This preprint describes two mechanistically different routes by which pancreatic ductal adenocarcinoma (PDAC) cells breach endothelial barriers during extravasation. Using in vitro and in vivo models, the authors show that different PDAC cell lines employ separable strategies to cross the endothelium: a protrusion-driven mode and an endothelial retraction–driven mode. These observations indicate heterogeneity in how tumor cells interact with and disrupt vascular barriers during metastatic dissemination.
One described route is a protrusion-driven mechanism exemplified by MIA PaCa-2 cells. In this mode, tumor cells form filopodia-like protrusions that penetrate endothelial junctions. These protrusions allow the cancer cells to access the basal extracellular matrix (ECM) beneath the endothelium and to spread across it. The process is characterized by active protrusive behaviour at cell–cell contacts with endothelial cells rather than by wholesale removal or retraction of adjacent endothelial cells.
The protrusion-driven route thus represents a direct, invasion-like crossing of the endothelial barrier in which the tumor cell physically extends into intercellular endothelial junctions to reach and travel along the subendothelial ECM.
By contrast, AsPC-1 cells display a markedly different phenotype during extravasation. These cells remain rounded on the luminal surface of the endothelium and do not visibly extend filopodial protrusions into junctions. Instead, AsPC-1 cells trigger rapid retraction and detachment of neighbouring endothelial cells from the basal lamina, creating gaps through which the tumor cells can cross. The authors observed that AsPC-1 arrest and survival on the vessel wall precede endothelial detachment and subsequent extravasation via this retraction-mediated pathway.
AsPC-1–mediated barrier breach therefore relies on alteration of endothelial behaviour and integrity rather than on a direct protrusive invasion of intercellular junctions by the tumor cell.
The two extravasation modes were not limited to cultured endothelial monolayers: similar behaviours were observed in zebrafish larvae, providing an in vivo corroboration of the distinct phenotypes. In the mouse lung, AsPC-1 cells were documented to arrest, survive on the vasculature, induce endothelial detachment from the basal lamina, and extravasate through the retraction mechanism prior to metastatic outgrowth. These model systems support the biological relevance of both modes of endothelial barrier crossing in different experimental contexts.
Mechanistic studies reported in the preprint indicate that AsPC-1 cells secrete factors sufficient to destabilize endothelial monolayers. AsPC-1 cells also induce endothelial apoptosis, which may contribute to barrier disruption. Importantly, the authors note that blocking apoptosis does not prevent the endothelial barrier disruption caused by AsPC-1 cells, implying that apoptosis is not the sole or essential driver of the retraction process.
Conversely, pharmacologic inhibition of Src-family kinases with saracatinib protected endothelial barriers in the reported experiments. Treatment with saracatinib limited early vascular disruption in the lung and delayed metastatic outgrowth, suggesting that Src-dependent signalling contributes to the endothelial responses that permit tumor cell extravasation, at least in part.
The demonstration of two mechanistically distinct extravasation modes has implications for anti-metastatic strategies. Because PDAC cells can either actively invade through junctions or induce endothelial retraction and detachment, therapies aimed at a single molecular pathway may be insufficient to block all routes of vascular escape. The protective effects of the Src-family kinase inhibitor saracatinib on endothelial barrier integrity in the models studied indicate that targeting endothelial signalling pathways can reduce vascular disruption and delay metastatic progression. However, effective clinical strategies may require combination approaches that address both protrusive invasion mechanisms and tumour-driven endothelial destabilization.
This work is presented as a preprint. The authors declare funding sources and institutional affiliations in the manuscript. A competing interest statement notes that one author (L.M.C.) has consulted for a pharmaceutical company and mentors a team unrelated to this work; other authors declared no competing or financial interests. Detailed experimental methods, quantitative results, and potential limitations of the models used are reported in the preprint; specific experimental parameters and numerical outcomes were not restated here and should be consulted directly in the source for complete data and methods.