To improve the efficacy of DNA vaccination, a pDNA vaccine made of a mix of 4 plasmids encoding 2 melanoma TAAs and 3 neoantigens (Additional file 1) was injected and electroporated in the tibialis muscle of mice and combined with a CpG-enriched OAd virus [21], administered intratumorally (IT) (Fig. 1a). Neoantigen mutations have been detected in the genomic DNA (gDNA) of the B16F1 cells at different passages to verify the stability of the mutations for the therapeutic vaccination. These neoantigens were specifically redesigned according to the mutations found in B16F1 cell line (Table 1).Fig1Fig. 1

In vivo pDNA and OAd combination. a Therapeutic DNA vaccination protocol in combination with oncolytic virus therapy. B16F1 cells were injected at day 0; pDNA vaccine was intramuscularly (IM) injected and electroporated 2, 9 and 16 days after tumor injection, while 10⁹ OAd virus particles (VP) were IT injected 10, 12 and 14 days after tumor injection. Mice were sacrificed at day 17. b Evolution of tumor volume (mm³) after B16F1 challenge as a function of time (days) (mean ± SEM). All the groups were statistically compared to the others using two-way ANOVA, column factor (p < 0.05, n = 8). c-f Tumor growth measurement for the single mouse and for each group of mice. In red, mice that developed a tumor >300mm³ in volume; in green, mice with a tumor volume < 300 mm³. The a, b and c letters indicate statistical differences: the presence of two different letters in two groups indicate a statistical difference (p < 0.05) between them; the same letter in two different groups indicates the absence of a statistical difference between these two groups

To study the mechanism underpinning the synergy between the plasmid and the oncolytic virus therapy (pDNA+OAd group), infiltration of immune cells was assessed in the tumors of mice 17 days after the tumor injection (Fig. 2). We first assessed the NK cells at the tumor, given their important role in innate rejection of tumor. [25] Interestingly, we found that the total amount of NK cells was increased in all the treated groups and, significantly, in the groups treated with the virus (OAd and pDNA+OAd, Fig. 2a). The number of NK cells that express the CD335 activating receptor [26] (active NK) was significantly higher only in the combination group compared to the mock (Fig. 2b). This data is particularly interesting as it has been so far very poorly studied in the context of oncolytic virus treatment. The number of CD4 and CD8 adaptive immune cells also increased in the treated groups (Fig. 2c and d), as well as the number of IFNg-secreting CD8 T cells (Fig. 2e). As the pDNA vaccine encoded the TRP2 melanoma antigen, the number of TRP2-specific CD8 T cells was evaluated to test the ability of the vaccine to produce an antigen-specific immune response. As TRP2 is encoded in 3 of the four plasmids, it was chosen as the representative epitope to test the antigen-specific T cell infiltration. When mice were treated with the single therapies (pDNA or OAd), the number of TRP2-specific T cells was not increased compared to the mock group. Only when the two therapies were combined, the amount of TRP2-specific T cells was found to be significantly higher compared to all the other groups (Fig. 2f). Different correlation analyses were performed to find the contribution of the different immune populations in the tumor growth. Almost no TRP2-specific T cells were found in bigger tumors (mock group), while a higher infiltration was found in smaller tumors (Fig. 2g). Furthermore, a linear correlation (R² > 0.90) was observed between the number of CD8 T cells and active NK for all the treated groups (Fig. 2h), but also between the TRP2-specific T cells and active NK cells for the combination group (Fig. 2i).Fig2Fig. 2

Immune cell infiltration in the TME. a Number of total NK cells/mm³ tumor. b Number of CD335+ (active) NK cells/mm³ tumor. c Number of non-Treg CD4 T cells/mm³ tumor. d Number of CD8 T cells/mm³ tumor. e Number of IFNg-secreting CD8 T cells/mm³ tumor. f Number of TRP2 antigen-specific CD8 T cells/mm³ tumor. g Correlation between the tumor volume and the number of TRP2 antigen-specific CD8 T cells in the TME. h Correlation between the number of active NK and CD8 T cells. R² was calculated by using a linear regression analysis. i Correlation between active NK and TRP2-specific CD8 T cells. The results in a, b, c, d, e, f and i are expressed as mean ± SEM (n = 4–6). a and b letters on the graphs indicate significantly different results when the superscript letters are different. The annotation “a,b” indicates no significant differences compared to a and b statistical groups. The presence of two different letters in two groups indicate a statistical difference (p < 0.05) between them; the same letter in two different groups indicates the absence of a statistical difference between these two groups

