Opportunities and challenges of natural killer cell-derived extracellular vesicles

1 Introduction

Cancer immunotherapy has gained widespread attention as a clinically proven therapeutic strategy, and the relay transfer of natural killer (NK) cells has emerged as a promising approach to controlling the immune system against cancer. As the first line of defense against tumors and viral infections, NK cells can induce antigen-independent immune responses against malignant cells. A growing number of scientific reports and clinical studies have demonstrated that NK cell-based immunotherapy has promising antitumor effects (Laskowski et al., 2022). Furthermore, NK cell therapy or chimeric antigen receptor (CAR) NK cell therapy has unique advantages over existing hot T-cell immunotherapy (Biederstadt and Rezvani, 2021; Rafei et al., 2021). NK cells recognize and kill tumors by combining signals generated by independent inhibitory and activating receptors, effectively inhibiting tumor escape through antigen downregulation (Basar et al., 2020). Currently, a wide spectrum of research is being conducted on NK cell-related cancer therapies, including CAR NK, engineered NK cells, and allogeneic natural killer cell infusion (Tang et al., 2018; Liu et al., 2021). However, despite several clinical trials, the prospects for NK cell-based therapies for solid tumors are not optimistic. Challenges include difficulty in ex vivo expansion meeting clinical grade, tumor immune escape, limited in vivo persistence, and limited infiltration into solid tumors (Oh et al., 2019; Bald et al., 2020). Moreover, the tumor microenvironment (TME) inhibits NK cell responses (Fang et al., 2018; Lian et al., 2021). These factors directly hinder or limit the use of NK cells in solid tumor therapy.

Extracellular vesicles (EVs) can be divided into three subgroups based on their biological origin, including exosomes (30–150 nm in diameter), microvesicles (150–1,000 nm), and apoptotic vesicles (50–2,000 nm). The role of EVs secreted by immune cells in antitumor therapy has received more attention in recent years, with many studies confirming their great potential (Shen and Ren, 2018; Choi et al., 2022a). Among them, NK cell-derived EVs (NEVs) have gained more attention for their unique biological properties. NEVs possess NK cell surface receptors and cytotoxic proteins that function similarly to parental cells, enabling them to kill tumors directly in the TME. Unlike cells, nanoscale NEVs can easily diffuse and infiltrate solid tumors and own natural targeting and biocompatibility properties. Furthermore, there have been few types of research showing that immunosuppressive TME affects NEVs. Therefore, the emergence of NEVs may overcome the limitations of NK cells in immunotherapy. Numerous studies have thoroughly investigated the biological basis for tumor killing by NEVs and found that they carry a variety of bioactive molecules, including membrane toxicity receptors, cytotoxic proteins, cytokines, and microRNAs (miRNAs). This is the biological basis for NEVs natural tumor-killing and tumor-targeting properties, as well as the ability to interact with immune cells such as tumor-associated macrophages (TAM) and cytotoxic T lymphocyte (CTL) (Federici et al., 2020). Due to the above advantages, engineered NEVs have received a lot of attention to enhance their tumor-killing capabilities. Current research focuses on the use of engineered modifications to enhance the functionality of NEVs (Yang et al., 2021a). Nevertheless, many challenges remain in the development of NEVs, such as cell source production methods and interaction mechanisms. This review aims to summarize the latest research on the production, application, mechanism, and modification of NEVs (Figure 1).

FIGURE 1. The cell sources of NEVs could be mainly divided into peripheral blood mononuclear cells (PBMC) and other cell lines. In the purification process of NEVs, the more mature isolation methods include ultracentrifugation (UC), ultrafiltration, and size exclusion chromatography (SEC). Purified NEVs have anti-tumor activity and immunomodulatory effects, and engineered modifications can confer new functions on NEVs. Common NEVs engineering techniques can be classified as exogenous and endogenous modifications. Exogenous modifications: (A) NEVs can be used to prepare the PTX-NEVs drug delivery system through electroporation (Han et al., 2020). (B) The therapeutic potential of doxorubicin-loaded NEVs shows promising antitumor activity in vivo against the MCF-7 induced tumor model (Pitchaimani et al., 2018). (C) Light-activatable silencing NK-derived exosomes (LASNEO) are orchestrated by engineering the NEVs with hydrophilic small interfering RNA (siRNA) and hydrophobic photosensitizer Ce6 (Zhang et al., 2022). (D) NEVs are used as a versatile toolkit to synergistically improve adoptive T-cell therapy for solid tumors (Nie et al., 2021). (E) NEVs are in combination with their biomimetic core–shell nanoparticles for tumor-targeted therapy (Wang et al., 2019). Endogenous modifications: (F) The NK cell is lentivirally transduced to express and load BCL-2 siRNAs (siBCL-2) into NEVs (20).

