presence of HSP-containing EVs released from cancer samples has been reported by several groups (75–80). For instance, a report by Gastpar et al. showed that HSP70 was present on the membrane of tumor-derived exosomes from pancreatic and colon cancer cell lines (81). The authors also showed that these exosomes were able to stimulate migration and HSP70 reactivity in NK cells (81). HSP70 has been reported to be released by tumor cells upon external stress such as radio or chemotherapy (77, 82). Therefore, it is presumed that under stressful conditions, the expression of HSP70 on exosomes is also increased. Lv et al. showed that there was indeed, a difference in the expression of HSP60, HSP70 and HSP90 in exosomes derived from HepG2 cells after treatment with chemotherapeutic drugs (83). Similar to the previous study, these HSP-containing exosomes were able to increase NK cell cytotoxic ability (83). A similar study by Elsner et al. suggested that HSP70-positive exosomes from melanoma cells were able to enhance NK cells cytolytic activity against YAC-1 cells (84). The increase of HSP70 in exosomes has also been shown upon induction by heat stress in murine models (85). Cho et al. showed that these heat-induced exosomes containing HSP70 elicit a stronger T helper type 1 (Th1) immune response (85). The presence of HSP70 in tumor-derived EVs has also been reported elsewhere. A study by Xie et al. showed that exosomes containing HSP70 stimulate anti-tumor immunity by enhancing the maturation of DCs and Th1 cells (86). A recent pilot study by Chanteloup et al. reported that exosomal HSP70 can be used to detect and monitor metastatic solid tumors such as breast and ovarian cancer (75, 76). HSP60 has also been shown to be released by tumor-derived exosomes (87, 88). A study by Wyciszkiewicz et al. showed that certain HSPs such as AlphaB-crystallin and HSP22 are present in exosomes from gynecological cancers (89). The authors showed that although these HSPs were present in both exosomes and serum, there is no correlation between the two sources (89). Similar to HMGB1, extracellular HSPs are not representative of exosomal/EV-derived HSPs in terms of abundance and function. According to vesiclepedia (58), HSPs, such as HSP90 were reported in cancer-derived EVs such as bladder cancer cells (90) and breast cancer (59). Almost all the reported studies show that HSPs are present within exosomes and not in other types of EVs. However, these studies report different techniques of isolation and characterization of exosomes and may not be conclusive enough to state that HSPs are exclusively found in exosomes. Nevertheless, though the presence of HSPs has been reported in EVs, the exact mechanism as to how these proteins induce an inflammatory/immune response is still elusive. The localization of HSPs as to whether it is present internally or on the surface of EVs warrants more studies (91). A study by Tang et al. showed that HSP90α is present on the surface of tumor-derived exosomes and is able to mediate communication with other cells (92). However, an earlier study by Clayton et al. showed that HSPs are also present in the lumen of exosomes and may not interact with target cells through cell surface receptors (93). Therefore, more in-depth studies are needed to determine whether HSPs are able to act as DAMPs and activate inflammation through certain receptors.

S100 are a class of proteins known to bind to calcium and regulate intracellular and extracellular processes (94, 95). There are around 24 types of S100 protein members that can be divided into three main subclasses depending on their function (96). S100 proteins have long been recognized as DAMPs due to their ability to elicit an inflammatory response (97, 98). In cancer, the S100 proteins have been reported to be involved in carcinogenesis. In a study done by Hiratsuka et al., S100A8 and S100A9 proteins are found to be involved in lung cancer invasion and myeloid cell recruitment (99). S100 proteins have been reported to be present in EVs as well. A study by Prieto et al. showed that in chronic lymphocytic leukemia (CLL), S100A9 protein was present in the plasma exosomes (100). The authors showed that the exosomes containing S100A9 were able to activate the nuclear factor-kappa-light-chain-enhancer of activated B cells (NF-κB) pathway in leukemic cells (100). Not only that, in a different study by Li et al., the authors demonstrated that the S100A9 protein was also present in exosomes derived from follicular fluid of polycystic ovary syndrome patients (101). These exosomes were also able to promote inflammation via the NF-κB pathway (101). Although the molecular mechanism of the activation was still unmapped, this study, however, displayed an interesting finding in which the levels of NF-κB pro-inflammatory cytokines were increased upon incubation with S100A9-enriched exosomes (101). According to vesiclepedia (58), the presence of members from the S100 family was reported to be present in EVs from various sources. For instance, the S100A7A protein was found in EVs from colorectal cancer cell lines (62) and T cells (102). S100A5 protein was also found in colorectal cancer cell lines (103), and S100A12 protein was found in EVs from brain cancer cells, colorectal cancer cells, melanoma cells, kidney cancer cells and more (59). Similar to other DAMPs, the presence of S100 proteins is also not exclusive to one type of EV. Although the presence of S100 proteins has been reported in EVs, the actual function of S100 as DAMPs within EVs remains to be elucidated. Generally, free or extracellular S100 proteins are able to act as DAMPs by binding to receptors such as RAGE or TLR, but the mechanism of S100 within EVs still needs to be investigated. The presence of S100 proteins in EVs and how this affects the pathway leading to inflammation is still unknown.

Besides the abovementioned molecules, other components within EVs that are also able to elicit an immune response is nucleic acid. It is well-established that microRNAs (miRNAs), short-lengthed nucleic aids, can be encapsulated within EVs. Some studies have shown that these EV-bound miRNAs were able to induce an immune response via the intracellular TLR pathway in several diseases (104–107). In rheumatoid arthritis, for instance, exosome-containing let-7b was able to differentiate macrophages into the M1 phenotype via TLR7 (108). A different study was able to show that miR-21 encapsulated in EVs was able to induce neurotoxicity through TLR7 signaling as well (109). In cancer, a study by Fabbri et al. demonstrated that exosome-derived miRNAs from lung cancer cells were able to bind to TLR8 on macrophages and activate the NF-κB pathway (110). This, in turn, led to the release of pro-inflammatory cytokines such as TNF-α and interleukin-6 (IL-6) (110). Although the presence of miRNA in EVs such as exosomes and microvesicles is well-established, there are still limited studies on whether these encapsulated miRNAs are able to stimulate TLR pathway, and subsequently activate inflammation. Additionally, most of the reported studies had purified EVs from sources that did not undergo any cellular stress such as chemotherapy or radiation, and thus the role of miRNA-EVs as DAMPs needs to be further determined. Figure 1 demonstrates the overall schematic representation of how DAMPs are released within EVs and subsequently interact with target cells.