To evaluate the activity and the contribution of the NK and T cells in the TME, an evaluation of the cytokine and perforin/granzymeB expression was performed. A general increase in the cytokine expression was observed when mice were treated with both pDNA and OAd (Fig. 3). In particular, a higher expression of proteins related to NK, Th1 and CTL activity was observed, such as Granzyme B [27] (Fig. 3a), perforin [28] (Fig. 3b), TNFa [29] (Fig. 3C) and IL2 [30] [31] (Fig. 3d). Many other interleukins were overexpressed, such as IL15, IL12 and IL1b (Fig. 3e-g). Finally, an increased CCL5 expression was observed in the OAd-treated groups, OAd and pDNA+OAd (Fig. 3h). Generally, all the cytokines, chemokines and enzymes were also overexpressed in the OAd group, but without observing significant differences compared to the mock group. Interestingly, IL10 was significantly overexpressed only in the OAd group (Fig. 3i). This was the only cytokine expressed in higher amount in OAd group but not in the pDNA+OAd group.Fig3Fig. 3

Cytokine expression in the TME. Granzyme B (a), Perforin (b), TNFa (c), IL2 (d), IL15 (e), IL12 (f), IL1b (g), CCL5 (h), IL10 (i). All the results are expressed as mean ± SEM (n = 4–6). a and b letters on the graphs indicate significantly different results when the superscript letters are different. The presence of two different letters in two groups indicate a statistical difference (p < 0.05) between them; the same letter in two different groups indicates the absence of a statistical difference between these two groups. The annotation “a,b” indicates no significant differences compared to a and b statistical groups

Next, we wanted to evaluate the systemic immune activity and compare it with the response in the TME. To this end, splenocytes were collected 17 days after the tumor challenge and analyzed (Fig. 4). All the treatments significantly increased not only NK infiltration but also the number of active NK in the spleen, compared to the mock group (Fig. 4a and b). In addition to that, the number of CD8 and non-Treg CD4 T cells was higher in the treated groups (Fig. 4c and d). In particular, OAd and pDNA+OAd group showed the highest number of CD8 T cells (Fig. 4c). ELISPOT analysis revealed high TRP2 specificity for pDNA and pDNA+OAd conditions, which was significantly different compared to the other groups (Fig. 4e and f). pDNA+OAd group showed a significantly higher number of IFNg-secreting and TRP2-specific splenocytes compared to pDNA alone (Fig. 4e and f).Fig4Fig. 4

Immune cell analysis in the spleen. a Percentage of total NK. b Percentage of active NK. c Percentage of CD8 T cells. d Percentage of non-Treg CD4 T cells. e-f ELISPOT analysis of the splenocytes stimulated with TRP2 peptide. All the results are expressed as mean ± SEM (n = 4–6). a, b and c letters on the graphs indicate significantly different results when the superscript letters are different. The presence of two different letters in two groups indicate a statistical difference (p < 0.05) between them; the same letter in two different groups indicates the absence of a statistical difference between these two groups

Previously, we have addressed the short-term efficacy of the combination between pDNA and OAd, showing a significant decrease in the tumor growth and a higher CTL infiltration and activity when the two therapies are combined. To better understand the o, we performed a new experiment to follow-up the long-term survival. The vaccine and the virus have been administered following the same protocol used to evaluate the tumor growth, but this time, mice were followed-up until day 45 (Fig. 5a). Our results showed that the median survival time (MST) was significantly longer in the group treated with pDNA and OAd. Furthermore, this combination cured almost 30% of mice (2/7 mice), compared to 0% in the other groups (Fig. 5b).Fig5Fig. 5

Evaluation of the long-term survival. a Therapeutic DNA vaccination protocol in combination with oncolytic virus therapy. B16F1 cells were injected at day 0; pDNA vaccine was intramuscularly (IM) injected and electroporated 2, 9 and 16 days after tumor injection, while 10⁹ OAd virus particles (VP) were IT injected 10, 12 and 14 days after tumor injection. b Survival curves representing the percentage of alive mice (%) as a function of time (days); MST = median survival time. Statistical analysis: Log-Rank (Mantel–Cox) test (p value < 0.05; n = 7–9). The presence of two different letters in two groups indicate a statistical difference (p < 0.05) between them; the same letter in two different groups indicates the absence of a statistical difference between these two groups

