Interactions of ADF/cofilin, Arp2/3 complex, capping protein and profilin in remodeling of branched actin filament networks

BACKGROUND: Cellular movements are powered by the assembly and disassembly of actin filaments. Actin dynamics are controlled by Arp2/3 complex, the Wiskott-Aldrich syndrome protein (WASp) and the related Scar protein, capping protein, profilin, and the actin-depolymerizing factor (ADF, also known as cofilin). Recently, using an assay that both reveals the kinetics of overall reactions and allows visualization of actin filaments, we showed how these proteins co-operate in the assembly of branched actin filament networks. Here, we investigated how they work together to disassemble the networks. RESULTS: Actin filament branches formed by polymerization of ATP-actin in the presence of activated Arp2/3 complex were found to be metastable, dissociating from the mother filament with a half time of 500 seconds. The ADF/cofilin protein actophorin reduced the half time for both dissociation of gamma-phosphate from ADP-Pi-actin filaments and debranching to 30 seconds. Branches were stabilized by phalloidin, which inhibits phosphate dissociation from ADP-Pi-filaments, and by BeF3, which forms a stable complex with ADP and actin. Arp2/3 complex capped pointed ends of ATP-actin filaments with higher affinity (Kd approximately 40 nM) than those of ADP-actin filaments (Kd approximately 1 microM), explaining why phosphate dissociation from ADP-Pi-filaments liberates branches. Capping protein prevented annealing of short filaments after debranching and, with profilin, allowed filaments to depolymerize at the pointed ends. CONCLUSIONS: The low affinity of Arp2/3 complex for the pointed ends of ADP-actin makes actin filament branches transient. By accelerating phosphate dissociation, ADF/cofilin promotes debranching. Barbed-end capping proteins and profilin allow dissociated branches to depolymerize from their free pointed ends.     

Arp2/3 complex and actin depolymerizing factor/cofilin in dendritic organization and treadmilling of actin filament array in lamellipodia.

The leading edge (approximately 1 microgram) of lamellipodia in Xenopus laevis keratocytes and fibroblasts was shown to have an extensively branched organization of actin filaments, which we term the dendritic brush. Pointed ends of individual filaments were located at Y-junctions, where the Arp2/3 complex was also localized, suggesting a role of the Arp2/3 complex in branch formation. Differential depolymerization experiments suggested that the Arp2/3 complex also provided protection of pointed ends from depolymerization. Actin depolymerizing factor (ADF)/cofilin was excluded from the distal 0.4 micrometer++ of the lamellipodial network of keratocytes and in fibroblasts it was located within the depolymerization-resistant zone. These results suggest that ADF/cofilin, per se, is not sufficient for actin brush depolymerization and a regulatory step is required. Our evidence supports a dendritic nucleation model (Mullins, R.D., J.A. Heuser, and T.D. Pollard. 1998. Proc. Natl. Acad. Sci. USA. 95:6181-6186) for lamellipodial protrusion, which involves treadmilling of a branched actin array instead of treadmilling of individual filaments. In this model, Arp2/3 complex and ADF/cofilin have antagonistic activities. Arp2/3 complex is responsible for integration of nascent actin filaments into the actin network at the cell front and stabilizing pointed ends from depolymerization, while ADF/cofilin promotes filament disassembly at the rear of the brush, presumably by pointed end depolymerization after dissociation of the Arp2/3 complex.     

Solution structure and dynamics of glia maturation factor from Caenorhabditis elegans

BackgroundThe GMF class of the ADF-H domain family proteins regulate actin dynamics by binding to the Arp2/3 complex and F-actin through their Site-1 and Site-2, respectively. CeGMF of C. elegans is analogous to GMFγ of human and mouse and is 138 amino acids in length.MethodsWe have characterized the solution structure and dynamics of CeGMF by solution NMR spectroscopy and its thermal stability by DSC.ResultsThe solution structure of CeGMF shows canonical ADF-H fold with two additional β-strands in the β4-β5 loop region. The Site-1 of CeGMF is well formed and residues of all three regions of Site-1 show dynamic flexibility. However, the β4-β5 loop of Site-2 is less inclined towards the C-terminal, as the latter is truncated by four residues in comparison to GMF isoforms of human and mouse. Regions of Site-2 show motions on ns-ps timescale, but dynamic flexibility of β4-β5 loop is low in comparison to corresponding F-loop region of ADF/cofilin UNC-60B. A general difference in packing of α3 and α1 between GMF and ADF/cofilins was noticed. Additionally, thermal stability of CeGMF was significantly higher than its ADF/cofilin homologs.ConclusionWe have presented the first solution structure of GMF from C. elegans, which highlights the structural differences between the Site-2 of CeGMF and mammalian GMF isoforms. Further, we have seen the differences in structure, dynamics, and thermal stability of GMF and ADF/cofilin.General significanceThis study provides a useful insight to structural and dynamics factors that define the specificity of GMF towards Arp2/3 complex.     

