CD73 controls Myosin II–driven invasion, metastasis, and immunosuppression in amoeboid pancreatic cancer cells

Pancreatic ductal adenocarcinoma (PDAC) has a very poor prognosis because of its high propensity to metastasize and its immunosuppressive microenvironment. Using a panel of pancreatic cancer cell lines, three-dimensional (3D) invasion systems, microarray gene signatures, microfluidic devices, mouse models, and intravital imaging, we demonstrate that ROCK–Myosin II activity in PDAC cells supports a transcriptional program conferring amoeboid invasive and immunosuppressive traits and in vivo metastatic abilities. Moreover, we find that immune checkpoint CD73 is highly expressed in amoeboid PDAC cells and drives their invasive, metastatic, and immunomodulatory traits. Mechanistically, CD73 activates RhoA–ROCK–Myosin II downstream of PI3K. Tissue microarrays of human PDAC biopsies combined with bioinformatic analysis reveal that rounded-amoeboid invasive cells with high CD73–ROCK–Myosin II activity and their immunosuppressive microenvironment confer poor prognosis to patients. We propose targeting amoeboid PDAC cells as a therapeutic strategy.

The PDF file includes:

Supplementary Text Figs. S1 to S7 Legends for movies S1 to S5 References
Other Supplementary Material for this manuscript includes the following:

Cell culture
Cells were grown at 37⁰C and 5% CO2 in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin (all from Gibco).Panc1, PaTu8988T and Cfpac1 cell lines were a kind gift from Professor Tatjana Crnogorac-Jurcevic (Barts Cancer Institute, London, UK); SW1990, Colo357, Capan2 and Patu8902 cell lines were from Doctor Claire Wells (King's College, London, UK), and PaTu8988S cell line was purchased from the German Collection of Microorganisms and Cell Cultures (DSMZ, Leibniz Institute, Germany).KPC-claus cells were from Doctor Claus Jorgensen's lab.All cell lines were regularly tested for mycoplasma contamination and cell phenotypes were verified routinely.

Transfection and siRNA
250K cells/well were seeded on 6-well plates and transfected with 20nM SmartPool siRNA oligonucleotides, using Optimem-I and Lipofectamine 2000 (Invitrogen).Nontargeting siRNA was used as control.Transfected cells were incubated for 24 h, after which they were harvested and re-seeded for conditioned media or western-blot experiments.

Confocal fluorescence microscopy and image quantification
Cells were seeded on top of a collagen I matrix and immunostained as described.
Imaging was carried out on a Zeiss LSM 510 Meta confocal microscope (Carl Zeiss) with C-Apochromat × 40/1.2NA (water) or a Plan Apochromat × 63/1.4NA (oil) objective lenses and Zen software (Carl Zeiss).p-MLC2 fluorescence signal was quantified calculating the pixel intensity in single cells relative to the cell area.

3D invasion assays
Experiment was performed as described in (21).Briefly, cells were suspended in serum-free bovine collagen I at 2.3 mg/ml to a final concentration of 1,5x10 4 cells per 100 μl of matrix, seeded on 96-wel(21)l plates and spun down to the bottom of the well.After the matrix was polymerized, 10% FBS-containing media was added on top of the matrix, allowing the cells to invade upwards for 24 h as previously described.
Then plates were fixed with 4% formaldehyde, stained with 5 µg/ml Hoechst 33258 (Life Technologies) and imaged on a Zeiss LSM 510 Meta confocal microscope (Carl Zeiss) with Zen software.Confocal z-slices were collected from each well at the bottom of the well and at 50µm.The 3D migration index was calculated as number of invading cells at 50 µm divided by the total number of cells at the bottom.

Spheroid invasion assays
Cells were resuspended in 1ml of low viscosity medium (800 µL of DMEM 10% FBS with 200 µL of low viscosity methylcellulose solution) at a final concentration of 4x10 5 cells/ml.40 droplets (25 µL each) were plated on the lid of a tissue culture dish for 48 hours at 37ºC -5% CO2, allowing cells to cluster into compact spheroids.Then, spheroids were collected, centrifuged (300 rpm, 15 seconds), and resuspended in a collagen I solution (600µl, 1.7mg/mL in DMEM).Spheroids were seeded in 2 wells of a 24-well plates (300µl each) and left in the incubator at 37ºC -5% CO2 for 4 hours to allow matrix polymerization.Medium with 10% FBS (supplemented with treatments if needed) was added on top and phase contrast pictures were taken every 24 hours, during 4 days.For spheroid invasion quantification, the increase on the area occupied by the spheroids between day 0 (when spheroids were embedded into the collagen I matrix) and day 4 was calculated by using ImageJ software.If needed, plates were fixed with 4% formaldehyde, stained with Hoechst 33258 (Life Technologies) and Alexa Fluor 647-phalloidin (Life Technologies) and imaged on a Zeiss LSM 510 Meta confocal microscope (Carl Zeiss) with Zen software.

