Poster Presentation 27th Lorne Cancer Conference 2015

PAI-2 (SerpinB2) Inhibits Pancreatic Cancer Cell Invasion. (#175)

Nathanial Harris 1 2 3 4 , J Conway 1 , R Shearer 1 , K L Vine 2 3 4 , M Pajic 1 , M Pinese 1 , P Timpson 1 , D N Saunders 1 , M Ranson 2 3 4
  1. Kinghorn Cancer Centre, Garvan Medical Research Institute, Sydney, NSW, Australia
  2. Center for Medical Bioscience , University of Wollongong, Wollongong, NSW, Australia
  3. Illawarra Health and Medical Research Institute, University of Wollongong, Wollongong, NSW, Australia
  4. Biological Sciences, University of Wollongong, Wollongong, NSW, Australia

The serine protease urokinase (uPA) coordinates the tissue-remodelling plasminogen activation system, with overexpression of uPA recognised as an important biomarker of metastatic disease and a druggable anticancer target. Conversely, elevated co-expression of the uPA inhibitor, PAI-2 (SerpinB2), improves survival in breast and other cancers [1]. Expression analysis of uPA, PAI-1 and PAI-2 in the patient cohort from the Australian Pancreatic Cancer Genome Initiative [2] corroborates this finding in pancreatic cancer. Further, analysis of genome data using the cBio portal indicates that the SerpinB2 gene is either mutated, amplified or deleted in 12% of pancreatic tumours [3]. Clinicopathological and functional genomics data to date suggests that the protective effect of PAI-2 may be related to differential tumour/stromal cell expression levels during cancer progression, related to its serpin protease inhibitor function [4].

To define the tumour versus host-specific effects of PAI-2, and the role of its serpin and/or other potential functions in the metastatic process, we are using organotypic assays, utilising fibroblast-contracted collagen 1 dermal equivalent matrices as malleable platforms for the assessment of tumour/stromal interactions in a 3D context [5]. In these assays, the consequences of PAI-2 up- or down-regulation in either tumour cells or fibroblasts on matrix formation and cell invasion can be functionally assessed.

PAI-2 significantly inhibits the invasion of pancreatic ductal adenocarcinoma cells (PDAC) over dermal fibroblast generated matrices compared to controls (p < 0.0001). An inactive PAI-2 mutant, R380APAI-2, had no significant effect on PDAC invasion, confirming uPA-dependent inhibition of invasion. Using second harmonic generation and SEM, we confirm that PAI-2 null murine embryonic fibroblasts (MEFs) significantly affect collagen I fibrillogenesis, cross-linking ability, matrix formation and structure compared to wild-type MEFs (p < 0.0001)and that the 3D organotypic matrices incorporating null fibroblasts are more permissive to tumour cell invasion (p = 0.001). The concomitant migration mode of uPA-overexpressing/PAI-2 under-expressing PDAC cells varied between the different matrices, with mesenchymal migration predominantly occurring through wild-type matrices, while amoeboidal migration was the principle migratory mode observed through PAI-2 null matrices. Altogether, this suggests that while the tumour microenvironment influences stromal-cell derived PAI-2 expression, at a particular point during cancer progression, down-regulation of PAI-2 in cancer-associated fibroblasts significantly impacts on tumour cell invasiveness.  We are currently generating tumour cell lines in which PAI-2 expression is modified using inducible overexpression and silencing approaches for use in the 3D assays.

In summary, this is the first study to provide evidence for both tumour and stromal cell specific effects of PAI-2 on PDAC invasion.

  1. Croucher, D.R., et al., Revisiting the biological roles of PAI2 (SERPINB2) in cancer. Nature Reviews Cancer, 2008. 8(7): p. 535-545.
  2. Biankin, A.V., et al. Pancreatic cancer genomes reveal aberrations in axon guidance pathway genes. Nature, 2012. 491(7424): p. 399-405.
  3. Cerami, E., et al. The cBio Cancer Genomics Portal: An Open Platform for Exploring Multidimensional Cancer Genomics Data. Cancer Discovery, 2012. 2(5): p. 401-404.
  4. Cochran, B.J., et al., Dependence on endocytic receptor binding via a minimal binding motif underlies the differential prognostic profiles of SerpinE1 and SerpinB2 in cancer. J Biol Chem, 2011. 286(27): p. 24467-75.
  5. Timpson, P., et al., Organotypic collagen I assay: a malleable platform to assess cell behaviour in a 3-dimensional context. J Vis Exp, 2011(56): p. e3089.