B16F1 cells, a melanoma cell line from C57BL/6 mice, were purchased from American Type Culture Collection (ATCC; Manassas, Virginia) and cultured in MEM complete medium, containing 10%FBS, 100 μg/ml streptomycin and 100 U/ml penicillin (Life Technologies, California). The human lung carcinoma cell line A549 were purchased from ATCC and cultivate in DMEM supplemented with 10% FBS, 1% glutamine, 100 μg/ml streptomycin and 100 U/ml penicillin.

B16F1 neoantigens were selected from immunogenic B16F10 mutations described in the literature, according to their low MHC I score, which indicates a high binding affinity, the response to RNA vaccination and the reactive T cell subtype (CD4 or CD8) [12]. The presence of three selected neoantigens was verified in B16F1 cell line. Specific mutations were detected in the gDNA of the cells at passage number 7, 16, and 27 to verify the presence of the mutations at different time points (passage number 1 is when cells were purchased from ATCC). gDNA was extracted using the PureLink™ Genomic DNA Mini Kit (Thermo Fisher, Massachusetts). Briefly, 20 μl of Proteinase K and 20 μl of RNAseA were added to 200 μl of cells (10⁶ cells) and incubated for 2 min at room temperature (RT). Then, 200 μl of the Lysis/binding buffer were added to the cells and incubate at 55 °C for 10 min, before adding 200 μl of pure ethanol. The lysate was purified using a spin column after 2 steps of washings with the washing buffers provided in the kit. gDNA was collected using 50 μl of elution buffer and stocked at − 20 °C, before performing the PCR amplification (see section 5.8). Sanger sequencing was performed by Genewiz (United Kingdom) to verify the presence of specific mutations.

Four different plasmids encoding melanoma antigens TRP2, Pbk, Kif18b, Cpsf3l and human gp100 were designed (Additional file 1). Among the chosen antigens, 3 are recognized by MHC class II (Gp100, Kif18b and Cpsf3l), to stimulate CD4 T cell response, while the others belong to MHC class I epitopes (TRP2 and Pbk). The antigens TRP2 and gp100 are already known melanoma antigens [49, 50]. Their presence has been verified in B16F1 cell line (Additional file 2). The other three are neoantigens described in the literature for the B16F10 melanoma cell line [12]. They have been chosen among the B16F10 immunogenic mutations, based on their MHC class I binding (low score for higher affinity) and their response after RNA and peptide vaccination in B16F10 tumor model [12]. These neoantigens were specifically redesigned according to the mutations found in B16F1 cell line (Table 1). The mutations were verified at different B16F1 cell passages and compared with the gene expression in the spleen of C57Bl/6 mice as a non-mutated control. They have been cloned in a pVAX2 vector encoding the VSV-G viral protein. Two antigens for each plasmid, one CD4 and one CD8, were inserted inside the VSV-G sequence, as described in [23]. Four plasmids were obtained: pTRP2-Gp100; pTRP2-Cpsf3l; pTRP2-Kif18b; pGp100-Pbk (Additional file 1). The mix of the four plasmids in a 1:1:1:1 proportion (1 μg for each plasmid) was called pDNA. In Table 1, the nucleotide and peptide sequences of each antigen are shown. The underlined amino acids indicate the presence of a mutation specific in B16F1 melanoma cell line.

As an additional control for the survival experiment, an irrelevant plasmid has been used, encoding two OVA epitopes in the VSV-G sequence (one CD4 and one CD8 epitopes). This plasmid has been called “pTOP-OVA CD4-OVA CD8”.

Ad5D24-CpG is an OAd bearing a CpG-enriched genome in the E3 gene. It was generated, propagated, and characterized using standard protocols, as previously described [51].