Cellular motility driven by assembly and disassembly of actin filaments

Motile cells extend a leading edge by assembling a branched network of actin filaments that produces physical force as the polymers grow beneath the plasma membrane. A core set of proteins including actin, Arp2/3 complex, profilin, capping protein, and ADF/cofilin can reconstitute the process in vitro, and mathematical models of the constituent reactions predict the rate of motion. Signaling pathways converging on WASp/Scar proteins regulate the activity of Arp2/3 complex, which mediates the initiation of new filaments as branches on preexisting filaments. After a brief spurt of growth, capping protein terminates the elongation of the filaments. After filaments have aged by hydrolysis of their bound ATP and dissociation of the gamma phosphate, ADF/cofilin proteins promote debranching and depolymerization. Profilin catalyzes the exchange of ADP for ATP, refilling the pool of ATP-actin monomers bound to profilin, ready for elongation.     

Arp2/3 complex interactions and actin network turnover in lamellipodia

Cell migration is initiated by lamellipodia-membrane-enclosed sheets of cytoplasm containing densely packed actin filament networks. Although the molecular details of network turnover remain obscure, recent work points towards key roles in filament nucleation for Arp2/3 complex and its activator WAVE complex. Here, we combine fluorescence recovery after photobleaching (FRAP) of different lamellipodial components with a new method of data analysis to shed light on the dynamics of actin assembly/disassembly. We show that Arp2/3 complex is incorporated into the network exclusively at the lamellipodium tip, like actin, at sites coincident with WAVE complex accumulation. Capping protein likewise showed a turnover similar to actin and Arp2/3 complex, but was confined to the tip. In contrast, cortactin-another prominent Arp2/3 complex regulator-and ADF/cofilin-previously implicated in driving both filament nucleation and disassembly-were rapidly exchanged throughout the lamellipodium. These results suggest that Arp2/3- and WAVE complex-driven actin filament nucleation at the lamellipodium tip is uncoupled from the activities of both cortactin and cofilin. Network turnover is additionally regulated by the spatially segregated activities of capping protein at the tip and cofilin throughout the mesh.     

Turnover of branched actin filament networks by stochastic fragmentation with ADF/cofilin

Cell motility depends on the rapid assembly, aging, severing, and disassembly of actin filaments in spatially distinct zones. How a set of actin regulatory proteins that sustains actin-based force generation during motility work together in space and time remains poorly understood. We present our study of the distribution and dynamics of Arp2/3 complex, capping protein (CP), and actin-depolymerizing factor (ADF)/cofilin in actin "comet tails," using a minimal reconstituted system with nucleation-promoting factor (NPF)-coated beads. The Arp2/3 complex concentrates at nucleation sites near the beads as well as in the first actin shell. CP colocalizes with actin and is homogeneously distributed throughout the comet tail; it serves to constrain the spatial distribution of ATP/ADP-P(i) filament zones to areas near the bead. The association of ADF/cofilin with the actin network is therefore governed by kinetics of actin assembly, actin nucleotide state, and CP binding. A kinetic simulation accurately validates these observations. Following its binding to the actin networks, ADF/cofilin is able to break up the dense actin filament array of a comet tail. Stochastic severing by ADF/cofilin loosens the tight entanglement of actin filaments inside the comet tail and facilitates turnover through the macroscopic release of large portions of the aged actin network.     

GMFβ controls branched actin content and lamellipodial retraction in fibroblasts

The lamellipodium is an important structure for cell migration containing branched actin nucleated via the Arp2/3 complex. The formation of branched actin is relatively well studied, but less is known about its disassembly and how this influences migration. GMF is implicated in both Arp2/3 debranching and inhibition of Arp2/3 activation. Modulation of GMFβ, a ubiquitous GMF isoform, by depletion or overexpression resulted in changes in lamellipodial dynamics, branched actin content, and migration. Acute pharmacological inhibition of Arp2/3 by CK-666, coupled to quantitative live-cell imaging of the complex, showed that depletion of GMFβ decreased the rate of branched actin disassembly. These data, along with mutagenesis studies, suggest that debranching (not inhibition of Arp2/3 activation) is a primary activity of GMFβ in vivo. Furthermore, depletion or overexpression of GMFβ disrupted the ability of cells to directionally migrate to a gradient of fibronectin (haptotaxis). These data suggest that debranching by GMFβ plays an important role in branched actin regulation, lamellipodial dynamics, and directional migration.     

