Alpha-actinin binding kinetics modulate cellular dynamics and force generation

Abstract

The actin cytoskeleton is a key element of cell structure and movement whose properties are determined by a host of accessory proteins. Actin cross-linking proteins create a connected network from individual actin filaments, and though the mechanical effects of cross-linker binding affinity on actin networks have been investigated in reconstituted systems, their impact on cellular forces is unknown. Here we show that the binding affinity of the actin cross-linker α-actinin 4 (ACTN4) in cells modulates cytoplasmic mobility, cellular movement, and traction forces. Using fluorescence recovery after photobleaching, we show that an ACTN4 mutation that causes human kidney disease roughly triples the wild-type binding affinity of ACTN4 to F-actin in cells, increasing the dissociation time from 29 ± 13 to 86 ± 29 s. This increased affinity creates a less dynamic cytoplasm, as demonstrated by reduced intracellular microsphere movement, and an approximate halving of cell speed. Surprisingly, these less motile cells generate larger forces. Using traction force microscopy, we show that increased binding affinity of ACTN4 increases the average contractile stress (from 1.8 ± 0.7 to 4.7 ± 0.5 kPa), and the average strain energy (0.4 ± 0.2 to 2.1 ± 0.4 pJ). We speculate that these changes may be explained by an increased solid-like nature of the cytoskeleton, where myosin activity is more partitioned into tension and less is dissipated through filament sliding. These findings demonstrate the impact of cross-linker point mutations on cell dynamics and forces, and suggest mechanisms by which such physical defects lead to human disease.

Keywords: Alpha-actinin; actin; cell mechanics; kidney disease; traction force.

Fig. 1.

ACTN4 binding face is altered in the K255E mutation. (A) Illustration of the binding interaction between WT (Upper) and K255E (Lower) ACTN4 with actin filaments. Principle actin binding domains are depicted in blue, with the third putative site in yellow. The K255E mutation removes the bridge (green top) allowing the third putative actin binding domain (yellow) to be exposed, which increases the cross-linker affinity for actin. (B) Representative FRAP experiment of photobleaching a local region of GFP-ACTN4. (Scale bar: 20 µm.) Cells are transfected to express GFP-conjugated WT or K255E ACTN4, and a 1-μm spot is photobleached within 1 s. The time-dependent recovery is monitored (Supporting Information). (C) Representative FRAP data collected from WT (blue) and K255E (red) GFP-conjugated ACTN4 constructs. Time-dependent fluorescence recovery of GFP-conjugated WT or K255E ACTN4 are monitored and then fit with a single exponential, I(t) = C − A * exp(−t/τ). The time constant (τ) of the exponential quantifies the average time required for the GFP-ACTN4 to dissociate from actin. (D) Histogram of time-constant exponents (τ) from fitting FRAP data. WT exponents are measured to be 29 ± 13 s, and K255E are 86 ± 29 s. (E) Histogram of immobile fraction, C, from fitting FRAP data. WT ACTN4 are measured to have a lower immobile fraction (0.17 ± 0.06), and K255E display a larger immobile fraction (0.53 ± 0.14). Values are quoted in mean ± SD.

Fig. 2.

The K255E mutation slows intracellular movement. Endocytosed 100-nm fluorescent particles are tracked in the fibroblast cell lines using confocal microscopy. The trajectories reveal the mobility of objects within the cytoplasm. MSD of the tracked particles is plotted as a function of time, demonstrating that particles in the WT (hollow blue square, solid squares are mean) cells are significantly more mobile than those in the K255E (hollow red triangles, solid triangles are mean) cells. At short times less than 0.1 s, beads in both WT and K255E cells display similar mobility within the cytoplasm; at longer timescales this bead mobility diverges; beads within WT cells begin to appear diffusive-like and approach a slope of 0.82 after 0.69 ± 0.09 s (blue circle). This transition to diffusive-like movement of beads in the K255E cells takes approximately three times longer; only after 2.21 ± 0.45 s (red circle) do beads in K255E cells appear diffusive-like and approach a slope of ∼0.48. The dashed black line provides a slope of unity as a guide. This delayed transition from local fluctuations to diffusive-like mobility in the K255E cytoplasm is consistent with previous rheology of reconstituted ACTN4-actin systems that has shown that a viscoelastic transition from solid-like to fluid-like behavior is delayed by the increased binding affinity of K255E.

Fig. 3.

Cells expressing K255E ACTN4 are more spread, slower, exert larger forces, and do more work than WT cells. (A) Representative traction force maps from WT (Left) and K255E (Right) cells calculated from 26-kPa polyacrylamide substrates. (B) Mean spread area of WT (blue) and K255E (red) cells. We find that, on average, K255E cells spread to cover roughly twice as large an area as WT (WT ∼3,090 ± 981 μm²; K255E: 5,860 ± 326 μm²; mean ± SD). (C) Mean speeds of WT (blue) and K255E (red) cells. We find the speed of the K255E cells is significantly less than WT cells; over 24 h, WT cells move on average 1.16 ± 0.15 μm/h, and K255E cells move 0.55 ± 0.09 μm/h (mean ± SD). (D) Mean rms traction stresses of WT (blue) and K255E (red) cells. We measure that WT ACTN4 cells exerted traction stresses of 1.8 ± 0.7 kPa, whereas K255E ACTN4 cells exert 4.7 ± 0.5 kPa. (E) Mean strain energy of WT (blue) and K255E (red). The strain energy measures the total elastic energy stored in the substrate, and is a measure of work done by the cell on the substrate. We measure that WT ACTN4 cells exert a strain energy of 0.4 ± 0.23 pJ, whereas K255E ACTN4 cells exert 2.1 ± 0.4 pJ (values quoted in mean ± SD).

Fig. 4.

K255E ACTN4 cells exert larger forces for longer durations. (A) Histogram of local traction stresses in WT (blue) and K255E (red). (Inset) Difference between distributions reveals that WT cells have a larger number of traction forces at 1.2 kPa, but that K255E are more pronounced from 1.4 to 2.2 kPa. (B) Histogram of traction stress durations in WT (blue) and K255E (red). (Inset) Difference between distributions reveals that WT cells traction stresses are briefer than K255E, and that K255E stresses are significantly more likely to last longer than 5 h. (C) 2D histogram of persistence of traction vs. traction force in K255E-expressing cells. The main locus of persistent forces is found at ∼50 min and 1 kPa, yet the distribution extends beyond 2 kPa and 15 h. (D) 2D histogram of persistence of traction vs. traction force in WT expressing cells. The main locus of persistent forces is found at ∼50 min and 1 kPa, and the distribution extends to ∼1.7 kPa and 10 h.