Unravelling Receptor and RGD Motif Dependence of Retargeted Adenoviral Vectors using Advanced Tumor Model Systems

Recent advances in engineering adenoviruses are paving the way for new therapeutic gene delivery approaches in cancer. However, there is limited knowledge regarding the impact of adenoviral retargeting on transduction efficiency in more complex tumor architectures, and the role of the RGD loop at the penton base in retargeting is unclear. To address this gap, we used tumor models of increasing complexity to study the role of the receptor and the RGD motif. Employing tumor-fibroblast co-culture models, we demonstrate the importance of the RGD motif for efficient transduction in 2D through the epithelial cell adhesion molecule (EpCAM), but not the epidermal growth factor receptor (EGFR). Via optical clearing of co-culture spheroids, we show that the RGD motif is required for transduction via both receptors in 3D tumor architectures. We subsequently employed a custom-designed microfluidic model containing collagen-embedded tumor spheroids, mimicking the interplay between interstitial flow, extracellular matrix and adenoviral transduction. Image analysis of on-chip cleared spheroids indicated the importance of the RGD motif for on-chip adenoviral transduction. Together, our results show the interrelationship between receptor characteristics, the RGD motif, the 3D tumor architecture and retargeted adenoviral transduction efficiency. The findings are important for the rational design of next-generation therapeutic adenoviruses.

a Petri dish lid, according to a hanging-drop protocol for spheroid preparation [2]. Following an overnight incubation, spheroids were used in the experiments.

Luciferase assay
The luciferase assay was performed with the firefly luciferase reporter assay system (Promega Corp.) according to the manufacturer's instructions. A total of 1.0 × 10 4 cells were seeded per well in a 96well cell culture plate (Corning) and incubated overnight. Following the overnight culture, cells were transduced as described and incubated for 48 h. Luciferase activity was measured using a Synergy™ 2 Multi-Mode Microplate Reader (BioTek Instruments, Inc.). All assays were carried out in biological duplicates. The number of independent experiments is indicated in the figure caption. The data was normalized over the sum of all signal intensities per experiment.

Flow cytometry
The investigation of viral transduction efficiency and specificity by flow cytometry was done via the detection of iRFP670 in cells pre-labelled with cytoplasmic dyes using the following lasers: 488 nm for carboxyfluorescein (detection: 525/50 nm) and CellTrace Yellow (detection: 585/40 nm), and 635 nm for iRFP670 (detection: 655-730 nm). Cells cultured in 2D were trypsinized while spheroids were dissociated by extensive washing in PBS and trypsinization for 15 min at 37°C. Cells were fixed in 4% PFA in PBS for 10 min and resuspended in PBS. Flow cytometry was performed using a MacsQuant

Confocal microscopy
For confocal microscopy, cells in 2D were seeded, incubated and analyzed in µ-slide 8-well chambers (Ibidi, Gräfelfing, Germany). For imaging of 3D spheroid co-cultures, spheroids were collected, washed in PBS and fixed as described above, and embedded in a 15-µl drop of 4 mg/ml collagen in an 8-well Ibidi chamber. Collagen gelation was achieved by incubation at room temperature for 10 min followed by incubation at 37°C for 30 min. The embedded spheroids were optically cleared as described above, were processed and quantified using Fiji image analysis software [3].
The method used for quantification of transduced tumor cells and fibroblasts from confocal microscopy images was validated against results from flow cytometry. Co-cultures were transduced and analyzed via flow cytometry as described above. Corresponding confocal images were analyzed with Fiji software: all channels from the microscopy were subjected to a Gaussian blur, followed by thresholding to create masks of 'tumor cells', 'fibroblasts' and 'transduced cells'. The fraction of overlapping areas of either 'fibroblasts' or 'tumor cells' with 'transduced cells' was determined and taken as an approximate measure of the percentage of transduced cells.

Statistical analysis
All statistical analyses were carried out using GraphPad Prism version 5.03. P values < 0.05 were considered significant. For comparisons between two groups, two-tailed unpaired t-tests were Supplementary Figure S5. Transduction of a co-culture spheroid consisting of MCF-7 and C5120 cells. The control adapter E2_5 (which blocks receptor binding) was used as an adapter on WT HAdV5. It can be observed that transduction (red, iRFP670) is preferentially localized in the core of the spheroid in C5120 fibroblasts (green), whereas MCF-7 cells (cyan) remain mostly not transduced. Viral transduction is depicted in red. A slice in the middle of an optically cleared spheroid is shown. The scale bar corresponds to 100 m. A representative image is shown from two independent experiments with three spheroids analyzed per condition per experiment. Figure S6. Designs of microfluidic chip drawn to scale. (a) Layout of the microfluidic chip for capturing spheroids. Circles indicate in-and outlets for perfusion and/or loading collagen-embedded spheroids. Four 300 m-sized bowl-shaped elements were included for capturing spheroids in the device. A lower layer of 20 m (height) containing cross channels (length: 100 m, width: 20 m) separated by 50 m-sized pillars was fabricated to connect the main middle compartment, containing the spheroid-capturing elements, with the side channels. A subsequent layer of 180 m (height) was fabricated, yielding a total channel height of 200 m. (b) Layout of the resistance channel chip that was connected via tubing to the spheroid-capturing chip in order to control the hydrodynamic resistance and thereby the interstitial flow. Through punching a hole at the desired location in the resistance channel chip (i.e. by not utilizing the full length of the channel system), the hydrodynamic resistance can be controlled, thereby affecting the interstitial flow rate.