2 Preparation of NEVs

2.1 NK cell source

For both NK cells and NEVs used in therapeutic studies, the optimal source of NK cells is currently a controversial topic. Peripheral blood mononuclear cells (PBMC) and cell lines are the two main sources of NK cells used in therapeutic studies. PBMC can be obtained from patient blood or blood collected from healthy adult volunteers. Among them, PBMC-induced NK cells from cancer patients had fewer clinical adverse effects and a higher safety profile (Sakamoto et al., 2015; Fang et al., 2018; Fang et al., 2022). However, their cytotoxic properties are compromised, with significantly reduced expression of activating receptors such as NKG2D, DNAM-1, and NKp46, which directly affects their derived EVs (Cianga et al., 2021). In contrast, healthy population-derived allogeneic NK cells have high yield and cytotoxicity but a low safety profile (Federici et al., 2020). It is worth noting that PBMC-derived NK cells are difficult to use in tumor therapy on a large scale and for a long time due to the restriction of donor and blood groups (Shah et al., 2015). Another important source is NK cell lines, such as NK-92 or NK-92MI, are important sources of NK cells and have become one of the important alternatives to autologous NK cell biologics. Moreover, the NK-92 cell line is the only human NK cell line approved for clinical use by the FDA. NK-92 can be expanded in culture in the presence of cytokines (NK-92MI amplification in vitro is not cytokine-dependent). It is inexpensive to administer, and there is substantial evidence that it is relatively safe (Gong et al., 1994). A study comparing the distribution of cytolytic proteins in NEVs from primary NK and NK-92 cells and found strong similarities and the same satisfactory tumor-killing effect (Aarsund et al., 2022).

Furthermore, umbilical cord blood (UCB), hematopoietic stem and progenitor cells (HSPCs), induced pluripotent stem cells (iPSCs), and CAR NK cells are valuable sources of NK cells (Li et al., 2018; Kundu et al., 2021; Boyd-Gibbins et al., 2022). CAR-NK and iPSC-NK cells could benefit from advances in manufacturing and genome engineering techniques to create NK cells and NEVs with context-dependent functions and enhanced potency and specificity. Future research is required to confirm the differences in the composition and effects of EVs produced by NK cells from different sources. The common cellular sources of NEVs production and their advantages and disadvantages are summarized in Table 1.

TABLE 1. Advantages and disadvantages of the mainstream NK cell sources.

2.2 Production of NEVs

The most common method for isolating EVs is cell culture supernatant collection. Since there is no “gold standard protocol” for the preparation of pure EVs, the properties and functions of EVs may vary depending on the culture and isolation methods. Differences in cell culture oxygen content and inoculation surface have been shown to affect EVs production (Zhang et al., 2017; Liu et al., 2020). Differences in cultural media also have an impact on EVs extraction. It has been shown that exogenous proteins introduced into cell culture can affect the type and characteristics of exosomes (Whitford and Guterstam, 2019; Chen et al., 2020). However, it has also been shown that the emergence of serum-free media modifies the biology of EVs (Mendt et al., 2018).

Obtaining EVs through natural secretion is hampered by the low yield. The advent of exosome-mimetic vesicles (EMs) with higher yield is expected to resolve this issue (Ou et al., 2021). Large-scale production of EMs could be an alternative to conventional EVs production. When cytoplasmic membranes are forced to rupture, they reassemble into smaller vesicles. Thus, EMs with diverse sizes can be produced using diverse filter membranes and micro-extruders. A study reported that the production of EMs using this method was 250 times higher than naturally secreted EVs (Lee et al., 2020). More recently, similar methods have been used to produce NK EMs, with tumor-killing abilities and impressive stability under physiological conditions, which could also be loaded with chemotherapeutic drugs for targeted cancer therapy (Pitchaimani et al., 2018; Zhu et al., 2018). However, more research is needed to investigate the differences in efficacy and safety with naturally secreted EVs.

