In vivo transfection of cytokine genes into tumor cells using a synthetic vehicle promotes antitumor immune responses in a visceral tumor model

The tumor microenvironment (TME) strongly affects the clinical outcomes of immunotherapy. This study aimed to activate the antitumor immune response by manipulating the TME by transfecting genes encoding relevant cytokines into tumor cells using a synthetic vehicle, which is designed to target tumor cells and promote the expression of transfected genes. Lung tumors were formed by injecting CT26.WT intravenously into BALB/c mice. Upon intravenous injection of the green fluorescent protein‐coding plasmid encapsulated in the vehicle, 14.2% tumor‐specific expression was observed. Transfection of the granulocyte‐macrophage colony‐stimulating factor (GM‐CSF) and CD40 ligand (L)‐plasmid combination and interferon gamma (IFNγ) and CD40L‐plasmid combination showed 45.5% and 54.5% complete remission (CR), respectively, on day 60; alternate treatments with both the plasmid combinations elicited 66.7% CR, while the control animals died within 48 days. Immune status analysis revealed that the density of dendritic cells significantly increased in tumors, particularly after GM‐CSF‐ and CD40L‐gene transfection, while that of regulatory T cells significantly decreased. The proportion of activated killer cells and antitumoral macrophages significantly increased, specifically after IFNγ and CD40L transfection. Furthermore, the level of the immune escape molecule programmed death ligand‐1 decreased in tumors after transfecting these cytokine genes. As a result, tumor cell‐specific transfection of these cytokine genes by the synthetic vehicle significantly promotes antitumor immune responses in the TME, a key aim for visceral tumor therapy.


| INTRODUCTION
Dendritic cells (DCs) are the most potent antigenpresenting cells that help differentiate naive T cells into antigen-specific cytotoxic T cells (CTLs).Therefore, DCs induced from monocytes and bone marrow precursors and pulsed with tumor antigens in vitro have been used in cancer immunotherapy.However, the immune responses observed are transient, and overall clinical outcomes have not been particularly successful. 1 This limitation is primarily because of the degradation and dysfunction of DCs, 2 resulting from the lack of activators in the tumor microenvironment (TME).Agonists or ligands (L) of CD40 induce maturation, enhance the function of DCs, 3 and elicit the generation of antitumoral M1 macrophages from progenitors or pro-tumoral M2 macrophages. 4Additionally, interferon gamma (IFNγ) activates an antitumoral immune response mediated by mature DCs.6][7] This approach substantially enhanced the tumor-specific immune response, leading to effective therapeutic outcomes, including 67% complete remission (CR) in subcutaneous tumor models. 8Subcutaneous solid tumors are convenient for observing the therapeutic effect over time and can be extracted individually for analysis; however, the findings in subcutaneous tumors are difficult to apply to therapy for visceral tumors which are difficult to obtain tumor lysates for DCs to present antigens.Furthermore, DCs cannot be directly administered to tumors, and almost all deaths due to solid tumors are caused by visceral metastases. 9n this study, to apply our strategy involving the genes producing CD40L and IFNγ to treating visceral tumors, we transfected the gene of granulocyte-macrophage colony-stimulating factor (GM-CSF), which induces DC differentiation from monocytes. 10We attempted to obtain DCs endogenously differentiated from tumor-infiltrating monocytes by GM-CSF, followed by their maturation and activation and generation of M1 macrophages by CD40L, and finally to enhance the immune responses against tumors by IFNγ in the TME.

| Animals and cell line
Female BALB/c mice (6-8 weeks old) (RRID:MGI:2161019) were purchased from Japan SLC, Inc. (Shizuoka, Japan).The mice were maintained under specific pathogen-free conditions.After initiating the experiments, the mice were examined daily for weight loss, labored respiration, and signs of discomfort.No unexpected deaths occurred during the study period.The experimental endpoint was 60 days after the injection of the tumor cells.In the tumor models, the mice were immediately euthanized when they exhibited strong signs of distress, such as hunching in the corner of the cage with crumpled hair, rapid weight loss, a lack of movement, or abnormal respiration.Mice were humanely euthanized by anesthesia with sodium pentobarbital (200 mg/kg, intraperitoneal injection), followed by cervical dislocation.All research staff involved in animal experiments received training in animal care and were approved by the Animal Experiment Committee of Osaka Prefecture University.The study protocols were approved by the Animal Experiment Committee of Osaka Prefecture University (Approval Nos.20-56 and  21-57).
alternate treatments with both the plasmid combinations elicited 66.7% CR, while the control animals died within 48 days.Immune status analysis revealed that the density of dendritic cells significantly increased in tumors, particularly after GM-CSF-and CD40L-gene transfection, while that of regulatory T cells significantly decreased.The proportion of activated killer cells and antitumoral macrophages significantly increased, specifically after IFNγ and CD40L transfection.Furthermore, the level of the immune escape molecule programmed death ligand-1 decreased in tumors after transfecting these cytokine genes.As a result, tumor cell-specific transfection of these cytokine genes by the synthetic vehicle significantly promotes antitumor immune responses in the TME, a key aim for visceral tumor therapy.

