Cell lines and cell culture
Patient-derived primary GBM cells (GBM12 and GBM28) were provided by Mayo Clinic, Rochester, MN. Human glioma LN229 (RRID:CVCL_0393) and U87ΔEGFR (RRID:CVCL_JY78) cells were kindly provided by Dr. Erwin G. Van Meir and previously described [22, 23]. African green monkey-derived kidney epithelial cells, Vero (RRID:CVCL_0059) were purchased from the ATCC. All cells were passaged with Trypsin-EDTA and maintained as adherent monolayers in DMEM with 2% FBS (GBM12 and GBM28) or 10% FBS (LN229, U87ΔEGFR), containing penicillin, streptomycin, and Plasmocin. Murine RAW 264.7 macrophages (RRID:CVCL_0493) were cultured in Petri dishes with 10% FBS RPMI-1640 medium supplemented with penicillin, streptomycin, and Plasmocin. sDLL4-expressing tumor cells were generated by transfecting the respective parental GBM12 and LN229 cells with a lentiviral vector encoding the extracellular domain of human DLL4 (sDLL4) (pLV[Exp]-mCherry:T2A:Puro-CMV>(hDLL4s designed in Vector Builder, Chicago, IL)), using Lipofectamine™ 3000 Transfection Reagent (#L3000001, Invitrogen, Carlsbad, CA) as per manufacturer’s instructions. Stable cells were selected and maintained in puromycin dihydrochloride (#A1113803, Gibco, Carlsbad, CA). All cells are maintained up to 30 passages (after STRS profiling) and routinely tested every month for Mycoplasma using the Mycoplasma PCR Detection Kit (#G238, Applied Biological Materials, Richmond, BC, Canada).
PBMCs were isolated from healthy human donors using buffy coat (obtained from Shepeard Community Blood Center, Augusta, GA) by Ficoll gradient centrifugation using the SepMate™ PBMC Isolation Tubes (# 85460, STEMCELL Technologies Inc., Vancouver, BC, Canada). Monocytes were purified using CD14 microbeads (#130-050-201, Miltenyi Biotec, Bergisch Gladbach, Germany).
Viruses
Generation of OncoD (OD), the UT1-strain oncolytic virus, is described in our previous study [24] and was used as the control unarmed oHSV. OVsDLL4 was generated by inserting the human DLL4 ECD sequence (encoding amino acids 1–529) under the regulation of a viral immediate-early promoter (IE4/5 promoter), as described. Briefly, cDNA encoding sDLL4 (amino acids 1–529), regulated by a viral immediate-early promotor were recombined into an ICP34.5-deleted and UL39-disrupted BAC, as described [24]. All viruses express RFP and were propagated in Vero cells grown in 2% FBS DMEM, and the plaque-forming units per milliliter (pfu/mL) used in this study were determined by virus titration in Vero cells as previously described [24].
Animal studies
Ethical statement
All experiments involving animals were conducted after approval by Augusta University Institutional Animal Care and Use Committee (IACUC, Protocol # 2022-1080) and are compliant with ethical standards. Athymic nude mice (RRID:IMSR_JAX:002019) were purchased from The Jackson Laboratory (Bar Harbor, ME) and housed according to Augusta University’s Division of Laboratory Animal Services (DLAS) guidelines and IACUC housing policy. Five animals per cage were housed in a 12-h light /12-h dark cycle with ad libitum access to standard chow pellet diet and water. Both male and female mice were used for each in vivo experiment. Previously published deidentified transcriptomic data from patients was mined for DLL4 expression changes after virotherapy [25]. All methods were performed in accordance with the relevant guidelines and regulations. All animal experimental procedures were approved by the Institutional Animal Care and Use Committee of Augusta University. Deidentified gene expression data from patient biopsies was obtained from clinical investigation of CAN-3110 for recurrent high-grade brain tumors [25]. The data was obtained by informed research consent for the collection and scientific analyses of each biopsied GBM core in the original study.
