The issues of transplanted organ rejection and graft versus host disease (GVHD) reaction remain topical despite modern methods for antigenic matching of donor and recipient tissues and immunosuppressive drug therapy. For example, according to published data, 60 out of 315 kidney allografts developed an acute rejection reaction [1], and 1921 out of 2813 bone marrow cell transplantations were associated with an acute graft versus host disease reaction [2]. Allograft rejection is controlled by two main immunological mechanisms: innate nonspecific reactions with proinflammatory signals and donor-specific adaptive responses. The adaptive response results from the presentation of alloantigens by antigen-presenting cells, mainly dendritic cells (DCs), and their allorecognition by recipient T-cells [3, 4, 5, 6].

Currently available immunosuppressive drugs do not exhibit sufficient selectivity and their use may cause side effects including suppression of IFN-gamma production, hematopoiesis inhibition leading to the development of leukopenia, thrombocytopenia, anemia, and even pancytopenia, impairment of defense functions of the body, the activation of secondary infection, septicemia, and neoplasia [7]. Therefore, reducing the dependence on immunosuppressive drug therapy and inducing tolerance to donor antigens are the main purposes of studies on transplantation, and technologies for inducing antigen-specific tolerance are now considered innovative strategies to promote durable engraftment [8]. Various cell types are used to induce suppression, including mesenchymal stem cells in bone marrow transplantation [9], regulatory T cells [10], myeloid regulatory cells (macrophages, dendritic cells, myeloid suppressors) [11], regulatory B cells [12], and apoptotic donor cells [13]. Dendritic cells link innate and acquired immune responses and control immunity and tolerance, and therefore are key components for the modulation of graft rejection. To induce antigen-specific immunological tolerance, dendritic cells are loaded with an antigen to which tolerance is to be generated. Currently, a donor cell lysate [14] or vesicles derived from donor tissue [15] are used to load tolerogenic DCs with alloantigens. However, this method does not ensure the long-term presence of an antigen in dendritic cells; therefore, DC-induced antigen-specific tolerance may disappear when transferred to a recipient [16]. In this study, we investigated the potential of generating antigen-specific tolerance in stimulator–responder pairs using DNA constructs encoding major histocompatibility complex (MHC) epitopes of stimulators as an antigen source for dendritic cells as responder cells, as well as the induction of non-specific suppression associated with the expression of the immunoregulatory cytokine IL-10.

A tolerogenic response can be induced by suppressive cytokines, such as IL-10 and TGF-beta. One technique to preserve the immature DC phenotype after in vivo delivery is the genetic modification of DCs with viral vectors expressing IL-10 and TGF-beta, because the continuous expression of these cytokines prevents the maturation of DCs. In our study, the pIL-10 plasmid was used to induce a stable, nonspecific, and tolerogenic phenotype of dendritic cells. The supernatants of cultured dendritic cells were collected at 1, 2, 3, 6, 24, and 48 h after transfection to determine the content of IL-10 using an enzyme-linked immunosorbent assay. IL-10 production upon cultivation of transfected DCs increased as early as the first hours after transfection and up to 24 h. Cytokine production reached a plateau between 24 and 48 h (Fig. 1). The control plasmid p5 did not induce IL-10 production. Similarly, we previously showed that the data obtained in mouse models were consistent with previous studies using the plasmid encoding murine IL-10 [19].

Results of flow cytometry assessment of intracellular IL-10 in dendritic cells in 1, 2, 3, 6. 24 and 48 h after transduction (Fig. 2) showed that the relative number of cells expressing intracellular IL-10 increased significantly on hour 2 after transduction. The relative number of dendritic cells transfected by control p5 plasmid was under 5% during 48 h after the transfection.

The phenotype of DCs transfected with the IL-10-encoding plasmid was evaluated by the expression of surface markers including CD83, CD86, CD80, and CCR7 at 6, 24, and 48 h after transfection. CD80 marker expression was significantly increased on the surface of transfected DCs. CCR7 expression was increased on the surface of transfected DCs 6 h after transfection, after which it decreased but remained elevated compared with non-transfected DCs. Expressions of CD83 and CD86 were decreased on the surface of transfected DCs and remained stable for 48 h of cultivation. Therefore, a stable tolerogenic phenotype with preserved cell migration potential was produced in transfected dendritic cells (Fig. 3).