Activation of the Arp2/3 complex by the Listeria acta protein. Acta binds two actin monomers and three subunits of the Arp2/3 complex

ActA is a bacterially encoded protein that enables Listeria monocytogenes to hijack the host cell actin cytoskeleton. It promotes Arp2/3-dependent actin nucleation, but its interactions with cellular components of the nucleation machinery are not well understood. Here we show that two domains of ActA (residues 85-104 and 121-138) with sequence similarity to WASP homology 2 domains bind two actin monomers with submicromolar affinity. ActA binds Arp2/3 with a K(d) of 0.6 microm and competes for binding with the WASP family proteins N-WASP and Scar1. By chemical cross-linking, ActA, N-WASP, and Scar1 contact the same three subunits of the Arp2/3 complex, p40, Arp2, and Arp3. Interestingly, profilin competes with ActA for binding of Arp2/3, but actophorin (cofilin) does not. The minimal Arp2/3-binding site of ActA (residues 144-170) is C-terminal to both actin-binding sites and shares sequence homology with Arp2/3-binding regions of WASP family proteins. The maximal activity at saturating concentrations of ActA is identical to the most active domains of the WASP family proteins. We propose that ActA and endogenous WASP family proteins promote Arp2/3-dependent nucleation by similar mechanisms and require simultaneous binding of Arp2 and Arp3.     

Cofilin produces newly polymerized actin filaments that are preferred for dendritic nucleation by the Arp2/3 complex.

One of the earliest events in the process of cell motility is the massive generation of free actin barbed ends, which elongate to form filaments adjacent to the plasma membrane at the tip of the leading edge. Both cofilin and Arp2/3 complex have been proposed to contribute to barbed end formation during cell motility. Attempts to assess the functions of cofilin and Arp 2/3 complex in vivo indicate that both cofilin and Arp2/3 complex contribute to actin polymerization: cofilin by severing and Arp2/3 by nucleating and branching. In order to determine if the activities of cofilin and Arp2/3 complex interact, we employed a light microscope-based assay to visualize actin polymerization directly in the presence of both proteins. The results indicate that cofilin generates barbed ends to increase the mass of freshly polymerized F-actin but does not directly affect the activity of Arp2/3 complex. However, while ADP, ADP-Pi, and newly polymerized ATP-filaments are all capable of supporting Arp2/3-mediated branching, newly polymerized F-actin supports most of the Arp2/3-induced branch formation. The results suggest that, in vivo, cofilin contributes to barbed end formation by inducing the initial increase in the number of barbed ends leading to increased ATP-F-actin, which in turn supports higher levels of dendritic nucleation by active Arp2/3 complex.     

NMR solution structures of actin depolymerizing factor homology domains

Actin is one of the most conserved proteins in nature. Its assembly and disassembly are regulated by many proteins, including the family of actin‐depolymerizing factor homology (ADF‐H) domains. ADF‐H domains can be divided into five classes: ADF/cofilin, glia maturation factor (GMF), coactosin, twinfilin, and Abp1/drebrin. The best‐characterized class is ADF/cofilin. The other four classes have drawn much less attention and very few structures have been reported. This study presents the solution NMR structure of the ADF‐H domain of human HIP‐55‐drebrin‐like protein, the first published structure of a drebrin‐like domain (mammalian), and the first published structure of GMF β (mouse). We also determined the structures of mouse GMF γ, the mouse coactosin‐like domain and the C‐terminal ADF‐H domain of mouse twinfilin 1. Although the overall fold of the five domains is similar, some significant differences provide valuable insights into filamentous actin (F‐actin) and globular actin (G‐actin) binding, including the identification of binding residues on the long central helix. This long helix is stabilized by three or four residues. Notably, the F‐actin binding sites of mouse GMF β and GMF γ contain two additional β‐strands not seen in other ADF‐H structures. The G‐actin binding site of the ADF‐H domain of human HIP‐55‐drebrin‐like protein is absent and distorted in mouse GMF β and GMF γ.     

