Plasmids and viral vectors
PD-L1-P2A-VEGF-B186 and VEGF-B186-P2A-PD-L1 constructs, with and without codon optimization, were synthesized by GenScript (Piscataway, NJ, USA). Coding DNA sequences were optimized using the GeneArt algorithm (Thermo Fisher Scientific, Waltham, MA, USA). Constructs were further subcloned into the pAdApt expression vector (Thermo Fisher Scientific) for characterization. For subsequent cloning into the pAAV-ITR backbone, muSEAP-P2A-PD-L1, muSEAP-P2A-PD-L2, VEGF-B186-P2A-PD-L1, and VEGF-B186-P2A-PD-L2 were synthetized (GeneWiz, South Plainfield, NJ, USA) in pUC57 backbone surrounded by SpeI and HpaI restriction sites. The muSEAP coding sequence was not codon-optimized; however, PD-L1 and PD-L2 sequences in muSEAP constructs were identical to the optimized sequences in other constructs. muSEAP and VEGF-B186 plasmids without ligands, and the vector production was described by Eriksson et al. [29]. All constructs were under a CMV promoter. The schematic presentation of each construct is shown in Supplementary Fig. 1.
Transfection and transduction assays
293T cells (ECACC 12022001, RRID: CVCL_0063) were transfected with polyethylenimine (PEIpro, Polyplus transfection, Illkirch-Graffenstaden, France) and expression plasmids at a 1:1 ratio to produce PD-L1, PD-L2, VEGF-B186, and muSEAP proteins. Cell culture supernatants were collected at 48 h post-transfection, and cells were lysed with RIPA lysis and extraction buffer (Thermo Fisher Scientific) or NP-40 buffer (150 mM NaCl, 1% NP-40, 50 mM Tris-HCl, pH 8.0). Both lysis buffers were supplied with protease inhibitor cocktail (Hoffmann-La Roche, Basel, Switzerland).
rAAV6 transductions were carried out at a multiplicity of infection of 20000. An hour before transductions, cells were treated with 5 µM Dorsomorphin (Sigma-Aldrich, Saint Louis, MO, USA) to enhance transduction efficacy. Cell culture supernatants and cell lysates were collected at 72 h post-transduction.
Immunoblotting
PD-L1, PD-L2, VEGF-B186, and muSEAP proteins were analyzed in 293T cell RIPA lysates and culture media using immunoblotting. Primary antibodies used were human PD-L1 antibody (cat# 13684, Cell Signaling Technologies, Danvers, MA, USA), human PD-L2 antibody (cat# 82723, Cell Signaling Technologies), human VEGF-B167/186 antibody (cat# AF751, R&D Systems, Minneapolis, MN, USA), human alkaline phosphatase antibody (cat# sc-398461, Santa Cruz Biotechnology, Dallas, TX, USA), and GAPDH (cat# 2118, Cell Signaling Technologies). Secondary antibodies used were Goat anti-Rabbit HRP (Bio Rad, Hercules, CA, USA, and Invitrogen, Waltham, Massachusetts, USA), Donkey anti-Goat HRP (R&D Systems), and Goat anti-Mouse HRP (Bio Rad). Restore™ PLUS Western Blot Stripping Buffer (Thermo Fisher Scientific) was used to remove bound primary and secondary antibodies from membranes before reprobing with another primary antibody.
Cell growth/survival assay
BaF3-R1 cells [30], received as a generous gift from Dr. Kari Alitalo, were used to analyze the biological activity of VEGF-B186. Cells, normally dependent on recombinant mouse interleukin-3 for growth and survival, were incubated with media from VEGF-B186-transfected 293T cells under recombinant mouse interleukin-3 free conditions for 48 h. Cell Titer 96 Aqueous One Solution Cell Proliferation Assay reagent (Promega, Madison, WI, USA) was added to each well and incubated for two more hours. Absorbance was measured at 490 and 700 nm using Varioskan Lux with SkanIt Software. The 700 nm values were subtracted from the 490 nm values to reduce the possible background signal.
SEAP reporter gene assay
A Phospha-LightTM System chemiluminescence assay (Applied Biosystems, Foster City, CA, USA) was used to measure muSEAP protein in transfected or transduced 293T cell media and mouse sera. The assay was conducted according to the manufacturer’s instructions.
Pulldown assay with recombinant PD-1-His
The binding of PD-L1/2 to PD-1 was analyzed using a pulldown assay. NP-40 lysates of transfected 293T cells were precleared with HisPur Ni-NTA Magnetic Beads (Thermo Scientific). Lysates were collected, supplied with 2.5 µg of PD1-His (cat# 8986-PD-100, R&D Systems) recombinant protein, and incubated at 4 °C overnight. Controls were incubated without PD1-His. The next day, the protein complexes were precipitated using HisPur Ni-NTA Magnetic Beads. Beads were washed with TBS containing 0.1% Tween, and protein complexes were eluted with PBS containing 500 mM imidazole (Sigma-Aldrich) at pH 8.0. Eluates were analyzed by immunoblotting with human PD-L1 and PD-L2 antibodies. Non-specific binding was controlled using the same blots and detection with a human Alkaline Phosphatase antibody and a human VEGF-B167/186 antibody. In addition, a human PD-1 antibody (cat# AMAb91197, Atlas Antibodies, Bromma, Sweden) was used to monitor recombinant PD-1 protein levels in samples.
Mice
A total of seventy 9- or 10-week-old female C57BL/6JOlaHsd mice were obtained and kept in the National Laboratory Animal Center of The University of Eastern Finland, Kuopio, Finland. No sample size calculation was performed for the experiments. Animal work was conducted according to the animal experimentation license approved by the National Animal Experiment Board of the Regional State Administrative Agency of Southern Finland and carried out following the guidelines of the Finnish Act on Animal Experimentation (62/2006).