Either artificially generated EMs or naturally secreted EVs typically contain multiple types of biological impurities. Therefore, it is essential to ensure that the purified products are inherently EVs without other contaminants before performing any functional analysis of the EVs. The current EVs isolation and purification methods and their advantages and disadvantages are summarized in Table 2.

TABLE 2. The advantages and disadvantages of the mainstream EVs isolation method.

2.3 Storage of NEVs

As a promising cell-free therapy, achieving long-term and stable storage of NEVs plays a key role in their clinical application. Therefore, it is necessary to explore preservation techniques to protect the biological activity of NEVs for transport and clinical applications. Common conservation techniques include freezing, lyophilization, and spray drying (Kusuma et al., 2018; Charoenviriyakul et al., 2019; Zhang et al., 2020a). Any frozen storage may “frostbite” the EVs, and the use of antifreeze may extend their shelf life. The traditional approach of adding DMSO during cryopreservation can protect the biological activity of EVs (Wu et al., 2015). Furthermore, alginate is considered as the most effective disaccharide antifreeze agent and prevents EVs aggregation and increases its stability without changing EVs morphology (Bosch et al., 2016; Charoenviriyakul et al., 2018). A recent study reported that PBS supplemented with human albumin and trehalose buffer significantly improved the short and long-term preservation of EVs samples stored at −80°C, and maintained stability over multiple freeze-thaw cycles (Gorgens et al., 2022). Moreover, the storage of EVs also varies from different sources and modifications (Agrawal et al., 2017). Multiple studies on NEVs have analyzed the impact of storage on NEVs and concluded that the existing technology could effectively ensure the storage stability of NEVs (Jong et al., 2017; Farcas and Inngjerdingen, 2020). In summary, the rational use of various EVs storage methods can significantly improve the storage stability of EVs and provide greater application benefits.

3 Function mechanisms of NEVs

In recent years, as research on NEVs have continued, knowledge about the mechanisms underlying their function has been gained. Several studies have confirmed the ability of NEVs to target and kill tumor cells, which have been summarized in Table 3. This section highlights the characteristics and mechanisms of the currently known NEVs in oncology therapy.

TABLE 3. Existing studies for NEVs.

3.1 NEVs exert antitumor effects through their contents and membrane proteins

NEVs contain many substances acting as tumor killers, such as membrane proteins, toxic proteins, and miRNAs. In this section, we will elaborate on each section individually (as shown in Figure 2).

FIGURE 2. The surface receptors of NEVs, loaded cytotoxic proteins, and functional miRNAs induce apoptosis in tumor cells. In addition to stimulating the polarization of macrophages towards M1 and activating T cells directly or via activated monocytes, NEVs can also activate resting NK cells, thereby augmenting their tumor-killing ability.

Several studies have shown that NEVs can express NK cell surface receptors (Zhu et al., 2019; Choi et al., 2020). These receptors include Natural Killer Lytic-Associated Molecule (NKLAM), Fas-L, DNAX accessory molecules-1 (DNAM-1/CD226), and NKG2D (CD314) (Enomoto et al., 2021). The expression of the natural cytotoxic receptor (NCR), NKp44 (CD336), NKp30 (CD337), NKp46 (CD335), and CD16 varies according to cell sources and activation status (Lugini et al., 2012). Moreover, NEVs induce apoptosis through a classical ligand/receptor interaction between Fas-L on the membrane surface and Fas on the target cell membrane. Fas-L binding to the membrane receptor results in the formation of the death-inducing signaling complex (DISC), which activates the extrinsic apoptotic pathway by activating the caspase pathway (Lavrik and Krammer, 2012; Sparrow and Bodman-Smith, 2020). The CD47 expressing on the NEVs surface interacts with its receptor on macrophages, SIRP-α, to inhibit the elimination of NEVs by macrophage via phagocytosis, thus enabling longer cycle times (Jaiswal et al., 2009; Wang et al., 2019).