Cytokine array
Secreted media from PaTu8988T, PaTu8988S or PaTu8988T transfected with a nontargeting or a NT5E-targeting siRNA were collected after 48h in serum-free conditions and incubated with Human Cytokine Antibody Array (RayBiotech, Inc., C4000), following the manufacturer's protocol.Membranes were incubated with biotinylated detection antibody cocktail, with HRP-conjugated streptavidin and with detection buffers.Images were obtained using a chemoluminescent imaging system and densitometry analysis was performed using the Protein Array Analyser plugin for ImageJ (http://image.bio.methods.free.fr/ImageJ/?Protein-Array-Analyzer-for-ImageJ.html).

Gene set enrichment analyses (GSEA)
Normalized gene expression microarray data of epithelial versus mesenchymal PDAC cells was analysed by comparing PaTu8988T and PaTu8988S cell lines (GSE14701 Gene sets for epithelial-to-mesenchymal transition, tumour zone peripheral versus central signature, metastasis signature and pancreatic cancer gene signature were downloaded and analysed using Gene Set Enrichment Analysis (GSEA) software (http://www.broadinstitute.org/gsea/index.jsp).For the transcriptional signature of melanoma cells with high myosin II activity, genes upregulated in high myosin II activity compared to low myosin II activity melanoma cells (cells treated with ROCKi and blebbistatin) were selected using a fold change >1.5 and a p value <0.01.GSEA analysis was performed as described above.Enrichment plot (green line) show upregulation of gene signature in indicated samples (resistant, non-responders or ontreatment).Analysis was performed with the specific settings: permutations-1,000, permutation type-gene set, metric for ranking genes-t-test.
Biological processes enriched in gene lists were identified using g:Profiler web server (75).

Analysis of gene expression data and immunostainings from human databases
Gene expression data, clinical information and Cytokeration-19 IHC pictures of human pancreatic cancer samples (176 patients) from The Cancer Genome Atlas (TCGA) database were downloaded from The Human Protein Atlas website (https://www.proteinatlas.org/).Comparison of gene expression between normal and tumoural pancreas tissue from TCGA, as well as correlation between genes from TCGA, were performed using OncoDB website (76) (https://oncodb.org/index.html).

Immune cell distribution estimation using CIBERSORT
Normalized gene expression data of tumoral pancreas (GSE71729 (51)) were used for immune cell proportion estimation.Samples were sorted according to ROCK1 and NT5E median expression and their RNA signatures were analysed using CIBERSORT analytical tool (https://cibersort.stanford.edu/),using LM22 signature gene file (100 permutations for statistical analysis).

FACS analysis of KPC Tumours
Pancreas was collected in ice-cold PBS and washed in Hank's balanced salt solution (HBSS) solution before mincing using scalpels.The pieces were then incubated in 2 mg/mL collagenase (Sigma) in HBSS with 50 mg/mL DNase (Sigma) for 20 min at

