Viral vector generation

Adenovirus vectors were generated using genetic modifications by AdZ homologous recombineering using previously described methods [14]. Replication-deficient vectors based on a wild type Ad5 genome were modified to generate the Ad5NULL-A20 platform. Ablation of CAR binding was achieved via the KO1 mutation in the AB loop of the L5 fiber knob gene; ablation of binding to coagulation factor 10 via a mutation in hypervariable region 7 of the L3 hexon gene; ablation of αvβ3/5 integrin binding via RGD-RGE mutation in the L2 penton base gene. Retargeting of the modified Ad genome to αvβ6 integrin was achieved by insertion of an A20 peptide sequence from FMDV (NAVPNLRGDLQVKVART) into the viral fiber knob HI loop (between residues G546 and D547). Gene synthesised FCY1 and FCU1 (codon optimised) PCR fragments were inserted under the control of a CMV promoter replacing a sacB cassette selectable marker.

Viruses were produced in either T-REx-293 (Ad5) or HEK293-β6 (β6-expressing) cell lines. DNA was amplified using a maxiprep kit as per the manufacturer’s guidelines. Virus particles were generated by transfection in a T25 CellBIND flask (Corning) of T-REx-293 (Ad5) or HEK293-β6 cells (β6-expressing) cells. When cytopathic effect (CPE) was observed in cells, cells were collected, and virus was further amplified in expanded cells (10 X T150 CellBIND flasks). Caesium chloride (CsCl) two-step purification method was used to extract purified viruses. Viral particles/mL (vp/mL) were quantified by micro-BCA assay (Thermo Fisher, Loughborough, UK) using the equation 1 µg protein = 4 × 109 viral particles (vp)

Cell lines and culture

Human pancreatic cancer cells, CFPAC1, MIA PaCa-2, PT45, (American Type Culture Collection (ATCC)) were maintained in Dulbecco’s modified Eagle’s medium (DMEM, Sigma-Aldrich, New York, NY, USA) supplemented with 10% heat-inactivated foetal bovine serum (FBS, Gibco, Grand Island, NY, USA), 1% penicillin/streptomycin (P/S, Gibco, Paisley, UK), 1% L-glutamine (Gibco). ASPC1, BxPC3 and PANC10.05 (ATCC) cells were maintained in Roswell Park Memorial Institute (RPMI) 1640 Medium (Sigma) supplemented with 10% FBS, 1% P/S, 1% L-glutamine. All cultures were maintained in a 5% CO2 humidified atmosphere at 37 °C. All cell lines were routinely tested for mycoplasma using MycoAlert Mycoplasma detection kit (Lonza).

Cell viability assay

Cells were seeded at a density of 5000 cells per well in triplicate in white clear-bottomed 96 well plates. Following 24 h in culture, cells were transduced with indicated viral vectors, within a range of 500–5000 viral particles (vp) per cell. Viruses were diluted in serum-free growth medium and cells were transduced for 24 h at 37 °C. Cells were then washed with PBS and incubated in full growth medium containing a dilution of 5-FC (Sigma-Aldrich, Gillingham, UK) or PBS vehicle control for 3 days. Cell viability was measured using CellTiter-Glo® Luminescent Cell Viability Assay (Promega, Madison, WI, USA) as per the manufacturer’s instructions, and luminescence was measured with a multimode plate reader (Bio-Rad, Hertfordshire, UK). Relative luminescence units were normalised to the vehicle control, with FCU1- and FCY1- viruses normalised to parental virus controls.