Perforin, granzyme A, B, granulysin, and tumor necrosis factor α (TNF-α) are all found in NEVs (Korenevskii et al., 2018; Cochran and Kornbluth, 2021). The perforin in NEVs can penetrate the cell membrane and allow cytotoxic proteins (granzyme A, B, granulysin) to enter the target cell and induce apoptosis by disrupting the outer mitochondrial membrane potential and cleaving caspases (Leon et al., 2017; Wu et al., 2019). Among them, granzyme B targets and cleaves cystathionine-3 and -7 directly, leading to the rapid initiation of apoptosis. It also induces an intrinsic apoptotic pathway by cleaving Bid to tBid (BH3 interacting domain death agonist protein), which disrupts the outer mitochondrial membrane potential and releases cytochrome C (MacDonald et al., 1999). The specific target of granzyme A in the apoptotic pathway is the SET complex, an ER-associated complex whose cleavage causes single-stranded DNA damage (Lieberman, 2010). Granulysin can induce apoptosis by binding to target cell membranes through electrostatic interactions based on its positive N-terminal charge. This process can disrupt cell membranes, active Caspase-9, and Caspase-12 by damaging mitochondria, as well as damaging the endoplasmic reticulum and activating Caspase-7 (Sparrow and Bodman-Smith, 2020). One quantitative analysis study demonstrated that NK-92 EVs revealed higher levels of perforin and Fas-L than NK cells and performed more effective inhibition of tumor proliferation (Zhu et al., 2017).

Regulatory miRNAs found in NEVs demonstrate tumor-killing and immunomodulation ability. Among these, miR-3607-3p encapsulated in NEVs inhibits cancer cell migration and invasion; miR-3607-3p-enriched NEVs may inhibit the malignant transformation of pancreatic cancer by directly targeting IL-26, and decreased miR-3607-3p levels were associated with poor prognosis and tumor metastasis (Sun et al., 2019). Another study demonstrated that miR-186-5p in NEVs can inhibit the growth and spread of neuroblastoma and induce apoptosis, and miR-186-5p containing NEVs was also taken up by NK cells to reduce the inhibition of cytotoxicity by the TME (Neviani et al., 2019). A recent study suggests that miR-10b-5p, miR-92a-3p, and miR-155-5p found in NEVs play a crucial role in immune regulation (Dosil et al., 2022). In addition, more information was summarized in Table 4.

TABLE 4. The functions of miRNAs contained in NEVs.

3.2 Immunomodulatory effects of NEVs

EVs secreted by numerous immune cells can be used to regulate innate and acquired immune responses (Lugini et al., 2012; Hong and Kim, 2022). NEVs possess similar immunomodulatory functions in the immune system (Figure 2). On the one hand, NEVs can effectively reduce the number of pre-tumor M2 macrophages or induce tumor-killing M1 macrophage polarization, which attenuates TAM-mediated CTL inhibition via the change of TAM -secreted cytokines and membrane surface proteins, and induces a direct antitumor effect of M1 macrophage (Bellora et al., 2014; Jia et al., 2020; Nie et al., 2021). NEVs can also act directly on T-cell activation or indirectly by stimulating monocytes to positively influence T-cell activation (Figures 3A–E) (Federici et al., 2020). Analyzing the miRNA types in NEVs reveals that they promote T-cell activation and induce DC expression of MHC-II and CD86 (Dosil et al., 2022). Furthermore, NEVs contribute to NK cell activation. A study demonstrated that NEVs pre-exposed to tumor cells could activate resting NK cells in humans, leading to higher levels of NCR and acquiring greater tumor-killing capacity (Figures 3F–H) (Shoae-Hassani et al., 2017). However, the immunomodulatory function of NEVs is still unknown due to the lack of in-depth mechanistic studies.

FIGURE 3. (A) NEVs activate monocytes. Flow cytometry analysis of CD80–CD86 geo mean fluorescence intensity (gMFI) of gated CD14⁺ cells in PBMCs cultured in the presence or absence of NEVs, and/or lipopolysacharide (LPS) for 24 h. Upper panels: Representative dot plots showing CD80–CD86 expression in the presence of NEVs, lower panel: Flow cytometry of human leukocyte antigen DR isotype (HLA-DR) gMFI of CD14⁺ gated monocytes (Federici et al., 2020). (B) Flow cytometry analysis of CD25 expression by CD3⁺ gated T cells in PBMCs evaluated after 72 h of culture with NEVs (Federici et al., 2020). (C) The graph shows the results obtained with PBMCs of different healthy donors (n = 3), in the presence or absence of transforming growth factor beta (TGFβ)/interleukin (IL)-10 (10 ng/ml each) (Federici et al., 2020). (D) NEVs affect the interaction between monocytes and T cells. Flow cytometry analysis of 72 h proliferation and CD25 expression by CD3, CD4, and CD8 T cells cultured in the presence of monocytes (medium), monocytes preconditioned with NEVs (Federici et al., 2020). (E) NEVs induce the release of cytokines by PBMCs. Cytometric bead array-measured cytokine production of 72 h PBMCs cultured (Federici et al., 2020). (F–H) Activation of resting NK cells by NEVs affects the expression of natural cytotoxicity receptors on their surface and tumor-killing viability (Shoae-Hassani et al., 2017). (F) NK cells were stained with specific NCRs monoclonal antibodies, a resting NK cell expresses different levels of NCRs. (G) The NEVs induce the expression of NCRs especially NKp44 similar to cytokine-activated NK. (H) In vitro cytotoxicity of peripheral blood natural killer cells against neuroblastoma (NB) cells. NEVs strongly stimulated NK activity in the presence of IL-21.

