Article Text
PDF
Abstract
Background With the successful development of modern immunotherapy, immune checkpoint inhibitors (ICIs) are currently considered potential therapeutic options for patients with cancer. However, the therapeutic potential of ICIs in human cancer is mainly limited by their systemic toxicity and low response rate, which suggests the necessity of local drug delivery with an effective vector and reshaping the immunosuppressive tumor microenvironment (TME) to enhance ICI therapy. Here, we constructed a novel double-gene recombinant oncolytic adenovirus named RCAd-LTH-shPD-L1 based on the RCAd virus platform armed with a DNA fragment encoding an anti-VEGF antibody and shRNA to inhibit PD-L1 expression.
Methods The correct assembly of RCAd-LTH-shPD-L1 was characterized by analyzing its secretion, antigen specificity, and replication using western blotting, ELISA and quantitative PCR, respectively. The in vitro effects of RCAd-LTH-shPD-L1 on cell proliferation, vasculogenic, and cell migration were assessed. Antitumor effects and therapeutic mechanisms were evaluated in vivo using immunodeficient and humanized immune system mouse models. The TME was studied by ELISA, immunohistochemistry and flow cytometry.
Results RCAd-LTH-shPD-L1 cells secreted anti-VEGF antibodies and inhibited the expression of PD-L1 in cancer cells. Moreover, RCAd-LTH-shPD-L1 exerted a specific cytotoxic effect on human cancer cells, but not on murine cancer cells or normal human cells. RCAd-LTH-shPD-L1 elicited a more potent antitumor effect in an immunodeficient mouse model and a humanized immune system mouse model than RCAd-shPD-L1, as demonstrated by the significant decrease in tumor growth. Furthermore, RCAd-LTH-shPD-L1 modulated the TME, which led to lymphocyte infiltration and alteration of their immune phenotype, as characterized by downregulation of anoxic factor HIF-1α and angiogenesis marker CD31, upregulation of cytokine such as IFN-γ, IL-6 and IL-12.
Conclusions In summary, our data demonstrated that the localized delivery of anti-VEGF antibodies and shPD-L1 by engineered RCAd-LTH-shPD-L1 is a highly effective and safe strategy for cancer immunotherapy. Moreover, the data underscore the potential of combining local virotherapy and anti-angiogenic therapy with ICIs as an effective TME therapy for poorly infiltrating tumors.
- CD8-positive T-lymphocytes
- immune checkpoint inhibitors
- oncolytic virotherapy
- tumor microenvironment
- lymphocytes, tumor-infiltrating
Data availability statement
All data relevant to the study are included in the article or uploaded as online supplemental information.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
- CD8-positive T-lymphocytes
- immune checkpoint inhibitors
- oncolytic virotherapy
- tumor microenvironment
- lymphocytes, tumor-infiltrating
WHAT IS ALREADY KNOWN ON THIS TOPIC
Immune checkpoint inhibitors (ICIs) are promising antitumor agents, while their therapeutic effect was limited by drug delivery methods and immunosuppressive tumor microenvironment (TME).
WHAT THIS STUDY ADDS
Our results demonstrated the potential of a double-gene recombinant oncolytic adenovirus, which can secrete anti-VEGF antibodies and inhibit the expression of PD-L1, could promote antitumor immunity by increasing lymphocyte infiltration and reshaping the TME.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
These findings encourage the use of combined local virotherapy and anti-angiogenic therapy with ICIs as an effective TME therapy for poorly infiltrated tumors.
Background
Immune checkpoint inhibitors (ICIs) are cancer immunotherapies that improve the outcomes of patients with many types of cancer.1 However, ICIs have several limitations. First, systemically administered ICIs cause severe side effects in numerous organs.2 3 Previous studies have shown that dose-dependent adverse events of cytotoxic T lymphocyte antigen 4 (CTLA-4) blockade ranging from mild to moderate occurred in more than 70% of patients.4 Compared with CTLA-4, the adverse events related to programmed death 1 (PD-1)/PD1 ligand (PD-L1) blockade are fewer, occurring in 16%–37% of patients receiving PD-1 inhibitors, and 12%–24% of those who received PD-L1 inhibitors.5 The use of appropriate delivery platforms can limit drug exposure to specific tissues and minimize off-target effects, which may be an effective measure to reduce the systemic toxicity and side effects of ICIs.2 Second, due to the low infiltration of immune cells and the immunosuppressive tumor microenvironment (TME) in solid tumors, the low response rate and drug resistance to ICI are still troubling for the effective treatment of most patients.6 On average, only approximately 12% of patients gain benefits across all tumor types, and patients with low or lack of antitumor immunity respond poorly to ICI treatment.7 8 Such immunotherapy-insensitive tumors are considered “cold” tumors that lack or have poor T-cell infiltration in TME. In contrast, “hot” tumors imply greater sensitivity to immunotherapy, whose characteristics include an immune-cell-rich TME with high percentage of CD8+ T cells and immune-stimulatory cytokine production.6 Therefore, increasing immunological heat and remodeling the TME can be a potential way to expand the benefits of ICI treatment and enhance its therapeutic effect.