| Preparation of genes for transfection
A plasmid containing cDNA encoding green fluorescent protein (GFP) (pApGFP1-C1) was purchased from Clontech Laboratories (Mountain View, CA, USA).Mouse IFNγ and CD40L cDNA were the same as those used in our previous study 8 and were inserted into the expression plasmid pcDNA™3.1/myc-His(−) (pcDNA) (Invitrogen, Waltham, MA, USA, Cat# V855-20).The mouse GM-CSF gene was prepared in the same way as previously described. 8Briefly, mouse GM-CSF cDNA in the cDNAs of activated lymphocytes was amplified using polymerase chain reaction (PCR).Primers were designed to amplify the specific nucleotide sequence of mouse GM-CSF cDNA (NM_009969).The sequences recognized by XhoI and BamHI were attached to the ends of forward and reverse primers, respectively.The PCR products were then inserted into a PCR-Blunt vector (Invitrogen) and amplified in transformed Escherichia coli DH5α.The cloned sequence was inserted into pcDNA using the designed restriction sites.The genes producing mouse IFNγ and CD40L used in this study were those prepared in the previous study. 8

| Preparation of a synthetic vehicle for gene transfection
][7] Briefly, a solution of egg yolk phosphatidylcholine in chloroform (Nippon Yuka Kogyo, Yokohama, Japan) was added to 3-methylglutarylated polyglycidol (MG-luPG), and a lipid/polymer thin membrane was formed by evaporation of organic solvents using a rotary evaporator (Tokyo Rikaki, Tokyo, Japan).After dissolving in a 5 mM HEPES-5% glucose solution (pH 8.0), the liposomes were stabilized by repeated freeze and thaw cycles, and the size was adjusted by passing the liposomes through a 50 nm-diameter pore membrane (AVESTIN, Mannheim, Germany).MGluPG was bound to transferrin (Sigma-Aldrich, St Louis, MO, USA) to target tumor cells via its corresponding receptor. 11The plasmid was diluted with a 20 mM Tris-HCl solution (pH 7.4) and mixed with Mul-tiFectam (Promega, Madison, WI, USA).After allowing them to stand at room temperature for 30 min, the transferrin-conjugated liposomes were diluted with a 5% glucose solution (pH 7.4), mixed well, and maintained on ice for 10 min.Opti-MEM medium (Invitrogen) was then added to adjust the volume and maintained on ice for more than 10 min before use.The N/P ratio (number of amines in cationic lipid/phosphate contained in DNA) and C/P ratio (carboxylates of MGluPG/phosphate contained in DNA) influence the transfection efficiency.The optimal ratio was detected as C/P = 5 and N/P = 8 in preliminary in vitro experiments.

| Detection of in vivo gene expression after transfection
BALB/c mice were injected with CT26.WT cells (1 × 10 5 cells) into the tail vein.Eight days after tumor injection, the synthetic vehicle encapsulating plasmids with the green fluorescent protein (GFP) cDNA (2.5 μg) were i.v.administered through the lateral tail vein.Two days after injection, the lungs were excised from euthanized mice, fixed in 10% neutral buffered formalin for 24 h, embedded in paraffin, and sectioned.GFP expression was detected by immunohistochemistry (IHC) using a rabbit-anti GFP tag antibody (Ab) (Thermo Fisher Scientific, San Diego, CA, USA, Cat# A-6455, RRID:AB_221570) and peroxidase-labeled goat anti-rabbit IgG F (ab') Ab (Nichirei Bioscience, Tokyo, Japan), as described previously. 8Cells in three or four highpower (×400) fields were counted in each section.For each tumor, >1000 cells were counted.Transfection efficiency was calculated as the number of GFP-expressing cells divided by the total number of cells.As a negative control, a synthetic vehicle encapsulating the same amount of the pcDNA plasmid was injected.Livers and intestines were collected from GFP-plasmid-injected mice and examined for GFP expression in the same manner.
2.5 | Tumor therapy regimen BALB/c mice were i.v.injected with CT26.WT cells (2 × 10 5 cells in 200 μL phosphate buffered saline (PBS)) into the tail vein.After the injection, a synthetic vehicle encapsulating a plasmid-coding cytokine gene was i.v.administered to BALB/c mice four or five times at sevenday intervals.The treatment was set to inject the following plasmids: [CD40L], a pcDNA plasmid encoding CD40L; [GM-CSF], a plasmid encoding GM-CSF; [GM-CSF + CD40L], both the GM-CSF-and CD40L-plasmids; and [IFNγ + CD40L], both the IFNγ-and CD40L-plasmid.For treatment with [GM-CSF + IFNγ + CD40L], the mice were injected alternately with [GM-CSF + CD40L] or [IFNγ + CD40L].For [None] treatment, mice were injected with pcDNA that did not contain any cytokine cDNA.The total amount of DNA in one administration was 5 μg (2.5 μg of each cytokine plasmid when two cytokine plasmids were used).The therapeutic effect was evaluated by measuring the survival rate of mice for 60 days (endpoint).