4 Characteristics of NEVs for therapeutic

4.1 Penetration

The TME and biological barrier, which are difficult to overcome in traditional tumor therapy, are important reasons that affect the therapeutic effect. NEV has a smaller molecular diameter and greater tissue penetration into solid tumors compared to whole cells. For example, the small molecular size of NEVs allows them to easily cross the blood-brain barrier (BBB) and enter the cancer reservoir to kill or deliver the drugs to central nervous system (CNS) tumors (Neviani et al., 2019; Weng et al., 2021; Choi et al., 2022b).

4.2 Natural targeting

Multiple studies have confirmed the ability of NEVs to target tumors (Lugini et al., 2012; Zhu et al., 2019; Sayitoglu et al., 2020). This characteristic may be attributed to the membrane proteins CXC receptors (CXCR3 and CXCR4), NCR, NKG2D, and DNAM-1 on the surface of NEVs, which can induce cancer cell lysis while targeting tumor cells (Deng et al., 2018; Wang et al., 2019; Di Pace et al., 2020; Sayitoglu et al., 2020; Aarsund et al., 2022). However, there is still some controversy about the targeting mechanism of NEVs.

4.3 Biocompatibility

NEVs enter target cells through micropinocytosis, and the number of internalized NEVs correlates with tumor cell cytotoxicity (Azarmi et al., 2020; Di Pace et al., 2020; Enomoto et al., 2021). It was discovered that co-incubating NEVs with target cells for approximately 30 min resulted in detection in target cells and induced maximal cytotoxic effects after 8–14 h (Zhu et al., 2017; Di Pace et al., 2020). Furthermore, it has been demonstrated that the acidic tumor microenvironment promotes the uptake of NEVs by tumor cells (Parolini et al., 2009; Fais, 2013).

4.4 Security

All current NEVs research has addressed the safety of NEVs at the cellular or animal level. Some studies point out that NEVs derived from PBMC have cytolytic activity against cancer cells but not against normal resting PBMC cells or normal cells (Lugini et al., 2012; Groot Kormelink et al., 2018). Furthermore, numerous studies have concluded that NEVs do not pose serious safety concerns in animal studies (Han et al., 2020; Kaban et al., 2021; Zhang et al., 2022). Existing research can provide initial confirmation of the safety of NEVs. However, because most of them are in vitro experiments, safety issues will become an issue that has to be addressed for the future development of NEVs.

4.5 Adjustability

EVs-secreting behavior of NK cells is independent of cell activation status (Fais, 2013; Aarsund et al., 2022). However, the killing activity of NEVs is closely related to the state of cell activation (Lugini et al., 2012). This process is regulated by many factors; for example, NEVs exhibited stronger cytotoxic effects and elevated levels of cytotoxicity-related molecules in hypoxic environments (Jiang et al., 2021). NEVs pre-exposed to the tumor environment may have higher cytotoxicity (Shoae-Hassani et al., 2017). Furthermore, the cytotoxicity of NEVs can be modulated by various cytokines (Markova et al., 2021; Aarsund et al., 2022). Namely, higher quality NEVs would be produced by appropriate regulation of NK cells based on the above characteristics in the future.

5 Engineering strategy for NEVs

Endowing EVs with a variety of fascinating capabilities via common engineering strategies has become a crucial direction for much applied research. Common EVs engineering techniques can be classified as endogenous and exogenous modifications. The primary objective is to increase the targeting ability or transform them into drug carriers (Gudbergsson et al., 2019; Zhang et al., 2020b). Existing engineering studies of NEVs are few but have shown satisfactory results.