Oncolytic viruses (OVs) could be a potential choice to reduce the systemic toxicity and side effects of ICIs owing to their advantage of delivering drugs as viral vectors. The large backbone of OVs can be manipulated by inserting therapeutic transgenes and exploiting selective viral replication for the concentrated delivery of ICIs within the tumor.6 Furthermore, OVs are regarded as a powerful novel form of immunotherapy not only because of their ability to specifically target and lyse tumor cells, but most importantly their capacity to break down the immune suppressive environment to create an immunologically “hot” environment that enhances antitumor immunity.9 Preclinical studies have demonstrated the remodeling of the TME by OVs, with an influx of CD8+ T cells and NK cells, upregulation of inflammatory genes, and a switch to an immune-stimulatory TME in mouse breast and renal cancer models. In addition, ICI treatment combined with OVs has shown enhanced therapy in GBM tumor-bearing mice, and similar trends have been observed in other studies of poorly immunogenic mouse colon and ovarian cancer models, suggesting a synergistic effect of combination therapy on TME remodeling.6 10 11
PD-L1 has been widely studied in clinical trials as a biomarker for ICI-based immunotherapy, and its expression is related to ICI efficacy.3 High expression of PD-L1 in tumors can bind to PD-1 on the surface of T cells, limiting T cell activation and inducing its depletion, leading to immune escape from tumors.12 Moreover, increasing evidence has shown that degradation of the PD-L1 protein could lead to enhanced immunotherapy for cancer.13 Furthermore, PD-L1 on the tumor endothelium can inactivate T cells within the tumor vascular lumen, which become functionally anergic before migrating across the vessel wall and entering the TME.7 Hence, inhibiting the expression of PD-L1 is not only an important strategy in ICI treatment, but also a potential approach to remodel the TME by overcoming immunosuppression and activating T cells, thus leading to enhanced ICI therapy.
Abnormal tumor angiogenesis is a major obstacle to T cell infiltration and an important component of the immunosuppressive TME.7 Among the various angiogenic molecules, excessive levels of vascular endothelial growth factor (VEGF) in the TME induce tumor-associated immunosuppression via diverse pathways. Increased VEGF has been reported to directly inhibit cytotoxic T lymphocyte trafficking, proliferation, and effector function, suggesting the possibility of using VEGF as a therapeutic target to enhance antitumor immunity.14 In addition to improving the immunosuppressive TME, targeting VEGF can also enhance blood flow in the main vessels and reduce hypoxia within the tumor mass, leading to significant improvements in drug delivery.15
In this study, we constructed a double-gene recombinant oncolytic adenovirus, RCAd-LTH-shPD-L1, with a DNA fragment encoding an anti-VEGF antibody and shRNA to inhibit PD-L1 expression. After assessing the biological activity of RCAd-LTH-shPD-L1 in vitro, we examined its antitumor effect in vivo, not only in nude mice, but also in human peripheral blood mononuclear cells (PBMCs) engrafted in severely immunodeficient M-NSG mice (humanized immune system mouse model) to more closely mimic human antitumor immune responses. Further dissection of the underlying mechanism of RCAd-LTH-shPD-L1 treatment showed that RCAd-LTH-shPD-L1 promoted the infiltration and activation of tumor-infiltrating T cells (TILs) by remodeling the immunosuppressive TME. Our study showed that RCAd-LTH-shPD-L1 is not only an effective tool with potential clinical applications for cancer treatment, but also a promising way to enhance ICI therapy by increasing immunological heat.
Methods
Cell culture
Human embryonic kidney (HEK) 293 cells and the human gastric adenocarcinoma cell line SGC-7901 were obtained from the Chinese Academy of Sciences, Shanghai Institutes for Biological Sciences (Shanghai, China). The poorly differentiated human mucinous adenocarcinoma cell line MGC-803 was purchased from Shanghai Fuheng Biotechnolog (Shanghai, China). The human astroglioma cell lines U87MG and U251MG and human retinal pigment epithelial cell line RPE1 were purchased from Wuhan Procell Life Science & Technology (Shanghai, China). The mouse glioma cell line GL261 was donated by the laboratory of Nanjing University. Human umbilical vein endothelial cells (HUVECs) were purchased from ScienCell Research Laboratories (Carlsbad, California, USA) and cultured in endothelial cell medium supplemented with 5% fetal bovine serum (FBS), 1% endothelial cell growth supplement, and 1% penicillin-streptomycin solution. These cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Hyclone, Logan, Utah, USA) supplemented with 10% FBS) (Sera Pro SA), and 1% penicillin-streptomycin in an incubator at 37℃ with 5% CO2.
Plasmid construction
Oncolytic adenovirus plasmids derived from human adenovirus serotype 5 were purchased from the MiaoLing Plasmid Platform (Wuhan, China). The adenovirus, RDAd, lacked the E1 and E3 regions required for replication. The recombinant adenovirus was constructed by cloning the hTERT promoter and the E1A gene into the E1 region of the plasmid, named RCAd. shRNAs targeting human PD-L1 (NM_001267706.1; GenBank) were synthesized by GENEZIE (Suzhou, China) and cloned into RDAd. The sequences of PD-L1 shRNA were as follows: PD-L1 shRNA (–), 5′-AA TCCCTAATTTGAGGGTCAG-3′; PD-L1 shRNA1, 5′-AA TCCCTAATTTGAGGGTCAG-3′; PD-L1 shRNA2, 5′-CCTACTGGCATTTGCTGAACGC-3′; PD-L1 shRNA3, 5′-GACCTA TATGTGGTAGAGTAT-3′; PD-L1 shRNA4, 5′-CGAA TTACTGTGAAAGTCAA T-3′; PD-L1 shRNA5, 5′-CTGACA TTCATCTTCCGTTTA-3′; The most effective sequence of shPD-L1 was selected and cloned with U6 promoter and SV40 terminator to construct RCAd-shPD-L1 and RCAd-LTH-shPD-L1. The sequence encoding anti-VEGF antibodies (bevacizumab) was obtained from the National Center for Biotechnology Information database and was inserted after E1A and linked by a T2A sequence to construct RCAd-LTH and RCAd-LTH-shPD-L1.
Cell transfection
The plasmid was mixed with polyether imide (PEI) of the same quality in Opti-MEM (Gibco, Grand Island, New York, USA) for 20 min. The mixture was then added to the cell supernatant for transfection. The cells were collected 48 hours after transfection for total RNA and protein extraction.
Adenovirus production and purification
The recombinant adenoviruses were packaged and produced in HEK 293 cells using PEI. They were purified via cesium chloride double-density gradient centrifugation and concentrated via dialysis bag activation. Titers of the amplified viruses were determined using a classical plaque assay.