Mouse pancreatic organoids

Pdx-1 CreERT [19], LSL-KrasG12D/+ [20]; LSL-Trp53R172H/+ [21]; Rosa26LSL-tdRFP [22] (KPC) mouse lines have all been previously described. KPC lines were used to generate pancreatic tumour-derived organoids as described previously [23, 24]. Animals were housed in conventional pathogen-free animal facilities and all procedures were conducted in accordance with the UK Home Office regulations (ASPA 1986 & EU Directive 2010) under Home Office approved Project licence granted to CH and under the guidelines of Cardiff University Animal Welfare and Ethics Committee. Mice were genotyped by PCR analysis using sequences as described in [24]. Cre recombinase was induced by administering a single intraperitoneal injection of tamoxifen (1 µg/40 g) to 6–8-week-old mice [24]. Typically, KPC mice develop PDAC tumours at 20–22 weeks post induction of Cre recombinase [25]. Tumour-bearing pancreas was harvested and routinely dissected for downstream analyses or organoid culture. For generation of organoids, mouse pancreas was mechanically dissociated before digestion in collagenase Type 1 (Sigma-Aldrich) and Dispase II (Sigma-Aldrich) to a concentration of 0.125 mg/mL at 37 °C. Following several washes in HBSS supplemented with 5% FBS, tissue was passed through a 40 μm cell strainer. The washed cells were then resuspended in Matrigel (Corning, Bedford, MA, USA) and seeded within individual domes in 24 well plates. Once polymerised, pancreatic cells were overlaid with expansion medium containing Advanced DMEM F12 (Gibco, Grand Island, NY, USA) supplemented with HEPES (1%, Gibco, Paisley, UK), GlutaMAX (1% Gibco, Paisley, UK)), 1% penicillin/streptomycin(P/S, Gibco), B27 (1 X ThermoFisher Scientific, MA, USA), N2 (1 X, ThermoFisher Scientific), 1.25 mM n-acetyl-L-cysteine (Sigma-Aldrich, MO, USA), R-spondin 1 conditioned medium (5%) or recombinant R-spondin (1 µg/ml; Peprotech, CA, USA), 10 mM Nicotinamide (Sigma-Aldrich), 10 nM recombinant human [Leu15] Gastrin I (Sigma-Aldrich), 50 ng/mL recombinant EGF (Peprotech, CA, USA), 100 ng/mL recombinant human FGF10 (Peprotech, NJ, USA) and 25 ng/mL recombinant human Noggin (Peprotech, NJ, USA) and incubated under standard tissue culture conditions (37 °C, 5% CO2).

Immunohistochemical staining

Tumour sections from tissues were mounted on slides prior to serial washes in xylene and graded ethanol (100%, 90% and 70%). Protease antigen retrieval was carried out for αvβ6 integrin by adding protease 2 (0.1 mg/ mL, Roche, Basel, Switzerland) to sections at 37 °C for 12 min. Slides were washed and incubated with 1% H2O2 for 15 min at room temperature, washed, and blocked using 2.5% horse serum for 30 min at room temperature. αvβ6 integrin primary antibody was added to slides (1:750; EM05201, Absolute Antibody) in 1% BSA/PBS at 4 °C overnight. Primary antibody was removed and replaced with diluted ImmPACT DAB chromogen (Vector Labs, Newark, CA, USA) for 4 min at room temperature. Slides were submerged in Mayer’s haematoxylin prior to rinsing in ddH2O. Slides were dehydrated prior to mounting with DPX mountant before imaging.

Organoid viability assay

Organoids established in culture were mechanically disaggregated into fragments, with a representative population further digested to single cells by incubation in TrypLE™ Express (Gibco, Grand Island, NY, USA) at 37 °C for counting purposes. Single cell counts were carried out using an automated cell counter (Cell Drop, DeNovix) and used to estimate cell numbers within organoid fragments. Organoid fragments were incubated with viruses at a dose of 5000 vp/cell for 30 min at 37 °C. Following incubation, tubes were transferred to ice and the medium containing transduced organoids (10% final volume) was supplemented with Matrigel (90% final volume), mixed and seeded in 5–10 µL drops in triplicate in white clear-bottomed 96 well plates. Plates were incubated at room temperature for 5 min, prior to inverting plates and incubating for 1 one hour at 37 °C to enable Matrigel polymerisation. Growth medium containing 10 µM ROCK inhibitor, Y-27632, (BD Biosciences, CA, USA) was overlaid on Matrigel domes and incubated under standard tissue culture conditions for 24 h. Following a 24 h incubation, medium was removed from wells and replaced with growth medium containing a dilution of 5-FC or PBS vehicle control for a further 5–6 days as required. Cell viability was measured as indicated after drug treatment using Cell Titer Glo® (Promega) assay. Cell Titer Glo® reagent was added to a total volume of 50 µL per well and placed on a shaker platform (595 rpm) for 1 h at room temperature, protected from light. Luminescence was measured on a multimodal platereader (BioRad) and values expressed as a percentage viability relative to vehicle control cells.