5.1 Exogenous modifications

NEVs are suitable for drug delivery systems (DDS) due to their strong penetration, antitumor activity, and natural targeting. The current strategy is transporting exogenous drugs into NEVs, with engineering methods mainly including electroporation, ultrasonication, extrusion, freeze-thaw cycles, and saponin treatment (Thakur et al., 2022). NEVs were used in a study to enhance the antitumor effects of the drug by encapsulating paclitaxel via electroporation (Figure 1A) (Han et al., 2020). Another study reported that NEVs loaded with doxorubicin by ultrasonication demonstrated excellent antitumor activity against MCF-7 human breast cancer cells both in vitro and in vivo (Figure 1B) (Pitchaimani et al., 2018). A recent study using NEVs loaded with hydrophilic siRNA and the hydrophobic photosensitizer Ce6 showed obvious tumor-killing effects due to not only the anti-tumor property of the NEVs but also the combination of the powerful gene silencing effect by the delivery of siRNA and significant photodynamic therapeutic effects with reactive oxygen species (ROS) generated after laser irradiation (Figures 1C, 4A) (Zhang et al., 2022). The use of NEVs in drug delivery overcomes most of the drawbacks of conventional nanomaterial drug delivery systems.

FIGURE 4. (A) Schematic illustration of light-activatable silencing NK-derived exosomes (LASNEO) mediated synergetic tumor eradication. (B) Binding of NEVs to CLTs via click chemistry reaction. (C) Schematic design of the NN/NKEXO cocktail for tumor targeting and drug delivery. (D) Loading of BCL-2 siRNAs (siBCL-2) in NK-92MI-derived EVs by lentiviral transfection.

Surface engineering is another type of exogenous modification of interest. Surface engineering of NEVs can improve their targeting or binding to other substances, increasing their stability and duration of action in vivo. Introducing nanomaterials and inserting lipophilic components into the membrane by fusion with liposomes or adsorbing molecules is the main approach for the surface engineering of NEVs (Yang et al., 2021a). The linkage can also be formed through covalent bonds on the vesicle surface through azide-alkyne cycloaddition reactions (Richter et al., 2021). In a recent study, NEVs were modified with dibenzocyclooctynes (DBCO), and CTL was modified with azide groups, respectively, which were subsequently linked via biorthogonal chemistry. Due to the pH-responsive structure, the NEVs could be released at low PH, exploiting their ability to target tumors during circulation and promote CTL to kill tumors (Figures 1D, 4B) (Nie et al., 2021). Another study reported using cocktail therapy by combining NEVs with dendrimer core loaded with therapeutic miRNAs for tumor-targeted therapy (Figures 1E, 4C) (Wang et al., 2019).

5.2 Endogenous modifications

Endogenous modification is also an essential method for the functionalization of NEVs. It is intended to engineer the membrane and contents of NEVs by genetically modifying the parent cells expressing the specific target product or chimeric protein. A genetic engineering study of the NK-92MI using lentiviral transduction to express BCL-2 siRNA (siBCL-2), which is enriched in NEVs and successfully enhanced tumor-killing ability by inhibiting overexpression of BCL-2 in breast cancer (Figures 1F, 4D) (Kaban et al., 2021). A study reported that EVs isolated from mesothelin-targeted CAR-T cells maintained most of the parental cells’ characteristics and had the same therapeutic potential without significant side effects (Yang et al., 2021b). Moreover, the administration of CAR cell-derived EVs is relatively safer than CAR cell therapy (Fu et al., 2019). However, there still exists no detailed study using CAR NK-derived EVs. Take as a whole, genetic engineering enables good control over the generated EVs; once the corresponding cell line is established, no further work is required to generate the modified EVs, making it an ideal method for the mass production of engineered EVs in the future.

6 Discussion

This review highlights the challenges and potential of NEVs in cancer therapy, which has demonstrated tremendous advantages in recent years as an emerging cell-free therapy in cancer immunotherapy, including smaller size, greater tissue penetration, lower acquisition costs, and independence from inhibitory TAM compared to conventional NK cell therapies. The NEVs inherit the tumor-killing and natural targeting abilities of their parent cells. It is associated with relatively few immune side effects due to the absence of cellular involvement. Therefore, in addition to cell therapy, NEVs have the potential to play a crucial role in future tumor immunotherapy.