Quantitative PCR
The concentration and quality of the RNA and DNA were determined using a Nano Photometer N50 Touch (Implen, München, Germany). RNA was reverse transcribed to cDNA using the TaKaRa PrimeScript RT Master Mix (RR036A; TaKaRa Bio, Shiga, Japan). Quantitative PCR (qPCR) was performed on a LightCycler 96 Real-Time System (Roche, Basel, Switzerland). The corresponding primer pairs used were as follows: PD-L1, primer-F-5′-ACTGGCATTTGCTGAACG-3′ and primer-R-5′-TCCTCCATTTCCCAATAGAC-3′; E1A, primer-F-5′-ACCGGAGGTGATCGATCTTA-3′ and primer-R-5′- GGGTGCTCCACA TAA TCTAACA-3′; GAPDH, primer-F-5′-TGACTTCAACAGCGACACCCA-3′ and primer-R-5′- CACCCTGTTGCTGTAGCCAAA-3′; and anti-VEGF antibody, primer-F-5′- CCAAGGTGGAGATCAAGAGAAC-3′ and primer-R-5′-GTACACCTTGTGCTTCTCGTAG-3′. The relative expression of the blank samples was defined as 1.
Western blot analysis
Total protein was isolated from cells or animal tissues using a radio immunoprecipitation assay buffer (Beyotime Institute of Biotechnology, Shanghai, China). Protein was quantified using the Wanleibio BCA Protein Assay Kit (WLA004; Wanleibio Company, Shenyang, China). Equal amounts of protein were separated via sodium dodecyl-sulfate polyacrylamide gel electrophoresis and electrophoretically transferred onto polyvinylidene fluoride membranes. To avoid non-specific binding, the membranes were blocked with 5% non-fat milk for 45 min at room temperature. Subsequently, the membranes were incubated with anti-PD-L1 (Proteintech, Rosemont, Illinois, USA), anti-h CD-31 (Wuhan Servicebio Technology, Wuhan, China), VEGF-A (Cell Signaling Technology, Danvers, Massachusetts, USA), or anti-HIF-1α (Abcam, Cambridge, UK) antibody (all diluted 1:5000) for 2 hours at room temperature. After washing, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies (diluted 1:5000; Proteintech) for 2 hours at room temperature. After washing thrice with Tris-buffered saline with Tween 20, the immunoblots were visualized using an Enhanced Chemiluminescence (ECL) Kit (Tanon Science & Technology, Shanghai, China) and scanned using Tanon AllDoc_x software. The integrated density values were quantified using ImageJ software (National Institutes of Health, USA).
ELISA analysis
ELISA was performed to analyze the anti-VEGF-binding ability of antibodies and quantify the protein expression of interferon (IFN)-γ, interleukin (IL)-6, and IL-12 (Ruixin Biotech). The standard was diluted 1:1 with phosphate-buffered saline (PBS) to an initial concentration of 1000 pg/mL. The samples were plated in an hVEGF (1 mg/mL; Z02689; GenScript Biotech, Piscataway, New Jersey, USA)-coated immunotransparent standard plate (468667; Thermo Fisher Scientific, Waltham, Massachusetts, USA) at a density of 100 µL per well. Incubation was performed at room temperature in a tabletop shaker for 2 hours. Then, 100 µL of 10 µg/mL detection antibody was added to each well. The cells were incubated at room temperature on a shaker for 45 min. Finally, 100 µL of the chromogenic substrate 3,3' 5,5'-tetramethylbenzidine was added to each well, and the plate was incubated at room temperature for 5–30 min in the dark. Absorbance was measured at 405 nm using a spectrophotometer (Tecan Life Sciences, Männedorf, Switzerland).
Isolation of PBMCs
PBMCs were isolated from whole blood via Ficoll density gradient centrifugation. A commercially available lymphocyte separation medium was added to whole blood, and the suspension was centrifuged at 800×g for 30 min. The cells were plated in a six-well plate containing complete RPMI 1640 medium at a density of 5×106 cells/mL and were incubated in a 37°C incubator for 2 hours.
Generation of cytokine-induced killer cells
Cytokine-induced killer (CIK) cells were generated from PBMCs obtained from volunteers using various animal PBMC separation liquid kits (Solarbio Life Sciences, Beijing, China) according to the manufacturer’s protocols and as previously described. To generate CIK cells, non-adherent PBMCs were prepared in complete RPMI 1640 medium with 1000 U/mL IFN-γ (Solarbio Life Sciences). After 24 hours of incubation, 50 ng/mL mouse anti-human CD3 monoclonal antibody (ACROBiosystems, Newark, Delaware, USA) and 1000 U/mL IL-2 (Solarbio Life Sciences) were added. The CIK cells were incubated at 37°C in a humidified atmosphere of 5% CO2 and were subcultured every 3 days with cytokine replenishment.
Coculture experiment
Tumor cells treated with RDAd-shPD-L1 and RDAd-shRNA (–) treatment were used as target cells, and CIK cells were used as effector cells. The effector and target cells were inoculated into the culture plate at ratios of 10:1, 20:1, or 40:1. Some wells containing effector or target cells were used as controls. The culture plates were placed in a humidified incubator with 5% CO2 at 37°C for 48 or 72 hours, after which 100 µL of medium was collected and mixed with 20 µL of 3-(4,5-dimethylthiazol-2-yl)-2,5 diphenyl tetrazolium bromide (MTT) solution in the incubation for another 4 hours. The optical density (OD) was measured at 450 nm using a microplate reader. Survival was calculated using the following formula: Survival (%)=(cocultured group−effector group)/target group×100%.
