The stimulatory action of amphiphysin on dynamin function is dependent on lipid bilayer curvature
2004; Springer Nature; Volume: 23; Issue: 17 Linguagem: Inglês
10.1038/sj.emboj.7600355
ISSN1460-2075
AutoresYumi Yoshida, Masahiro Kinuta, Tadashi Abe, Shuang Liang, Kenta Araki, Ottavio Cremona, Gilbert Di Paolo, Yoshinori Moriyama, Tatsuji Yasuda, Pietro De Camilli, Kohji Takei,
Tópico(s)Alzheimer's disease research and treatments
ResumoArticle19 August 2004free access The stimulatory action of amphiphysin on dynamin function is dependent on lipid bilayer curvature Yumi Yoshida Yumi Yoshida Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Department of Biochemistry, Faculty of Pharmaceutical Sciences, Okayama University, Okayama, Japan Search for more papers by this author Masahiro Kinuta Masahiro Kinuta Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Search for more papers by this author Tadashi Abe Tadashi Abe Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Search for more papers by this author Shuang Liang Shuang Liang Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Search for more papers by this author Kenta Araki Kenta Araki Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Department of Biochemistry, Faculty of Pharmaceutical Sciences, Okayama University, Okayama, Japan Search for more papers by this author Ottavio Cremona Ottavio Cremona DIBIT-Scientific Institute San Raffaele Universita' Vita – Salute San Raffaele, Milano, Italy Search for more papers by this author Gilbert Di Paolo Gilbert Di Paolo Department of Cell Biology and Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, CT, USA Search for more papers by this author Yoshinori Moriyama Yoshinori Moriyama Department of Biochemistry, Faculty of Pharmaceutical Sciences, Okayama University, Okayama, Japan Search for more papers by this author Tatsuji Yasuda Tatsuji Yasuda Department of Cell Chemistry, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Search for more papers by this author Pietro De Camilli Pietro De Camilli Department of Cell Biology and Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, CT, USA Search for more papers by this author Kohji Takei Corresponding Author Kohji Takei Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Search for more papers by this author Yumi Yoshida Yumi Yoshida Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Department of Biochemistry, Faculty of Pharmaceutical Sciences, Okayama University, Okayama, Japan Search for more papers by this author Masahiro Kinuta Masahiro Kinuta Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Search for more papers by this author Tadashi Abe Tadashi Abe Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Search for more papers by this author Shuang Liang Shuang Liang Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Search for more papers by this author Kenta Araki Kenta Araki Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Department of Biochemistry, Faculty of Pharmaceutical Sciences, Okayama University, Okayama, Japan Search for more papers by this author Ottavio Cremona Ottavio Cremona DIBIT-Scientific Institute San Raffaele Universita' Vita – Salute San Raffaele, Milano, Italy Search for more papers by this author Gilbert Di Paolo Gilbert Di Paolo Department of Cell Biology and Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, CT, USA Search for more papers by this author Yoshinori Moriyama Yoshinori Moriyama Department of Biochemistry, Faculty of Pharmaceutical Sciences, Okayama University, Okayama, Japan Search for more papers by this author Tatsuji Yasuda Tatsuji Yasuda Department of Cell Chemistry, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Search for more papers by this author Pietro De Camilli Pietro De Camilli Department of Cell Biology and Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, CT, USA Search for more papers by this author Kohji Takei Corresponding Author Kohji Takei Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan Search for more papers by this author Author Information Yumi Yoshida1,2,‡, Masahiro Kinuta1,‡, Tadashi Abe1, Shuang Liang1, Kenta Araki1,2, Ottavio Cremona3, Gilbert Di Paolo4, Yoshinori Moriyama2, Tatsuji Yasuda5, Pietro De Camilli4 and Kohji Takei 1 1Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan 2Department of Biochemistry, Faculty of Pharmaceutical Sciences, Okayama University, Okayama, Japan 3DIBIT-Scientific Institute San Raffaele Universita' Vita – Salute San Raffaele, Milano, Italy 4Department of Cell Biology and Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, CT, USA 5Department of Cell Chemistry, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan ‡These authors contributed equally to this paper *Corresponding author. Department of Neuroscience, Okayama University Graduate School of Medicine and Dentistry, 2-5-1 Shikata-cho, Okayama 700-8558, Japan. Tel.: +81 86 235 7120; Fax: +81 86 235 7126; E-mail: [email protected] The EMBO Journal (2004)23:3483-3491https://doi.org/10.1038/sj.emboj.7600355 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Amphiphysin is a major dynamin-binding partner at the synapse; however, its function in fission is unclear. Incubation of large unilamellar liposomes with mice brain cytosol led to massive formation of small vesicles, whereas cytosol of amphiphysin 1 knockout mice was much less efficient in this reaction. Vesicle formation from large liposomes by purified dynamin was also strongly enhanced by amphiphysin. In the presence of liposomes, amphiphysin strongly affected dynamin GTPase activity and the recruitment of dynamin to the liposomes, but this activity was highly dependent on liposome size. Deletion from amphiphysin of its central proline-rich stretch dramatically potentiated its effect on dynamin, possibly by relieving an inhibitory intramolecular interaction. These results suggest a model in which maturation of endocytic pits correlates with the oligomerization of dynamin with either amphiphysin or other