. Related to Figure 1 .A,Supplementary Figure 2 .Supplementary Figure 3 . Related to Figure 3 .A,Supplementary Figure 4 .
GSEA plots showing enrichment of "epithelial-to-mesenchymal" gene signature in melanoma amoeboid cells compared to amoeboid cells treated with ROCK and myosin inhibitors.B, Representative immuno-blots of CD44, Nanog, KLF4, Oct4, CD13 and GAPDH for PaTu8988T and PaTu8988S cells.C, F-actin confocal images of individual, doublet, cluster and colony-forming cells.Scale bar 10 µm.D, Representative bright field pictures of PaTu8988T and Panc1 cells plated on plastic or on top of a collagen gel.Scale bar 25 µm.E, Quantification of normalized pMLC2 levels from (1F) immunoblots (n=5 to 10).F, GSEA plots showing enrichment of "tumour periphery" and "metastasis" gene signatures in PaTu8988T cells compared to PaTu8988S cells.G, Venn diagram showing genes upregulated in PaTu8988T cells compared to PaTu8988S cells, in Panc1 cells compared to Capan2 cells, and commonly upregulated.H, Schematic of the invasion assay protocol (upper panel); representative merged images of PaTu8988T, Cfpac1 and PaTu8988S cells at the bottom (green) and invading at 50 µm (red).Scale bar 100 µm (lower panel).I, Quantification of 3D invasion index of PDAC cell lines (n≥3, normalized to PaTu8988T cell line, each dot represents the average index in an independent experiment).(E, I) graphs show mean +/-SEM.P-values to compare pMLC2 levels (E) were calculated using one-way ANOVA with Tukey's multiple comparison tests.Related to Figure 2. A, pMLC2 confocal images in PaTu8988T, Cfpac1 and PaTu8988S cells treated with DMSO (control) or 1 µM GSK269962A for 24 hours.Scale bar 10 µm.B, Quantification of pMLC2 immunofluorescence signal normalized by cell area (n≥3, intensity relative to the same cell line treated with DMSO, each value represents average intensity in an independent experiment).C, Representative merged images of PaTu8988T, Panc1, SW1990 and Cfpac1 cell lines treated with 1 µM GSK269962A or DMSO (control) for 24 hours at the bottom (green) and invading at 50 µm (red).Scale bar 50 µm.D, Quantification of 3D invasion index of A and A/E PDAC cell lines treated with 1 µM GSK269962A or DMSO (control) for 24 hours (n≥3, normalized to DMSO index, each dot represents the average invasion index in an independent experiment).(B,D) graphs show mean +/-SEM.P-values to compare invasion index (D) were calculated using two-way ANOVA with Sidak's multiple comparison test.Representative pMLC2 immunostainings of KPC mouse normal pancreas, PanIN and PDAC (left panel, scale bar 100 µm) and quantification of the proportion of pMLC2 positive cells within each compartment (right panel, n≥7).B, Representative pMLC2 immunostainings of KPC mouse tumours and corresponding Qupath map classifier, showing cancer cells (blue), fibroblasts (green), immune cells (purple) and pMLC2-positive cells (red).Scale bar 100 µm.C, Quantification of the proportion of cancer cells and cancer-associated fibroblasts (CAF) within KPC tumour bulk, invasion front and liver metastasis (n≥4).D, Representative Ecadherin and E-cadherin/pMLC2 pseudo-coloured multiplex images of KPC tumour bulk (TB) and invasion front (IF) (left panel, scale bar 100 µm) and quantification of the proportion of Ecadherin positive cancer cells within KPC tumour bulk and invasion front (right panel, n≥5).E, Representative immuno-blots of E-cadherin, Cytokeratin-19, CD44, Vimentin and GAPDH of PaTu8988T and KPC-claus cells.F, Representative bright field (left panel, scale bar 500 µm) images of KPC-claus spheroids treated with 1 µM GSK269962A or DMSO (control) at day 4. G, Representative merged images of KPC-claus cells treated with 1 µM GSK269962A or DMSO (control) for 24 hours at the bottom (green) and invading at 50 µm (red) (left panel, scale bar 100 µm), and quantification of 3D invasion index of KPC-claus cells treated with 1 µM GSK269962A or DMSO (control) for 24 hours (right panel, n=4 technical replicate, normalized to DMSO index).H, Origin of KPC-claus-LivMet cells (left panel); pMLC2, Ecadherin and CD44 confocal immunofluorescence images (right panel, scale bar 40 µm).I, quantification of pMLC2 immunofluorescence signal normalized by cell area (right panel, n≥30, each dot represents a cell).J, Representative immuno-blots of E-cadherin, Cytokeratin-19, CD44, Vimentin, pMLC2, total MLC2 and GAPDH of KPC-claus and KPC-claus-LivMet cells.K, Representative pMLC2 immunostaining of an intravital PaTu8902-GFP tumour (upper panel, scale bar 1mm; inset scale bar 100 µm) and quantification of the proportion of pMLC2 positive cells within tumour bulk and invasion front (lower panel, n=3).(A, C, D, G, I, K) graphs show mean +/-SEM.P-values to compare percentage of pMLC2 positive cells (B) were calculated using one-way ANOVA with Tukey's multiple comparison test.P-values to compare E-cadherin positive cells (D) and invasion index (G) were calculated using Student's t-test.Pvalue to compare p-MLC2 intensities (I) was calculated using Mann-Whitney test.P-value to compare percentage of pMLC2 positive cells (K) was calculated using paired Student's t-test.Related to Figure 4. A, Pie chart showing biological process enriched with secreted factors upregulated in PaTu8988T cells.B, Quantification of normalized p-Akt levels from (4H) immunoblots (n=4 to 7).(B) graphs show mean +/-SEM.P-values to compare p-Akt levels were calculated using Student's t-tests.