Tube formation and vasculogenic mimicry formation assay
Growth factor-reduced Matrigel (250 µL; Corning, New York, USA) was added to the wells of a 96-well plate, which was incubated at 37°C for 30 min. HUVECs and U87MG cells were cultured in serum-free medium for 12 hours before adding the virus. After culturing for 12 hours, the cells were seeded in the Matrigel matrix layer at a density of 1×105 cells/well (HUVEC) or 1×103 cells/well (U87MG). Images were acquired for quantification after 4 hours of incubation. The area covered by the pipeline network was determined using optical imaging. Pictures of the pipeline were scanned using Adobe Photoshop and quantified using the Image-Pro Plus software (Media Cybernetics, Rockville, Maryland, USA).
Transwell migration assay
The Transwell migration assay was conducted using Transwell chamber (Corning). For the migration assay, 2×104 cells suspended in 100 µL of serum-free DMEM were seeded in the upper compartment of the chamber, and 800 µL of DMEM with 10% FBS was added to the lower compartment. After incubation for 48 hours, the cells were fixed with 4% paraformaldehyde for 30 min and stained with 0.1% crystal violet. Non-migrating cells in the upper chamber were carefully removed using cotton swabs. The migrated cells on the lower surface were photographed under a microscope in five randomly selected visual fields and quantified using Image-Pro Plus V.6.0 software (Media Cybernetics).
Immunodeficient mice model
Female BALB/c nude mice aged 4 weeks old were purchased from Weitong Lihua (Beijing, China). All mice were maintained in a specific pathogen-free facility at China Medical University. Tumor cells (2×106 /mouse) in 0.1 mL of PBS were injected subcutaneously into the right inguinal region of nude mice. Tumor volume was determined using the formula: tumor volume=length×width2×0.5 mm3. When the tumor volume reached 50 mm³, the animals were randomly divided into each group (n=6 per group). The tumor volume was measured every 3 days using a caliper. Tumors were harvested 21 days after cell engraftment.
Immunofluorescence
The paraffin sections were deparaffinized in water and subsequently subjected to antigen repair. The blocking solution was blocked by dropping BSA for 30 min, and the blocking solution was dried and incubated on the sections with primary antibodies prepared in PBS dropfold at 4°C overnight. The cells were shaken thrice for 5 min each in PBS (pH 7.4). After drying the sections, they were treated with DAPI staining solution and incubated in the dark for 10 min. After washing thrice, autofluorescence quenching was performed, and the slides were sealed after washing with running water for 10 min. Finally, the images were captured under a microscope (Nikon AX).
Humanized immune system mouse model
Female NOD-PrkdcscidIl2rgem1Smoc (M-NSG) mice aged 6 weeks old were purchased from Shanghai Model Organisms Center (Shanghai, China). NSG mice were intraperitoneally injected with 1×107 freshly isolated PBMCs from healthy donors, and 2×106 SGC-7901 cells were implanted subcutaneously into the right flank of the mice. After 8 days, the mice were treated with PBS (mock), RCAd, RCAd-shPD-L1, RCAd-LTH, or RCAd -LTH-shPD-L1. Tumor volumes and body weights were recorded every 2 days.
In vitro T cell activation assay
Human peripheral PBMCs (1×106) obtained from a healthy donor were stimulated with anti-CD3/CD28-coated beads in 24-well plates for 48 hours. U87MG cells were plated in 96-well-plates (1×104 cells/well) and infected for 48 hours. Preactivated PBMCs (1×105 cells/well) were then added and cocultured with U87MG for 24 hours. Flow cytometry was used to detect cell activation.
Flow cytometry
To obtain single-cell suspensions, tumor tissues were cut into small fragments and incubated in lysis buffer (1 mg/mL collagenase IV (Absin Bioscience, Shanghai, China) and 10 unit/mL DNase I (New Englang Biolabs, Ipswich, Massachusetts, USA) in RPMI 1640 medium) for 2 hours at 37℃. Tumors and spleens were filtered through a 70 µm Falcon cell strainer (Corning). The cells were then subjected to red blood cell lysis using 2 mL of red blood cell lysis buffer (Absin Bioscience). Cell suspensions were stained using the Fixable Viability Kit (BioLegend, San Diego, California, USA) for 20 min to eliminate dead cells, incubated for 15 min with a fragment crystallizable block antibody, and subsequently stained with the corresponding anti-human antibodies for 30 min at 4°C. The stained cells were analyzed using a BD Fortessa (BD Biosciences, San Jose, California, USA). Flow cytometry data were analyzed using FlowJo software (TreeStar, Ashland, Oregon, USA). The antibodies and fluorescent dyes (all from BioLegend unless otherwise stated) used were as follows: Zombie NIR Fixable Viability Kit, Human BD fragment crystallizable Block (BD Biosciences), Pacific Blue anti-human CD45, Brilliant Violet 605 anti-human CD3, FITC anti-human CD4, PerCP/Cyanine5.5 anti-human CD8, PE anti-human CD366 (Tim-3), PE anti-human CD8a, PE anti-human CD11c, APC anti-mouse/human CD11b, and PE anti-human CD107a.
Statistical analysis
Statistical analyses were performed using GraphPad Prism software (V.7; GraphPad Prism Software, La Jolla, California, USA). Differences were considered significant when p<0.05.
Results
Construction and validation of RCAd-LTH-shPD-L1
As shown in figure 1A, we constructed the recombinant oncolytic adenovirus, RCAd-LTH-shPD-L1, armed with shPD-L1 and a DNA fragment encoding an anti-VEGF antibody “Light-T2A-Heavy (LTH)”. The sequence of “LTH” was the same as bevacizumab in this study. Tumor-selective adenovirus RCAd was used as a control. RCAd-shPD-L1 and RCAd-LTH, armed with shPD-L1 and anti-VEGF antibodies, respectively, were used as negative controls (NCs). To verify the correct assembly of RCAd-LTH-shPD-L1, we infected SGC-7901 cells with the four adenoviruses. Compared with the control group, there was a significant decrease in PD-L1 expression in the RCAd-shPD-L1 and RCAd-LTH-shPD-L1 groups at the RNA and protein levels (figure 1B–D). In addition, the coculture experiment of PBMCs and infected cells showed the RCAd-LTH-shPD-L1 group had better immune activation effect compared with other groups (online supplemental figure S2A and B). We performed western blotting and ELISA to detect the expression and antigen specificity of anti-VEGF antibodies secreted by the infected cells. Specific bands were detected in the RCAd-LTH and RCAd-LTH-shPD-L1 groups, similar to those detected in the bevacizumab group (figure 1E). In addition, the secretion of anti-VEGF antibodies was only detected in the RCAd-LTH-shPD-L1 group and increased over time (figure 1F,G).