Subcutaneous studies
To evaluate in vivo tumor growth, 2 × 106 LN229 (control or sDLL4-expressing) cells were injected into the right flank of male and female athymic nude mice. Tumor volumes were assessed twice weekly until mice reached a maximum tumor volume of 2000 mm3. To evaluate the antitumor efficacy of OVsDLL4, mice were randomized when their tumors grew to 150–200 mm3 (LN229 in athymic nudes) to be treated intratumorally with 100 μL of sterile saline (PBS) or 1 × 105 pfu (LN229) of the indicated virus diluted in PBS. Tumor volume was calculated as (length × width2)/2, where the length is the longest dimension and the width is the shorter dimension of the tumor.
Survival studies
For intracranial tumor cell injections, anesthetized mice (with ketamine HCl (100 mg/kg) and xylazine (10 mg/kg) diluted in distilled water) were fixed in a stereotactic device (David Kopf Instruments, Tujunga, CA). The surgical site was sterilized with 70% ethyl alcohol rub, and a skin incision was performed over the midline. After localizing the Bregma, a burr hole was created 2 mm lateral (right) and 1 mm anterior relative to it. A volume of 2 μL (cells or virus diluted in PBS) was then injected intracranially using Hamilton syringes (Hamilton™ 701N Microliter Syringes for cell implantation and Hamilton™ Microliter 700, Model 1701N for virus) at a depth of 3.5 mm and a rate of 0.4 μL/min using autoinjectors (KD Scientific Inc, Holliston, MA). Athymic nude mice were injected with human GBM12 or U87ΔEGFR (1 × 105 cells/mouse) and randomly allocated to PBS or 1 × 105 pfu of virus/mouse treatment groups. Incisions were closed using 5-0 nylon sutures. Experimental animals were observed daily, monitored for standardized mice body condition scoring, and humanely euthanized when they exhibited a body score of 2 (thin) or 1 (emaciated) and based on appearance (hunched back) and behavioral assessment (inactive) [26]. The investigators were not blinded to treatment group allocation and data collection.
Tumor cell proliferation and virus replication
Cell viability was assessed by plating 1 × 104 of the indicated tumor cells in a 96-well plate. A standard MTT assay using the tetrazolium salt MTT was used to assess cell viability using the kit (#11465007001, Roche, Basel, Switzerland) per manufacturer’s instructions. GBM12 and GBM12-sDLL4 cell counts were determined by counting viable cells using Trypan Blue exclusion over a 96 h time course.
Viral infection and replication assays were performed in the indicated cancer cells infected with the indicated viruses at MOIs 0.005-0.1. Briefly, 5 × 104 or 2 × 104 cells per well were seeded in 24- or 96-well plates, respectively (Greiner). The next day, the indicated cells were infected with the indicated virus. Virus-infected cells were quantified as RFP+ tumor cells by the Incucyte® SX5 Live-Cell Analysis System from Sartorius (Sartorius, Göttingen, Germany), where nine to sixteen images/well were taken every 2 h.
qRT-PCR
To assess the mRNA expression levels of Notch receptors, ligands, or target genes, 1 × 105 parental or stable tumor cells per well were seeded in 24-well plates for 24 h (n = 4). Total RNA was isolated using TRIzol reagent (#15596026, Invitrogen, Carlsbad, CA), and cDNA was synthesized using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA) using random hexamers per manufacturer’s instructions. Species-specific primers were obtained from Integrated DNA Technologies, Inc. (IDT, Coralville, IA), and sequences are shown in Supplementary Table S1 and following [5]. qPCR was performed using Fast SYBR™ Green Master Mix with ROX pre-mixed as a passive reference dye (#4385612, Applied Biosystems) using a QuantStudio 3 Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA). Cycling conditions were as follows: initial denaturation at 95 °C for 2 min; amplification cycle: denaturation at 95 °C for 15 s, annealing at 60 °C for 20 s, and extension at 72 °C for 5 s. 18S rRNA was used as a housekeeping control, and the 2(−ΔΔCt) method was used to determine differential fold change in gene expression.