Transfected DCs were co-cultured with mononuclear cells at a 1:10 ratio 6 h after transfection, and the production of IL-10 and the number of T-regulatory cells were evaluated after 48 h of co-culture. Co-cultivation of MNCs and DCs transfected with the pIL-10 plasmid was accompanied by the stable production of IL-10, which promoted the maintenance of tolerogenicity. Co-cultivation of MNCs and DCs transfected with the IL-10 plasmid increased the number of T-regulatory cells, which promoted the maintenance of tolerogenicity (Fig. 4).

Therefore, the use of a DNA construct encoding IL-10 enabled stable cytokine production in DC cultures and the generation of DCs with a tolerogenic phenotype. To assess the maximal and minimal allogeneic responses, we conducted mixed lymphocyte cultures from 10 healthy donors and generated a panel of donor genotypes. To evaluate the level of response, stimulation index (SI) was calculated that is a result of dividing the average cpms of every donor to the average control cpms of the corresponding donor (Fig. 5). Control cpm was a result for the control probe that was lymphocytes without stimulation. A SI greater than 2 was considered as a max response. SI levels between 1.55 and 2 were considered as a medium response; SI levels below 1.55 were considered as a minimum response.

To identify which HLA genotypes were most active regarding the stimulation and response to alloantigens, genotype frequencies were determined in the first third of the values, and to determine the weakest genotypes in MLC, the genotype frequencies were determined in the last third of the values (Tables 1,2). Thus, we selected the best responder genotypes (ID 1, 2, and 3) and the best stimulator genotypes (ID 6 and 7). They formed potential responder–stimulator pairs: 1–7, 1–6, 2–7, 2–6, and 3–6, the DNA sequences of which were further used for the calculation and analysis of antigenic epitopes and subsequent synthesis of DNA constructs.

The results of DNA sequencing are presented as laboratory conclusions on the donor HLA genotype (summarized in Table 3) and as FASTQ data files from an Illumina MiSeq sequencer. The DNA sequence data were used to develop a software product, analyze sequences, and select epitopes for incorporation into genetic constructs.

A software product was developed for the generation of constructs, which enabled the personalized selection of major histocompatibility complex epitopes chosen based on a comprehensive analysis of HLA sequencing of the patient and the donor, and the theoretical design and synthesis of personalized DNA constructs for their transfection into dendritic cells to suppress the graft rejection reaction were performed. Then, artificial genes encoding target constructs were designed, the artificial genes were cloned into the pmax-Ub vector, and genetic constructs encoding target protein products were generated, followed by the amplification of genetic constructs encoding target protein products and their purification from endotoxins. Thus, personalized DNA constructs were generated for five responder–stimulant (recipient–donor) pairs, based on these sequences of the major histocompatibility complex.

Responder dendritic cells were transfected with: (1) DNA plasmids encoding epitopes of stimulator histocompatibility complex molecules; (2) an IL-10-encoding plasmid to induce nonspecific tolerance; and (3) two plasmids simultaneously. After 96 h, significantly higher FoxP3 expression in CD4${}^{+}$CD25${}^{+}$ cells was found in the culture of MNCs and transfected DCs compared with the initial mononuclear cell culture. Significantly higher IL-10 production was found in the culture of MNCs and transfected DCs than that of control MNC cultures not cultured with experimental DC groups. These findings indicate that the dendritic cells performed a tolerogenic function through the activation of regulatory T cells and their secretion of IL-10 (Fig. 6).

To assess the ability of cultures of transfected DCs and MNCs (MNCs+DCs) to restrict the activation of autologous MNCs in response to stimulation by MHC alloantigens, mixed lymphocyte cultures were performed. The cultures of mononuclear cells from responder donors were co-cultured with MNCs from stimulator donors pretreated with mitomycin C. After 96 h of co-cultivation, the proliferation level in cultures was evaluated, and the proliferation suppression index was calculated ([proliferation of MNC culture without allogeneic stimulation]/[proliferation of culture upon addition of mitomycin C-treated allogeneic MNCs from stimulator donors]). Therefore, the higher the suppression index, the lower the proliferative response of the culture to stimulation with alloantigens. The results showed that cultures generated using tolerogenic DCs transfected with MHC epitopes had a significantly greater ability to inhibit the proliferation of autologous MNCs in response to an allogeneic MHC stimulus. The results are presented in Fig. 7.