Supplemental material
Construction and validation of RCAd-LTH-shPD-L1. (A) Structure of RCAd, RCAd-LTH, RCAd-shPD-L1, and RCAd-LTH-shPD-L1. The recombinant oncolytic adenovirus RCAd-LTH-shPD-L1 was constructed with the inverted terminal repeat ITR of the adenovirus vector, the E1A sequence driven by the hTERT promoter, the anti-VEGF antibody sequence connected with a T2A after E1A, and the shPD-L1 sequence driven by the U6 promoter. ITR: Inverted Terminal Repeat. hTERT: human Telomerase Reverse Transcriptase. (B) mRNA expression of PD-L1 was detected via qPCR after 48 hours of infection (MOI=1). Data are expressed as the mRNA level relative to that of gapdh. MOI: Multiplicity Of Infection. (C) Protein expression of PD-L1 was detected via western blotting after 48 hours of infection (MOI=1). Data are expressed as the protein level relative to that of gapdh. (D) Quantification of relative PD-L1 protein level. (E) Anti-bevacizumab antibodies secreted by infected tumor cells in the cell supernatant were detected via western blotting after 48 hours of adenovirus infection (MOI=1). (F) The binding ability of anti-VEGF antibody to h-VEGF in the supernatant of infected tumor cells was detected via ELISA. The commercial bevacizumab treatment was used as a positive control. (G) Secretion of anti-VEGF antibody in the supernatant of RCAd-LTH- or RCAD-LTH-shPD-L1-infected cells at different time points was determined by ELISA. Data are presented as the mean±SEM. Statistical significance was obtained using Student’s t-test (n=3 per group). NS, not significant; **, p<0.01; ***, p<0.001; ****, p<0.0001.
One advantage of oncolytic adenoviruses is their ability to replicate specifically in tumor cells.16 To verify the specificity of the recombinant oncolytic adenovirus, we evaluated replication of RCAd-LTH-shPD-L1 in different cell lines. There did no replication of RCAd-LTH-shPD-L1 in the murine glioma cell line GL261 or the human retinal cell line RPE1, whereas its replication was significantly increased in human cancer cell lines, proving the specificity of RCAd-THL-shPD-L1 in human tumor cells (online supplemental figure S3A–F).
In vitro testing of RCAd-LTH-shPD-L1
After assessing the ability of RCAd-LTH-shPD-L1 to infect and replicate human tumor cells, we evaluated its antitumor efficacy in vitro. We first evaluated the cytotoxic effects of RCAd-LTH-shPD-L1 on cell proliferation. Similar to RCAd, RCAd-LTH-shPD-L1 specifically inhibited human cancer cell proliferation, but did not have a significant effect on GL261 or RPE1 (figure 2A). To evaluate the effect of RCAd-LTH-shPD-L1 on angiogenesis, tube formation was induced in the HUVEC. Cells treated with bevacizumab were used as positive controls. As shown in figure 2B, long thin tubular structures similar to reticular capillaries were significantly damaged in the bevacizumab group. The damage to junction points and decrease in length in the RCAd-LTH-shPD-L1 group were comparable to those in the bevacizumab group, and both were better than those in the RCAd group, indicating the inhibitory effect of RCAd-LTH-shPD-L1 on tube formation (figure 2C,D). Furthermore, vasculogenic mimicry (VM) formation analysis, which may be a potential target for improving anti-angiogenic strategies,17 18 suggested an inhibitory effect of RCAd-LTH-shPD-L1 on VM formation in U87MG cells (online supplemental figure S6A and B). In addition, compared with the RCAd and bevacizumab groups, there was a remarkable reduction in the number of migrated cells in the RCAd-LTH-shPD-L1 group, indicating efficient inhibition of RCAd-LTH-shPD-L1 on the cell migration (online supplemental figure S6C and D).
Effect of RCAd-LTH-shPD-L1 on cancer cells in vitro. (A) Proliferation of different cell lines was detected via CCK-8 analysis after 48 hours of infection (MOI=1). MOI: Multiplicity Of Infection. (B) Representative images of tube formation in HUVECs infected for 48 hours. HUVECs: Human Umbilical Vein Endothelial Cells. (C) Quantification of the junction point number. (D) Quantification of the total length. Data are presented as the mean±SEM. Statistical significance was obtained using Student’s t-test (n=3 per group). NS, not significant; *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001.
RCAd-LTH-shPD-L1 could suppress tumor growth by improving TME in immunodeficient mouse model
After assessing the ability of RCAd-LTH-shPD-L1 in vitro, we evaluated its antitumor efficacy in vivo. Female immunodeficient BALB/c nude mice aged 4 weeks old (n=6/group) at an age of 4 weeks were implanted with 2×106 U87MG cells. The animals were separated into six groups and injected with a single dose of PBS (0.1 mL) or virus (1×108 pfu) three times into the right flank when their tumor volume reached 50 mm3 (figure 3A). The results showed that both the virus and the combination of RCAd-shPD-L1 with bevacizumab delayed tumor growth. RCAd-LTH-shPD-L1 was significantly more effective than RCAd at suppressing tumor volume and weight. Compared with RCAd-LTH-shPD-L1, the combination of RCAd-shPD-L1 with bevacizumab demonstrated the most powerful therapeutic effect, with no significant differences (figure 3B,C). Repeated experiments with SGC-7901 cells showed similar results (figure 3D,E).