| Analysis of immune status
Immune status in the treated mice was examined in two independent ways: three-dimensional (3D) tissue fluorescence imaging after optical tissue clearing (the location of tumor and immune cells in the lungs was grossly recognized) and IHC (immune cells in tumor tissues are microscopically analyzed).BALB/c mice were i.v.injected with CT26.WT cells (2 × 10 5 cells in 200 μL PBS) into the tail vein.On the next day and day 8, the same plasmids, including the cytokine gene described in the preceding section, were i.v.administered using the synthetic vehicle.The lungs were excised from mice in each treatment group 7 days after receiving treatments twice.The collected lungs were perfused with 10% neutral buffered formalin through the trachea and immersed in the same solution for 24 h.
For IHC, formalin-fixed lung tissues were embedded in paraffin and cut into thin sections.CD11c-, granzyme B-, or FoxP3-expressing cells infiltrated in tumors were detected using the same Abs as CD11c, granzyme B, and Foxp3, as mentioned above.In addition, lung tissues were incubated with anti-mouse CD80 mAb (clone 1E2F10, Proteintech, Cat# 66406-1-Ig, RRID:AB_2827408) to detect M1 macrophages and rabbit anti-human CD163 Ab (Proteintech, Cat# 16646-1-AP, RRID:AB_2756528) to detect M2 macrophages.The expressing cells were visualized using peroxidase-labeled goat IgG F(ab') Ab (Nichirei Biosciences) against primary antibodies.Experiments were performed using three mice from each treatment group.Three sections were prepared from the tumors of each mouse.Cells in three or four high-power (×400) fields were counted in each section.More than 3000 cells were counted for each tumor.The percentage of positive cells was evaluated.

| Cytotoxicity assay
The cytotoxicity assay was performed as described previously with minor modification. 14The activity of the natural killer (NK) cells and tumor-specific CTLs within the splenic cell populations was evaluated by a 51 Crrelease assay with YAC-1 (RIKEN, Tsukuba, Japan; Cat# RBC1165, RRID:CVCL_2244) and CT26.WT as the targets to detect the activities of NK and CTL, respectively.A B16 mouse melanoma line (RIKEN, Cat# RBC1283, CVCL_ F936) derived from C57BL/6 mouse (H-2 b ) was used as a target to detect activity against allogeneic tumors.Spleen cells were collected from mice in each treatment group 7 days after receiving treatments twice.In the NK activity assay, the spleen cells collected were incubated as effector using a 96-well plate (AGC Techno Glass Co., LTD., Shizuoka, Japan) for 6 h in 0.2 mL of RPMI medium with 10 4 YAC-1 cells labeled with 51 CrO 4 (Perkin Elmer, Waltham, MA, USA) at various effector/target (E/T) ratios.After the 6 h incubation, the supernatant (0.1 mL) was collected, and the radioactivity of the released 51 Cr was measured using a gamma counter (AccuFLEX γ7001, Aloka-Hitachi, Tokyo, Japan).In the CTL activity assay, the spleen cells (5 × 10 6 /2 mL RPMI medium) were cultured with 20 ng/ mL recombinant mouse interleukin (IL)-2 (BioLegend, San Diego, CA, USA) using a 24-well plate (AGC Techno Glass).After the 6-day incubation, the cultured cells (as effector) were washed repeatedly and, using a 96-well plate, incubated for 6 h in 0.2 mL of the culture medium with 10 4 CT26.WT or B16 cells (as target) labeled with 51 CrO 4 (Perkin Elmer, Waltham, MA, USA) at various E/T ratios.Following this incubation, radioactivity of the released 51 Cr was measured as described above.The results are calculated as the percentage of lysis according to the formula: To determine the minimum release, labeled tumor cells were incubated in the culture medium.For maximum release, labeled cells were incubated in a detergent, 10% NP-40 (Nacalai Tesque).

| Statistics
Statistical analyses were performed using the Statcel3 software (OSM, Saitama, Japan, RRID:SCR_016753).Kaplan-Meier survival curves were compared using the log-rank test.Positive cell rate differences in the experimental groups were compared using the Tukey-Kramer test.Differences between the groups were considered significant at p < .05.