Dendritic cells are a heterogeneous population classified in humans into myeloid DCs, plasmacytoid DCs, and Langerhans cells, based on ontogenesis, phenotype, and functionality [20, 21]. Mature dendritic cells mediate immune responses in inflammatory conditions, while immune tolerance is induced by tolerogenic DCs with an immature phenotype.

The development of antigen-specific tolerance, especially for donor antigens, is an urgent problem of transplantology, the solution of which might reduce the immunosuppressive drug load and toxic effects on recipients as well as increase the overall survival of patients after transplantation [22, 23].

The HLA locus is the most polymorphic region in the human genome and is associated with various diseases. HLA matching is a key factor in improving patient outcomes after transplantation. HLA matching has traditionally focused on HLA-A, -B, and -DR loci in solid organ transplantation and on HLA-A, -B, -C, and -DRB1 loci in hematopoietic cell transplantation [24]. Because of the high diversity of HLA haplotypes, the search for a suitable donor usually begins with close relatives, in particular siblings. However, ideal donors are found only for 15%–30% of recipients, and unrelated donors are often involved in the search. Previously, matching for HLA-A, -B, -C, and -DRB1 (8/8) alleles was demonstrated to correlate with better survival rates, while mismatches were associated with an increased risk of acute GVHD and patient death [24]. Single -DQ mismatches probably have no significant effect on the outcome of transplantation, whereas -DP mismatches slightly increased the risk of acute GVHD without compromising overall survival [25].

Based on the mixed lymphocyte culture, donor pairs with a maximum response to alloantigens and donors whose cells maximally stimulated responses to alloantigens, were identified. Among these donors, stimulator and responder pairs with a maximum difference in HLA loci were selected. These donor and recipient haplotypes would normally have an extremely high risk of transplant rejection in the donor during the transplantation of solid organs or hematopoietic cells. A software product for the personalized selection of HLA epitopes was developed, which provides a comprehensive analysis of the recipient and donor HLA sequences, selection and comparison of main histocompatibility antigens of the donor and recipient, and, based on the selected different antigens, the selection of epitopes of these antigens for the generation of a personalized DNA construct. The generated DNA construct should ensure the recipient’s tolerance to donor cells; in this case, there was no direct contact with allogeneic material, which reduced the risk of infection and uncontrolled composition (lysates or vesicles).

Various agents (IL-10, vitamin D3, dexamethasone and rapamycin) have been used to induce sufficient numbers of ex vivo tolerogenic DCs for the effective suppression of GVHD [26, 27]. These DCs provide tolerance in culture without reference to a specific antigen. In this study, we compared the effects of DNA constructs on the development of nonspecific immunosuppression (DNA construct encoding IL-10) and antigen-specific tolerance (DNA constructs encoding major histocompatibility complex antigens), as well as their joint action. The stage of interaction with mononuclear cells in vitro suggests that tolerogenic DCs do not become immunogenic after their introduction in vivo. Using five genetic constructs (three responders and two stimulators), respondent tolerogenic DCs transfected with epitopes of the stimulator main histocompatibility complex decreased the proliferative activity of recipient cells in response to alloantigen (by almost 50%). In this case, the use of a construct encoding only alloantigens without the use of IL-10 suppressed proliferation in stimulator cells, and the maximum suppression of proliferation in this group was observed when cells from stimulator No. 6 were used. An analysis of IL-10 production in co-cultures of MNCs and DCs transfected with plasmids encoding IL-10 and major histocompatibility complex antigens revealed a similar level of cytokine production in all groups, regardless of the antigen specificity. An analysis of the T-regulatory cell number showed that only the simultaneous application of the IL-10-encoding plasmid and the plasmid encoding major histocompatibility complex antigens led to a significant increase in target cells compared with groups where only one of the plasmids was used. Tolerogenic DCs control the differentiation of T-regulatory cells that, in turn, modulate the phenotype and function of DCs. This feedback between tolerogenic DCs and T-regulatory cells is supported by the fact that weak, if any, CD80/CD86 co-stimulation is required to induce Foxp3+ T-regulatory cells, and that IL-10 and TGF-beta secreted by T-regulatory cells inhibit DC maturation and promote the maintenance of DCs in the tolerogenic state [28]. In our study, transfection of DCs with the IL-10-encoding plasmid led to the production of the target cytokine after 3 h and an increase during the first days after transfection. In this case, there was a decrease in the expression of costimulatory molecules CD83 and CD86, an increase in the number of T-regulatory cells, and suppression of mononuclear cell proliferation. Furthermore, IL-10 production by transfected DCs slightly exceeded the reference values provided by the manufacturer of kits for PHA-stimulated cultures, which indicates the stimulation of cytokine production within physiological levels without a potential destructive effect on organs and tissues of the potential recipient.