Antitumor effect of RCAd-LTH-shPD-L1 in an immunodeficient mouse model. (A) Experimental flow chart of treatment in an immunodeficient mouse model. (B) Tumor growth curve of a U87MG-engrafted immunodeficient mouse model. (C) Tumor weight of a U87MG-engrafted immunodeficient mouse model. (D) Tumor growth curve of a SGC-7901-engrafted immunodeficient mouse model. (E) Tumor weight of a SGC-7901-engrafted immunodeficient mouse model. Statistical significance was obtained using two-way ANOVA for (B, D). Data are presented as the mean±SEM. Statistical significance was obtained using Student’s t-test (n=6 per group). NS, not significant; **, p<0.01; ***, p<0.001; ****, p<0.0001. ANOVA, analysis of variance.
High expression of PD-L1 in tumors could promote immune escape, and CD31 is primarily a marker of tumor angiogenesis, both of which are abnormally expressed in the TME.12 ,19 We first evaluated the expression of PD-L1 and CD31 by western blotting. Compared with the control group, there was a remarkable reduction in PD-L1 and CD31 expression in the RCAd-LTH-shPD-L1 group, indicating a significant improvement in the TME (figure 4A–C). In addition, there was significantly reduction of VEGF-A and HIF-1α expression at protein level in the RCAd-LTH-shPD-L1 group, suggesting the normalization of vasculature in tumor (figure 4D–G). Moreover, based on the immunodeficient mouse model, we confirmed by immunofluorescence staining that RCAd-LTH-shPD-L1 could replicate and express anti-VEGF antibodies in tumors (online supplemental figure S4).
Improvement of TME by RCAd-LTH-shPD-L1 in an immunodeficient mouse model. (A) Detection of PD-L1 and CD-31 expression in an immunodeficient mouse model via western blotting. (B) and (C) Quantification of (A). (D) Detection of VEGF-A expression in an immunodeficient mouse model via western blotting. (E) Quantification of (D). (F) Detection of HIF-1α expression in an immunodeficient mouse model via western blot. (G) Quantification of (F). Data are presented as the mean±SEM. Statistical significance was obtained using Student’s t-test (n=3 per group). NS, not significant; *, p<0.05; **, p<0.01; ***, p<0.001. TME, tumor microenvironment.
RCAd-LTH-shPD-L1 could enhance the antitumor immunity by activating TILs in the humanized immune system mouse model
Despite the remarkable efficacy of RCAd-LTH-shPD-L1 in an immunodeficient mouse model, its ability to overcome immunosuppression in vivo has not been tested. To detect the effect of RCAd-LTH-shPD-L1 on the immunosuppressive TME, we constructed a humanized immune system mouse model by intraperitoneally injecting human PBMCs into severely immunodeficient M-NSG mice at the age of 6 weeks. We evaluated the tumor-infiltrating lymphocytes and mouse splenocytes in a humanized immune system mouse model using flow cytometry and found that most of the tumor-infiltrating lymphocytes were T cells, indicating the successful construction of a humanized immune system mouse model (figure 5A, online supplemental figure S5A and B). Female humanized immune system mice were implanted with 2×106 SGC-7901 cells. The animals were separated into five groups and injected with PBS (0.1 mL) or virus (1×108 pfu) into the right flank (figure 5B). Compared with PBS treatment, recombinant oncolytic adenovirus treatment significantly reduced tumor volume. However, RCAd failed to effectively inhibit tumor growth over time, probably because of immunosuppression of the TME. The remarkable inhibitory effect of RCAd-shPD-L1 on tumor growth suggests that RCAd can improve the ability of shPD-L1 to overcome the immunosuppressive TME. Moreover, RCAd-LTH-PD-L1 demonstrated the most effective inhibitory effect on tumor growth, even in a humanized immune environment (figure 5C, online supplemental figure S7A and B).
Antitumor effect of RCAd-LTH-shPD-L1 in a humanized immune system mouse model. (A) Gating strategy of CD3+, CD4+, or CD8+ T cells in the tumor and spleen of a humanized immune system mouse model. Left plots gated on live-cell population and right plots gated on the left. (B) Experimental flow chart of treatment in a humanized immune system mouse model. Human immune system was reconstituted in NSG mice via intraperitoneal injection with 1×107 PBMCs. At 8 days postengraftment of 2×106 SGC-7901 cells, the mice were treated with PBS, RCAd, RCAd-shPD-L1, RCAd-LTH, or RCAd-LTH-shPD-L1. (C) Tumor growth curve of the SGC-7901 cell-engrafted humanized immune system mouse model. (D) Weight of the SGC-7901-engrafted humanized immune system mouse model. Statistical significance was obtained using two-way ANOVA (n=5 per group). NS, not significant; ****, p<0.0001. ANOVA, analysis of variance.
The in vitro expression of cytokines in SGC-7901 cells, such as IFN-γ, IL-6 and IL-12, was detected by ELISA. Compared with other group, RCAd-LTH-shPD-L1 significantly upregulated the expression of IFN-γ, IL-6 and IL-12, indicating the in vitro immunity activation effect of RCAd-LTH-shPD-L1 on cancer cells (figure 6A–C). Furthermore, we performed flow cytometry analysis of TILs and mouse splenocytes to detect the effect of RCAd-LTH-shPD-L1 on T cell populations. The results showed that the proportion of CD3+ T cells in TILs increased in the RCAd-LTH-shPD-L1 group compared with that in the RCAd group, including an increased proportion of CD4+ T cells and CD8+ T cells. The CD4+/CD8+ ratio increased in the RCAd-LTH-shPD-L1 group; however, the difference was not statistically significant (figure 6D). In addition, RCAd-LTH-shPD-L1 promoted the expression Tim-3 in CD3+ T cells, including CD4+ and CD8+ T cells (figure 6E). These results demonstrate that RCAd-LTH-shPD-L1 enhanced the function of T cells in killing tumor cells.
