Analysis of actin folding, thermal stability and nucleotide exchange

(A) Homology model of human cytoskeletal γ–actin (based on PDB–ID: 2BTF). Subdomains (SDs) of actin are colored in grey (SD1), dark grey (SD2), light green (SD3) and green (SD4). The hinge region encompassing residues 137– 145 and 333–338 is colored in blue. The mutant residue E334 is shown as an orange sphere. (B) Inhibition of DNase–I activity by monomeric γ–actin WT and p.E334Q. Data is the mean of three individual experiments ± SEM. A Hill equation was fitted to the data, which yields the half–maximal inhibitory concentration (IC50, Table 1). (C) The protein denaturation temperature of monomeric and filamentous p.E334Q and WT actin were determined by DSF. Representative experimental traces are shown. The protein denaturation temperature is derived from the peak of the melting curve (Table 1) (p.E334Q (G-actin), N=3; WT (G-actin), N=5; p.E334Q (F–actin), N=5; WT (F–actin), N=5). (D) Nucleotide exchange rates (k-T, k-D) were determined for monomeric p.E334Q and γ–actin WT using fluorescently labeled ATP (ε–ATP) (p.E334Q, N=18; WT, N=21) and ADP (ε–ADP) (p.E334Q, N=11; WT, N=11) Representative experimental traces are shown. Rates were determined by fitting a single exponential function to the data (Table 1).

Comparison of biochemical parameters, velocities, rate and equilibrium constants measured in experiments with human cytoskeletal γ–actin and variant p.E334Q.

Analysis of the polymerization capacity of WT and p.E334Q γ–actin in the absence and presence of human Arp2/3 complex.

(A) Polymerization of p.E334Q and WT γ–actin (1 µM, 10% Atto– 655 labeled) was induced by salt–shift and the progression of the reaction was tracked by TIRF microscopy. Shown are representative micrographs at the indicated points in time. Scale bar corresponds to 10 µm. (B) The elongation rates of individual filaments were determined by manual tracking of the elongating barbed ends of the filaments. Every data point represents an individual filament (WT: n = 128 filaments, p.E334Q: n = 114 filaments). Data is shown as the mean ± SEM from 4 individual experiments. (C) Nucleation efficiencies were determined by monitoring changes in the number of filaments over the time of the polymerization experiments. The solid lines and shades represent the mean ± SEM of four individual experiments. (D) TIRF microscopy–based observation of human Arp2/3 complex mediated differences in the salt-induced polymerization of p.E334Q and WT γ–actin (1 µM, 10% Atto–655 labeled). Shown are representative micrographs at the indicated points in time. For both WT and p.E334Q γ–actin, the formation of the first branch points was observed after 250 s in the presence of 0.01 nM Arp2/3. Scale bar corresponds to 10 µm. N=3 for each Arp2/3 concentration (E) Pyrene–based polymerization experiment of 2 µM γ–actin WT, p.E334Q actin, or a 1:1 mixture of both (5% pyrene–labeled) in the absence and presence of Arp2/3. Shown are representative traces. The solid lines/shades represent the mean ± SEM of at least three individual experiments. (F) Bulk–polymerization rates determined from the experiments shown in (E). Every data point corresponds to an individual polymerization experiment, as shown in (E). Data is shown as the mean ± SEM. Individual values and differences with statistical significance are summarized in Table 1.

In silico analysis of the effects of the E334Q mutation on the actin–cofilin interaction.

(A) Human homology model of a cytoskeletal γ–actin filament partially decorated with cofilin–1 (model based on PDB–ID:6UC4). Cofilin–1 is colored magenta, actin monomers are depicted in light blue and the actin residue E334 is colored orange. (B) Close up showing the details of the Van der Waals interaction between E334 and cofilin residue L111 on the pointed end side. The length of the dotted line between the side chains of actin-E334 and cofilin-L111 corresponds to 3.1 Å in our model. (C) The change in surface charge at the actin-cofilin interaction interface introduced by mutation E334Q is visualized by coulombic surface coloring.