All animal experiments were reviewed and approved by the Experimental Animal Committee of the University of Helsinki and the Provincial Government of Southern Finland and the Belgian national regulations guidelines in accordance with EU Directive 2010/63/EU. The animal experiments were approved by the ethical committee for animal care of the faculty of medicine of the Université catholique de Louvain (UCL/MD/010/2019). Five weeks-old C57BL/6 mice were purchased from Envigo (Harlan, Netherlands) or by Janvier (France). Water and food were provided ad libitum.

ELISpot was performed according to the manufacturer’s instruction (Immunospot, The ELISPOT source, Germany). Briefly, 3 × 10⁵ fresh splenocytes diluted in 100 μl CTL-Test medium (Immunospot, The ELISPOT source, Germany) were cultured overnight at 37 °C in anti-IFNg coated 96 well plate. For stimulation, 10 ng/μl TRP2 peptide was added to the splenocytes and incubated for 3 days. As positive control for splenocyte activation, Cell Stimulation Cocktail (Invitrogen, California) were used; PBS was used as negative control. The development of the ELISpot plate followed the manufacturer’s instruction (Immunospot, The ELISPOT source, Germany). Spots were counted by using an ELISPOT reader system (Immunospot).

NK, CD4 and CD8 T cell populations were analyzed by FACS. Tumors and spleen were surgically removed 17 days after B16F1 cell injection and FACS analysis of tumors and spleens T and NK population was performed. To prepare single cell suspensions, cells were passed through a 70 μm cell strainer (BD Falcon, New Jersey). Then, they were collected, counted using an automatic cell counter (Invitrogen, California) and washed with PBS, before adding the blocking solution with anti-CD16/CD32 antibody for 10 min on ice (clone 93, Biolegend, San Diego, California). Cells were washed and incubated for 60 min at 4 °C with the following antibodies: CD49-APC, CD335-FITC, CD11b-PerCP-Cy5.5 (for NK detection), CD3-PerCPCy5.5, CD4-PeCy7, CD8-FITC (for CD4 and CD8 detection). For staining with anti-FoxP3-PE (clone FJK-16 s, eBioscience, Thermo Fisher, Waltham, Massachusetts) or anti-IFNg-PE (clone XMG1.2, Biolegend, San Diego, California), cells were previously incubated overnight at 4 °C with a permeabilization/fixation solution (eBioscience™ Foxp3 / Transcription Factor Staining Buffer Set, Thermo Fisher, Waltham, Massachusetts). Cells were then incubated with anti-CD16/CD32 antibody for 10 min on ice (Biolegend, San Diego, California), washed and incubated for 60 min at 4 °C with anti-IFNg-PE or antiFoxP3-PE diluted in the permeabilization/fixation solution. Samples were washed with PBS fixed for 10 min with 4% formalin and, then, suspended in PBS. Sample data were acquired with FACS Fortessa or FACS Accuri (BD bioscience, Franklin Lakes, New Jersey) and analyzed with FlowJo software (FlowJo LLC, Ashland, Oregon). For the tumor analysis, the number of cells was normalized by the tumor volume (mm³).

Tumors extracted at day 17 were analyzed by qPCR. Total RNA was isolated using TRIzol reagent (Thermo Fisher, Waltham, Massachusetts) and phenol separation, as previously described [14]. The quality and quantity of RNA were evaluated using a nanospectrophotometer (NanoDrop 2000, Thermo Fisher, Waltham, Massachusetts). Extracted RNA was considered pure if the 260/280 absorbance ratio of the sample was approximately 2 and the 260/230 absorbance ratio was 1.8–2.2. One microgram of RNA was reverse transcribed using a first-strand synthesis system (SuperScriptTM, Thermo Fisher, Waltham, Massachusetts) and oligo(dT) primers (Eurogentec, Liege, BE) according to the supplier’s protocol. The resulting cDNA was used as template for 40 cycles of PCR amplification. SYBR™ green real-time qPCR (GoTaq qPCR MasterMix kit, Promega, Fitchburg, Winsconsin) was conducted on a StepOne Plus Real-Time PCR System (Thermo Fisher, Waltham, Massachusetts) to detect different interleukins, chemokines, perforin and granzyme B. mRNA expression in the tumors and antigen gDNA and mRNA expression in the B16F1 cells. Analysis of the melting curves was performed to ensure purity of PCR products. The results were analyzed with StepOne Software V2.1. The mRNA expression of the cytokines was calculated relative to the corresponding expression of β-actin (reference gene) according to the delta-delta Ct method. The results were normalized compared to the mock control group. A complete list of the primers used in this study is shown in Table 2.Tab2