Western blot
Whole-cell lysates of the indicated cells were harvested using RIPA buffer (#R0278-500ML, Sigma-Aldrich, St. Louis, MO) and Protease & Phosphatase Inhibitor Cocktail (100X) (#1861284, Thermo Scientific, Waltham, MA). After the cell lysates were sonicated and cleared of cell debris (centrifugation at 10,000 rpm for 10 min), protein was quantified using BCA (#23225, Thermo Scientific, Waltham, MA). Equal amounts of protein were loaded into 4–20% PROTEAN® TGX™ Precast Protein Gel from Bio-Rad Laboratories, Inc. (Hercules, CA) and transferred to a nitrocellulose membrane. The membrane was blocked with a 5% Blotting-Grade Blocker (#1706404, Bio-Rad, Hercules, CA) for 1 h at room temperature (RT) and then incubated in diluted primary antibodies overnight at 4 °C. The following antibodies were used: GAPDH (#2118, Cell Signaling Technology, Danvers, MA), ICP4 (#6514, Abcam, Cambridge, United Kingdom), and DLL4 (#MA5-17069, Thermo Fisher, Waltham, MA). Membranes were washed three times using TBS with 0.1% Tween-20 (TBS-T) and then incubated with HRP-conjugated secondary antibodies (goat anti-rabbit IgG; #7074 and horse anti-mouse IgG; #7076 from Cell Signaling Technology, Danvers, MA) for 1 h at RT. Membranes were washed with TBS-T [3×] and visualized using the Cytiva Amersham ECL Prime Detection Reagents (#RPN2236, Amersham, United Kingdom) and the ChemiDoc MP Imaging System (Bio-Rad, Hercules, CA).
Seahorse analysis of oxygen consumption rate
Parental or sDLL4-expressing GBM28 and U87ΔEGFR cells were seeded in XFe24 cell culture microplates (Agilent Technologies, Santa Clara, CA) at a density of 8 × 104 cells per well (n ≥ 3 replicates) in 180 μL Seahorse Basal media pH7.4 (DMEM, 5 mM HEPES, 2 mM Glutamine, 1 mM Pyruvate, and 10 mM Glucose) and incubated at 37 °C incubator (CO2-less) for 1 h. Parameters for the MitoStress test were: 1.5 μM oligomycin-A, 2 μM trifluoromethoxy carbonylcyanide phenylhydrazone (FCCP), and 0.5 μM rotenone/antimycin A. OCR was analyzed using the XFe 96 Flux Analyzer (Agilent, Santa Clara, CA).
Reactive oxygen species quantification
For measurement of changes in cellular reactive oxygen species (ROS), GBM28 and U87ΔEGFR cells were infected with 0.1 MOI of the indicated virus. 24 h post-infection, cell pellets were resuspended in 5 µM CellROX™ Green Reagent (#C10444, Invitrogen, Carlsbad, CA) and incubated for 30 min at 37 °C. Cells were washed and analyzed for oxidative stress by flow cytometry using the NovoCyte Quanteon instrument from Agilent Technologies (Santa Clara, CA).
Binding assay
Purified, recombinant mouse Notch1 receptor (#50654-M08B, Sino Biological, China) was immobilized on 96-well plates at 100 ng/mL in carbonate buffer (0.05 M carbonate/bicarbonate, pH 9.6) overnight at 4 °C. Recombinant, Fc-conjugated mouse and human DLL4 proteins (#10185-D4-050, Bio-Techne, USA) were diluted in carbonate buffer at the indicated concentrations and added to the plates after washing off excess, unbound Notch1 and blocking for 1 h in reagent diluent. After washing off excess, unbound ligand, the plates were incubated with goat anti-hIgG-Fc-HRP (1:50,000, #OKGD00001, Aviva, USA) for 30 min at RT, followed by washing. The substrate was incubated for 20 min, followed by stop solution. The concentration of Fc was determined by absorbance at 450 nm, indicating binding of DLL4 to Notch1.
