Background
Comprising mRNA or DNA antigen precursors, nucleotide vaccines are taken up by host cells, where their nucleotide sequences are translated into antigens intracellularly. Then, the antigens could be degraded by the proteasome in the cytoplasm to generate epitope peptides for MHC-I molecules and to activate cytotoxic responses. The extracellularly secreted antigens are phagocytized by antigen-presenting cells (APCs) to trigger T helper cell or humoral immunity simultaneously. The characteristics of nucleotide vaccines endowed them with the ability not only in prophylactic but also in therapeutic settings to treat diseases such as cancer. Many clinical trials of tumor nucleic acid vaccines are in progress, of which a therapeutic DNA vaccine for the treatment of cervical intraepithelial neoplasia (CIN), targeting the E6 and E7 antigens of the HPV 16 and 18 strains, has shown a positive effect in patients in a phase 2b trial [
1]. Besides, an individualized RNA mutanome vaccine successfully immunized melanoma patients, resulting in a good T-cell response and prolonging progression-free survival [
2]. The immunogenicity of nucleic acid vaccines is the key to success. Therefore, further improvement is highly desirable.
Improvement of immunogenicity can be achieved by nucleotide sequence optimization to increase mRNA stability and the expression of antigens [
3]. Epitope optimization could promote cross-recognition of wild-type antigens and break immune tolerance to wild-type antigens [
4]. Introducing MHC class I trafficking signals (MITD) or lysosomal / endosomal localization signals to assist the process of MHC class I or class II epitope presentation by dendritic cells (DCs) has also been applied to increase cellular immune responses [
5,
6]. Gene fusion to combine tetanus toxoid fragments with antigens has been widely used in both DNA and mRNA vaccine design, aiming at strengthening the immune response [
7,
8]. However, this approach may lead to a more potent tetanus toxoid fragment response since antigen dominance hierarchies shape CD8 T-cell phenotypes [
9]. As antigen expression of the nucleic acid vaccine is extremely low, selective targeting of antigens to APCs is the key to increase effective biological distribution and immunogenicity. APCs have specific receptors that could be targeted by receptor-specific monoclonal antibodies (mAbs) or natural ligands [
10]. Delivering antigens together with the interaction between natural ligands such as chemokines and specific receptors on the surface of APCs could realize APC activation, with the advantage of no or weak induction of immune responses of the chemokines themselves and potential adjuvant effects.
Chemokines and chemokine receptors contribute to leukocyte trafficking and recruitment to sites of inflammation. Indeed, XCL1, CCL3 (MIP1α), CCL5 (RANTES), CCL7, CCL20, CCL21, CCL22, CCL25, CCL27, CCL28, CXCL10 and CXCL13 have been demonstrated to significantly promote cellular and humoral responses when fused or co-delivered with antigens from viruses or tumors [
11‐
17]. However, systematic evaluation of the immune response of chemokines after fusing with antigens has not been reported. It remains unclear whether chemokines have different preferences for inducing cellular and humoral immunity. Thus, understanding the full landscape of the immune response shaped by chemokine-fused antigens will provide useful information for the development of chemokine-based preventive or therapeutic vaccines against viruses or tumors and promote the understanding of the coordination of different immune cell subpopulations.
In the present study, we used our DNA tech platform to individually fuse chemokines with the HPV16 E6 and E7 antigens and evaluated the cellular and humoral immune responses in mice after immunization. CCL11 was found to induce the strongest cellular immune response and displayed superior antitumor activity. Different from previous chemokine-fused DNA vaccines, a single immunization of 25 µg CCL11-E6E7 DNA vaccine removed large established tumor masses in multiple experiments, achieving an average tumor clearance rate of over 90%. Furthermore, this technique is fully applicable to mRNA vaccine platforms. Chemokines that are suitable for inducing humoral immune responses were also identified. Our results provided an important foundation for the use of chemokines in vaccine design.
Methods
Mice and in vivo electroporation
Eight-week-old C57BL/6N female mice were purchased from Charles River Laboratories Co., Ltd. and raised at Peking Union-Genius Pharmaceutical Technology Company (Beijing, China). The entire experimental process received approval and supervision from the Peking Union-Genius Institutional Animal Care and Use Committees (IACUC). The ethic number is JY23001. Mice were ordered at least one week before the experiment to adapt to the new environment. The mice were first injected with plasmid into the tibialis anterior (TA) muscle followed by an electroporation with 60 V voltage, 10 Hz frequency and 50 ms interval (TERESA).
