Cancer vaccines, typically given through exogenous administration of selected tumor antigens combined with adjuvants that activate dendritic cells (DCs), ultimately aim to stimulate the immune system against tumor antigens. In recent years, the success of therapeutic cancer vaccines has dramatically revolutionized cancer treatments and has provided more alternative approaches to patients with cancer (1, 2). Several clinical trials have also confirmed the potential of this strategy, including those for melanoma, lung cancer, and prostate cancer (3–5). With the discovery of neoantigen target being continually advanced to improve the identification of immunogenic neoepitopes that can be recognized by CD8+ T cells, personalized therapeutic cancer vaccines have been considered in clinical investigation (6). The potential of therapeutic cancer vaccines in tumor immunotherapy has led to growing interest in developing more effective vaccine strategies to achieve tumor regression. However, recent years have seen excellent advances in self-assembled nanoparticle (SANP) platforms for therapeutic cancer vaccines. SANPs can evoke the epitope-specific B/T-cell immune responses through epitope folding similar to microorganisms that can protect payloads from enzymatic degradation, deliver to lymphoid organs, and target the tumor (7, 8). Additionally, antigen loading is also increased with the improved performance of SANPs. Because of the advantages from SANPs, they have naturally become the focus of current vaccine research and have appeared in various medical fields.

As early as the 1950s, rod-shaped particles have been found in tobacco mosaic virus (TMV); this is one of the earliest discoveries in history about self-assembled particles (9). In 1981, Drexler proposed the molecular engineering of proteins, being the monomeric building blocks, through self-assembly (10). Since then, materials based on self-assembled peptides such as nanofibers have been produced (11). Indeed, with the development of nanoparticle (NP) technology, including poly(lactic-coglycolic acid) (PLGA) (12), polymersomes (13), and liposomes (14), the design and manufacture of self-assembled nanovaccines are an empirical process. SANPs are composed of multi-component homologous NPs through non-covalent bonds or weak covalent bond interactions aimed at achieving a stable and balanced state (15). The assembly of different or similar molecules driven by intermolecular forces, including van der Waals interactions, electrostatic interactions, hydrogen bonding, coordination, hydration forces, and solvation interactions, occurs if there is equilibrium between them (11, 16). While these forces are indeed weak when considered individually, mixtures of different interaction types can result in structural and chemical stabilization, and these conformations formed by the free assembly of molecules are easy to accept. Of course, the modified module can also be selected as the substrate, which can control the direction of self-assembly to a certain extent and can directly act between the molecules or the module by external stimuli, including pH, temperature, solvent polarity, electromagnetic radiation, and light, to guide self-assembly (17). NPs designed through synthetic self-assembled technology (SANPs) can be given certain physiological functions. They can be used as platforms to display the arrangement of specific immunogens and orderly matrices to mimic the folding and complex structure of natural microbial surfaces (18). Therefore, SANPs can offer potential advantages such as improved stability due to the co-delivery of an antigen with adjuvant, and constant release and/or specific activation of the immune system (19). To a certain extent, we believe that the limitations of clinical application caused by the defects of the conventional vaccine platforms can be alleviated by SANPs.

The development of SANPs has led to many vaccine platforms and drug delivery systems (17), including broad platforms such as peptides/polymers (20). Generally, antigen loading is chemically defined, and SANPs based on peptide antigens are linked to hydrophobic carriers like peptide, lipids, and polymers. Of course, we also introduce a new way of antigen loading here. In this review, we discuss the characteristics of the internal structure of vaccines based on various self-assembled NPs, such as polymers and lipids for inducing anticancer T-cell immunity, and some SANP-based delivery platforms. Additionally, we also discuss current challenges of therapeutic cancer vaccines and their potential applications when combined with other treatments for overcoming tumor resistance and promoting clinical efficacy.