The role of NEVs is still poorly understood, and researchers continue to investigate it. One study showed that NK-92MI cells-derived EVs could inhibit TGF-β1-induced HSC proliferation and activation, preventing liver fibrosis by carrying miR-223 (Wang et al., 2020a; Wang et al., 2020b). Another study demonstrated that miR-207-containing NEVs alleviated symptoms of chronic mild stress in mice, suggesting that NEVs may also have a role in the treatment of depression (Wang et al., 2020a; Wang et al., 2020b). Moreover, NEVs ameliorated lung injury in a mouse model of Pseudomonas aeruginosa lung infection by promoting M1 macrophage polarization. This suggests that NEVs may play a protective role in inflammation, especially in diseases with an imbalanced M1/M2 macrophage ratio (69). As the in-depth functions of NEVs have not yet be investigated, following research may need to focus on the functional contained biomolecules and the critical roles in the immune regulation process.

Despite the large number of studies demonstrating the efficacy of NEVs in cancer therapy, the development of NEVs still faces significant challenges. Not only do NEVs face these challenges, but all therapeutic EVs developments must also overcome them. The first is the heterogeneity of EVs, which complicates quality control and hinders a comprehensive understanding of their function. The main reason comes from the cell source and isolation methods for the production of EVs. The optimal cell source and isolation method for EVs is still under investigation. Ultracentrifugation is the most widely-used EVs isolation method, which needs to be integrated with another isolation method to improve the separation purity. The second challenge is selecting designs that improve the cycling stability and the cytotoxicity of NEVs. As mentioned above, although various endogenous and exogenous modification methods are used, there is no effective method to improve the loading efficiency of bioactive molecules without compromising the integrity of EVs, and most modification methods may cause clustering. Furthermore, it is necessary to evaluate the need for these modifications and their improvement in therapeutic efficacy. A reliable method to determine whether the loaded EVs contain active molecules is still urgently needed (Wahlgren et al., 2012). Finally, due to the insufficient number of studies and inconsistent experimental conditions, there is no uniform standard on how numerous NEVs and how long it will take to achieve the desired anticancer effect, as well as what delivery method and treatment regimen should be employed during treatment to achieve improved clinical outcomes. Therefore, there is still much work to be done before the utilization of NEVs in clinical settings. Although the full-scale mechanism and function of NEVs need to be addressed, NEVs are a highly promising cell-free therapeutic option, which are easily to be obtained, modified, and stored in comparison with cells. The increasing number of studies on engineering NEVs have also proved it as an excellent vector for personalized modification. The unique anti-tumor properties of NEVs convince us that the anti-tumor strategy based on NEVs is worthy of comprehensive and in-depth study. Future research should take full advantage of NEVs and integrate it with multiple therapeutic strategies including sonodynamic therapy, photodynamic therapy, photothermal therapy and radiotherapy, so as to achieve more powerful tumor-killing effects. We hope this review will contribute to the promotion of multidisciplinary research on NEVs in a concerted effort to make NEVs the next-generation of cancer therapeutic strategy.

Author contributions

YQ and XZ, conception, design, and inviting co-authors to participate. YQ, YD, MW, JW, ZF, and QW, writing an original manuscript draft. JL, HY, and XZ reviewed and edited the manuscript critically for important intellectual content and provided comments and feedback for the scientific contents of the manuscript. All authors contributed to the article and approved the submitted version.

Funding

This work was supported by the National Natural Science Foundation of China (82002250), Sichuan Science and Technology Program (2022YFS0201), General Program of Chongqing Natural Science Foundation (cstc2021jcyj-msxmX0687).

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Abbreviations

BBB, blood-brain barrier; CAR, chimeric antigen receptor; CNS, central nervous system; CTL, cytotoxic T lymphocyte; DBCO, dibenzocyclooctynes; DDS, drug delivery systems; EMs, exosome-mimetic vesicles; EVs, Extracellular vesicles; HSPCs, hematopoietic stem and progenitor cells; iPSCs, induced pluripotent stem cells; miRNAs, microRNAs; NCR, natural cytotoxic receptor; NEVs, natural killer cell-derived EVs; NK, natural killer; NKLAM, Natural Killer Lytic-Associated Molecule; PBMC, peripheral blood mononuclear cells; ROS, reactive oxygen species; siRNA, small interfering RNA; TAM, tumor-associated macrophages; TME, tumor microenvironment; TNF-α, tumor necrosis factor α; UCB, umbilical cord blood.

References