Abstract
Dendritic cells are professional antigen-presenting cells and the most potent stimulators of various immune responses, such as antitumor responses. Modern studies have not shown an effective antitumor immune response development in patients with malignant tumors. The major cause is the decrease in functional activity of dendritic cells in cancer patients through irregularities in the maturation process to a functionally active form and in the antigen presentation process to naive T lymphocytes. This review describes the main stages of cellular antitumor immune response induction in vitro, aimed at resolving the problems that are blocking the full functioning of dendritic cells, and additional stimulation of antitumor immune response.
Introduction
Currently, progress in the treatment of oncological patients is associated with a significant breakthrough in molecular biology, immunology, and biotechnology, as well as with an enhanced understanding of the causes of tumor growth and the pathogenetic mechanisms of malignant transformation. 1 The body has a number of mechanisms and systems to suppress the emergence and development of tumors. These include the immune system that recognizes and eliminates cells expressing antigenic markers that differ from normal human tissues.
Recent studies have reported that protective antitumor immune responses do not develop in response to malignant cells. 2 This may be related to several factors: low immunogenicity of tumor cells, ability of the tumor to induce local or systemic immunosuppression by various factors (interleukin-10 (IL-10), transforming growth factor β (TGF-β), vascular endothelial growth factor (VEGF), indoleamine-2,3-dioxygenase (IDO), and prostaglandin E2 (PgE2), decreased T cells activity, and the disruption of tumor-associated antigen (TAA) presentation.2–4
Currently, one of the most promising approaches for the treatment of oncological patients is the selective activation of T cell antitumor immunity using immune cells. Cellular immune response efficiency depends on antigen capture, processing, delivery to lymph nodes, and presentation to effector cells of the adaptive immune system. The main problem of research investigating the delivery of tumor antigens to immune cells to modulate their response is how to predict the amount of antigen captured by immune cells that is required to trigger antitumor responses. This primarily applies to the separate use of recombinant peptides and proteins, tumor cell lysates, and DNA/RNA constructs. In this regard, a delivery method that allows control of these processes, and, therefore, control of the immune response, is of great importance. Dendritic cells (DCs) stimulate immune responses by effectively inducing tumor-specific effector T cells (Figure 1), which reduce the tumor size and generate immunological memory, thus controlling possible tumor recurrence. 5

Putative effects of DCs on the development and functions of antigen-specific T cells.
The functional activity of DCs in oncological patients is significantly reduced.5,6 DCs showed abnormal maturation and presentation of antigen to naive T cells.7,8 Therefore, the generation of functionally active DCs in vitro constitutes a promising approach in the development of DC-based anticancer vaccines to mobilize patient’s defense systems, because their activation by tumor-specific antigens to induce cytotoxic responses and their increased efficiency of antigen presentation to induce cytotoxic T lymphocytes (CTLs) via costimulatory molecules and cytokines can be controlled.9,10 Therefore, the above mentioned research areas form separate stages that underlie technologies for the induction of a cellular antitumor immune response in vitro and include the generation of mature DCs, delivery of tumor antigens to DCs, stimulation of a cellular antitumor immune response, and evaluation of the response efficiency. This review will compare the different methods of tumor antigen delivery and the evaluation of cytotoxic response induction efficiency in oncological patients.
Types of tumor antigens
The role of the innate immune response in the development of spontaneous tumors has been known for a long time. However, the immune system is currently believed to be able to control the cancer process, both in mice and in humans.11,12 The presence of preexisting cellular (cytotoxic CD8+ T cell and CD4+ (T helper) Th1 responses) and humoral (antibodies) antitumor immune responses against tumor cells13,14 as well as the discovery of TAAs have initiated research in the field of immune response modulation to develop various antigen-specific immunotherapeutic approaches for cancer treatment.
An effective strategy for antitumor immunotherapy depends on the selection of an antigen, its form, and a method of antigen delivery. The tumor, as a biological object, is not limited to parenchymal components (tumor cells and tumor stem cells) because it contains a large group of normal cells, including stromal and immune cells, extracellular matrix, vascular components, mesenchymal stem cells, and different precursors.15,16 Together, they create the tumor microenvironment and provide, in concert with each other, conditions for tumor growth and progression. Therefore, the tumor, being a non-homogeneous structure, contains multiple antigens. The structure of an antigen used to load DCs is important when planning an immunotherapeutic vaccination strategy because it determines whether the antigen can be presented in a complex with major histocompatibility complex (MHC) I and II class molecules to induce CD8+ and CD4+ T cell responses. However, other factors should be taken into account including the amount of antigen for loading, load efficiency, and the length of antigen persistence and presentation. 17
TAAs are tumor cell antigens that are absent or weakly expressed on normal tissues in the adult organism.12,18 Modern methods of molecular identification and characterization have been used to characterize the following subclasses of TAAs:19,20
Oncofetal antigens are highly expressed during fetal development but suppressed after birth. Expression of these molecules is restored in tumors, while normal cells remain negative for these antigens. This group includes the oncofetal antigen/immature laminin receptor, glypican 3, alpha-fetoprotein, and carcinoembryonic antigen (CEA).
