(229o) E-Selectin-Mediated Rolling and Firm Adhesion of Pancreatic Cancer Cells in Shear Flow

Authors: 
Shea, D., Johns Hopkins University
Li, Y. W., Johns Hopkins University
Konstantopoulos, K., Johns Hopkins University

E-Selectin-mediated Rolling and Firm Adhesion of Pancreatic Cancer Cells
in Shear Flow

Daniel J Shea1, Yi Wai Li1and  Konstantinos Konstantopoulos1

1Department of Chemical and Biomolecular Engineering, Johns
Hopkins University

Introduction:
Metastasis is a highly-regulated, multistep process, in which tumor cells migrate
from the primary tumor and enter the circulatory system where they interact
extensively with host cells before extravasating and
colonizing a distal organ. Extravasation of a tumor cell from a blood vessel first
requires the rolling and arrest of the cell to the vessel wall through the
formation of specific bonds [1,2].  Our
lab recently identified the sialofucosylated surface protein
Podocalyxin (PODXL) to be overexpressed in metastatic pancreatic cancer cells
and to bind to E/L-selectin [3]. PODXL-E-selectin mediated rolling velocities were found to depend on single bond kinetics and could be
predicted by the single
receptor-ligand off-rate and the
number of bonds [4].  We looked to build on this and develop
a quantitative understanding of how selectin-ligand tethering facilitates firm
adhesion at elevated levels of shear stress.  We also determined the rate-limiting
parameters in this cascade of events, such as the lengths of coated regions
with selectins and hyaluronic acid, necessary to support downstream firm
adhesion.  The quantitative analysis of the selectin-ligand binding
kinetics of known pancreatic cancer surface proteins in the presence of shear
flow will enable us to further our understanding of the metastatic process.

Materials and Methods: Using standard microfabrication
principles we patterned a glass
slide with E-selectin and hyaluronic acid (HA) patches with lengths varying
from 10-160μm in geometrically defined
regions with a gap distance between the two regions of between 30-120μm (Fig. 1A). Using
a microfluidic chamber, PODXL expressing metastatic pancreatic cancer cells
(Pa03c) were then perfused over the patches and all cells were tracked using a modified
code package produced by Eric M. Furst (University of
Delaware).  Using a custom Matlab program, we identified both the rolling and firm
adhesion regions and separate these regions into ten binned sections of
increasing patch length. The program then analyzes each cell track with respect
to these regions and indicates the position and velocity of the cell, while
also determining if the cell binds in either the rolling or firm adhesion
regions.

Results and Discussion: Experiments were performed with Pa03c pancreatic
cancer cells using devices coated with E-selectin (rolling) and HA (firm adhesion).  Cells were tracked as they pass over the
rolling, gap and firm adhesion regions, respectively (Fig. 1A). Cells interacting with E-selectin showed a clear rolling
behavior. The extent of interacting cells decreased with shear stress.  However, it increased with longer E-selectin patch lengths due to
increased ligand-selectin contact times (Fig. 1B). We determined that cells
bind infrequently to hylaronic acid-coated regions,
however the binding frequency increases when a cell first rolls on E-selectin (Fig. 1C)
Interestingly, for rolling cells, firm adhesion is most frequent when
cells roll on shorter E-selectin patch lengths (<40μm), likely because these cells are bound immediately
before entering the gap thus having a slower velocity. Together these data
indicate that rolling on E-selectin can facilitate firm adhesion with HA
and can do so even at small E-selectin patch lengths.


Figure 1. (A) Device schematic with rolling region (red) and firm adhesion region (green). (B) Frequency of rolling events at increasing patch lengths. (C) Frequency of firm adhesion events for cells that roll (blue) or that do not roll (red) before firmly adhering. 30μm gap.

 

BMES1 (1).tif

Conclusions: This work impacts a broad range of areas
including cell-biophysics, engineering and cancer research and is critically
important as it advances the knowledge of receptor-ligand processes as they
pertain to cancer metastasis.  Knowledge of the properties of
receptor-ligand-mediated tumor cell adhesion, in physiologically relevant
environments is critical to the design parameters necessary to engineer
sensors, to target biological entities based on recognition, and to design
molecules to interrupt adverse or pathological adhesion events.

References: (1) Cheung,
L. S. L. et al., Phys
Biol
, 2011, 8(1), 015013. (2) Stroka, K. M., and Konstantopoulos, K. Am J Physiol
Cell Physiol
, 2014, 306(2), 98-109. (3) Dallas, M. R. et al., Am
J Physiol Cell Physiol
, 2012, 303, C616–624. (4) Shea, D. J. et al.,
Oncotarget, 2015, 6(28), 24842.