Control of actin assembly and disassembly at filament ends

The most important discovery in the field is that the Arp2/3 complex nucleates assembly of actin filaments with free barbed ends. Arp2/3 also binds the sides of actin filaments to create a branched network. Arp2/3's nucleation activity is stimulated by WASP family proteins, some of which mediate signaling from small G-proteins. Listeria movement caused by actin polymerization can be reconstituted in vitro using purified proteins: Arp2/3 complex, capping protein, actin depolymerizing factor/cofilin, and actin. actin depolymerizing factor/cofilin increases the rate at which actin subunits leave pointed ends, and capping protein caps barbed ends.     

Stochastic severing of actin filaments by actin depolymerizing factor/cofilin controls the emergence of a steady dynamical regime

Actin dynamics (i.e., polymerization/depolymerization) powers a large number of cellular processes. However, a great deal remains to be learned to explain the rapid actin filament turnover observed in vivo. Here, we developed a minimal kinetic model that describes key details of actin filament dynamics in the presence of actin depolymerizing factor (ADF)/cofilin. We limited the molecular mechanism to 1), the spontaneous growth of filaments by polymerization of actin monomers, 2), the ageing of actin subunits in filaments, 3), the cooperative binding of ADF/cofilin to actin filament subunits, and 4), filament severing by ADF/cofilin. First, from numerical simulations and mathematical analysis, we found that the average filament length, 〈L〉, is controlled by the concentration of actin monomers (power law: 5/6) and ADF/cofilin (power law: −2/3). We also showed that the average subunit residence time inside the filament, 〈T〉, depends on the actin monomer (power law: −1/6) and ADF/cofilin (power law: −2/3) concentrations. In addition, filament length fluctuations are ∼20% of the average filament length. Moreover, ADF/cofilin fragmentation while modulating filament length keeps filaments in a high molar ratio of ATP- or ADP-P_i versus ADP-bound subunits. This latter property has a protective effect against a too high severing activity of ADF/cofilin. We propose that the activity of ADF/cofilin in vivo is under the control of an affinity gradient that builds up dynamically along growing actin filaments. Our analysis shows that ADF/cofilin regulation maintains actin filaments in a highly dynamical state compatible with the cytoskeleton dynamics observed in vivo.     

The Three Mouse Actin-depolymerizing Factor/Cofilins Evolved to Fulfill Cell-Type–specific Requirements for Actin Dynamics

Actin-depolymerizing factor (ADF)/cofilins are essential regulators of actin filament turnover. Several ADF/cofilin isoforms are found in multicellular organisms, but their biological differences have remained unclear. Herein, we show that three ADF/cofilins exist in mouse and most likely in all other mammalian species. Northern blot and in situ hybridization analyses demonstrate that cofilin-1 is expressed in most cell types of embryos and adult mice. Cofilin-2 is expressed in muscle cells and ADF is restricted to epithelia and endothelia. Although the three mouse ADF/cofilins do not show actin isoform specificity, they all depolymerize platelet actin filaments more efficiently than muscle actin. Furthermore, these ADF/cofilins are biochemically different. The epithelial-specific ADF is the most efficient in turning over actin filaments and promotes a stronger pH-dependent actin filament disassembly than the two other isoforms. The muscle-specific cofilin-2 has a weaker actin filament depolymerization activity and displays a 5-10-fold higher affinity for ATP-actin monomers than cofilin-1 and ADF. In steady-state assays, cofilin-2 also promotes filament assembly rather than disassembly. Taken together, these data suggest that the three biochemically distinct mammalian ADF/cofilin isoforms evolved to fulfill specific requirements for actin filament dynamics in different cell types.     