Analysis of the effect of cofilin–1 on the spontaneous polymerization of WT and p.E334Q γ–actin.

(A) TIRF microscopy–based observation of cofilin–1 mediated differences in the salt-induced polymerization of p.E334Q and WT γ–actin (1 µM, 10% Atto–655 labeled). Shown are representative micrographs 450 s after induction of polymerization. Scale bar corresponds to 10 µm. (B) Time dependence of the increase in filaments observed by TIRF microscopy. The solid and dotted lines and shades represent the mean ± SEM of 3 individual experiments.

Analysis of the interaction of filamentous WT and p.E334Q γ–actin with cofilin–1.

(A) Pyrene-based dilution–induced depolymerization experiments (5 % pyrene–labeled actin) were used to quantify the impact of cofilin–1 on the disassembly of filaments formed by p.E334Q, WT and a 1:1 mixture of p.E334Q and WT γ–actin. Depolymerization was induced by adding G–buffer (pH 7.8). Every data point represents an individual experiment. Data is shown as the mean ± SEM. (B, C) In addition, cofilin-1 mediated filament disassembly at different pH values (pH 6.5, pH 7.8) was investigated in co– sedimentation experiments. Shown are the means ± SEM of at least three individual experiments. (D) The dissociation equilibrium constant KD for binding of cofilin-1 to pyrene-labeled WT or p.E334Q filaments was determined by monitoring the change in fluorescence amplitude upon cofilin-1 binding. Shown are the means ± SEM of at least three individual experiments.

Analysis of the interaction of WT and p.E334Q γ–actin with eGFP–cofilin–1

(A) The disassembly of 300 nM “aged” WT and p.E334Q F–actin (10% Atto–655 labeled, capped by CP) in the presence of 100 nM eGFP–cofilin–1 was visualized using two–color TIRF microscopy. F–actin is shown in magenta, eGFP–cofilin in cyan. Shown are representative micrographs at the indicated time points. Scale bar corresponds to 10 µm. (B) Time dependence of the change in the number of eGFP–cofilin–1 clusters on actin filaments. The solid lines and shades represent the mean ± SEM of three individual experiments. (C) Time dependence of the decrease in filament count. The solid lines and shades represent the mean ± SEM of three individual experiments. (D) Temporal tracking of actin filament disassembly by monitoring changes in total actin fluorescence per image frame. The solid lines and shades represent the mean ± SEM of three individual experiments. (E) Visualization of temporal changes in the abundance of eGFP–cofilin–1 clusters based on monitoring total eGFP–cofilin–1 fluorescence per image frame. The solid lines and shades represent the mean ± SEM of three individual experiments.

In silico analysis of the effects of the E334Q mutation on the actin–myosin interaction.

(A) Contact region between the CM–loop of NM2C (blue) and residue E334 (orange) in cytoskeletal γ–actin (PDB–ID: 5JLH) (Von Der Ecken et al, 2016). Actin interacts with a core contact triad in the myosin CM-loop, which in the case of NM2C consists of lysine K429 and hydrophobic residues I420 and V427. (B) Multiple sequence alignment of the CM-loop of selected cytoskeletal myosin isoforms. The alignment shows complete sequence identity for NM2C, NM2A and conservation of residues contributing to the core contact triad in class 5, class 19 and selected members of class 1 myosins (blue boxes). Residue numbering refers to the NM2C primary structure. (C) Mutation E334Q leads to a weakening of complementary electrostatic interactions. The interaction interface involving the CM-loop of NM2C and the region affected by the mutation on cytoskeletal γ–actin is visualized by coulombic surface coloring. Positive potentials are shown in blue, and negative potentials in red.

Analysis of the interaction of human NM2A with WT and p.E334Q γ–actin.

