Fig 2: Effects of selective OPN receptor blocking on metabolic activity in cultured ONH astrocytes.Metabolic cell activity of ONH astrocytes incubated with an anti-CD44 antibody or a synthetic RGD peptide for 36 and 72 hours compared to untreated cells (co). Values represent mean ± SD of 11 independent experiments (n = 11). Statistical significance is calculated by student’s t-test (*p<0.05; **p<0.01; ***p<0.001).

Fig 3: TGF-ß2 and OPN effects on metabolic activity in cultured ONH astrocytes.Cell viability was assessed by MTS assays. Quantification of the metabolic cell activity of ONH astrocytes treated with TGF-ß2 (1 ng/ml) and different OPN concentrations (0, 250, 1000, 2000 ng/ml) for 48, 96 and 144 hours. Values represent mean ± SD of 11 independent experiments (n = 11). Statistical significance was calculated by student’s t-test (*p<0.05; **p<0.01; ***p<0.001).

Fig 4: OPN loss impairs growth of basal-like PDA.(A) CRISPR-Cas9-mediated knockout of SPP1 in KP4 cells leads to a decrease in mRNA expression and secreted Osteopontin in two independent clones. (B,C) Immunoblot of KRT14 and E-cadherin (B) and GATA6 (C) levels following KO of SPP1 in KP4 cells. (D) Brightfield images of KP4 cell morphology when grown in 2D monolayer following KO of SPP1. Note the switch to a more epithelial phenotype in the KO cells. Scale bars, 100 μm. (E) In vitro growth rate of KP4 control versus SPP1 KO cell lines. Error bars represent s.e.m. (F) Growth rate of control and SPP1 KO KP4 cells following subcutaneous injection into the flank of NOD/SCID mice. Error bars represent s.e.m. (G) Relative size of control and SPP1 KO KP4 xenografts resected at day 25 (left) and comparative tumor weight (graph at right). (H) Representative H and E stained sections of lungs at endpoint following tail vein injection (TVI) of 2 × 106 KP4 control cells (WT; top) or SPP1 KO (bottom) into NOD/SCID mice. Scale bars, 400 μm. (I) Quantification of metastatic tumor burden in the lungs as a percentage of total lung area per mouse (n = 5 CTRL; n = 10 SPP1 KO). (J) Kaplan-Meier graph depicting survival of NOD/SCID mice following TVI of KP4 WT (black) or SPP1 KO cells (red). (K) High expression of SPP1 in basal-like PDA patient tumors from TCGA predicts shorter overall survival (SPP1 high n = 17, SPP1 low n = 48). Data from 65 patients. (L) Model depicting a classical to basal-like switch mediated by a GLI2-OPN signaling axis and in response to KRASG12D ablation. The classical and basal-like subtypes are marked by GLIlo/Hhhi and GLIhi/Hhlo expression, respectively. p-Value calculated by Log-rank test and by two-tailed unpaired t test. *p<0.05; **p<0.01; ***p<0.001.10.7554/eLife.45313.022Figure 7—source data 1.OPN loss impairs growth of basal-like PDA.

Fig 5: Localization of most prevalent OPN receptors on cultured human ONH astrocytes.Immunofluorescence signals (green) indicate expression of CD44 (A) and integrin receptor subunits IntaV (B), Intß3 (C) and Intß5 (D). The nuclei (blue) are counterstained with DAPI. Pictures are representative for three independent experiments. Scale bar 50 µm.