Enhancement of antitumor immunity by RCAd-LTH-shPD-L1 in a humanized immune system mouse model. (A–C) Detection of cytokine (A) IFN-γ, (B) IL6, and (C) IL-12 expression in SGC-7901 cells via ELISA. (D) Proportions of CD3+, CD4+, and CD8+ T cells among tumor-infiltrating T cells in a humanized immune system mouse model. (E) Flow cytometry analysis of Tim-3 expression in tumor-infiltrating T cells. Representative histograms (top) and median fluorescence intensity (MFI) bar graphs (bottom). Statistical significance was determined using Student’s t-test (n=3–4 per group). NS, not significant; *, p<0.05; **, p<0.01.
Safety evaluation of RCAd-LTH-shPD-L1 in humanized immune system mouse model
To evaluate the safety of RCAd-LTH-shPD-L1 in vivo, we measured the body weight of each mouse in the different groups. However, there were no significant differences in mouse weights between the different groups, even in the PBS treatment group (figure 5D). Flow cytometric analysis of splenocytes in SGC-7901 engrafted humanized immune system mouse model indicated that RCAd-LTH-shPD-L1 did not change the proportion of CD3+, CD4+ T cells or CD8+ T cells in mouse splenocytes (figure 7A–C). The expression of cytokines IFN-γ, IL-6 and IL-12 was detected in the spleen of SGC-7901 engrafted humanized immune system mouse model by ELISA. As shown in figure 7D–F, there was no significant difference between the PBS and RCAd-LTH-shPD-L1 groups, suggesting that RCAd-LTH-shPD-L1 enhanced antitumor immunity rather than systemic immunity, showing high safety in vivo.
Safety evaluation of RCAd-LTH-shPD-L1 in a humanized immune system mouse model. (A–C) Proportions of (A) CD3+, (B) CD4+, and (C) CD8+ T cells in the spleen of a humanized immune system mouse model. (D–F) Detection of cytokine (D) IFN-γ, (E) IL6, and (F) IL-12 expression in splenocytes in a humanized immune system mouse model via ELISA. Statistical significance was determined using Student’s t-test (n=3–4 per group). NS, not significant; *, p<0.05; **, p<0.01. PBS, phosphate-buffered saline.
Discussion
In this study, we developed a recombinant oncolytic adenovirus to reduce toxicity and enhance the efficacy of ICIs treatment, RCAd-LTH-shPD-L1, by arming with a DNA fragment encoding an anti-VEGF antibody and shRNA to inhibit PD-L1 expression. This strategy combines direct oncolysis of malignant cells with inhibition of immunosuppressive factor expression. PD-L1, an ICs, was chosen as a suitable pan-cancer marker because it is widely expressed in tumors, not only in tumor cells but also in immunosuppressive cells, which are important mediators of antitumor immune inhibition and tumor progression.20 VEGF is not only an important factor in the abnormal vasculature caused by tumors, but is also an important component of the immunosuppressive microenvironment, making inhibition of VEGF expression a potential approach to enhancing ICIs therapy.7 14 Indeed, the combination of PD-L1 inhibitors and anti-VEGF/VEGFR2 monoclonal antibodies is regarded as a promising treatment for cancer immunotherapy, showing efficacy in clinical research.21–23 However, PD-L1 inhibitors and anti-VEGF antibodies exhibit short biological activities in clinical cancer therapy because of their short half-life of antibody drugs.24 The large backbone of OVs can be manipulated by inserting therapeutic transgenes, exploiting selective viral replication for concentrated delivery of drugs within the tumor.6 In this sense, OVs are promising candidates for the intratumoral delivery of PD-L1 inhibitors and anti-VEGF antibodies, and synergies among these three have not been reported.
The recombinant oncolytic adenovirus with only the E1A sequence retained and arranged with shPD-L1 and anti-VEGF antibody sequences was named RCAd-LTH-shPD-L1 (figure 1A). Adenoviral E1 enables infected cells to enter the S phase of the cell cycle and prevents premature apoptosis. Therefore, E1 is the preferred target for generating tumor-selective replicating adenoviruses.25 E1 includes E1A, E1B19KD, and E1B55KD, among which only E1A is essential for the replication of OVs.26 To enable the virus to replicate specifically in human tumor cell lines, we used the human telomere reverse transcriptase promoter to target E1A.27 Anti-angiogenic drugs have been proven to cause side effects such as pneumonitis and cardiovascular events in patients; therefore, we aimed to minimize their distribution in the systemic system by releasing anti-VEGF antibodies specifically in the tumor mass.28 Therefore, to reduce the genome size and specifically express the antibody in the tumor, we linked the anti-VEGF antibody with T2A behind E1A. We used short hairpin RNA (shRNAs) rather than antibodies to inhibit PD-L1 expression for two reasons. On the one hand, Brachtlova et al have shown that shRNA can be long-termed expressed from the genome of an oncolytic adenovirus to silence a target gene in cancer cells.29 However, the molecular weight of the PD-L1 antibody is too high for arm oncolytic adenoviruses.