| Tumor cell-specific gene transfection by synthetic vehicle
To investigate whether the synthetic vehicle elicited specific and efficient gene transfer to visceral tumors, we i.v.administered the vehicle encapsulating the GFP plasmid into BALB/c mice bearing lung tumors of CT26.WT cells and examined the expression of GFP by IHC.As shown in Figure 1A, brown-tinted GFP-positive cells were found only in tumors in the GFP-plasmid [GFP]-injected mice.In contrast, no GFP expression was observed in normal lung tissue.Additionally, no transfection product was found in samples from tumor-bearing mice injected with the vehicle encapsulating the noncoding plasmid [None].The efficiency of GFP expression in the tumors was 14.2% (Figure 1B).
Transfection specificity was also examined for the transfected cytokine genes.Plasmids encoding the genes for GM-CSF ([GM-CSF]), CD40L ([CD40L]), or IFNγ ([IFNγ]) were encapsulated in the synthetic vehicle and i.v.injected into BALB/c mice bearing CT26.WT tumors.As shown in Figure 2, cytokine expression was detected in tumor cells but not in normal lung cells of mice transfected with the corresponding gene.survived by the set endpoint, but the survival was not significantly longer compared to the [None]-treated mice.All surviving mice showed CR.

| Effect of in vivo cytokine gene transfection on immune status
We investigated the immune status of tumor-injected mice 7 days after two treatments with combinations of cytokine plasmids, which elicited prolongation in survival after five administrations.In this study, 3D fluorescence imaging after optical tissue clearing was performed to visualize the location of tumors and immune cells in the lung.Previously, it was confirmed by conventional hematoxylin-eosin staining that the area stained with PI in this imaging was a tumor. 12In contrast, the immune cells in the tumor tissue were microscopically analyzed by IHC.
Patients with a TME containing significantly higher numbers of DCs and killer cells, including CTLs, but with fewer immunosuppressive Tregs, had a better prognosis. 15,16Therefore, we first investigated these cells in the lungs of treated mice.As shown in Figure 4, even at 3 weeks after tumor cell injection, most of the area in the lung of the [None]-treated mice were occupied by the tumor, as represented by the area intensively stained with PI.In contrast, a few small tumors were observed in the lungs of the cytokine plasmid-treated mice.The expression of CD11c, a marker of DCs, in the lungs of the [None]-treated mice was significantly lower than in the PI-stained area.In contrast, the expression of CD11c in the lungs of cytokine plasmid-treated mice mostly matched the location of PI-stained areas.Concerning CD11c-expressing cells in the tumor area (Figure 5), cytokine-plasmid-treated mice had a significantly greater number than [None]-treated mice.In particular, the number of DCs in the [GM-CSF + IFNγ + CD40L]-treated mice was significantly higher compared to other groups (p < .01 vs. [None] and [IFNγ + CD40L] and p < .05 vs. [GM-CSF + CD40L]).In addition, the [GM-CSF + CD40L]treated mice had significantly more DCs than the [IFNγ + CD40L]-and [None]-treated mice.The number of cells in the [IFNγ + CD40L]-treated mice was lower than that in the [GM-CSF + IFNγ + CD40L]-and [GM-CSF + CD40L]-treated mice but significantly higher than that in the [None]-treated mice.As shown in Figure 6, the expression of granzyme B, a marker of activating killer cells including CTLs, was observed in the PI-stained tumors in all groups.However, as shown in Figure 7, the Tregs and helper T cells/Tregs, in lungs of the [None]treated mice were expressed in PI-stained areas.In the lungs of the cytokine-plasmid-treated group, expression was also observed in the PI-stained area, but at reduced levels.As shown in Figure 9, the [GM-CSF + CD40L]-and [GM-CSF + IFNγ + CD40L]-treated mice had significantly fewer FoxP3-expressing cells than the [None]-treated mice.
CD163 + M2 macrophages, which are dominant in tumor-associated macrophages, support tumor growth by suppressing immune responses against tumors. 17The agonist/ligand of CD40 elicits the generation of CD80 + M1 macrophages from progenitors or pro-tumoral M2 macrophages and promotes immune responses against tumors. 4Moreover, IFNγ enhances the immune response.As shown in Figure 10, cytokine gene-transfected mice, [IFNγ + CD40L]-treated mice, had significantly more M1 macrophages in the tumor than [None]treated mice.In contrast, the cytokine gene-transfected mice had significantly fewer M2 macrophages (Figure 11).
Finally, the effect of cytokine gene transfection on immune-related molecules in tumor cells was investigated.9][20] As shown in Figure 12, PDL-1 expression in the lungs of [None]-treated mice was almost consistent with the PI-stained tumor area; most tumor cells expressed PDL-1.However, in the lungs of cytokine gene-treated mice, the PDL-1-expressing area was significantly less than the PI-stained tumor area, and the PDL-1 expression of tumor cells was significantly inhibited.PD-1 expression in the lungs of [None]-treated mice was observed in some parts of the tumor area.However, the tumor-related stained area of PD-1 was not found in the lungs of the cytokine gene-treated mice.
In addition to immune status in tumors, we investigated systemic tumor immunity using spleen cells from cytokine gene-treated mice.NK activity represented by lysis of YAC-1 cells was significantly higher in the [GM-CSF + CD40L]-, [IFNγ + CD40L]-, and [GM-CSF + IFNγ + CD40L]-treated mice than the [None]treated mice (Figure 13A).Further, CT26WT-specific CTL activity in the cytokine gene-treated groups was significantly higher than that in the [None]-treated mice (Figure 13B).There was a marginal immune response against allogeneic tumor cell B16 in mice of all treatment groups but without any significant difference among those groups (Figure 13C).