Differentiation antigens are tissue-specific proteins or glycoproteins synthesized in normal tissues during histogenesis at certain developmental stages, including tyrosinase, glycoprotein 100, melan-A/melanoma antigen recognized by T cells (Melan-A/MART-1), prostate-specific antigen, and prostate-specific membrane antigen.
Testicular cancer antigens are molecules that are expressed by many histological types of tumors, but not by normal tissues, except the testes (spermatogonia and primary spermatocytes). Because male germinal cells do not express human leukocyte antigen (HLA) molecules, these cells cannot present cytotoxic epitopes to T cells and cannot be a target for an antitumor CD8+ immune response. These antigens include MAGE, G melanoma antigen (GAGE), and B melanoma antigen (BAGE).
Overexpressed gene products and universal tumor antigens are antigens that are identified via the overexpression of the appropriate genes. In normal tissues, these genes are usually expressed at low levels. Examples include p53, Her-2, CEA, and MUC-1.
Oncogenic viral antigens include E6 and E7 antigens of human papillomavirus serotypes 16 and 18 in cervical cancer, the Epstein–Barr virus in Burkitt’s lymphoma, and hepatitis B and C viruses, which are most often associated with cancerous transformation.
Unique antigens are the products of mutations and rearrangements of expressed antigens. Therefore, tumor antigens may be allocated into two groups based on their origin: unique (mutated) antigens (tumor-specific antigens) and common non-mutated self antigens (TAAs). The differences between these two antigen groups may be considered for immunity stimulation (mutated antigens) and tolerance management and autoimmune process stimulation (self antigens). 21 The development of tolerance to tumor antigens may be explained by the physiological mechanisms of immunological self-tolerance that prevent immune responses to autoantigens. However, each autoantigen has isolated determinants that are incapable of inducing the development of tolerance. For this reason, the determinants can contain immunogenic properties. If a tumor antigen is presented at a sufficiently high level, subdominant peptides of the antigen are also presented at a high level by MHC I molecules on the tumor cell surface. Therefore, peptides that overcome the T cell activation threshold can be recognized by the immune system. However, the abundance of subdominant epitopes presented in the MHC complex can cause a high expression of autologous proteins, which leads to the development of an autoimmune process upon triggering a T cell response. The current research goal is to determine the conditions to overcome immune tolerance without stimulating an autoimmune process.
Sources of tumor antigens to prime DCs
The discoveries of tumor antigens as well as a preexisting antitumor immune response have helped us to develop various immunotherapeutic approaches to treat cancer, including immune response modulation. It should be noted that the tumor antigen profile is not qualitatively or quantitatively stable at different tumor stages, and metastases may have a set of tumor antigens different from the primary tumor. This is a challenge for immune response modulation. The simultaneous use of several antigens as well as a large concentration of immunogenic TAA epitopes may help overcome immune tolerance and the potential antigenic variability of tumor cells.
Tumor vaccines are delivered using various methods, including the use of free peptides, tumor lysates, DNA or RNA vaccines, as well as DCs primed with various tumor antigens.22–27 All active immunotherapy strategies aim to induce a specific T cell immune response. 28 In this review, we describe the main sources of tumor antigens used to prime DCs, followed by activation of the cytotoxic phase of the immune response (Table 1).
Methods to deliver antigen material to DCs: advantages and disadvantages.
DC: dendritic cells; MHC: major histocompatibility complex; TAA: tumor-associated antigen; HLA: human leukocyte antigen; mRNA: messenger RNA.
DCs capture and process tumor antigens from various origins, 29 which enables them to effectively present antigens 30 and stimulate the differentiation and functional activity of tumor-specific effector T cells.31–33 The first step in this process is providing DCs with tumor antigens. 34
Regardless of the method of tumor antigen delivery, there is convincing evidence for the induction of a protective and therapeutic antitumor immune response without specific toxicity.
