Figure 1.
Schematic of the microfluidic device for 3D microrheology experiments.
a) Cells in gel are loaded into the gel region via pipetting. The PDMS posts and surface tension keep the gel localized in the gel region. Media is then added to the media reservoirs via pipetting and allowed to diffuse into the gel region. b) Image of the device with collagen gel loaded into and kept contained in the gel region. The device channel height is 200 µm and the inter-post separation is 330 µm. (c) Bright field and (d) fluorescence images of cells embedded in 3D inside the device with ballistically injected 500 nm-diameter fluorescent nanoparticles. The scale bars are 100 µm in (b) and 20 µm in (c) and (d).
Figure 2.
Intracellular microrheology of a single cell in 3D in a microfluidic device.
a) Fluorescence image of an MDA-MB-231 cell in 2.5 mg/mL collagen. Actin is green and mitochondria are red. b) 1D MSDs in the two orthogonal directions with the maximum ratio in displacement fluctuations. Blue is the average of the MSDs in the directions of maximum fluctuation and green is the average of the MSDs in the directions of minimum fluctuation. Maximum and minimum directions are determined on a particle by particle basis. Yellow curves represent the 1D MSDs of each individual mitochondrion tracked. c,d) The complex shear modulus G in the directions with maximum (c) and minimum (d) fluctuations. Green is the elastic modulus and red is the loss modulus. e,f) Spatial plots of the log of the magnitude of the shear modulus G in the direction of (e) maximum and (f) minimum fluctuations at 10 Hz. Arrows point in the directions with the maximum (e) or minimum (f) fluctuations. Colors represent the magnitudes of log|G|. This experiment was performed at ambient conditions. The scale bar is 20 µm.
Figure 3.
Intracellular fluctuations in 2D vs. 3D.
a,b) 1D mitochondria MSDs in the two orthogonal directions with maximum ratio in displacement fluctuations for cells in (a) 2D and (b) 3D. Larger slopes (more fluid-like behavior) and magnitudes are exhibited by cells in 2D at longer time intervals. c,d) Logarithmic time derivative β of MSDs for cells in (c) 2D and (d) 3D in the corresponding 1D directions. Cells in 3D have lower β's, indicating more solid-like behavior. Error bars are s.e.m. The color code of the curves is the same as in Figure 2b.
Figure 4.
Cytochalasin D treated cells in 2D and 3D.
a,b) MSDs of cells treated with Cytochalasin D in (a) 2D and (b) 3D. c,d) β's of cells treated with Cytochalasin D in (c) 2D and (d) 3D. e) Comparison of MSDs at 50 ms of untreated and Cytochalasin D treated cells in 2D and 3D. f) Comparison of β's at 1s of untreated and Cytochalasin D treated cells in 2D and 3D. Max and min indicate the 1D direction of maximum and minimum fluctuations, respectively. “untr” and “CD” indicate untreated and Cytochalasin D treated cells, respectively. Cells in 2D exhibit a more pronounced effect under treatment than cells in 3D. Error bars are s.e.m. * indicates p<0.05. The color code in (a) and (b) is the same as in Figure 2b.
Figure 5.
Images of MDA-MB-231 cells in 2D and 3D with and without Cytochalasin D treatment.
a–h) Bright field and fluorescence images of (a,b) untreated cells in 2D, (c,d) Cytochalasin D treated cells in 2D, (e,f) untreated cells in 3D, and (g,h) Cytochalasin D treated cells in 3D. Green and red represent actin and mitochondria, respectively. The scale bar is 20 µm.
Figure 6.
Brownian dynamics simulations of active actin networks.
a) A 3D 3×3×3 µm3 (in the x, y, and z directions, respectively) domain with periodic boundary conditions on all sides, mimicking the cytoskeletal network in 3D. Actin filaments (blue) are polymerized from G-actin monomers and allowed to bind to myosin II motors (orange) and crosslinkers (yellow). Motors walk along two filaments toward their barbed ends, generating internal stress. b) A 3D 3×3×1 µm3 domain with periodic boundary conditions in the x and y directions and fixed boundaries in the z-direction, mimicking a 2D cell configuration. The actin network is more planarized, as the filaments align along the x-y plane. c,d,e) Average internal stresses in the x (blue), y (green), and z (red) directions for (c) a 3D network, (d) a 2D network with 1 µm height in the z-direction, and (e) a 2D network with 500nm height in the z-direction. 2D networks have larger stress fluctuations from the mean in the x and y directions (the plane of interest in experiments), but reduced fluctuations in z, as compared to 3D networks. 3D networks have comparable average internal stresses in all 3 dimensions. 2D networks have reduced average stresses in the z-direction (perpendicular to the plane of filament alignment), generating mechanical anisotropy. The stresses for each time point for each simulation were calculated by summing the tensional forces of filaments crossing a total area of 9 µm2 in each direction. Motors start walking and generating tension along filaments at time = 100 s.
Figure 7.
Simulated MSDs and β's for 2D and 3D systems using the effective temperature model to account for active stress fluctuations.
a,b) Simulated (a) MSDs and (b) corresponding β's for cells in 2D (blue) and 3D (green). Cells in 2D have a larger stress fluctuation term A (3 times larger in this case). c) The Laplace transform of the MSDs based on the effective temperature model for different A's (from 0.001 to 106 [1/s]) and fixed s0 = 0.001 Hz. The simulations in (c) assume an elastic material (G is constant and equal to 20 Pa) with 70 nm-radius tracer particles at T = 300 K and power-law trends are attributed to the effective temperature. The bottom and top dotted lines have slopes of 1 and 2, indicating thermal and super-thermal (with motor activity) spectral trends, respectively.