Considering that different shRNA sequences have different efficacies, we screened shRNA sequences before constructing a recombinant adenovirus. To screen for the shRNA sequence with the greatest effect on the inhibition of PD-L1 expression, we first transfected plasmids carrying shPD-L1 into the SGC-7901 cancer cell line and detected the expression of PD-L1 by qPCR and western blotting. Compared with the PD-L1 mistranslated sequence in the NC group, plasmids carrying different shPD-L1s significantly knocked down the expression of PD-L1 without differences (online supplemental figure S1A–C). Next, we infected tumor cells using a non-replicating adenovirus (RDAd) with different recombinant plasmids. RDAd is a non-replicable OV that excludes the effects of oncolytic replication on shRNA expression. As shown in online supplemental figure S1D–F, all five RDAd-shRNAs inhibited the expression of PD-L1, among which RDAd-shRNA3 performed better. It has been reported that knockdown of PD-L1 expression can improve cytotoxic sensitivity to CIK therapy.30 After assessing the inhibitory effect of shRNA3 on PD-L1 expression, we evaluated its efficacy in CIK cells in vitro. To this end, we performed a coculture experiment and evaluated the effect of RDAd-shRNA3 on CIK cell therapy using the MTT assay. SGC-7901 cells were infected with RDAd-shRNA3 or RDAd-shRNA (−), which was used as the NC. As the proportion of CIK cells increased, the percentage of viable cells decreased at 48 or 72 hours postinfection at a ratio of 10:1 or 20:1, respectively. Compared with the NC, the cells infected with RDAd-shRNA3 showed more cytotoxic sensitivity to CIK therapy (online supplemental figure S1G and H). As the results show, the expression of IFN-γ increased in SGC-7901 cells infected with RDAd-shRNA3 compared with the NC, indicating the activation effect of RDAd-shRNA3 on CIK cells (online supplemental figure S1I). Similar results were obtained for MGC-803 cells (online supplemental figure S1J–L). Taken together, we chose shRNA3 as the optimum shPD-L1 sequence to construct the recombinant OV because of its inhibitory effect on the expression of PD-L1 and activation effect on immune cells in vitro.
After assessing the correct assembly of RCAd-LTH-shPD-L1, we evaluated its antitumor efficacy by cell proliferation, tube formation, VM formation, and cell migration assays in vitro. VM is a recently discovered method of angiogenesis in many malignant tumors and provides a new strategy for the clinical treatment of tumor angiogenesis.31 VM has been described in a plethora of tumors, including carcinomas, sarcomas, glioblastomas, astrocytomas, and melanomas. VM is associated with high tumor grade, short survival, invasion, and metastasis.32 Although there were no significant differences in cell proliferation between RCAd-LTH-shPD-L1 and its parental virus RCAd, RCAd-LTH-shPD-L1 had better inhibitory effects on VM formation and cell migration, suggesting that RCAd-LTH-shPD-L1 may have efficient efficacy in metastatic cells (online supplemental figure S6).
In an immunodeficient mouse model, RCAd-LTH-shPD-L1 showed remarkable improvement in immunosuppressive factor expression and abnormal tumor vasculature. However, immunodeficient mouse models cannot be used to evaluate the enhancement of antitumor immunity or the promotion of lymphocyte infiltration because of their lack of an immune system. Therefore, we reconstructed the human immune system in M-NSG mice by the intraperitoneal injection of PBMCs. Previous studies have shown that T cell activation is accompanied by the upregulation of immune checkpoint molecules, such as TIM3, which plays an important role in regulating T cell responses and maintaining immune homeostasis.33 In our study, RCAd-shPD-L1 promoted T cell infiltration with an increased proportion of CD8+ T cells, accompanied by the upregulation of TIM3, indicating activation and enhancement of T cell cytotoxicity. IFN-γ plays a key role in activation of cellular immunity and subsequently, stimulation of antitumor immune response. Moreover, IFN-γ may inhibit angiogenesis in tumor tissue, induce regulatory T-cell apoptosis, and stimulate the activity of M1 proinflammatory macrophages to overcome tumor progression.34 IL-6 and IL-12 play vital roles in dendritic cell, natural killer and T cell maturation, and increase IFN-γ levels in the TME.35 Therefore, RCAd-LTH-shPD-L1 may promote T-cell infiltration and immunosuppressive TME remodeling through the upregulation of cytokines and immune checkpoint molecules. The low infiltration of immune cells and immunosuppressive TME in solids are the reasons for the low response rate and drug resistance to ICI, which are still troubling for the effective treatment of most people.6 Therefore, according to our data, RCAd-LTH-shPD-L1 can not only be used as an ICI therapy but can also be combined with other ICI therapies to enhance their efficacy.
In summary, these findings serve as a proof of principle that localized immunotherapy with RCAd-LTH-shPD-L1 can promote infiltration of T cells and improve the immunosuppressive TME by normalizing tumor vasculature and up-regulating immune-promoting set point such as IFN-γ, IL-6 and IL-12. Furthermore, our study provides mechanistic insights and provides preclinical data to test the combination of oncolytic adenoviruses, ICIs, and anti-VEGF therapy.
Data availability statement
All data relevant to the study are included in the article or uploaded as online supplemental information.
Ethics statements
Patient consent for publication
Ethics approval
The protocols for animal experiments were approved by the Animal Care and Use Committee and were in compliance with the Guidelines on Animal Welfare of the China National Committee for Animal Experiments. All animal procedures were performed according to the animal protocols approved by the Medical Ethics Committee of China Medical University (CMU2021185).
Acknowledgments
We thank Dr Chiyuan Ma and Dr Feng Yuan for their helpful suggestions.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
YM and HL contributed equally.
Correction notice This article has been corrected since it was first published online. In the original article, Ying Liu was not listed as a corresponding author. In addition to this, the correspondence email of Guangzuo Luo has been updated. In addition to this, the author affiliations have been updated.
Contributors GL accepts full responsibility for the work and/or the conduct of the study, had access to the data, and controlled the decision to publish. Conception and design: YL and GL. Development of methodology: YM, HL and HZ. Acquisition of data: YM, HL, HZ, DS and WZ. Analysis and interpretation of data: YM, HL, HZ and DS. Drafting of the manuscript: YM, HL and GL. Review and/or revision of the manuscript: YM, HZ, YL and GL. Administrative, technical, or material support: YL and GL. Study supervision: YL and GL.
Funding This work is supported by the National Natural Science Foundation of China, China (NO.82070826) and Foundation of Liaoning Educational Committee of China (ZD2020005).
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.