Tumor lysates
Because tumors have a heterogeneous structure, the surface of cells comprising the tumor bears an individual set of TAAs. The use of tumor lysate as a source of tumor immunogens has the potential advantage of stimulating a response against a variety of known and unknown TAAs in both the particular tumor type and the particular patient. This method enables the induction of a polyclonal immune response, stimulating both helper CD4+ and cytotoxic CD8+ immune responses, thereby reducing the risk of the tumor escaping immune surveillance. The use of a tumor lysate reduces the time and effort spent identifying and synthesizing immunodominant peptide epitopes, enabling DCs to process tumor antigens naturally. The disadvantage of this method is the limited amount of tumor material available for lysate preparation as well as the limited suitability of tumor cells derived from patients. 35 Vaccine efficacy depends on the concentration of immunogenic and immunosuppressive antigens in the tumor material. 27 Furthermore, the use of tumor cell lysate antigens does not cover the changing repertoire of tumor antigens, which occurs during metastasis as well as during specific chemotherapy and/or radiotherapy. Furthermore, this activation method is associated with the risk of autoimmune reactions because tumor materials may contain normal cells and antigens. However, autologous tumor cell lysate antigens presented to DCs stimulate specific CTL responses.31,32,36–38
Proteins
Another source of antigens for priming DCs is proteins, both synthesized in vitro and isolated from a tumor tissue. Direct cultivation of a full-length protein with DCs eradicates the need to choose a protein based on the MHC/HLA haplotype and the identification of individual epitopes. 39 This method has been successfully used in clinical trials of cell vaccines for lung and kidney cancers, lymphoma, and myeloma, where it has been reported to produce antigen-specific CTL responses.40,41 However, it should be noted that priming DCs with an extracellular antigen is not accompanied by the effective presentation of antigen epitopes in a MHC I complex or the subsequent stimulation of CD8+ CTLs. For this reason, “fusion proteins” are now used in addition to native proteins. Fusion proteins can promote more effective processing and even increase the immunogenicity of a protein. For example, fusion proteins containing the human immunodeficiency virus TAT (trans-activator of transcription) protein improve cellular penetration. 42
Peptides
The development of synthetic peptides has improved immune response against tumor tissues. The use of synthetic peptides enables the direct loading of antigens into MHC I or II class complexes, depending on the epitope, and the induction of epitope-specific T cell responses. 43 This method also reduces the risk of autoimmune reactions because it only uses tumor epitopes that lack cross-reactivity to self tissues. A significant disadvantage of this method is the mandatory knowledge of antigen epitopes, the relevant human HLA type, and the amino acid sequence of a peptide. When using a peptide, it is impossible to predict the accurate processing by MHC molecules, therefore the presentation of essential antigens in their native form often fails. Furthermore, the use of peptides as an antigen source is limited to the size of molecules capable of penetrating cells or binding to MHC on the surface of antigen-presenting cells. 44 To date, peptide-DC therapies are limited to the stimulation of CD8+ cytotoxic responses.23,45–49 The problem of inducing CD4+ T cell responses that is also required to treat tumors remains unresolved.50,51
Therefore, when using lysate/protein/peptide-loaded DCs, the expression duration is limited not only by the affinity of a peptide to a MHC molecule but also by the half-life of a peptide–MHC complex and the turnover of MHC molecules. 52 Although priming DCs with short peptides can lead to their direct loading into MHC molecules on the cell surface, the use of full-length proteins or lysates requires their capture, processing, and presentation, which does not guarantee the successful loading of immunogenic epitopes to MHC molecules to trigger effective protective responses.