| DISCUSSION
In this study, we attempted to suppress visceral tumors by creating an immune environment that induces and activates DCs and immune cells to attack the tumor cells.Therefore, we transfected genes encoding GM-CSF, CD40L, and IFNγ into tumor cells in mouse lungs using a synthetic vehicle.
First, we examined the tumor specificity and expression efficiency of the transfected genes using cDNA of GFP not synthesized in the mouse body.We previously observed the expression of transfected genes in tumor cells that grew subcutaneously after i.v.injection of the synthetic vehicle. 8However, it is difficult to demonstrate tumor specificity in the subcutaneous model because tumors grow separately from subcutaneous tissue.Therefore, to clarify tumor-specific transfection and expression, we performed in vivo gene transfection in a lung tumor model and determined that the expression of the transfected gene was observed only in the tumor but not in the normal region of the same lungs (Figure 1A).The expression efficiency of the transfected gene was 14.2% in IHC analysis (Figure 1B).6][7] Transferrin binding of the vehicle to malignant tumors prone to express the corresponding receptor. 21When the liposomes are incorporated into the endosome, the pH-sensitive polymers fuse with the endosome membrane.By the fusion, the vehicle efficiently releases the encapsulating DNA into the cytoplasm (Figure S1B).6][7] In addition to GFP cDNA, cytokine cDNAs showed tumor-specific expression with high efficiency (Figure 2).Viral vectors have mainly been used for in vivo gene transfection.Treatment with an adenoviral vector coding for the gene encoding CD40L significantly reduced tumor extent in a mouse [22][23][24] However, adenoviral vector infection not tumor-specific and should be directly into tumors.As a result, adenoviral vectors are not suitable for the treatment of visceral tumors.Additionally, the efficacy of adenoviral vector infection is reduced by innate immunity. 25In contrast, non-viral vectors have recently been developed with specificity for tumors by i.v.administration. 26,27Bao et al. used cationic liposomes decorated with iRGD that bind to α v β 3 integrins and neuropilin-1, which are overexpressed on tumor cells and the tumor vascular endothelium, and succeeded in tumor-specific transfection of the pigment epitheliumderived factor (PEDF) gene in a metastatic lung tumor model with CT26WT. 27The expression efficiency of the viral vector after eight every days was approximately 9% in IHC These findings indicate that the synthetic gene used in this study is a promising tool for tumor gene therapy.
In a subcutaneous tumor model, we have obtained significant prolongation of survival and 67% CR by transfecting genes encoding CD40L and IFNγ, followed by administration of DCs into the tumor. 8Given that both CD40L and IFNγ synergistically activate DCs, 28 gene transfection changed the TME to activate the exogenously administered DCs and augment immune responses against tumor cells.However, in the visceral tumor model, which is more usable for clinical applications, the direct injection of exogenous DCs into tumors is impossible.Therefore, GM-CSF was transfected to induce DC differentiation from tumor-infiltrating monocytes.Moreover, by combining it with CD40L gene transfection, [GM-CSF + CD40L], we aimed to induce maturation and activation of the endogenously induced DCs.Four times-transfection starting 8 days after tumor cell injection when lung tumors had formed, however, did not elicit any therapeutic effect (Figure 3A).We then initiated the treatment on the day after tumor cell injection and determined that 20% of the [GM-CSF + CD40L]-and [GM-CSF]-treated mice survived with CR at the endpoint (Figure 3B).In addition, 40% of the mice which received the [CD40L + IFNγ]-treatment and 20% of mice which received the [CD40L]-treatment, both of which began the day following tumor injection, survived with CR at the endpoint.In immunohistochemical assays, a significant increase in DCs was obtained by the two-time transfection of the gene encoding GM-CSF (Figure 5).Additionally, M1 macrophage and granzyme B + -activated killer cells were significantly increased through transfection of genes encoding CD40L and IFNγ gene (Figures 7 and 10 respectively).These findings imply that induction of the immune cells, which effect therapeutic results, was elicited by the transfection of the corresponding cytokine gene.As a result, the 7-day lag, which caused the difference in therapeutic impact between two time-courses of treatment (treatment starting at 8 days or the day after tumor injection), may be the time required for inducing DCs from monocytes or precursors for converting the macrophage population from M2 to M1 and activating killer cells.