Macrophage co-culture with GBM cells
For co-culture experiments, 2 × 105 GBM12 cells were infected with the indicated virus at 0.02 MOI for 1 h, and the media was changed to 0.5% FBS DMEM. 24 h post-infection, 2 × 105 serum-starved macrophages were co-cultured with treated tumor cells for 6 h. Co-cultures were harvested for RNA isolation or flow cytometry for analysis of cell surface markers CD80-APC (#104713, BioLegend, San Diego, CA), CD206-FITC (#141703, BioLegend, San Diego, CA), and stained for viability with LIVE/DEAD™ Fixable Near IR (780) (#L24994, Invitrogen, Carlsbad, CA). RNA was isolated using the RNeasy mini kit (#74104, Qiagen, Germany) with on-column DNase digestion (#EN33-050, Qiagen, Germany) as per the manufacturer’s protocol for bulk mRNA sequencing and species-specific deconvolution.
Macrophage migration assay
The macrophage migration assay was performed utilizing a modified Boyden chamber as described [27]. In the bottom chamber, tumor cells were seeded in 2% FBS DMEM medium and allowed to attach overnight. The next day, they were treated with PBS or infected with 0.1 MOI of virus for 1 h, after which the medium was changed to 0.5% FBS DMEM. In parallel, murine RAW 264.7 macrophages were serum-starved overnight in 0.5% RPMI-1640. At 24 h post-infection, macrophages were seeded on the top Transwell insert of the Boyden chamber and allowed to migrate toward treated cells for 6 h. Images of fixed and crystal violet-stained RAW 264.7 macrophages that migrated toward tumor cells were taken and quantified in ImageJ.
Luminex analysis
Detection of cytokines in supernatants from human PBMCs co-cultured with PBS, OncoD, or OVsDLL4-treated GBM12 glioma cells (MOI = 0.05) for 24 h was performed using the following LEGENDplex Mix and Match Panels: Human Inflammation Panel1 13-plex (including IL-1β, IFN-α2, IFN-γ, TNF-α; Cat# 740809, BioLegend, San Diego, CA), according to the manufacturer’s instructions. In brief, samples were diluted 2-fold with Assay Buffer, and standards were mixed with media used for cell culture. Standards and samples were plated with capture beads and incubated overnight at 4 °C on a plate shaker (350 rpm). After washing the plate, detection antibodies were added to each well, and the plate was incubated on a shaker (600 rpm) for 1 h at room temperature. Finally, without washing, SA-PE was added and incubated for 30 min. Samples were acquired on a Novocyte Quanteon flow cytometer (Agilent Technologies, Santa Clara, CA). Standard curves and protein concentration were calculated using the R package DrLumi installed on R 3.6.1. The limit of detection was calculated as the average of background samples plus 3xSD. The assay was performed at Georgia Cancer Center’s Immune Monitoring Shared Resource at Augusta University (RRID: SCR_026590).
Magnetic resonance imaging (MRI)
MRI scans were carried out on a 7T Bruker Biospec 70/20 (Bruker Biospin, Billerica, MA) equipped with an 86 mm quadrature transmit coil and a dedicated 4-element phased array coil. Briefly, tumor-bearing mice were anesthetized using a mixture of isoflurane/medical air (3% for induction and 1–2% for maintenance) via a nose cone. The mice were placed in a prone position on a dedicated mouse bed with a circulating warm-water circuit to maintain body temperature. Respiration rate and rectal temperature were continuously monitored throughout the experiments (SA-instruments, Stony Brook, NY). MRI contrast agent, Gadolinium diethylenetriaminepentaacetic acid (Gd-DTPA), was administered i.v. at a dose of 0.2 mM/kg to obtain signal enhancement in the tumor. Multi-slice T1-weighted spin echo images were obtained in the coronal orientation using a repetition time of 1500 ms, echo time of 8 ms and an imaging matrix of 256 × 256 with a field of view of 19.2 × 19.2 mm2. To match the histological analysis, a slice thickness of 0.75 mm was used without slice gap. The number of signal averages was 4 for all the scans. T1-weighted spin echo imaging was done before and after administration of the contrast agent for each animal using the same imaging parameters.