RNAs
When producing vaccines based on tumor material, researchers faced problems regarding TAA standardization and the availability of tumor cells. However, DCs can be loaded with tumor RNA because it can be obtained even when the isolation of a sufficient amount of tumor cells is difficult. This method is also convenient in situations where a tumor-specific antigen cannot be identified on the tumor cell surface. The literature has repeatedly demonstrated the ability of RNA-primed DCs to induce strong antitumor CD4+ and CD8+ T cell responses.53–56 Transfection with messenger RNA (mRNA), compared with peptide loading, avoids the limitations associated with the use of known TAAs and a matching HLA phenotype. In addition, small interfering RNAs targeting components of the immunoproteasome or mRNAs encoding costimulatory molecules can be additionally introduced into RNA to increase the number of transcribed derivatives of TAA peptides, which eliminates the need for an extra maturation period. This latter approach was evaluated in melanoma patients using a vaccine based on DCs transfected with mRNAs encoding immunostimulatory molecules CD40L, CD70, and, constitutively, Toll-like receptor (TLR)4 in a combination with mRNAs encoding tyrosinase, melanoma-associated antigen (MAGE)-A3, MAGE-C2, and gp100. Functionally active antigen-specific CD8+ and CD4+ T cells formed in most patients. The cells were found both in the skin and in the system.57,58
DNA constructs
The introduction of DNA constructs into DCs has several advantages compared with ex vivo peptide/lysate-loaded cells and tumor cells. 59 Through the use of special databases and software, only the most immunogenic epitopes of a tested TAA are introduced into DNA constructs, while insignificant latent epitopes are excluded. DNA constructs enable the targeted modulation of an immune response against tumor cells expressing a particular antigen. 60 In this case, several genes (including those differing in specificity to an HLA haplotype) encoding epitopes of different TAAs can be introduced into a single plasmid DNA. 26 Transfection of DCs with DNA constructs ensures the long-term expression of a tumor antigen, which allows DCs to present antigens over a long period. Following the introduction of the genetic material into DCs, an antigen is endogenously processed to be presented in a complex with MHC I class molecules, which leads to the stimulation of a cytotoxic antitumor immune response. After processing, the cell expresses antigens in their native form, which facilitates their processing and presentation to the immune system. 61 The use of polyepitope constructs enables the activation of several T cell clones to ensure a stronger immune response against different cancer cells comprising the tumor and also enables to overcome the possible loss of expression of a particular TAA in tumor cells. Additional HLA typing allows for the selection of only those epitopes that increase the chance of developing a specific CD8+ or CD4+ T cell response or both simultaneously. 28 For example, our laboratory demonstrated that the transfection of DCs with a universal plasmid and allele-specific plasmid encoding immunogenic epitopes of the HER2 protein increased the cytotoxic activity of mononuclear cells against the MCF-7 human adenocarcinoma cell line. At the same time, transfection of DCs with a plasmid encoding the full-length protein HER2 did not affect the cytotoxic effect.32,62 Furthermore, we demonstrated that DCs transfected with a polyepitopic DNA construct encoding epitopes of TAAs (CEA, MUC4, and epithelial cell adhesion molecules) stimulated the cytotoxic activity of mononuclear cells in colorectal cancer patients against autologous tumor cells and facilitated perforin accumulation in peripheral blood mononuclear cells. 31 In addition, a plasmid carrying TAA polyepitopes acts as a natural adjuvant (because of the presence of non-methylated CG sequences) and stimulates innate immunity.
In addition to sequences of certain immunogenic epitopes from tumor antigens, a plasmid DNA can contain constructs that affect the various stages of signal transduction between DCs and T cells, including the following:
Genetic modifications that provide antigen delivery for T cell receptor stimulation—this is achieved by using HLA-specific epitope sequences, enhanced endogenous antigen expression by DCs, and a sufficient and continuous delivery of a naturally processed antigen.
Genetic modifications to enhance costimulatory signals—this is achieved either through enhancing costimulatory signals or through suppressing the expression of inhibitory molecules.
Genetic modifications to improve the immune microenvironment—this is achieved by stimulating the secretion of Th1 cytokines (tumor necrosis factor-alpha (TNF-α), interferon-gamma (IFN-γ), IL-2, and IL-18), suppressing the activity of regulatory cytokines (TGF-β and IL-10), or changing the chemokine secretion. 63
Limitations of this approach are related to the search for an optimal method to deliver a DNA construct to DCs. The transfer of nucleic acids encoding tumor antigens into DCs is implemented by plasmid transfection (liposome-mediated transfection and magnetic transfection) using viral vectors or by electroporation.61,64 The main disadvantage of genetic constructs is a low efficiency (about 5%−20%) of their direct capture by DCs. 56 The delivery method used for gene transfer can significantly affect the transfection efficiency. 52 A low transfection efficiency was associated with a reduced ability of DNA molecules to reach the nucleus where transcription occurs. 65 In addition, physical methods can disrupt and change the function and phenotype of DCs or are toxic. 66 However, the use of electroporation demonstrated an increased transfection efficiency and cell viability. 66 Although viral vectors have a much higher efficiency to penetrate DCs (about 90%−100%), their use in DC vaccine trials is limited because of the potential viral infection of cells and viral vector integration into the genome.
