Neutrophil morphology and migration are affected by substrate elasticity

Abstract

To reach sites of inflammation, neutrophils execute a series of adhesion and migration events that include transmigration through the vascular endothelium and chemotaxis through the vicinal extracellular matrix until contact is made with the point of injury or infection. These in vivo microenvironments differ in their mechanical properties. Using polyacrylamide gels of physiologically relevant elasticity in the range of 5 to 100 kPa and coated with fibronectin, we tested how neutrophil adhesion, spreading, and migration were affected by substrate stiffness. Neutrophils on the softest gels showed only small changes in spread area, whereas on the stiffest gels they showed a 3-fold increase. During adhesion and migration, the magnitudes of the distortions induced in the gel substrate were independent of substrate stiffness, corresponding to the generation of significantly larger traction stresses on the stiffer gels. Cells migrated more slowly but more persistently on stiffer substrates, which resulted in neutrophils moving greater distances over time despite their slower speeds. The largest tractions were localized to the posterior of migrating neutrophils and were independent of substrate stiffness. Finally, the phosphatidylinositol 3-kinase inhibitor LY294002 obviated the ability to sense substrate stiffness, suggesting that phosphatidylinositol 3-kinase plays a mechanistic role in neutrophil mechanosensing.

Figure 1

Comparison of neutrophil spreading on substrates of different stiffness. (A) DIC images show neutrophils spreading over a period of 100 seconds on 4 gels of different stiffness (5, 10, 20, and 50 kPa), each at 3 time points. (B) The average spread area for each stiffness condition is plotted over time. Neutrophils on the softer gels maintained a smaller area and spread significantly less than the neutrophils on the stiffer gels. The majority of spreading was completed within the first 120 seconds. Neutrophils on the softest substrate showed no statistical change in size, even after 10 minutes, whereas those on the stiffest substrate more than tripled in area in the same amount of time. The number of cells measured for each condition is indicated in the legend. Error bars represent SE measurements. *Significant increase for the 10-kPa gels (with respect to the t = 0 point). #Significant increase for the 20-kPa gels (with respect to the t = 0 point). **Significant increase for the 50-kPa gels (with respect to the t = 0 point). Inset: The areas of the 4 neutrophils shown in the DIC images are plotted for the full 10 minutes.

Figure 2

Comparison of root mean square gel deformation and traction stress exerted by neutrophils on substrates of different stiffness. (A) The root mean square value of the displacements of the fiducial markers embedded in the gel under the area of the cell is plotted as a function of time. There is no statistical difference between gels of different stiffness. (B) The root mean square stress applied by the neutrophil on the gel in the area occupied by the cell is plotted as a function of time for each stiffness. Cell outlines for calculations were determined from DIC images. Error bars represent SEM.

Figure 3

Migration plots showing 8 random neutrophil migration paths for gels of 5 different stiffnesses. On the softer gels (5 and 10 kPa), the neutrophils moved quickly and changed direction often. On the stiffer gels, the neutrophils were more efficient, moving slower but showing a much greater persistence and changing direction less often.

Figure 4

Analysis of mobility and directionality of neutrophils migrating on substrates of different stiffness. (A) The MSD is plotted as a function of time between steps. Neutrophils on the softer gels (5 and 10 kPa) had a slope of ∼ 1, indicative of diffusive behavior. As the stiffness increased, the slope of the MSD increased, indicating more persistent behavior. The MSD was also larger for neutrophils on softer gels. (B) The angular distribution of steps is shown for each stiffness based on a 10-second interval between frames. The stiffer gels showed a more peaked distribution ∼ 0°, confirming that corresponding motion was more directed. Inset: The persistence, defined in the inset as the percentage of steps for which −π/2 < θ < π/2, is shown for time steps of 10, 20, and 30 seconds. The cells were more persistent on the stiffer gels. Error bars represent SE. *P < .05 vs 10 kPa.

Figure 5

Comparison of chemokinetic and chemotactic behavior. (A) The MSDs for neutrophils on 10- and 100-kPa substrates are shown for both random and directed migration. The cells move more efficiently on the stiffer substrate. (B) The angular distribution of steps is shown for neutrophils undergoing both chemokinesis and chemotaxis on a 60-second interval. Neutrophils chemotaxing toward a pipette tip showed more peaked distributions than neutrophils randomly migrating. For both chemokinetic and chemotactic behavior, however, the angular distribution of steps showed a greater peak on the 100-kPa substrates compared with the 10-kPa substrate for each condition. Migration is thus more persistent on stiffer substrates.

Figure 6

Distribution of tractional stresses in a migrating neutrophil. (A) A time series of images showing a neutrophil migrating chemokinetically to the right on a 10-kPa gel. The DIC images show that the cell has moved over a full body length in 300 seconds (supplemental Video 2). The images to the right of the DIC images are the corresponding traction images for those time points. The white outline over the traction images represents the approximate boundary as determined from the DIC images. The largest stresses occurred in the posterior of the neutrophil, a feature independent of gel stiffness. The full series of images for this particular example can be seen in supplemental Video 3. (B) A larger version of the last traction image in the series from panel A is shown. The white arrows represent the overall direction and magnitude of the applied tractional stresses. The few large tractions in the center of the cell were balanced by more numerous small tractions in the surrounding areas.

Figure 7

LY294002 inhibits the ability of the neutrophil to sense stiff substrates. The root mean squared speed was calculated for neutrophils chemotaxing toward a pipette tip loaded with fMLP. Untreated cells and cells treated with 0.02% DMSO (vehicle control) and 20 μM LY294002 were recorded for a period of 30 minutes. No statistical difference was seen between the untreated cells and vehicle (DMSO)–treated cells. *Both the untreated and vehicle (DMSO)–treated cells showed a statistical difference between 10- and 100-kPa gels. Cells treated with LY294002 were indistinguishable on the 10- and 100-kPa gels. **The cells treated with LY294002 on the 100-kPa gels were significantly different from the untreated and vehicle cells on 100-kPa gels. In addition, cells treated with LY294002 showed limited ability to spread as shown in the representative DIC images below the plot. The DIC images are snapshots of single cells taken from supplemental Videos 3 through 6, respectively.