There are two ways to achieve nanotechnology: top-down and bottom-up (21). The top-down method dissociates NPs from larger structures while maintaining their original composition and properties; the bottom-up method uses assembly or self-assembly to design NPs at the basic molecular level. To design an appropriate antigen carrier that delivers the screened pathogen epitopes and can induce a corresponding response in animal immunity studies model (22, 23). For example, in the PGS-co-PEG NPs system, Ankur and his colleagues synthesized the PGS-co-PEG polymer through a two- step polycondensation reaction (24). The first step, polycondensation reaction of PEG and sebacic acid under stirring condition. In the second step, glycerol was added to obtain a block copolymer of PGSco-PEG (pre-polymer) with different ratio of PEG segments.. The size of NPs can be controlled by controlling the molar ratio of PEG, thereby synthesizing precise nanostructures. To a certain extent, the size of NPs determines the binding and activation of their membrane proteins and receptors (25). Different immune cells have their corresponding SANPs from different dimensions. SANPs of size > 100 nm are mostly absorbed by local antigen-presenting cells (APCs) at the site of injection, as well as tend to accumulate in highly permeable tumor tissue, when employed in prophylactic and therapeutic immunotherapy (26). In contrast, smaller nanostructures (40–50 nm) tend to enter lymphatic vessels and reach lymph nodes (LNs); they can be internalized with greater efficiency by LN-resident DCs, triggering an effective and long-lasting immune response against tumor (26, 27) ( Figure 1C ). The peptide self-assembly process and the size of its production are strongly influenced by amino acid sequences and physicochemical indicators, such as pH (from acidic to slightly alkaline), temperature (4–95°C), metal ion concentrations (such as calcium, potassium, magnesium and sodium), and salt concentrations (28). SANPs can become structurally stable if the environmental parameters are obtained (26). The self-assembled form can be achieved by adjusting the concentration of metal ions. Changing the MgCl₂ concentration, for example, could encourage carbonic anhydrase (BCA) linked with P114 peptide (BCA-P114) NPs to self-assemble (26). NPs can be obtained by using MgCl₂ and Tris-HCl as low as 5 mM and 10 mM at neutral pH. The maximum size of these NPs can be formed when MgCl₂ concentration is 25 mM. When the concentration of MgCl₂ exceeds 25 mM, the particle size decreases (28). In another study (29), a self-assembled nanovaccine based on synthetic peptide conjugation of IKVAV-PA and OVA shows that different concentrations have an effect on vaccine characteristics including size, zeta potential, and encapsulation efficacy. Therefore, optimization of the formulation that is used for chemical reaction is a critical step for producing the best-performing SANP.

To summarize, one of the important functions of SANPs as tumor antigen carriers is to effectively deliver tumor antigens to LNs and then induce durable and effective specific-tumor immunity and immune memory. Aside from the fact that the encapsulation process protects antigens from the effect of phagocytosis or protease degradation, facilitates longer circulation time, and can generate effective immune responses (30–32), modularization with surface-attached antigens often elicit superior immune responses in comparison to encapsulated antigens perhaps owing to intracellular processing (33). Thus, composition, size, shape, and surface characteristics are key factors in ensuring stable delivery and inducing an immune response. Better biocompatible materials need to be selected to prevent their rapid removal. Particle size, as well as surface area (related to the load of the antigen), plays a critical role in DC phagocytosis and immune induction, which can be determined by controlling reaction conditions. During the design, attention should be paid to the control of chemical reaction conditions, while avoiding the infiltration of impurities, which may cause toxicity or damage the stability of the structure.

We can classify SANPs for vaccine design from multiple perspectives. From a general point of view, we can classify them according to the method of self-assembly: dynamic self-assembly and static self-assembly (15). According to the constituent elements, SANPs can be divided into protein-based biological entities and non-protein-derived synthetic particles (23). In this review, we briefly introduce several vaccine designs based on the different components of SANPs ( Figures 1A, B ; Table 1 ).