Preparation of single cell suspension and flow cytometry analysis
The collected anticoagulated mouse peripheral blood was lysed the red blood cells by adding 1 × red blood cell (RBC) lysis buffer (Biolegend, cat. 420,301). The prepared peripheral blood mononuclear cells (PBMCs) were then subjected to stain and flow cytometric analysis. Single cell suspensions of the spleen and lymph nodes were prepared by direct grinding with a 100 µm cell sieve (Corning). For tumor tissues, they were minced into small pieces and then tissue dissociation solution was added according to the manufacturer's protocol (Miltenyi, cat. 130–095-929). Red blood cell lysis with 1 × RBC lysis buffer was required for the processed spleen. The presence or absence of red blood cells in the lymph node and tumor single cell suspension determines whether red blood cells were lysed. The processed single cell suspension was subjected to staining and analysis by flow cytometry. For the staining of intracellular transcription factors, the BD Fixation/Permeabilization Kit (cat. 554,715) was used. The staining scheme used to distinguish each subpopulation in this article was as follows: E749-57 specific CD8 + T cells (E749-57 tetramer + , CD8 +), MDSC (CD11b + Gr-1 + Ly-6G +), NK cells (CD45 + CD3 − NK1.1 +), CD8 + T cells (CD45 + CD8 +), CD4 + T cells (CD45 + CD4 +), Treg (CD45 + CD4 + Foxp3 +), M1 macrophages (CD45 + CD11b + F4/80 + CD206-), M2 macrophages (CD45 + CD11b + F4/80 + CD206 +). Flow cytometric data was acquired on a BD FACS Canto II flow cytometer (BD Biosciences) and analyzed with FlowJo 7.6.5 software.
Cell culture and subcutaneous transplantation tumor model construction
The 293 T (cat. CTCC-001–0188), TC-1 (cat. CTCC-400–0328), and B16-OVA (cat. CTCC-007–0623) cell lines were purchased from Zhejiang Meisen Cell Technology Co., Ltd. The 293 T and B16-OVA cells were cultured in DMEM medium supplemented with 10% fetal calf serum (FBS). TC-1 cells were cultured in RPMI1640 medium supplemented with 10% FBS. When the cells reached the logarithmic growth phase, TC-1 and B16-OVA cells were digested with trypsin into single-cell suspensions, and the cell concentration was adjusted to 2 × 105 per 100 µl and inoculated subcutaneously into mice.
Western blot assay
Twenty-four hours after the transfection of the plasmid into the 293 T cells, the cells were harvested and lysed with RIPA buffer (50 mM Tris–HCl, pH 8.0, 150 mM NaCl, 1.5 mM MgCl2, 0.1% SDS, 0.5% deoxycholate (DOC), 1% NP-40) supplemented with 1 mM PMSF, 1 × protease inhibitor mixture (Roche). After centrifugation at 12,000 rpm for 10 min at 4 °C, the supernatant was collected and SDS-PAGE loading buffer (5 ×) was added. The samples were boiled for ten minutes following a by separation on 10%SDS-PAGE gels. Then the samples were blotted using the anti-FLAG antibody (Sigma, #1804).
Cell depletion
After grouping mice according to tumor size, CD4 (BioXcell, cat. BE0119. 350 µg per mouse), CD8 (BioXcell, cat. BP0061. 350 µg per mouse), NK1.1 (BioXcell, cat. BE0036. 250 µg per mouse), and CCR3(BioXcell, cat. BE0316. 250 µg per mouse) blocking antibodies were administered intraperitoneally every three days for a total of seven times. The cell depletion effect was detected by FACS after the third administration.
Synthetic DNA fragments encoding the protein of interest were cloned into Genscript plasmid vectors containing sequences corresponding to the T7 promoter, a 5' untranslated region (UTR), a 3' UTR, and a 31 + 10 nt spacer + 71 nucleotide poly A tail. The maps and sequences are included in the supplementary materials. Quality control passed plasmids were linearized with the class -IIS restriction enzyme BspQI to generate a template with no additional nucleotides beyond poly A. Linearized plasmid DNA purified by ethanol precipitation was subjected to in vitro transcription (IVT) with T7 RNA polymerase (Genscript). In the presence of 10 mM N1-methylpsedouridine-5'-triphosphate, adenosine 5'-triphosphate, cytidine 5'-triphosphate and guanosine 5'-triphosphate. This IVT process also included a co-transcriptional capping reagent capable of forming a cap1 structure. The RNA was further purified using NGS magnetic beads. The RNA concentration and quality were evaluated by spectrophotometry and capillary gel electrophoresis systems. The preparation of the liposomes and the formation of lipoplex (LPX) were performed according to a previous study described [
18].