Cell surface receptor staining

To assess cell surface receptors by flow cytometry, cells were harvested, washed in 5% FBS/PBS and added at a density of 100,000 cells per well in a v-bottomed 96 well plate (Nunc) and incubated on ice for 1 h with the respective primary mouse mAb; Anti-CAR (RmcB, 3022487; Millipore, Watford, UK) and anti-αvβ6 (10D5, MAB2077Z; Millipore) and matched IgG control were used at a concentration of 1:500 or 1:1000 for a matched IgG control. Cells were then washed and incubated on ice for 30 min with 1:1000 dilution of Alexa-647 labelled goat anti-mouse F(ab’)2 (A-21237; Life Technologies, Paisley, UK) or Alexa-488 labelled goat anti-mouse F(ab’)2. Stained cells were fixed using 4% paraformaldehyde prior to measurement by flow cytometry on Accuri C6 (BD Biosciences). For flow cytometry, a minimum of 10,000 events were acquired. Analysis was performed using FlowJo v.10 (FlowJo, LLC) by sequential gating on cell population, singlets and Alexa-647 or Alexa-488 positive cells.

Patient derived organoids

Patient-derived organoids (PDOs) from pancreatic tumours HCM-CSHL-0079-C25 (ATCC® PDM30™); HCM-CSHL-0089-C25 (ATCC® PDM36™) HCM-CSHL-0091-C25 (ATCC® PDM38™); HCM-CSHL-0092-C25 (ATCC® PDM39™) were acquired from the ATCC repository and cultured according to manufacturer’s instructions. We used models and data derived by the Human Cancer Models Initiative (HCMI) https://ocg.cancer.gov/programs/HCMI; dbGaP accession number phs001486. PDOs were cultured in Matrigel (100% v/v, Corning, Beford, MA, USA) and maintained at a seeding density of 0.25–1 × 106 cells/ 100 µL of Matrigel per well of a 6-well plate. A complete medium change was carried out every 3-4 days in culture. Organoids were maintained in Advanced DMEM F12 (Gibco, Grand Island, NY, USA) supplemented with HEPES (1%, Gibco, Paisley, UK), GlutaMAX (1% Gibco, Paisley, UK)), 1% penicillin/streptomycin (P/S, Gibco), 1X B-27 (ThermoFisher Scientific, MA, USA), Wnt3A Conditioned medium (50%), R-spondin Conditioned medium (10%), 1.25 mM n-acetyl-L-cysteine (Sigma-Aldrich), 10 mM Nicotinamide (Sigma-Aldrich), 10 nM recombinant human [Leu15] Gastrin I (Sigma-Aldrich), 50 ng/mL recombinant EGF (Peprotech, CA, USA), 100 ng/mL recombinant human FGF10 (Peprotech, NJ, USA), 100 ng/mL recombinant human Noggin (Peprotech, NJ, USA) and 500 nM A 83-01 (Peprotech) and incubated under standard tissue culture conditions (37 °C, 5% CO2). Organoids were split by enzymatic digestion when the appropriate confluence was reached, using TrypLE Express dissociating agent. 10 µM ROCK inhibitor (BD) was added to growth medium for the first 3 days after splitting.

In vivo efficacy studies

Athymic nude mice were purchased from Charles River (UK) and allowed to acclimatise for a minimum of 7 days prior to experiments. All procedures were conducted in accordance with the UK Home Office regulations (ASPA 1986 & EU Directive 2010) under Home Office approved Project licence granted to TP and under the guidelines of Cardiff University Animal Welfare and Ethics Committee. Mice were housed in filtered cages. Subcutaneous CFPAC1 models were established by injecting 2 × 106 cells into both flanks. 7 days post engraftment, 3 × 1010 vp of Ads or PBS control were injected intratumourally. 24 h post I.T injection, 5-FC (200 mg/kg) or vehicle control (PBS) was administered daily by I.P injection for 8 days, then again at days 22 and 25. Tumours were measured with calipers every 1–3 days to ensure tumour volumes did not exceed 1.5 cm throughout the experiment, and the volume was calculated as V (mm3) = π/6 × W2 × L and normalized to day 0 post I.T treatment.

Viral genome copy number analysis of in vivo tumours

DNA was extracted from snap frozen tissues using DNeasy Blood and Tissue DNA extraction kit (Qiagen) according to the manufacturer’s protocol. DNA concentration was determined using a NanoDrop photospectrometer. 25 ng of total DNA was subjected to quantitative PCR using Fast SYBR Green Master Mix. Reactions were performed in technical triplicate, using primers for the hexon region of the genome. Standard curves were prepared by serial dilution of replication-deficient Ad5.

Statistical analysis

Statistical analyses were performed using GraphPad (San Diego, CA) Prism software. Data are presented as mean ± standard error of the mean unless otherwise specified. Experiments were performed to n = 3 independent experiments, unless otherwise stated. Statistical analysis was carried out as indicated and statistical significance is shown as follows; ns = p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Graphical depictions were created using BioRender.com