Tumor cells HepG2 (ATCC® HB-8065), PLC/PRF/5 (ATCC® CRL-8024) were obtained from ATCC. Huh7 cell lines were obtained from Antihela BioTech. 293T cell were obtained from TAKARA (Lenti-X™ 293T Cell Line, 632180). All cell lines were cultured using DMEM medium (Gibco, 10564011).
Lentivirus and CAR-T preparation
The G3CAR gene comprises GC33 scFv, CD8a hinge and transmembrane domain, and the signaling domains of 4-1BB and CD3zeta. To echo the gene structure of G3CAR, sequences corresponding to IL-7 and PH20 gene were placed in tandem, downstream of G3CAR, linked by a 2A sequence to form G3CAR-7×20. Both G3CAR and G3CAR-7×20 were cloned into a lentivirus vector.
The lentivirus vector and two helper vectors were transduced into 293T cell for 48 hours. The supernatant was then collected and purified using a Lenti-X Concentrator (TAKARA, 631232). The lentivirus titer was measured using a Lenti-X™ p24 Rapid Titer Kit (TAKARA, 632200), following manufacturer’s instructions.
To prepare CAR-T, PBMCs supplied by healthy donors were stimulated using CTS™ (Cell Therapy Systems) Dynabeads™ CD3/CD28 (thermofisher, 40203D), and cultured in T cell medium (X-VIVO 15 medium contained 5%FBS and 100IU/ml IL-2 (Peprotech, 200-02)). After 24 hours, lentivirus was added with MOI at 3. CAR-T cells were then cultured in T cell medium for 10 days.
Flow cytometry assay
To detect transduced CAR-T, CAR-T cells were labeled with Human Glypican 3 / GPC3 Protein, His Tag (ACRO, GP3-H52H4-1mg) for 30min. After washing with PBS, the cells were stained using APC anti-His Tag Antibody (biolegend, 362605).
For detection of memory phenotype, CAR-T cells were labeled with Human Glypican 3 / GPC3 Protein, His Tag, and then stained using APC anti-His Tag Antibody and PE anti-human CD197 (CCR7) Antibody (biolegend, 353203). The cells were analyzed using BD Conto II. Subsequent data analysis was performed using Flowjo V10.
WB analysis was performed as previously described. The primary antibodies used were PH20 (Abcam, ab196596) and GAPDH (Abcam, ab9485).
1*105 T cells were co-cultured, with or without equal quantities of tumor cells, for 18 hours and the supernatant collected. IL-2 and IFN-gamma were quantified using Human, Th1/Th2, Cytokine Kit II (BD, 551809), following the manufacturer’s instructions. IL-7 was quantified using a Human IL-7 Quantikine HS ELISA Kit (R&D HS750). For hyaluronic acid detection, CAR-T cells (G3CAR-7×20 and G3CAR) were co-cultured with Huh7 and Huh7-HSA2 respectively for 24 hours and the supernatant collected. Hyaluronan was detected using a Hyaluronan Quantikine ELISA Kit.
Cell counts were quantified using a Cellometer Auto 2000.
For the CFSE based assay, CAR-T cells were labeled using a CellTrace™ CFSE Cell Proliferation Kit (ThermoFisher C34554). Cells were co-cultured with or without tumor cells for 96 hours and detected using standard FACS, with analysis performed using Flowjo V10.1.
Cell invasion assay
Invasive capacity was assessed using Corning® BioCoat™ Matrigel® Invasion Chambers (Corning, 354480), following manufacturer’s instructions. The lower chamber contained 10% FBS medium as chemoattactant. As a control, the same experimental method was performed using 8.0 µm PET Membrane without a matrigel coating. The invasion percentage was calculated thus: (mean of cells invading through the Matrigel chamber membrane/mean of cells migrating through the control insert membrane) × 100.
The CAR-T cells were co-cultured with tumor cells in round bottom 96-well-plates at different ET ratios for 4 hours or 24 hours, as required. The supernatant was collected and analyzed using a CytoTox 96® Non-Radioactive Cytotoxicity Assay (Promega, G1780), following manufacturer’s instructions. The percentage of lysis was calculated as follows: [(experimental - spontaneous release) / (maximum load - spontaneous release) * 100(%)].
STAT5 report assay
CD3+ T cells, sorted using a Dynabeads™ FlowComp™ Human CD3 Kit (ThermoFisher 11365D), were electro-transfected with pGL4.52 [luc2P/STAT5 RE/Hygro] Vector (Promega E465A). CAR-T cells were simultaneously cultured without any cytokines. After 24 hours, the culture medium of the T cells transfected with a STAT report vector was replaced by that of the CAR-T cells. Cell culture continued for an additional 24 hours. Cells (50ul contains 1*105 cells) were then transferred into a black assay plate and supplemented with 50 µL of Dual-Luciferase® Reporter Media before incubation for 20min. Fluorescence was then measured using an Envision Multilabel Reader (PerkinElmer).
Xenogenic mouse models
6-8 week-old male NSG mice (Jackson Laboratory) were injected subcutaneously with 2.5*106 Huh-7-HSA2 cells. The tumor volume was measured every 4 days using a Vernier caliper (V= 1⁄2 x L (length) x W (width) x W). CAR-T cells were injected intravenously when the tumor volume exceeded 300mm3. Body mass was measured every 4 days.
Blood was collected from the tail vein. After centrifugation, plasma was collected for cytokine detection using BD™ CBA Flex Set (558334 and 561515). Cells were treated with Red Blood Cell Lysis Buffer (Beyotime, Shanghai, China) and stained using APC Mouse Anti-Human CD3ε (BD, 558257). T cells were counted and analyzed using CountBright™ Absolute Counting Beads.
For the tumor re-challenge model, 2*106 CAR-T cells were injected intravenously when the mean tumor volume in the right flank exceeded 180mm3. 2*106 Huh7-HSA2 was injected subcutaneously into the left flank. The tumor volume was measured every 4 days.
For the IHC assay, tumor tissue samples were fixed, processed, and stained according to standard IHC procedures, as previously reported. Sections were stained using an Anti-CD3 antibody (Abcam, ab5690). An HRP system was used for detection. Positive cells were scanned and counted.
GraphPad Prism 6.0 was used for the statistical analysis of data. One-way ANOVA, two-way ANOVA with Bonferroni post-test, or unpaired, two-tailed t-tests were used for as appropriate. Symbols indicate statistical significance as *P < 0.05; **P < 0.01; ***P < 0.001. Each experiment was performed at least three times.