DeActs: genetically encoded tools for perturbing the actin cytoskeleton in single cells

Abstract

The actin cytoskeleton is essential for many fundamental biological processes, but tools for directly manipulating actin dynamics are limited to cell-permeable drugs that preclude single-cell perturbations. Here we describe DeActs, genetically encoded actin-modifying polypeptides, which effectively induce actin disassembly in eukaryotic cells. We demonstrate that DeActs are universal tools for studying the actin cytoskeleton in single cells in culture, tissues, and multicellular organisms including various neurodevelopmental model systems.

Figure 1. Construction and characterization of DeActs

(a) Schematic of DeAct constructs and mechanism of action. (b) Expression of DeActs in rat embryonic fibroblasts, visualized with GFP and Alexa Fluor 594-phalloidin. CellMask Blue was used to reveal full cell morphology upon DeAct-SpvB expression. Scale, 20 μm. Representative micrographs from N = 4 independent experiments. (c–e) Quantification of DeAct effect on actin in rat embryonic fibroblasts showing (c) average phalloidin intensities (N = 7, 7, 3, 3 independent experiments, left to right), (d) phalloidin intensity relative to the DeAct-SpvB level (n = 207 GFP or 240 DeAct-SpvB cells), and (e) percent of cells with abnormal actin filament distribution (N = 3 independent experiments; see Supplementary Methods). (f) Velocities of single cell motility of rat embryonic fibroblasts transfected with DeActs or GFP control (see also Supplementary Video 2). n = 7, 8, 4, or 5 expressing cells per condition, left to right. (g) Oligodendrocyte-specific expression of GFP or DeAct-GS1 in primary rat oligodendrocytes using the OL-specific myelin basic protein promoter, quantifying phalloidin intensity relative to the DeAct-GS1 level (fluorescence intensities not directly comparable to panel d). n = 55 GFP or 103 DeAct-GS1 cells. Graphs show mean ± SEM. Trend lines in d and g show nonlinear (exponential) fit; each data point is one cell. Statistical significance: one-way ANOVA followed by Dunnett’s multiple comparison test, ****p<0.0001, ***p<0.001, *p<0.05.

Figure 2. DeActs markedly affect growth cones and neuronal migration

(a) Growth cones of DIV4 rat embryonic hippocampal neurons expressing MARCKS-eGFP +/− latrunculin B (LatB) or DeAct-GS1 or DeAct-SpvB and stained with Alexa Fluor 568-phalloidin to visualize actin filaments. Scale, 10 μm. (b) Overview of actin filament staining (magenta) in neuronal cultures expressing MARCKS-eGFP and treated with LatB (affects the whole culture), or expressing DeAct-GS1 (cell-specific). Scale, 40 μm. (c) Quantification of actin filament intensity in growth cones normalized to untransfected neighboring neurons; LatB affects all growth cones, therefore cannot be included (N = 4, 2, 2 independent experiments, n = 44, 19, 22 neurons, left to right). (d) Quantification of growth cone morphology upon MARCKS-eGFP or DeAct expression compared to LatB (N = 4, 2, 2, 2 independent experiments, n = 44, 23, 19, 22 neurons, left to right). (e–f) Dynamics of axonal branches are lost after addition of LatB or in neurons expressing DeAct constructs. A representative example is given in (f) from a 5 minute time lapse acquisition. Closed arrow heads indicate growth and open arrow heads represent no growth or retraction (see also Supplementary Video 3). Scale, 10 μm. (N = 4, 2, 2, 2 independent experiments, n = 37, 22, 17, 14 neurons, left to right). (g–j) Cortical neuronal migration after in utero electroporation with GFP, DeAct-GS1, or DeAct-SpvB. (g and i) transfected neurons (GFP signal). (h) Same slice as (g) combined with immunostaining against Satb2 (cortical layer II–IV), Ctip2 (cortical layer IV–V) and Neurofilament (axons in the IZ). CP, Cortical plate; IZ, intermediate zone; SVZ, subventricular zone; VZ, ventricular zone. Scale, g–h 100 μm, j 50 μm. (i) Quantification of cortical neuronal migration described in (g). (N = 3 embryos from 3 different litters, n = 3412, 2143, 847 cells GFP, DeAct-GS1 or DeAct-SpvB). (k) Quantification of neuronal morphologies of GFP-positive neurons which migrated past the IZ. 1, leading and trailing process; 2, leading process only; 3, no processes; 4, trailing process only (N = 3 embryos from 3 different litters, n = 104, 135, 38 cells GFP, DeAct-GS1, or DeAct-SpvB). Graphs show mean ± SEM. Statistical significance: one-way ANOVA and Dunnett’s multiple comparison post hoc test, and Wilcoxon test for paired data, *** p<0.001, **p<0.01, *p<0.05.

Figure 3. DeActs efficiently inhibit PVD neuron development in C. elegans

(a) Schematic representation of the C. elegans highly branched PVD neuron. (b) Co-expression of the moesinABD actin marker with GFP (control) or DeAct-GS1 in the PVD neuron. Other neurons expressing the actin marker are marked by *. Scale, 20 μm. (c) Representative images and schematic representation of the PVD neuron morphology upon cell specific DeAct expression. Scale, 50 μm. (d–g) Quantification of the DeAct-induced branching defects (d) and primary neurite outgrowth defects (c–d). Controls are siblings which lost the DeAct constructs. N = 34 for controls and N = 32 for DeAct animals (for DeAct-GS1 64 animals were analyzed and split in low and high DeAct expressing animals). Graphs show mean ± SEM. Statistical significance: one-way ANOVA and Dunnett’s multiple comparison post hoc test. Micrographs are representative, *** p<0.001.