With an increase in the frequency of the treatment from four times to five times, the therapeutic effect of the cytokine gene-treatment was significantly augmented, and the survival of the [GM-CSF + CD40L]-and [CD40L + IFNγ]treated mice was significantly prolonged compared to the [None]-treated and [CD40L]-treated mice, and 45.5% (5/11) and 54.5% (6/11) of them showed CR at the endpoint, respectively (Figure 3C).These findings indicate that an increase in the anti-tumor immunity created by the cytokine gene-treatments in TME significantly improves therapeutic outcomes.Furthermore, the cytokine gene treatments certainly convert the "cold" tumor, which is resistant to therapy with immune checkpoint inhibitors (ICIs), 29 to the ICI-sensitive "hot" tumor.Consequently, combining with ICI therapy, the cytokine gene treatments may elicit a successful outcome against grown-up tumors (in the case of this study, those at 8 days after injection).
We demonstrated that the expression level of PDL-1 in tumor cells was significantly diminished in the [GM-CSF + CD40L]-, [CD40L + IFNγ]-, and [GM-CSF + CD40L + IFNγ]-treated mice but not in the [None]treated mice (Figure 12).As a result, the expression of these cytokines by transfected genes may suppress the expression of PDL-1 in tumor cells, although the underlying mechanism remains unknown.PDL-1, expressed on tumor cells, plays a critical role in evading immune attacks by inducing suppressor cells. 18,19Indeed, the number of FoxP3 + Tregs was significantly lower in the tumors of cytokine gene-transfected mice (Figure 9).Therefore, the suppression of PDL-1 expression may be a critical mechanism for suppressing tumor growth by cytokine gene transfection, particularly in the early period of therapy before tumor formation.
Using the lung-tumor model, a visceral tumor model, we could not observe a change in tumor size with time in the same mouse.However, we previously found in the subcutaneous tumor model that survival significantly correlated with tumor size 8 .Furthermore, PI-staining of the lungs in the cytokine gene-treated mice revealed significantly fewer and smaller tumors in them compared to [None]-treated mice, even after receiving treatment twice (Figures 4, 6, 8, and 12).Therefore, evaluation of survival may be sufficient to assess the therapeutic effect in the visceral tumor.
Bao et al. transfected a plasmid encoding PEDF using a non-viral vector, resulting in inhibition of tumor angiogenesis and induction of apoptosis of CT26.WT tumors in the lungs of BALB/c mice. 27After every 2 days of injection of 0.25 mg/kg DNA eight times from the third day of tumor inoculation, they succeeded in significantly reducing CD31 + endothelial cells in tumors and obtained 20% survival at day 60.We injected ≤0.25 mg/ kg DNA of [GM-CSF + CD40L], [CD40L + IFNγ], or [GM-CSF + CD40L + IFNγ] into the same mouse model five times at 7-day intervals from the day following tumor inoculation and obtained more than 45% survival and CR at day 60.A direct comparison between their and our findings is impossible, but the transfection of the cytokine genes used in this study seemed to elicit greater therapeutic outcomes.
We measured GM-CSF and IFN-γ in serum 2 days after the second transfection of the corresponding genes and no difference in the of mice transfected with cytokine and [None] plasmids detection limit in the enzyme-linked immune sorbent assay).This may be due to efficient consumption of cytokines expressed in the TME.In addition, the injection of the vehicle with cytokine-encoding plasmids into normal mice showed no toxicity.However, systemic immune responses against tumor cells were significantly increased by the cytokine gene treatment.As a result, an increase in systemic immunity may be caused by the extension of the anti-tumor immune responses generated in the TME.
Overall, these results confirm that tumor-specific and efficient gene transfection can be achieved using synthetic vehicles, which promote the expression of the transfecting gene.Furthermore, transfection of genes encoding CD40L, GM-CSF, and IFN-γ into tumor cells efficiently induced both acquired and innate immunity against tumors in the TME, significantly decreased the immune escaping molecule PD-L1 on tumor cells, and successfully suppressed the growth of visceral tumors.Therefore, our method is a promising tool for future clinical applications.