(A) The interaction of surface–immobilized NM2A–HMM with WT γ–actin filaments, p.E334Q γ–actin filaments and heterofilaments (1:1 mixture) was analyzed using the unloaded in vitro motility assay. Representative velocity distributions are obtained from recorded trajectories of WT and mutant filaments in a single experiment. The average sliding velocity of the filaments in each experiment was determined by applying a Gaussian fit (black line) to the obtained velocity distributions. (B) Secondary plot of all measured sliding velocities. Each data point represents a single experiment (N=20 for WT filaments, N=18 for p.E334Q filaments, N=8 for heterofilaments) in which the sliding velocities of a minimum of 600 filaments have been analyzed. The average sliding velocity in each experiment was determined as described in (A). (C, D) Determination of the dissociation rate constant (k-A) for the interaction of NM2A–2R (C) or NM2C-2R (D) with p.E334Q or WT γ–actin in the absence of nucleotide. k-A was determined in displacement experiments where unlabeled α–skeletal actin was added to pyrene– labeled acto•NM2A–2R/NM2C-2R. Shown are representative traces. The shown k-A values are the results of 5–11 individual experiments using single–exponential fit functions. (E, F) The affinity of WT and p.E334Q γ–actin for NM2A–2R in the absence (KA) and in the presence of saturating ADP concentrations (KDA) was determined using the method developed by Kurzawa and Geeves (Kurzawa & Geeves, 1996). Increasing concentrations of myosin were dissociated from pyrene–labeled WT or p.E334Q F–actin by mixing the complex with excess ATP. The obtained fluorescence amplitudes of the individual transients were plotted against the corresponding myosin concentrations. A quadratic equation (see Material and Methods) was fitted to the data, which yields the respective affinity. Data is shown as the mean ± SD of three individual experiments.

Analysis of the interaction of human Myo5A-HMM with WT and p.E334Q γ–actin in the unloaded in vitro motility assay at different Myo5A surface densities.

(A) The interaction of surface– immobilized Myo5A–HMM with WT γ–actin and p.E334Q γ–actin filaments was analyzed using the unloaded in vitro motility assay. Representative velocity distributions were obtained from one individual experiment at the optimal Myo5A-HMM surface density for WT (770 µm−2) and p.E334Q filaments (3100 µm−2). The average sliding velocity of the filaments in each experiment was determined by applying a Gaussian fit (black line) to the obtained velocity distributions. (B) Semi–logarithmic plot summarizing all measured sliding velocities at different Myo5A–HMM surface densities. Each data point is the result of 3–10 individual experiments. The dependence of the average sliding velocity on surface motor density is best–described in the semi–logarithmic graph by lines up to approximately 1000 motors µm-2 for WT filaments and 7000 motors µm-2 for p.E334Q filaments.

CD–Spectra of γ–actin WT and p.E334Q protein. Shown are the smoothed (Savitsky–Golay) and averaged spectra with the averaged buffer spectrum subtracted. The depicted secondary structure composition of γ–actin WT and p.E334Q protein was determined by using the DichroWeb online platform.

Normal mode analysis of changes induced by mutation E334Q in the global mobility of bare, partially and a fully cofilin-1 decorated γ–actin mini-filaments

(A) Models of a bareγ–actin mini-filament, a partially and a fully cofilin-1 decorated γ–actin mini-filament were generated for normal mode analysis. Each mini–filament consists of nine actin monomers. Actin is colored light blue and cofilin–1 magenta. (B) Significant differences in the global mobility of the different types of mini–filaments shown in panel (A), containing either WT or mutant cytoskeletal γ–actin, were only observed in the case of partial cofilin-1 decoration. Large differences are observed for partially decorated filaments in modes 7, 8 and 9, which are the global motions with the lowest frequencies.

Exemplary traces of dilution–induced depolymerization experiments in the absence and presence of increasing human cofilin–1 concentrations.

(A) 300 nM “aged” WT and p.E334Q F–actin (10% Atto–655 labeled, capped by CP) was visualized using TIRF microscopy. Shown are representative micrographs at the indicated time points. Scale bar corresponds to 10 µm. (B) Quantification of the number of actin filaments over the time of the experiment. The solid lines and shades represent the mean ± SEM of 2–3 individual experiments.