Actin-based motility of pathogens: the Arp2/3 complex is a central player

Bacterial actin-based motility has provided cell biologists with tools that led to the recent discovery that, in many forms of actin-based motilities, a key player is a protein complex named the Arp2/3 complex. The Arp2/3 complex is evolutionally conserved and made up of seven polypeptides involved in both actin filament nucleation and organization. Interestingly, this complex is inactive by itself and recent work has highlighted the fact that its activation is achieved differently in the different types of actin-based motilities, including the well-known examples of Listeria and Shigella motilities. Proteins of the WASP family and small G-proteins are involved in most cases. It is interesting that bacteria bypass or mimic some of the events occurring in eukaryotic systems. The Shigella protein IcsA recruits N-WASP and activates it in a Cdc42-like fashion. This activation leads to Arp2/3 complex recruitment, activation of the complex and ultimately actin polymerization and movement. The Listeria ActA protein activates Arp2/3 directly and, thus, seems to mimic proteins of the WASP family. A breakthrough in the field is the recent reconstitution of the actin-based motilities of Listeria and N-WASP-coated E. coli (IcsA) using a restricted number of purified cellular proteins including F-actin, the Arp2/3 complex, actin depolymerizing factor (ADF or cofilin) and capping protein. The movement was more effective upon addition of profilin, alpha-actinin and VASP (for Listeria). Bacterial actin-based motility is now one of the best-documented examples of the exploitation of mammalian cell machineries by bacterial pathogens.     

Profilin enhances Cdc42-induced nucleation of actin polymerization

We find that profilin contributes in several ways to Cdc42-induced nucleation of actin filaments in high speed supernatant of lysed neutrophils. Depletion of profilin inhibited Cdc42-induced nucleation; re-addition of profilin restored much of the activity. Mutant profilins with a decreased affinity for either actin or poly-l-proline were less effective at restoring activity. Whereas Cdc42 must activate Wiskott-Aldrich Syndrome protein (WASP) to stimulate nucleation by the Arp2/3 complex, VCA (verpolin homology, cofilin, and acidic domain contained in the COOH-terminal fragment of N-WASP) constitutively activates the Arp2/3 complex. Nucleation by VCA was not inhibited by profilin depletion. With purified N-WASP and Arp2/3 complex, Cdc42-induced nucleation did not require profilin but was enhanced by profilin, wild-type profilin being more effective than mutant profilin with reduced affinity for poly-l-proline.Nucleation by the Arp2/3 complex is a function of the free G-actin concentration. Thus, when profilin addition decreased the free G-actin concentration, it inhibited Cdc42- and VCA-induced nucleation. However, when profilin was added with G-actin in a ratio that maintained the initial free G-actin concentration, it increased the rate of both Cdc42- and VCA-induced nucleation. This enhancement, also seen with purified proteins, was greatest when the free G-actin concentration was low. These data suggest that under conditions present in intact cells, profilin enhances nucleation by activated Arp2/3 complex.     

Activation of the Arp2/3 complex by the actin filament binding protein Abp1p

The actin-related protein (Arp) 2/3 complex plays a central role in assembly of actin networks. Because distinct actin-based structures mediate diverse processes, many proteins are likely to make spatially and temporally regulated interactions with the Arp2/3 complex. We have isolated a new activator, Abp1p, which associates tightly with the yeast Arp2/3 complex. Abp1p contains two acidic sequences (DDW) similar to those found in SCAR/WASp proteins. We demonstrate that mutation of these sequences abolishes Arp2/3 complex activation in vitro. Genetic studies indicate that this activity is important for Abp1p functions in vivo. In contrast to SCAR/WASp proteins, Abp1p binds specifically to actin filaments, not monomers. Actin filament binding is mediated by the ADF/cofilin homology (ADF-H) domain of Abp1p and is required for Arp2/3 complex activation in vitro. We demonstrate that Abp1p recruits Arp2/3 complex to the sides of filaments, suggesting a novel mechanism of activation. Studies in yeast and mammalian cells indicate that Abp1p is involved functionally in endocytosis. Based on these results, we speculate that Abp1p may link Arp2/3-mediated actin assembly to a specific step in endocytosis.     

Three regions within ActA promote Arp2/3 complex-mediated actin nucleation and Listeria monocytogenes motility

The Listeria monocytogenes ActA protein induces actin-based motility by enhancing the actin nucleating activity of the host Arp2/3 complex. Using systematic truncation analysis, we identified a 136-residue NH(2)-terminal fragment that was fully active in stimulating nucleation in vitro. Further deletion analysis demonstrated that this fragment contains three regions, which are important for nucleation and share functional and/or limited sequence similarity with host WASP family proteins: an acidic stretch, an actin monomer-binding region, and a cofilin homology sequence. To determine the contribution of each region to actin-based motility, we compared the biochemical activities of ActA derivatives with the phenotypes of corresponding mutant bacteria in cells. The acidic stretch functions to increase the efficiency of actin nucleation, the rate and frequency of motility, and the effectiveness of cell-cell spread. The monomer-binding region is required for actin nucleation in vitro, but not for actin polymerization or motility in infected cells, suggesting that redundant mechanisms may exist to recruit monomer in host cytosol. The cofilin homology sequence is critical for stimulating actin nucleation with the Arp2/3 complex in vitro, and is essential for actin polymerization and motility in cells. These data demonstrate that each region contributes to actin-based motility, and that the cofilin homology sequence plays a principal role in activation of the Arp2/3 complex, and is an essential determinant of L. monocytogenes pathogenesis.     