DC vaccines have demonstrated good results at the preclinical trial stage, but further improvements will be associated with the expansion of their therapeutic practicality (applicability). Clinical trials largely use TAA-loaded DCs matured under the influence of a specific cytokine cocktail. These DCs are sufficient to activate a T cell response, but do not always provide adequate costimulation to increase and maintain the proinflammatory immune environment and to recruit a sufficient effector cells. As mentioned above, the existing tumor creates an immunosuppressive immune environment. Therefore, successful DC vaccines should prime a strong and persistent immune response after immunization under conditions of strong immunosuppression induced by tumor cells. Genetic modifications of DCs should ensure the continuous feeding of a naturally processed antigen and immunostimulatory molecules, as well as provide a stronger and more persistent immune response in vivo under tumor microenvironment conditions.
Given the presented data, we believe that the rational choice of a tumor antigen source for DCs should be based on the following factors:
Antigen molecules that contain epitopes are specific to a particular HLA type. Therefore, their use limits the development of an effective immune response, dependent on a human HLA type.
The use of certain antigen molecules restricts the repertoire of T cell clones and thus restricts the ability of the immune system to develop a strong polyclonal antitumor response.
A limited set of well-characterized immunogenic TAAs specific for each cancer type.
Biosecurity of a tumor antigen source in clinical practice.
The efficiency of antitumor immune response induction
The main task when developing therapeutic technologies for cancer treatment is to induce an effective cytotoxic and Th1 immune response. The main indicator of the successful induction of a response in vitro is cell-mediated cytotoxicity when DC-activated effector cells directly destroy tumor target cells.
Cytotoxicity can be evaluated using methods to detect specific cytotoxic molecules. These methods are based on the fact that cytotoxic CD8+ T cells and natural killer cells primarily use two contact-dependent cytotoxic mechanisms. The first is the exocytosis of lytic granules released by cytotoxic effector cells that produce a pore-forming toxin (perforin) and pro-apoptotic serine proteases (granzymes) that synergistically kill target cells through the activation of various lytic mechanisms. 67 In the second mechanism, effector cells produce molecules of the TNF family, such as TNF-α, Fas ligand, and the TNF-related apoptosis-inducing ligand (TRAIL), that induce the multimerization of appropriate receptors on target cells, which leads to apoptosis. 68 Of great popularity in basic research are methods that enable the simultaneous evaluation of the amount and functional activity of cytotoxic T cells, for example, ELISpot used to detect IFN-γ and granzyme B producing cells. 69 Another method for determining cytotoxicity is the identification of metabolic factors released during target cell lysis or produced by living cells only. The most popular test for the evaluation of cell-mediated cytotoxicity is the release of 51 Cr. However, it has several disadvantages: It is a semi-quantitative analysis with a low sensitivity level and some target tumor cells are poorly labeled and characterized by a high level of spontaneous isotope release. An alternative to radioactive compounds is the detection of lactate dehydrogenase enzyme released during target cell lysis.33,70
Conclusion
Currently, there is an intensive development of approaches to induce antitumor immunity using DC-based vaccination technologies described in this review. The main purpose of these technologies is the transfer of key vaccination events occurring in the patient’s body that result in the generation of antigen-specific T cells, to cell culture. In the future, this transfer will probably involve not only the formation of antigen-specific effector T cells but also the in vitro production of memory T cells.
Undoubtedly, each of the described technologies for the in vitro induction of a cellular antitumor immune response is important and requires further modernization in relation to various cancers. In our opinion, efforts should be primarily concentrated on improving polyepitope DNA constructs, which are the most promising vehicles to deliver antigens to DCs to stimulate antitumor immune responses. In addition, the in vitro stimulation of an antitumor immune response, which involves the selection and use of various effective stimulators and inhibitors of immunosuppressive molecules to shift the response toward a Th1 response, is also important. The generation and reproduction of clones of antigen-specific cytotoxic T cells and the generation of memory T cells for use in cellular immunotherapeutic strategies for cancer treatment are at the developmental stage. To date, a large number of clinical trials using technologies to initiate DC-induced antitumor immune responses have been conducted. 71 These are technologies that mainly use DCs loaded with tumor antigens in the form of a lysate or RNA from tumor cells as well as TAAs and peptides. Concurrently, the first reports of data on clinical trials using a cell protocol containing a “training step” for naive T cells to respond to a specific tumor antigen have been published.72,73 This is a very extensive topic that requires a separate analysis and discussion.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The work was supported by the Federal Target Program “Research and Development in Priority Areas of Development of the Russian Scientific and Technological Complex for 2014−2020” (Agreement No. 14.607.21.0043; Unique identifier: RFMEFI60714X0043).