The regulatory mechanism of adjuvants is related to the direct or indirect stimulation of APCs (60). The maturation of DCs stimulated by adjuvants and the immune response to host antigens are two major prerequisites for therapeutic cancer vaccines to work (102) (Figure 1D). DCs are an important messenger between CD4+ T helper cells and CD8+ T cells, and are associated with triggering further adaptive immune responses (103). The mechanistic routes of adjuvants were summarized by Schijns (1): adjuvants facilitate antigen uptake, presence, and transport by antigen-capturing and -processing cells; (2) adjuvants prolong the time of antigen storage and release; (3) adjuvants activate innate immune cells to release cytokines through targeting the pattern recognition receptor (PRR); and (4) adjuvants mimic danger signals to stimulate the APCs (104). Toll-like receptors (TLRs) are a popular target for adjuvant research, because of their involvement in identifying pathogen-associated molecular patterns (PAMPs) (60). In the process of the adaptive immune response, adjuvants can fully stimulate CD4+ T cells, but the initiation of CD8+ T cells requires more complex immunological activities (60).

Conventional adjuvants like aluminum salt have a certain effect on inducing T helper 2 (Th2) cell responses, not Th1 or cytotoxic T lymphocytes. Therefore, it is assumed that this type of adjuvant would fail to activate the immune response required to cause tumor killing (105). Live-attenuated viral vaccines generally do not need an adjuvant because they can mimic a natural infection. Therefore, virus-like NPs can be regarded as an exogenous antigen to be presented by MHC class II molecules, which can also combine with MHC class I molecules by cross-presentation to activate humoral and cellular immunity (106). In general, self-adjuvanticity of SANPs relies on the formulation characteristics of the particulate’s structure. Because SANPs have a bio-particulate structure that is usually taken up efficiently by DCs, these SANPs with a perfect phenotype are often preferred as adjuvants of cancer vaccines and have been applied in multiple immunotherapy programs involving vaccines. A variety of SANPs have been reported to induce powerful antigen-specific T-cell responses through targeting TLRs. Yoshikawa et al. demonstrated that poly-γ-glutamic acid could activate APCs and induce a potent antigen-specific T-cell response through the TLR4 and MyD88-dependent signaling pathway (107). Mesa et al. found that TLR4 on DCs can be activated by small-sized proteoliposomes (VSSP), which are produced by the hydrophobic interaction of GM3 ganglioside with the meningococcal outer-membrane protein complex (108). Additionally, Luo et al. developed heterocyclic lipid SANPs with self-adjuvanticity that could activate type I interferon-stimulated genes by inducing STING activation but not the TLR or MAVS pathway (109). The potent T-cell activation with checkpoint inhibition showed great synergy with 100% survival over 60 days in a TC-1 tumor model. Furthermore, maintaining safety and reducing the vaccine dose are important factors for cancer patients who can benefit from the advancement of SANP technology.

Notably, the potential of SANPs can act both as delivery systems for vaccine antigens and as immunomodulators as previously reported (110). Several in situ vaccine (ISV) approaches have been suggested, with vaccine regimens including SANPs, Flt3L, and TLR agonists (111–113). However, the development of nanomedicine has fully demonstrated the advantages of SANP with both physiological activity and physical properties. A semiconducting polymer nano-immunomodulator (SPNI) has been previously reported (114), which is self-assembled by a polymer conjugated with a TLR7 agonist via an acid-labile linker. The semiconducting polymer contributed to tumor eradication and immunogenic cancer cell death through a near-infrared (NIR) absorbing to exert photodynamic effects. This way, the synergistic action of released immunogenic factors and acidic TME-activated TLR7 agonist can not only kill tumor cells directly but also serve as an in situ generated cancer vaccine to evoke strong antitumor activities. This SANP-based design refreshes the acquired route and presentation of traditional vaccines and improved therapeutic cancer vaccines.