List of the primers used in this study

The synergy vs additivity analysis has been performed by using the Spector’s formula, as described in [24] and in Kos et al. (revised version under review). Briefly, for an additive effect, the combination index (CI) is in between +/− 2 times the standard error (SE): -2SE < CI < + 2SE; while, for a synergic effect: CI > + 2SE. The SE is defined as the derived as the square root of the total variance divided by the number of samples. The CI is defined by the following formula: CI=lnX1¯+lnX2¯−lnX1+2¯−lnX0¯\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ CI=\ln \left(\overline{X1}\right)+\ln \left(\overline{X2}\right)-\ln \left(\overline{X1+2}\right)-\ln \left(\overline{X0}\right) $$\end{document}

Where: X1¯=mean of the effect of the control1inourcasethe mean of the tumor volumes for pDNA alone.X2¯=mean of the effect of the control2inourcasethe mean of the tumor volumes forOAdalone.X1+2¯=mean of the effect of the combinationinourcasethe mean of the tumor volumes for the pDNA+OAdgroup.X0¯=mean of the effect of the non‐treated groupinourcase the mean of the tumor volumes for the mock group.\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {\displaystyle \begin{array}{l}\overline{\mathrm{X}1}=\mathrm{mean}\ \mathrm{of}\ \mathrm{the}\ \mathrm{effect}\ \mathrm{of}\ \mathrm{the}\ \mathrm{control}\ 1\ \left(\mathrm{in}\ \mathrm{our}\ \mathrm{case},\mathrm{the}\ \mathrm{mean}\ \mathrm{of}\ \mathrm{the}\ \mathrm{tumor}\ \mathrm{volumes}\ \mathrm{for}\ \mathrm{pDNA}\ \mathrm{alone}\right).\\ {}\overline{\mathrm{X}2}=\mathrm{mean}\ \mathrm{of}\ \mathrm{the}\ \mathrm{effect}\ \mathrm{of}\ \mathrm{the}\ \mathrm{control}\ 2\ \left(\mathrm{in}\ \mathrm{our}\ \mathrm{case},\mathrm{the}\ \mathrm{mean}\ \mathrm{of}\ \mathrm{the}\ \mathrm{tumor}\ \mathrm{volumes}\ \mathrm{for}\ \mathrm{OAd}\ \mathrm{alone}\right).\\ {}\overline{\mathrm{X}1+2}=\mathrm{mean}\ \mathrm{of}\ \mathrm{the}\ \mathrm{effect}\ \mathrm{of}\ \mathrm{the}\ \mathrm{combination}\ \left(\mathrm{in}\ \mathrm{our}\ \mathrm{case},\mathrm{the}\ \mathrm{mean}\ \mathrm{of}\ \mathrm{the}\ \mathrm{tumor}\ \mathrm{volumes}\ \mathrm{for}\ \mathrm{the}\ \mathrm{pDNA}+\mathrm{OAd}\ \mathrm{group}\right).\\ {}\overline{\mathrm{X}0}=\mathrm{mean}\ \mathrm{of}\ \mathrm{the}\ \mathrm{effect}\ \mathrm{of}\ \mathrm{the}\ \mathrm{non}\hbox{-} \mathrm{treated}\ \mathrm{group}\ \left(\mathrm{in}\ \mathrm{our}\ \mathrm{case}\ \mathrm{the}\ \mathrm{mean}\ \mathrm{of}\ \mathrm{the}\ \mathrm{tumor}\ \mathrm{volumes}\ \mathrm{for}\ \mathrm{the}\ \mathrm{mock}\ \mathrm{group}\right).\end{array}} $$\end{document}

Statistical analyses were performed using GraphPad Prism 7 for Windows. Survival curves were compared using a Mantel–Cox (log-rank) test. p-values less than 0.05 were considered statistically significant and indicated with different letters on the graphs (a, b, c, d). In particular, the presence of two different letters in two groups indicate a statistical difference (p < 0.05) between them; the same letter in two different groups indicates the absence of a statistical difference between these two groups. The annotation “a,b” indicates no significant differences compared to a and b statistical groups.