Anisotropic rheology and directional mechanotransduction in vascular endothelial cells

Abstract

Adherent cells remodel their cytoskeleton, including its directionality, in response to directional mechanical stimuli with consequent redistribution of intracellular forces and modulation of cell function. We analyzed the temporal and spatial changes in magnitude and directionality of the cytoplasmic creep compliance (Gamma) in confluent cultures of bovine aortic endothelial cells subjected to continuous laminar flow shear stresses. We extended particle tracking microrheology to determine at each point in the cytoplasm the principal directions along which Gamma is maximal and minimal. Under static condition, the cells have polygonal shapes without specific alignment. Although Gamma of each cell exhibits directionality with varying principal directions, Gamma averaged over the whole cell population is isotropic. After continuous laminar flow shear stresses, all cells gradually elongate and the directions of maximal and minimal Gamma become, respectively, parallel and perpendicular to flow direction. This mechanical alignment is accompanied by a transition of the cytoplasm to be more fluid-like along the flow direction and more solid-like along the perpendicular direction; at the same time Gamma increases at the downstream part of the cells. The resulting directional anisotropy and spatial inhomogeneity of cytoplasmic rheology may play an important role in mechanotransduction in adherent cells by providing a means to sense the direction of mechanical stimuli.

Fig. 1.

Staining for different organelles and vesicles in the same VEC reveals that the endogenous particles used in our experiments are the mitochondria. (A) Phase image; (B) mitochondria marked with MitoTracker Green FM (Invitrogen); (C) lysosomes marked with Lysotracker Red DND-99 (Invitrogen); (D) merged image composed by using panels (A, B and C) in the blue, green and red channels, respectively. The images at the right of each panel show magnifications of the framed area in the corresponding left panel. The scale bars are 5 μm in left panels and 2 μm in right panels.

Fig. 2.

A_MSD (Eq. 1) of the mitochondria of VECs (n = 8), represented as a function of time separation τ. Open circles, particles undergoing active motion (r₊² = V²τ² + Dτ with V > 0); open triangles, particles undergoing passive motion (V = 0). The chain-dotted line indicates isotropy, A_MSD = 1, whereas the dashed line represents anisotropy increasing as A_MSD = 4 + 0.5 s⁻¹ τ. (Inset) Evolution of A_MSD for short τ. Error bars indicate S.D.

Fig. 3.

PMSD of the mitochondria of VECs (n = 8), represented as a function of time separation τ. Open triangles, PMSD along the direction of minimal compliance (r₋²); copen circles, PMSD along the direction of maximal compliance (r₊²). The chain-dotted, solid and dashed straight lines have power slopes β ≈ 0.45, 0.50 and 0.85, respectively. (Inset) Power slopes of the PMSD, β₊ and β₋, as a function of τ. Error bars indicate S.D.

Fig. 4.

Magnitude and directionality of creep compliance (Γ) at different positions of two VECs. (A) Before application of continuous LSS. (B) Twenty-four hours after starting LSS. Circles indicate the average position of each tracked marker; the segments are oriented parallel to the direction of maximum Γ at each point at τ = 1 s. The circles and segments are colored according to the logarithm of MSD of corresponding markers. Hot and cold colors indicate respectively low and high Γ.

Fig. 5.

Distribution of the angles, α₊, of the direction of maximum compliance in VECs under static condition for single cells (A) and averaged for a population (n = 8) (B). The data were obtained from PMSD at τ = 1 s. All distributions are normalized to yield the same area.

Fig. 6.

Distribution of α₊ for VECs at t = 0 h (dashed blue, n = 4) and t = 24 h (solid red, n = 16), for different time separations, τ. (A) static condition, τ = 1 s. (B–C) Continuous flow along the direction α = 0 at τ = 1 s (B) and 10 s (C). All distributions are normalized to yield the same area.

Fig. 7.

Time evolution of distribution of α₊ for single cells, p(α₊|t). Each line shows p(α₊|t) for the same cell at a different times, t. The black (filled triangles) curve represents the initial state (t = 0), while blue (filled circles), green (filled squares) and red (filled diamonds) curves were obtained 3, 6, and 9 h, respectively, after starting the experiment. The data were obtained from PMSD at t = 1 s. All distributions are normalized to yield the same area. (Insets) Higher-time-resolution, 2D plots of p(α₊|t) where the abscissa is time and ordinate is α₊.

Fig. 8.

Variation of average MSD in the upstream and downstream parts of VECs subjected to continuous LSS (Left) and under static condition (Right). *, P < 0.01.

Fig. 9.

Time evolution of PMSD of VECs subjected to continuous LSS (A) or under static condition (B). The curves for r₊² are plotted with solid symbols A and B Upper, and those for r₋² are plotted with open symbols in Lower. The black triangles represent the initial state (t = 0), whereas blue circles, green squares and red diamonds denote 4, 8, and 24 h, respectively, after starting the experiment. (Insets) 3D representations of r₊²(τ, t).