Reports in recent years have shown us a variety of strategies that try to improve the immunogenicity of vaccines by using the powerful immune stimulation of SANPs in the tumor microenvironment. However, the great obstacles to the production of the specific killing effect of cancer vaccines include the genomic and phenotypic heterogeneity of cancer and the immunosuppressive effect of the tumor microenvironment. Furthermore, intertumoral and intratumoral heterogeneity provided a basis for therapeutic resistance (115). The advent of personalized therapeutic cancer vaccines provides a strategy by targeting personal tumor antigens (neoantigen), and the success of the neoantigen platform was achieved by tumor mutational burden (TMB), which is the total number of mutations per coding area of a tumor genome (116, 117). Specifically, the clinical significance from immune checkpoint inhibitor therapy has also been shown to correlate with the higher TMB, within several tumors including NSCLC, small cell lung cancer, melanoma, and colorectal cancer (118–123). It is reasonable to assume that high TMB induces high densities of neoantigen-specific tumor-infiltrating lymphocytes, leading to tumor cell secretion of IFN-γ and upregulation of PD-L1, while their relationship across the entire human cancer spectrum remains unclear (124, 125). If the combination of the SANP platform and a neoantigen is feasible, then it will bring unexpected results. However, considerable challenges to this approach include the cost and time required to achieving this high degree of personalization.

Today, most of the existing production processes of self-assembled nanovaccines are synthetic methods or semibiological technologies based on certain specific cells (101). The potential advantages of these strategies include reduced production time and cost. However, chemical synthesis and semibiosynthesis will limit the preventive and therapeutic effects of vaccines, which must face the difficulty of displaying versatile antigens on these proteinaceous NPs, especially glycan antigens with complex structures. Of course, the toxicity of individual materials is still present. A series of local events are generally considered as tissue or cell response continuum, such as injury by injection or implantation, acute inflammation, chronic inflammation, formation of granulation tissue, severe cytotoxicity, foreign body reaction, and fibrosis following implantation of microspheres (59, 151, 152). The method of using protein-based fully biosynthetic vaccines is still being explored, and this strategy still faces many difficulties related to antigen display. The preparation process of some SANPs used for antigen and drug delivery involves a series of chemical reactions; interestingly, the molecular interaction between the drug and the carrier does not break its inherent activity with the self-assembly process (17, 153, 154), which is conducive to the carrier and cargo fully playing their respective roles in vivo.

Future research on therapeutic cancer vaccines will focus not only on individualization of antigen for every tumor, but even more on the proper vaccine platform to maximize their impact. Breakthroughs in antigen delivery platforms have increased our capacity to personalize vaccines. Several supramolecular assembled programmable nanomedicines reported in recent years to improve immunotherapy efficiency and the emerging technology have made it possible to dissect the TME in-depth (113, 155, 156). To further inform the design of future treatment platforms, mechanistic analyses can be collected from previously failed and some clinically significant vaccine trials. Many ongoing studies using a range of different vaccine delivery platforms and combination therapies to alleviate tumor resistance have the ultimate goal of inducing effective, long-lasting, tumor-specific immunity in cancer patients.

SANPs composed of each material have specific advantages. For example, polymers have good stability and excellent biocompatibility, and peptide NPs are expected to achieve comprehensive biosynthesis in the future. The high loading, low toxicity, and high biocompatibility of most SANPs make them more valuable in vaccine development than traditional adjuvant vaccines. At the same time, vaccines based on various NPs also have shown good results in animal models, which provides a huge impetus for the development of many vaccines against complex diseases in the future. Although many satisfactory experimental results have been produced in the field of self-assembled nanovaccines, the current understanding of the physical and chemical characteristics of nanomaterials and their interaction with physiological systems is limited, and the optimal synthesis, synthesis of nanomaterials, and chemical modification are unclear (157). To provide better protection, a self-assembled nanovaccine must be able to induce a series of immune responses. Although existing functionalized NPs can be internalized by immune cells, the effect of the sites of many functional proteins for the uptake of APCs is still unknown (158). It should be recognized that the idea of carrying antigens on the surface of SANPs to induce immune responses proves to be promising in the least. In the future, we are expected to load more antigens on the surface of multivalent NPs to drive specific and strong CD4+ T-cell and/or CD8+ T-cell responses. If an NP platform can resolve distribution barriers in the future, the age of enlightenment of nanotherapeutics is nearing its end.

TS and FM conceived the structure of this review. TS wrote and edited the manuscript. MS collected and prepared the related figures. FM and CL reviewed and made revisions to the manuscript. All authors contributed to the article and approved the submitted version.