Protein purification
The GST-tagged HPV16-E6E7 sequence was cloned into pGEX-6P-1 vector and transfected into E. coli BL21 (DE3) strain for expression under a final concentration of 500 μM Isopropyl-β-D-thiogalactopyranoside (IPTG) inducing at 16 °C for 18 h. Cells were then harvested by centrifugation at 6500 g for 15 min and suspended in PBS buffer for cell disruption using low temperature ultra-high pressure continuous flow cell disrupters (ATSHPH AH-NANO). The supernatant of cell lysis was collected by centrifugation at 11000 g for 60 min, followed by a one-step affinity chromatography using GST-tag purification resin (BeyoGold, Lot No. P2253) for purifying GST-tagged proteins.
ELISA
Five micrograms of GST-tagged HPV16-E6E7 protein suspended in PBS buffer was coated on the Corning ELISA plates at 4 °C overnight. Mouse plasma was obtained at 21 days after immunization. The serum was separated and incubated with the plates after blocking with 1 × ELISAPOT Diluent (Invitrogen, cat. 00–4202-56) for 2 h at room temperature (RT) with a 1:100 dilution overnight at 4 °C. After five times washings, the goat anti-mouse IgG2a-HRP (Southern Biotech, cat. 1081–05) were added into the plates with a 1:5000 dilution for 1 h at RT. Finally, after washing the plates, TMB 1-Component Peroxidase Substrate (Invitrogen, cat. 00–4201-56) was used to indicate the reaction which was stopped using a 2 M HCl solution. The absorbance at 450 nm was determined within 30 min using a Synergy HTX instrument (BioTek Instruments, Highland Park, VT).
RNA-Seq Library Construction and Sequencing
Total RNA (1 µg) from each treated or control group was used to enrich poly(A) mRNA using oligo(dT) magnetic beads (Invitrogen, USA). RNA-seq libraries were then prepared using the NEBNext® UltraTM RNA Library Prep Kit for Illumina® (NEB, USA) according to Illumina's library construction protocol. The libraries were sequenced to a depth of 20 million reads on the Illumina Novaseq platform. RNA library construction and next-generation sequencing were performed at Novogene. Raw reads were generated and sorted by index codes for further analysis. Low quality and adaptor sequences were trimmed using Trimmomatic v 0.39 software to obtain clean data for downstream analyses.
Gene expression analysis
Clean reads from each library were mapped to the reference genome using Hisat2 v2.2.1. FeatureCounts v1.5.0 was then used to count the number of reads mapped to each gene. Genes with less than 10 mapped reads in the total sample were excluded. Samples were analyzed by DESeq2 to obtain log2 fold change and corresponding p-value in R v4.2.0. Differentially expressed genes were identified from these transformed values using the criteria of log2 (fold change) > 0 and p adjustment value < 0.05. The Benjamini–Hochberg method was used to adjust p-values to control for false discovery rate. R studio was used to run custom R scripts to perform principal component analysis (PCA), volcano plots, and heat maps.
Enrichment analysis of differentially expressed genes
Gene Ontology (GO) enrichment analysis of differentially expressed genes was performed using the cluster Profiler R package. GO terms with a Bonferroni adjusted p-value of less than 0.05 were considered significantly enriched. R studio was used to execute custom R scripts to perform the bubble plot.
TCR clonality analysis
T cell receptor clonality was assessed from the RNAseq data using the MiXCR v3.0 tool. MiXCR applied the standard parameters described in the RNAseq workflow manual to obtain clonotypes from the raw fastq files. After obtaining the quantified clonotypes, the R package immunarch V1.0 was used to calculate sample diversity and counts, respectively. The diversity of T cell receptor clonotypes was determined using the chao1 index, which is a nonparametric asymptotic estimator of species richness.
Statistics
Statistical analyses, unless otherwise indicated, were performed using GraphPad Prism 5. P values below 0.05 were considered statistically significant. Data are shown as the mean ± SEM.