3. 2 |
Effect of in vivo cytokine gene transfection on tumor therapy In our previous study with a subcutaneous tumor model, we initiated treatment by transfection of IFNγ-and CD40L-plasmids along with DC administration 12 days %lysis = (experimental release − minimum release) ∕(maximum release − minimum release) × 100.

F I G U R E 1 F I G U R E 3
Expression of transfected genes.A plasmid coding the green fluorescent protein (GFP) gene ([GFP]) or noncoding plasmid ([None]) was encapsulated in the synthetic vehicle and intravenously (i.v.) injected into BALB/c mice bearing CT26WT tumors in the lungs.Two days after the injection, the lungs were excised, and the expression of GFP was examined by immunohistochemistry (IHC).(A) Right: typical photo of lung tissue collected from the [GFP]-transfected mouse.Left: typical photo of lung tissue collected from the [None] transfected mouse.Brown color products indicate expression of GFP.(B) Percentage of GFP-expressing cells in tumors in the mice injected with [None] or [GFP].Three mice from each injection group were used.Three sections were made from the tumor in each mouse.Cells in three or four high-power fields were counted in each section.More than 1000 cells were counted in each tumor.Results are expressed as the mean ± SEM of three experiments.aftertumor cell injection when the diameter of the tumors was 0.4-0.6 cm and obtained 67% CR by four treatments at 7 days intervals.8Given that lung tumors were detected at least 8 days after tumor cell injection in preliminary experiments, we first injected [CD40L], [GM-CSF], [GM-CSF + CD40L], and [IFNγ + CD40L] plasmids 8 days after the tumor injection using the synthetic vehicle.However, after four injections at 7 days intervals, none of the treatments elicited effects on survival, and all mice died 35 days after treatment (Figure3A).Then, treatments were initiated the day after tumor cell injection.As shown in Figure3B, 40% (4/10) of the [IFNγ + CD40L]-treated mice and 20% (2/10) of the [GM-CSF + CD40L], [CD40L], and [GM-CSF]-treated mice survived by the designated endpoint, and no tumor cells were observed in the surviving mice either grossly or by histological examination (CR).However, the difference in survival between these four groups and that of the [None]-treated group was not statistically significant (p > .09).We then increased the frequency of the treatment to five times (previously four times).Furthermore, in addition to [IFNγ + CD40L]-and [GM-CSF + CD40L]treatments, we examined [GM-CSF + IFNγ + CD40L]treatment.As shown in Figure 3C, 66.7% (8/12) of the [GM-CSF + IFNγ + CD40L]-treated group, 54.5% (6/11) of the [IFNγ + CD40L]-treated group, and 45.5% (5/11) of the [GM-CSF + CD40L]-treated group survived by the set endpoint, and the survival of these groups was significantly longer than that of the [None]-and [CD40L]treated groups.By the set endpoint, the survival of the [GM-CSF]-treated group was 20% (2/10), and the survival was significantly longer than that of the [None]treated mice.Ten percent (1/10) of the [CD40L]-treated F I G U R E 2 Expression of transfected genes.Plasmid coding the granulocyte-macrophage colony-stimulating factor (GM-CSF) ([GM-CSF]), CD40 ligand (CD40L) ([CD40L]), or interferon gamma ([IFNγ]) was encapsulated in the synthetic vehicle and i.v.injected into BALB/c mice bearing CT26.WT tumors in the lungs.Two days after the injection, the expression of the corresponding cytokine was examined by IHC.Typical photos of tumor and normal regions in the lungs.Brown color products represent expression of the indicated cytokine.Two experiments were performed, and reproductive results were obtained.Effect of cytokine gene transfection into tumors on therapy.Mice were treated with plasmids including the specified cytokine gene(s) using the synthetic vehicle.Kaplan-Meier survival curves of the treated mouse group are shown.Arrows under the horizontal axis indicate the schedule of plasmid injection.(A) Mice were treated four times.Treatments were initiated 8 days after tumor inoculation; n = 10.(B) Mice were treated four times.Treatments were initiated on the day following tumor inoculation; n = 10.(C) Mice were treated five times.Treatments were initiated on the day following tumor inoculation; n = 12 in the [GM-CSF + IFNγ + CD40L] and [None] groups; n = 11 in the [GM-CSF + DC40L] and [IFNγ + CD40L] groups; and n = 10 in the [GM-CSF] and [DC40L] groups.Survival between groups was compared by the log rank test.(a) p < .01versus [None] and [CD40L] and p < .05versus [GM-CSF]; (b) p < .01versus [None] and p < .05versus [DC40L]; (c) p < .05versus [None].
[GM-CSF + IFNγ + CD40L]-treated mice had significantly more granzyme B-expressing cells than those of the [GM-CSF + CD40L]-and [None]-treated mice, and the [IFNγ + CD40L]-treated mice had significantly more granzyme B-expressing cells than those of the [None]-treated mice.As shown in Figure 8, FoxP3 and DC4, a marker of F I G U R E 4 Effect of cytokine gene transfection into tumors on immune status concerning CD11c + cells.Lung tissues were collected from mice 7 days after the second injection with plasmids including the indicated cytokine gene.Collected lung tissues were cleared using tetraethylene glycol dimethacrylate (tetra-EGMA)-crosslinked 2-methacryloyloxyethyl phosphorylcholine (MPC) polymer hydrogel.The cleared tissues were stained with propidium iodide (PI) and an anti-CD11c antibody.Blue and green colors represent the 3D fluorescence images of nuclei and CD11c + cells, respectively.Arrows in [GM-CSF + CD40L], [IFNγ + CD40L], and [GM-CSF + IFNγ + CD40L] indicate the CD11c-expressing areas matched to the location of the PI-stained area.