RNA library construction and data analysis
Sequencing and RNA library preparations were performed at the Georgia Cancer Center Integrated Genomics Core. For bulk RNA sequencing (RNA-seq), total RNA was prepared from indicated cells using the RNeasy Mini Kit (#74104, Qiagen, Germany) and the Qiagen RNeasy Mini Kit with on-column DNase digestion (#74104, Qiagen, Germantown, MD) as per the manufacturer’s protocol for bulk mRNA sequencing experiments. RNA samples were then processed at the Integrated Genomics Core Shared Resources at the Georgia Cancer Center, Augusta University. RNA quality was assessed using the Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA), ensuring all samples had an RNA Integrity Number (RIN) above 7. Library preparation was performed with the SMARTer® Stranded Total RNA Sample Prep Kit (Takara Bio USA, Inc., cat# 634878). This process incorporated RiboGone™ and SMART® (Switching Mechanism at the 5′ end of RNA Template) technologies, along with PCR amplification, to generate Illumina-compatible libraries without adapter ligation. RiboGone efficiently depleted rRNA from total RNA before library construction. Library quality was assessed using the TapeStation 2000 (Agilent Technologies, Santa Clara, CA), and concentrations were measured with Qubit (Thermo Fisher Scientific, Waltham, MA) before sequencing. Dual-indexed i5/i7 paired-end libraries were sequenced on the NovaSeq 6000 system (Illumina) with a minimum coverage of 50-60 million reads per sample.
The RNA-seq data analysis started with quality control using FastQC (version: v0.12.1) to assess the raw reads, followed by Cutadapt (version: 2.8) to trim Illumina Universal Adapter sequences. Low-quality bases at the end of reads and the minimum length of the reads after trimming were set with parameters q = 30 and m = 20, respectively. The cleaned reads were then aligned to the Homo sapiens(human) genome assembly GRCh38 (hg38) reference genome using the STAR aligner (version: 2.7.3a), generating BAM files for further analysis. Using R (version: 4.4.1), read count data from the BAM files were merged, and genes with zero counts were excluded. Differential expression analysis was performed with DESeq2 (RRID:SCR_000154), identifying significant genes based on log2 fold changes and adjusted p-values. Functional enrichment analysis was carried out using clusterProfiler (RRID:SCR_016884) to identify overrepresented Gene Ontology terms and KEGG pathways. Visualization was done using ggplot2 (RRID:SCR_014601), including MA and volcano plots, and heatmaps were generated using the pheatmap package to illustrate gene expression patterns across conditions for each of the samples.
For human tumor cells-murine macrophage co-culture bulk sequencing, the RNA-seq data analysis started with quality control using FastQC (version: v0.12.1) to assess the raw reads, followed by Cutadapt (version: 2.8) to trim Illumina Universal Adapter sequences. The cleaned reads were then aligned to their respective reference genomes, GRCh38.p14 for Homo sapiens and GRCm39 for Mus musculus, using the STAR aligner (version: 2.7.3a). This step produced BAM files containing aligned reads, which were further processed for species-specific deconvolution. The XenofilteR package in R (version: 4.4.1) was employed to separate human and mouse reads, ensuring accurate downstream gene expression analysis. Following species-specific read separation, STAR was used to generate raw read count matrices. The read count data were merged in R, and genes with zero counts across all samples were excluded to enhance statistical robustness. Differential expression analysis was conducted using DESeq2, identifying significantly differentially expressed genes (DEGs) based on log2 fold changes and adjusted p-values. Functional enrichment analysis was carried out using clusterProfiler to identify overrepresented Gene Ontology terms and KEGG pathways. Visualization was done using ggplot2, including MA and volcano plots, and heatmaps were generated using the pheatmap package to illustrate gene expression patterns across conditions for each of the samples.
Statistical analysis
GraphPad Prism10 software (GraphPad, San Diego, CA; RRID:SCR_002798) was used for statistical analysis. One- or two-way analysis of variance was used to determine the difference between more than two groups and more than one categorical variable, respectively. An unpaired Student’s t-test was used to compare two groups. Kaplan–Meier curves were used for comparison of survival using the log-rank (Mantel–Cox) test. Data is represented as mean ± SD from at least three biological replicates per independent experiment. p-values of ≤0.05 were considered statistically significant. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.
