Multiphoton excited polymerized biomimetic models of collagen fiber morphology to study single cell and collective migration dynamics in pancreatic cancer

Abstract

The respective roles of aligned collagen fiber morphology found in the extracellular matrix (ECM) of pancreatic cancer patients and cellular migration dynamics have been gaining attention because of their connection with increased aggressive phenotypes and poor prognosis. To better understand how collagen fiber morphology influences cell-matrix interactions associated with metastasis, we used Second Harmonic Generation (SHG) images from patient biopsies with Pancreatic ductal adenocarcinoma (PDAC) as models to fabricate collagen scaffolds to investigate processes associated with motility. Using the PDAC BxPC-3 metastatic cell line, we investigated single and collective cell dynamics on scaffolds of varying collagen alignment. Collective or clustered cells grown on the scaffolds with the highest collagen fiber alignment had increased E-cadherin expression and larger focal adhesion sites compared to single cells, consistent with metastatic behavior. Analysis of single cell motility revealed that the dynamics were characterized by random walk on all substrates. However, examining collective motility over different time points showed that the migration was super-diffusive and enhanced on highly aligned fibers, whereas it was hindered and sub-diffusive on un-patterned substrates. This was further supported by the more elongated morphology observed in collectively migrating cells on aligned collagen fibers. Overall, this approach allows the decoupling of single and collective cell behavior as a function of collagen alignment and shows the relative importance of collective cell behavior as well as fiber morphology in PDAC metastasis. We suggest these scaffolds can be used for further investigations of PDAC cell biology.

Statement of significance

Pancreatic ductal adenocarcinoma (PDAC) has a high mortality rate, where aligned collagen has been associated with poor prognosis. Biomimetic models representing this architecture are needed to understand complex cellular interactions. The SHG image-based models based on stromal collagen from human biopsies afford the measurements of cell morphology, cadherin and focal adhesion expression as well as detailed motility dynamics. Using a metastatic cell line, we decoupled the roles of single cell and collective cell behavior as well as that arising from aligned collagen. Our data suggests that metastatic characteristics are enhanced by increased collagen alignment and that collective cell behavior is more relevant to metastatic processes. These scaffolds provide new insight in this disease and can be a platform for further experiments such as testing drug efficacy.