Figure 1.
(A) Schematic of Experimental Setup. Three inlets supply cell media to an epithelial sheet cultured in the wide,
high channels. This is done under no flow. (B) After cells have reached confluency, 15
l/min flow (
) is applied, leading to three separated laminar streams. One stream contains 0.05% trypsin, while the other two contain cell media. This cleaves cells from the channel as shown in (C). On average, cleavage takes 5 min. Images are acquired in brightfield.
Figure 2.
Composition of a Reconstituted Wound.
(A) Schematic of Experimental Setup in (B) and (C). Cell media and trypsin merge to selectively cleave cells from a confluent epithelial sheet. (B) A gradient in trypsin (0.05% or 21.4M) denudes the cell sheet, leaving cells viable as shown by a live stain (green). (C) Two-dimensional plot of trypsin concentration during denudation. The flow,
, is 15
l/min for an average of 5 min. The diffusion coefficient for trypsin,
= 2.1
m
/s was estimated with a molecular radius of 1.5 nm [50]. The white box refers to the region which is averaged to see a line plot perpendicular to the direction of flow in (G). (D) Schematic of the Experimental Setup in (E) and (F).
replaced trypsin after the cleavage of cells. (E) Delivery of
(10mM) after trypsin treatment creates an apoptotic border (1-4 rows of cells), and immerses the rest of the sheet in a non-apoptotic concentration. Green: alive, red: dead. (F) Concentration profile of
for
up to 15 min. The diffusion coefficient for
,
m
/s was taken at 25C [51]. This diffusivity creates a larger mixing zone than the mixing zone for trypsin. The white box refers to the line plot in (G). (G) Line plot of trypsin and peroxide concentrations averaged over the x dimension in parts (C) and (F). The calculation is taken 10
m from the coverslip, to approximate the concentration at the apical surface of the cells. (F) Shear calculation across the width of the channel (z) for the initial flow,
, used for delivery of trypsin and
, and final flow of pure media (no trypsin or peroxide) at
= 0.5
l/min. The
is maximum in the center of the channel, and equals 0.35 Pa for
, and 0.001 Pa for
.
Figure 3.
Denudation Disrupts the Balance of Tension in the Sheet.
(A,B) Displacement field of the retraction of an epithelial sheet during the enzymatic cleavage of cells from the left lane. Retraction is shown by the direction of the yellow vectors that are plotted over top of the submarginal region of the sheet. A low density sheet ( 2000 cells/
) is shown in (A) and a high density sheet (
cells/
) is shown in (B). In both cases, retraction occurs for multiple cell diameters. The retraction occurs perpendicular to the flow,
. Scale bar is 50
m. Images are in brightfield. The spacing between grid points is 8.6
m. (C) Schematic of the experimental setup in (D-F). Trypsin is delivered through the central channel or the left channel. We group the results in both cases (*). (D) During trypsin treatment, cell bonds break and cells elongate in the direction of movement. (E) After a few minutes, the sheet is cleaved, and there is a retraction of the border cells perpendicular to the flow. Images are in brightfield. (F) The retraction of the border cells is shown to be dependent upon density, where cells are grouped broadly into two regimes, low density (below 2000 cells/
) and intermediate density (
2000 cells/
). There is a clear density dependence on the retraction (
). (G) Schematic for the experimental setup in (H–J).
blebbistatin is delivered in the leftmost channel to disrupt actomyosin contractility (with cell media in the other channels). (H) Brightfield image of a confluent epithelial sheet during blebbistatin treatment. (I) Retraction at the blebbistatin/media interface at T = 44 min (Movie S3). Red Arrows show cell displacement. Scale bar is 50
m. (F) Density dependence of retraction with blebbistatin treatment (
). Images are in brightfield.
Figure 4.
Healing is a Composite of Spreading and Migration.
Timelapse of healing of enzymatically denuded sheets for (A) intermediate density (2000-3000 cells/) and (B) high density (
3000 cells/
). Scale bar is 100
m. Cells are stained with DiI. (C) Diagram of migration that increases the overall sheet area (
), which also increases the curvature of the leading edge border. The cell sheet immediately after cleavage by trypsin (left) has a linear profile (red arrows). Thus the angle between successive vectors is small. Over time, the cell edge becomes non-uniform (middle, right). This increases the angle between successive vectors. The mean cosine decays with a characteristic length,
. In comparing migration rates, we discount the initial acquired area, as it likely corresponds to the spreading regime, as
varies during this time. (D,E) Increase in acquired area,
(blue, closed circle) and characteristic length,
(red, open circle) for the (D) intermediate density and (E) high density cell sheet.
Figure 5.
Leading Edge Dynamics are Equivalent between Abrasively Denuded and Enzymatically Denuded Sheets.
Increase in area () during the healing of the cell sheet (N = 15). The rate of healing,
for cell sheets of different densities (inset). Solid symbols: microfluidic assay. Open symbols: scratch assay.
Figure 6.
Velocity and Persistence are Hindered in the Submarginal Region of Enzymatically Denuded Sheets.
Displacement Vectors (by cell tracking) for sheets after enzymatic (A) or scratched (B) denudations (red arrows). L.E.: leading edge, S.M.: submarginal cells. The leading edge is hand drawn in gray. Grid size is 20 m. Straightness (C) and Speed (D) of cell movement as a function of distance from the (initial) leading edge,
(N = 6). (E) Immunostaining of E-cadherin at T = 0, 4, and 12 hours post cleavage (N = 3). Scale bar is 50
m. (F) Fluorescence intensity as a function of
. Straightness (G) and Speed (H) as a function of
for sheets blocked with 200
g of anti-E-cadherin antibody (N = 2).
Figure 7.
Tension at E-cadherin junctions between the leading edge and submarginal cells.
Immunostaining of E-cadherin of epithelial cells 20 hrs after denudation by trypsin. (A) Two examples of E-cadherin localization at the leading edge cells that shows elongated cell-cell interface (red arrows), both in the direction of movement as well as perpendicular to the direction of movement. (B) E-cadherin in submarginal cells that does not have elongated interfaces. Scale bar is 10 m.
Figure 8.
ROS Inhibits Movement in Enzymatically Denuded Sheets.
Mean healing rates () for cells immersed in lysate or exposed to a gradient in
(all densities). ** indicates a p-value
. Anova Analysis in Fig S7.