Half of the mice (n = 36) were pre-immunized with rAAV6-LacZ by IM injection into the left gastrocnemius medialis two weeks prior to gene transfer. For injection, mice were anesthetized with isoflurane, and 5×109 vg in 12.5 µl was injected. For gene transfer, two weeks after pre-immunization, rAAV6-muSEAP-PD-L1 (n = 12), rAAV6-muSEAP-PD-L2 (n = 10), or rAAV6-muSEAP (n = 10) vectors were injected at 5 × 109 vg in 12.5 µl into the right gastrocnemius medialis and gastrocnemius lateralis, with one dose into each muscle. The negative control mice (n = 4) received an equal volume of PBS. The animals were assigned to different groups without randomization. Blood samples were collected from the vena saphena prior to pre-immunization and every other week after gene transfer, up to 5 or 12 weeks after gene transfer. All experiments were performed unblinded.
RNA isolation and RT-qPCR
RT-qPCR was used to quantify transgene products from mouse tissues. After sacrification, tissues were collected in liquid nitrogen and stored at –80 °C. Total RNA from approximately 50–100 mg of treated gastrocnemius muscle and liver was extracted using TRI Reagent (Invitrogen) according to the manufacturer’s instructions. QuantiTect Reverse Transcription kit (Qiagen, Hilden, Germany) was used to transcribe 1 µg of RNA into cDNA. Quantitative measurements of mRNA levels were performed using PowerUp SYBR Green Master Mix (Thermo Fisher Scientific) and 10 pmol of primers (Supplementary Table 1) with the StepOnePlus Real-Time PCR System (Thermo Fisher Scientific). HPRT amplification was used as a housekeeping control to standardize the amount of RNA in each sample.
Protein extraction and transgene detection from the muscles
Proteins were extracted with T-Per Tissue Protein Extraction Reagent (Thermo Fisher Scientific), and the total protein concentration was determined using the Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific). Equal amounts of protein were used for immunoblotting analysis with PD-L1, PD-L2, and Alkaline Phosphatase antibodies. GAPDH antibody was used as a housekeeping control.
Histopathology
Tissue samples were collected in 4% formaldehyde fixative (pH 7.4), processed, and embedded in paraffin. Cross-sections (5 µm) of the right gastrocnemius medialis injected with either rAAV6-muSEAP-PD-L1, rAAV6-muSEAP-PD-L2, or rAAV6-muSEAP were stained for CD3, CD4, and CD8 (Cell Signaling Technologies) detection to analyze inflammation state. Ki67 (Cell Signaling Technologies) was used as a proliferation indicator. The benefit of PD-L1/2 depends on PD-1 expression on immune cells, so PD-1 (Cell Signaling Technologies) staining was also conducted. The sections were counterstained with hematoxylin. Sections were imaged with a Nikon H550L microscope, and immune cells were quantified using ImageJ Fiji [31] image processing software.
Detection of AAV6 and PD-L1/2 IgG antibodies in mouse serum
Serum AAV6 antibody levels were quantified by ELISA. rAAV6 particles were diluted in PBS (pH 7.4) to a final concentration of 1 × 108 vg/μl. The wells of a 96-well Nunc MaxiSorp plate (Thermo Fisher Scientific) were coated with the vector solution and incubated at 4 °C overnight. PBS without rAAV6 was used as a non-coated control. The following day, the plate was washed with PBS and blocked with 5% BSA in PBS by incubation at 37 °C for 2 h. The plate was washed with PBS, and mouse serum, diluted threefold from 1:100 to 1:24300, was added and incubated at 37 °C for 90 min. After washing, mouse IgG peroxidase antibody (Sigma-Aldrich) was added, and the mixture was incubated at 37 °C for 1 h. After washing, tetramethylbenzidine substrate solution (Sigma-Aldrich) was added and incubated for 20 min at room temperature. The reaction was stopped with the stop reagent (Sigma-Aldrich), and absorbances were read at 450 and 650 nm using Varioskan Lux with SkanIt Software. 650 nm readings were subtracted from 450 nm readings to reduce the possible background signal.
PD-L1/2 antibodies were detected in mouse sera using the same method. For antibody capture, plates were coated with recombinant PD-L1 (cat# 9049-B7-100, R&D Systems) or PD-L2 (cat# 9075-PL-100, R&D Systems) proteins in sodium bicarbonate coating buffer (pH 9.5) overnight at room temperature.
Epitope prediction
Antigenic transgene product epitopes were predicted using the B and T cell epitope prediction tool [32]. B-cell epitopes were predicted using the Kolaskar-Tongaonkar method [33] and T cell epitopes were predicted using TepiTool [34]. For MHC I epitope prediction, 8–11 amino acid-long epitopes were included, and the peptide selection criterion cutoff was set at 1. For MHC II epitope prediction, 15 amino acid-long epitopes were included, and the peptide selection criterion cutoff was set at 10.
Statistical analysis
Statistical analysis was performed using GraphPad Prism 9.3.0 (San Diego, California, USA). Specific statistical analyses and data presentation are indicated in the figure legends. For normally distributed data, analysis of variance (ANOVA) was used as appropriate. For non-normally distributed data, the Mann-Whitney test, Kruskal-Wallis test, or Wilcoxon test was applied as suitable. Statistical significance is denoted by *(p < 0.05), **(p < 0.01), ***(p < 0.001), and ****(p < 0.0001).