Discussion
Although there has been some progress in tumor therapeutic nucleic acid vaccines, clinical experimental results have shown limited efficacy. To further improve the immunogenicity of antigens delivered by nucleic acid vaccines has become an urgent problem to be solved. In this study, we systematically evaluated the effect of chemokine-fused antigens on antigen immunogenicity. The results showed that there were differences in the ability of different chemokines to induce cellular and humoral immunity as summarized in Fig.
6G. Among them, CCL11 had the strongest capacity to induce cellular immunity. CX3CL1 induced the strongest humoral immune response. Both types of immune responses were generated by CXCL6, CXCL9, CXCL10, CXCL14 and CCL14. We further investigated whether CCL11 could enhance antitumor activity and found that CCL11-E6E7 had potent antitumor effects. In general, large tumors are difficult to be eradicated with therapeutic DNA vaccines in tumor animal models. This is due to the rapid tumor growth and DNA-induced immune responses requiring an onset time lasting more than ten days. However, the E6/E7 DNA vaccine fused to CCL11 could eliminate large established HPV16 E6/E7 tumors in multiple repeated experiments, confirming CCL11 as a powerful enhancer of immunogenicity. Further, this strategy was also found to be suitable for mRNA vaccine development, expanding the use scenarios of the future. CCL11-E6E7 induced infiltration of innate and adaptive immune cell subpopulations into tumors and generated significantly diverse T-cell clonal subpopulations within the tumor. CCL11-E6E7 also increased the number of stem-like CD8 + T cells, which are believed to be the cells responsible for response to immune checkpoint therapy, in lymph nodes. The antitumor activity of CCL11-E6E7 was enhanced by anti-CTLA-4 treatment, indicating a synergy produced by complementary mechanisms in promoting cellular immune responses. The current mechanistic studies revealed that the antitumor effects of the CCL11-E6E7 DNA vaccine are partially dependent on CCR3-positive cells, though the exact subset of CCR3-positive cells and the mechanism through which they induce these effects need to be further identified.
In the design of antigens for vaccine development, many pathogen protein sequences, such as those of diphtheria or tetanus toxins, have been incorporated to increase immunogenicity [
7]. Although immune-enhancing effects have been observed in mice, such approaches can be counterproductive because significant differences in the immunogenicity of antigens may offset the immune responses. The main response generated by high affinity antigens will suppress responses against insufficient MHC affinity antigens. The present strategy of fusion with chemokines can avoid this undesirable outcome, as chemokines are self-antigens and will not cause a response.
Antigen fused with CCL11 would artificially introduce the expression of CCL11, which may produce short-term or long-term pathophysiological effects through its receptor expressing cells. This requires a comprehensive analysis based on the pharmacokinetic characteristics of plasmid DNA and the tissue distribution patterns. Previous studies have extensively investigated the pharmacokinetic characteristics of plasmid DNA vaccines. The plasmids would be completely eliminated in the vast majority of organs within 14 days after injection. Even at the injection site, it does not exceed 28 days which is also applicable for the expressed antigens [
31,
32]. Therefore, CCL11-E6E7 would also show short-term expression in animals and cause no long-term effects on normal animal cells or CCR3 + cells in particular. CCL11 is a member of the eotaxin family which are chemotactic agents for eosinophils and participate innate immunity. CCL11 can selectively recruit eosinophils into inflammatory sites. In the inflammatory sites, T-helper 2 cytokines, such as interleukin-4 (IL-4) and IL-10 induce eosinophils, T cells, B cells and macrophages to produce CCL11. CCR3, the main receptor of CCL11 is expressed on mast cells, eosinophils, Th2 lymphocytes, and keratinocytes which are the effect cells for tissue inflammation [
33]. Therefore, CCL11 is mainly produced by the inflammatory site and promotes local inflammation acting as an effect factor rather than a prime factor. Our plasmid DNA is mainly translated in muscles, and local inflammation occurs at the injection site whose short-term effects are unlikely to cause chronic allergic diseases such as asthma. However, to apply CCL11 in vaccine preparation, systematic and comprehensive safety evaluations are still needed. Moreover, CCL11 may also regulate eosinophil migration in the tumor microenvironment through its interaction with CCR3, and its significant role has been confirmed in colorectal cancer, Hodgkin lymphoma, and oral squamous cell carcinoma [
34,
35]. Whether eosinophils function in tumor promotion or tumor elimination is still unclear. The antitumorigenic role of eosinophils has been well described in several studies [
36,
37]. Adoptive transfer of eosinophils into mice promoted lung metastasis in multiple tumor models [
38]. Our results suggest the involvement of eosinophils in the immunogenic enhancement caused by vaccination and the potential of targeting eosinophils for vaccine applications. However, due to the broad expression of CCR3 and the receptor diversity of CCL11, the exact role of eosinophils in vaccination still needs to be investigated, by using multiple gene knockout mouse models, which is a limitation of our study.