F I G U R E 5
Effect of cytokine gene transfection into tumors on immune status concerning CD11c + cells.Lung tissues were collected from mice with the indicated treatments 7 days after the second treatment and embedded in paraffin, and tissue sections were prepared.CD11c + cells in tumor areas were examined by IHC.(A) Typical IHC images of the indicated treatment group.Brown-colored cells (pointed with an arrow) are CD11c + cells.(B) Percentage of CD11c + cells in over 3000 cells in the tumor of the indicated treatment group.Results are expressed as the mean ± SEM of three experiments.(a) p < .01versus [None] and p < .05versus [IFNγ + CD40L]; (b) p < .05versus [None]; (c) p < .01versus [None] and [IFNγ + CD40L] and p < .05versus [GM-CSF + CD40L].F I G U R E 6 Effect of cytokine gene transfection into tumors on immune status concerning granzyme B + cells.Lung tissues were collected from mice 7 days after the second injection with plasmids containing the cytokine genes.Collected lung tissues were cleared using tetra-EGMA-crosslinked MPC polymer hydrogel.The cleared tissues were stained with PI and anti-granzyme B antibodies.Blue and green colors indicate the 3D fluorescence images of nuclei and granzyme B + cells, respectively.

F I G U R E 7
Effect of cytokine gene transfection into tumors on immune status concerning granzyme B + cells.Lung tissues were collected from mice (with the treatments specified in the text) 7 days after the second treatment and embedded in paraffin, and tissue sections were prepared.Granzyme B + cells in tumor areas were examined by IHC.(A) Typical IHC images of the indicated treatment group.Brown-colored cells are granzyme B + cells.(B) Percentage of granzyme B + cells in over 3000 cells in the tumor of the indicated treatment group.Results are expressed as mean ± SEM of three experiments.(a) p < .01versus [None]; (b) p < .01versus [None] and p < .05versus [GM-CSF + CD40L].F I G U R Effect of cytokine gene transfection into tumors on immune status concerning CD4 + and FoxP3 + cells.Lung tissues were collected from mice 7 days after the second injection of plasmids with the indicated cytokine gene.The excised lung tissues were cleared using tetra-EGMA-crosslinked MPC polymer hydrogel.The cleared tissues were stained with PI and anti-FoxP3 and anti-CD4 antibodies.Blue, purple, and green colors represent the 3D fluorescence images of nuclei, FoxP3 + cells, and CD4 + cells, respectively.F I G U R E 9 Effect of cytokine gene transfection into tumors on immune status concerning FoxP3 + cells.Lung tissues were collected from mice (with the treatments specified in the text) 7 days after the second treatment and embedded in paraffin, and tissue sections were prepared.FoxP3 + cells in tumor areas were examined by IHC.(A) Typical IHC images of the indicated treatment group.FoxP3 + cells have a brown-colored nucleus (indicated with arrows).(B) Percentage of FoxP3 + cells in over 3000 cells in tumor of the indicated treatment group.Results are expressed as the mean ± SEM of three experiments.(a) p < .05versus [GM-CSF + CD40L] and p < .01versus [GM-CSF + IFNγ + CD40L].

F I G U R E 1 0
Effect of cytokine gene transfection into tumors on immune status concerning CD80 + M1 macrophages.Lung tissues were collected from mice (with the treatments specified in the text) 7 days after the second treatment and embedded in paraffin, and tissue sections were prepared.CD80 + cells in tumor areas were examined by IHC.(A) Typical IHC images of the indicated treatment group.Brown-colored cells are CD80 + cells.(B) Percentage of CD80 + cells in over 3000 cells in the tumor of the indicated treatment group.Results are expressed as the mean ± SEM of three experiments.(a) p < .05versus [None].F I G U R E 1 1 Effect of cytokine gene transfection into tumors on immune status concerning CD163 + M2 macrophages.Lung tissues were collected from mice (with the treatments specified in the text) 7 days after the second treatment and embedded in paraffin, and tissue sections were prepared.CD163 + cells in tumor areas were examined by IHC.(A) Typical IHC images of the indicated treatment group.Brown-colored cells are CD163 + cells.(B) Percentage of CD163 + cells in over 3000 cells in the tumor of the indicated treatment group.Results are expressed as the mean ± SEM of three experiments.(a) p < .05versus [GM-CSF + CD40L], [IFNγ + CD40L], and [GM-CSF + IFNγ + CD40L].

F I G U R E 1 2
Effect of cytokine gene transfection into tumors on immune status concerning PD1 and PD-L1 expression.Lung tissues were collected from mice 7 days after the second treatment with plasmids including the specified cytokine gene.Collected lung tissues were cleared using tetra-EGMA-crosslinked MPC polymer hydrogel.The cleared tissues were stained with PI and anti-PD1 and anti-PDL-1 antibodies to visualize cell nuclei and cells expressing corresponding antigens.Blue, green, and purple colors indicate the 3D fluorescence images of nuclei and expressions of PDL-1 and PD-1.F I G U R E 1 3Effect of cytokine gene transfection into tumors on systemic tumor immunity.Spleen cells, as effector, were collected from mice 7 days after the second treatment with plasmids containing specified cytokine genes.Lysis activity of the spleen cells against (A) YAC-1, the target to detect natural killer (NK) activity; (B) CT26WT, the target to detect cytotoxic T lymphocyte (CTL) activity; and (C) B16, a target to detect lysis activity against allogeneic tumor.**p < .01,*p < .05versus [None] at the same effector/target (E/T) ratio.