Mapping the ADF/cofilin binding site on monomeric actin by competitive cross-linking and peptide array: evidence for a second binding site on monomeric actin

The binding sites for actin depolymerising factor (ADF) and cofilin on G-actin have been mapped by competitive chemical cross-linking using deoxyribonuclease I (DNase I), gelsolin segment 1 (G1), thymosin beta4 (Tbeta4), and vitamin D-binding protein (DbP). To reduce ADF/cofilin induced actin oligomerisation we used ADP-ribosylated actin. Both vitamin D-binding protein and thymosin beta4 inhibit binding by ADF or cofilin, while cofilin or ADF and DNase I bind simultaneously. Competition was observed between ADF or cofilin and G1, supporting the hypothesis that cofilin preferentially binds in the cleft between sub-domains 1 and 3, similar to or overlapping the binding site of G1. Because the affinity of G1 is much higher than that of ADF or cofilin, even at a 20-fold excess of the latter, the complexes contained predominantly G1. Nevertheless, cross-linking studies using actin:G1 complexes and ADF or cofilin showed the presence of low concentrations of ternary complexes containing both ADF or cofilin and G1. Thus, even with monomeric actin, it is shown for the first time that binding sites for both G1 and ADF or cofilin can be occupied simultaneously, confirming the existence of two separate binding sites. Employing a peptide array with overlapping sequences of actin overlaid by cofilin, we have identified five sequence stretches of actin able to bind cofilin. These sequences are located within the regions of F-actin predicted to bind cofilin in the model derived from image reconstructions of electron microscopical images of cofilin-decorated filaments. Three of the peptides map to the cleft region between sub-domains 1 and 3 of the upper actin along the two-start long-pitch helix, while the other two are in the DNase I loop corresponding to the site of the lower actin in the helix. In the absence of any crystal structures of ADF or cofilin in complex with actin, these studies provide further information about the binding sites on F-actin for these important actin regulatory proteins.     

How is actin polymerization nucleated in vivo?

Actin polymerization in vivo is dependent on free barbed ends that act as nuclei. Free barbed ends can arise in vivo by nucleation from the Arp2/3 complex, uncapping of barbed ends on pre-existing filaments or severing of filaments by cofilin. There is evidence that each mechanism operates in cells. However, different cell types use different combinations of these processes to generate barbed ends during stimulated cell motility. Here, I describe recent attempts to define the relative contributions of these three mechanisms to actin nucleation in vivo. The rapid increase in the number of barbed ends during stimulation is not due to any single mechanism. Cooperation between capping proteins, cofilin and the Arp2/3 complex is necessary for the development of protrusive force at the leading edge of the cell: uncapping and cofilin severing contributing barbed ends, whereas activity of the Arp2/3 complex is necessary, but not sufficient, for lamellipod extension. These results highlight the need for new methods that enable the direct observation of actin nucleation and so define precisely the relative contributions of the three processes to stimulated cell motility.     

Arp2/3 ATP hydrolysis-catalysed branch dissociation is critical for endocytic force generation

The Arp2/3 complex, which is crucial for actin-based motility, nucleates actin filaments and organizes them into y-branched networks. The Arp2 subunit has been shown to hydrolyse ATP, but the functional importance of Arp2/3 ATP hydrolysis is not known. Here, we analysed an Arp2 mutant in Saccharomyces cerevisiae that is defective in ATP hydrolysis. Arp2 ATP hydrolysis and Arp2/3-dependent actin nucleation occur almost simultaneously. However, ATP hydrolysis is not required for nucleation. In addition, Arp2 ATP hydrolysis is not required for the release of a WASP-like activator from y-branches. ATP hydrolysis by Arp2, and possibly Arp3, is essential for efficient y-branch dissociation in vitro. In living cells, both Arp2 and Arp3 ATP-hydrolysis mutants exhibit defects in endocytic internalization and actin-network disassembly. Our results suggest a critical feature of dendritic nucleation in which debranching and subsequent actin-filament remodelling and/or depolymerization are important for endocytic vesicle morphogenesis.