In the investigation of which specific immune cell subpopulations influence the therapeutic efficacy of the CCL11-E6E7 DNA vaccine, we found that CD8 + T cells play a critical role, which is consistent with their role in other types of vaccines. However, the removal of CD4 + T cells also resulted in a decrease in specific CD8 + T-cell levels. Although significant antitumor activity was maintained, the rate of tumor clearance was reduced, indicating both types of cells are needed for the complete eradication of tumors. Intriguingly, NK cells depletion not only reduced the specific CD8 + T-cell level but also decreased the tumor eradication rate. This result highlights the important role of NK cells in the recruitment of cDC1s into the tumor microenvironment for the initiation of the antitumor response [
39].
Multiple types of tumor immunotherapy are recommended to be combined to successfully activate the cancer-immunity cycle and overcome tumor immune escape and immunosuppressive effects [
40]. Thus, many tumor therapeutic vaccines are being combined with immune checkpoint therapy. We have explored the combined effects of multiple immune checkpoints with our DNA vaccines. In which, a combination of CCL11-E6E7 DNA vaccine with anti-CTLA-4 exhibited the greatest treatment benefit, [
41]. As an immunosuppressive checkpoint, targeting CTLA-4 to activate immunity for cancer treatment has been widely studied. Two CTLA-4 blocking monoclonal antibodies, ipilimumab (IgG1) and tremelimumab (IgG2), are clinically approved and have shown efficacy in a subset of solid tumor patients [
42,
43]. Since CTLA-4 competes with the costimulatory receptor CD28 for the CD80 and CD86 ligands to suppress T cell activation, the removal of CTLA-4-mediated negative co-stimulation to augment effector T-cell-mediated immune responses has been identified as the central mechanism of anti-CTLA-4 [
44]. In addition, a high level of CTLA-4 is expressed on the surface of regulatory T cells (Tregs) [
45]. Several murine studies have suggested CTLA-4 blockade may impair the suppressive activity of Tregs and deplete Tregs within the tumor microenvironment (TME) via antibody-dependent cell-mediated cytotoxicity (ADCC) resulting in reduced tumor immunosuppression and expansion of effector T cells [
46‐
48]. In fact, a previous study found that peptide vaccines combined with anti-CTLA-4 significantly reduced the proportion of intratumoral Tregs in the same TC-1 tumor model [
41]. In our study, we observed an increased number of E7 specific effector T cells in peripheral blood (Fig.
6F). Therefore, it is possible that both effects of CTLA-4 blockade played a role in enhancing the efficacy of DNA vaccines, although DNA vaccine treatment did not significantly increase the proportion of Tregs in tumors. Several vaccines have shown the effectiveness of combining PD-1 [
49]. PD-1 functions by reversing exhausted T cells. In our study, to determine the reason why the combination of the CCL11-E6E7 vaccine and PD-1 is not effective, further tests are needed to determine whether the tumor specific T cells induced by DNA vaccines are in a PD-1-responsive state. We will further evaluate whether anti-CTLA-4 also reduces the proportion of Tregs and the state of E6E7 specific T cells in tumors in future studies. Therefore, we encourage experimentally combining various immune checkpoint therapies with different vaccine platforms, to identify potential pharmacologically synergistic combinations, rather than simply superimposing the efficacy of two separate drug therapies.
In our study, electroporation was used for plasmid delivery. Electroporation has been demonstrated an adjuvant-like property for plasmid DNA delivery. Electroporation in skeletal muscle would release a danger signal to recruit antigen-presenting cells independently of plasmid DNA administration [
50]. The DNA vaccine delivered by electroporation showed levels of antibodies that were equivalent to those of cationic lipids and a stronger T cell immune response [
51]. The preventive COVID-19 vaccine and the therapeutic HPV vaccine delivered by electroporation have finished Phase II clinical trials and shown good safety [
1,
52]. However, it is indeed uncomfortable and painful which may lead to a poor compliance. Due to the convenience of use, there are difficulties in popularizing electroporation, especially when using preventive vaccines that are targeted to healthy individuals. In future applications, we note that needle-free injection can be used as an alternative solution, such as the Indian DNA vaccine approved for COVID-19 [
31].
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.