Therefore, the produced dendritic cells exhibited tolerogenic properties and might be used to develop antigen-specific tolerance. Since the formation of tolerogenic properties is not associated with a long and high-affinity interaction between DCs and T cells and actively involves feedback mechanisms, the use of a plasmid encoding MHC antigens and the absence of maturing stimuli leads to the development of tolerogenic DCs and to the manifestation of effects at a level comparable with the use of a construct encoding IL-10. Thus, the co-application of plasmids encoding IL-10 and MHC antigens does not lead to a dramatic change in the studied effects. On the basis of this study, we think that the generation of DNA constructs enabling antigen-specific tolerance is no less effective than constructs encoding only certain immunosuppression factors. Depending on the vector and promoter used, it is possible to generate a DNA construct with a very high production of the target cytokine, cell death, and immune response suppression caused by the hyperproduction of IL-10 or TGF-beta. Therefore, DNA constructs encoding antigen-specific epitopes will be free from this disadvantage.

In our previous study [19] using an in vivo mouse model, we reported the suppression of GVHD reactions in an allogeneic transplantation model and better engraftment of a transplanted allogeneic skin flap using nonspecific and specific DNA constructs encoding IL-10 or MHC antigens, respectively. In this study, we demonstrated that the effectiveness of antigen-specific DNA constructs encoding histocompatibility complex antigens from a donor can be preliminarily assessed in vitro, which, together with the data on in vivo use suggests the potential clinical use of this method. A preparation of transfected DCs and mononuclear cells can be produced before surgery or lymphadenopathy tested, and then used during further treatment. The main application of these DNA constructs is the transplantation of bone marrow, kidneys, and liver from a living donor, i.e., in cases where the donor is known and available, and surgery is elective rather than urgent.

We demonstrated in vitro that human dendritic cells transfected with DNA constructs against HLA stimulating epitopes exhibited tolerogenic properties and may be used to suppress proliferative activity against allogeneic cells. Thus, we proposed a perspective approach to the induction of antigen-specific tolerance, which should be further studied for possible use in clinical practice to prevent GVHD reactions and graft rejection.

DC, dendritric cells; GVHD, graft versus host disease; HLA, human leukocyte antigen; MHC, major histocompatibility complex; MNC, mononuclear cells (non-adherent PBMC fraction); PBMC, peripheral blood mononuclear cells.

JAS contributed to the experimental work (MLC assays and experiments with pIL-10-transfected DCs), data analysis and interpretation, and drafting the manuscript. JAL—experimental work (MLC assay to identifying the best stimulator-responder matches), JNK—experimental work (DC transfection), ANS—interpretation of the DNA sequencing data and calculations for the determination of the most active stimulator-responder pairs. MSK—data interpretation, drafting the manuscript, figure preparation. VVK—experimental work (flow cytometry). AZM—designed and produced plasmid DNA constructs. SVS—contributed to the conception, design, revision, and final approval of the manuscript.

Voluntary informed consent was obtained from all donors participated in the research. The study followed the principles outlined in the Declaration of Helsinki for all human and animal experimental investigations and was approved by the local Research Institute of Fundamental and Clinical Immunology (RIFCI) ethics committee (protocol №127 dated Feb 17, 2020).

The authors are grateful to the staff of the Dmitry Rogachev National Research Center of Pediatric Hematology, Oncology, and Immunology of the Ministry of Health of the Russian Federation (Moscow, Russia) for performing the DNA sequencing and to the Laboratory of molecular immunology, RIFCI, for all assistance provided. We thank Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript.

The research (MLC, culture work, functional assays) was conducted under the framework of State Assignment № 1021062512015-4 (FGMN-2021-0003). The steps of DNA transduction and HLA-A genotyping were sponsored by the grant of the Russian Science Foundation project no. 21-65-00004, https://rscf.ru/project/21-65-00004/.

The authors declare no conflict of interest. SS is serving as one of the Editorial Board members of this journal. We declare that SS had no involvement in the peer review of this article and has no access to information regarding its peer review. Full responsibility for the editorial process for this article was delegated to GP.