proteins with similar domain structure. Formation of these complexes is coupled to the activation of dynamin GTPase activity, thus explaining how deep invagination of the pit leads to fission. Introduction Clathrin-mediated endocytosis initiates with the recruitment of clathrin coat components to the plasma membrane, which is followed by the invagination of the membrane to form a clathrin-coated pit. It is completed by the fission reaction in which dynamin GTPase plays a key role (Slepnev and De Camilli, 2000; Takei and Haucke, 2001). Dynamin polymerizes into rings at the neck of clathrin-coated pits, but its precise mechanism of action in fission remains unclear. In vitro studies have shown that dynamin binds to lipids, and it can deform lipid bilayers into narrow tubules and fragment them in a GTP-hydrolysis-dependent way. Thus, one model proposes that dynamin acts as a mechanoenzyme, that is, by constricting the vesicle neck via a conformational change coupled to its catalytic action (Takei et al, 1995; Sweitzer and Hinshaw, 1998; Marks et al, 2001; Chen et al, 2004). Another model proposes that GTP-bound dynamin functions by activating a still unidentified downstream effector (Sever et al, 1999; Song and Schmid, 2003). The GTPase module of dynamin is located at the N-terminus of the protein. This domain is followed by the evolutionary conserved middle domain, a plecstrin homology (PH) domain, and the C-terminal proline/arginine-rich domain (PRD). The region between the PH domain and the PRD is referred to as the GTPase effector domain (GED), because it binds the GTPase module of adjacent dynamin molecules within dynamin polymers and may function as a GTPase activating protein (GAP) (Muhlberg et al, 1997). The PH domain binds phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2), while the PRD binds a variety of src-homology 3 (SH3) domain-containing proteins. These interactions also have been reported to stimulate dynamin GTPase activity (Gout et al, 1993; Zheng et al, 1996; Barylko et al, 1998). Dynamin 1 is the best-characterized dynamin isoform and is enriched at synapses. Amphiphysin, which is present in brain primarily as a dimer of two similar isoforms, amphiphysin 1 and 2, is a major binding partner of dynamin 1 in the vertebrate nervous system. Amphiphysin binds the PRD of dynamin via a C-terminal SH3 domain and co-assembles with dynamin into rings either in solution or on lipid tubules (Takei et al, 1999). This interaction is expected to be physiologically important in nerve terminals, because acute disruption of amphiphysin by peptide microinjection leads to a block of synaptic vesicle endocytosis (Shupliakov et al, 1997). The N-terminal portion of amphiphysin contains a BAR (BIN/amphiphysin/Rvs) domain that mediates homo- and heterodimerization (Wigge et al, 1997; Ramjaun et al, 1999). Via this domain, amphiphysin, like dynamin, binds to and tubulates liposomes (Takei et al, 1999), and can also generate narrow plasma membrane tubules in living cells when overexpressed (Peter et al, 2004). The structure of the BAR domain of amphiphysin was recently determined by crystallographic analysis and found to represent an evolutionary conserved protein module present in a variety of proteins (Habermann, 2004; Peter et al, 2004). It is represented by a triple helix arranged in an antiparallel dimer to form a crescent-like structure. It was proposed that lipid interactions mediated by the concave portion of this module, in cooperation with an amphipathic N-terminal helix (Farsad et al, 2001), may be responsible for its property to curve the bilayer (Peter et al, 2004). In addition, lipid-binding studies revealed preferential binding to small liposomes, suggesting that the BAR domain may also function as a curvature sensor, thus facilitating the recruitment of amphiphysin to pre-existing high curvature membranes (Peter et al, 2004). Curvature-sensitive properties of proteins may have an important role in orchestrating vesicle formation and fission, with the sequential recruitment and dissociation of cytosolic factor (Bigay et al, 2003). The central region of brain amphiphysin 1 and 2 contains binding sites for clathrin and for the clathrin adopter protein 2 (AP-2) (Ramjaun and McPherson, 1998; Slepnev et al, 2000; Miele et al, 2004). Via this multiplicity of interactions, which are regulated by an intramolecular binding (Farsad et al, 2003), cyclin-dependent kinase 5-dependent phosphorylation (Floyd et al, 2001; Tomizawa et al, 2003) and calcineurin-dependent dephosphorylation (Bauerfeind et al, 1997), amphiphysin is thought to play an important role in coordinating vesicle budding and fission (Farsad et al, 2003). Accordingly, defects in synaptic vesicle recycling were demonstrated in amphiphysin 1 knockout mice (Di Paolo et al, 2002). In these mice, not only expression of amphiphysin 1 is suppressed, but also the level of amphiphysin 2 is severely reduced in brain, probably due to the stabilizing effect of the amphiphysin heterodimer on the turnover of amphiphysin 2. Although these studies support the importance of the amphiphysin–dynamin interaction in endocytosis, it remains unclear whether amphiphysin is directly implicated together with dynamin in the fission reaction. The reported lack of significant amphiphysin expression in Drosophila neurons (Leventis et al, 2001; Razzaq et al, 2001; Zelhof et al, 2001) and of a major synaptic vesicle recycling defect in amphiphysin 1 knockout mice raise questions on the importance of this partnership. However, other proteins with similar domain structure and enriched at synapses, for example, endophilin and intersectin/pacsin, may functionally replace amphiphysin (Simpson et al, 1999; Modregger et al, 2000; Qualmann and Kelly, 2000; Farsad et al, 2001). In this study, we demonstrate the importance of amphiphysin for the vesicle-generating properties of dynamin 1 within the context of mouse brain cytosol in vitro. We also show a striking stimulatory effect of amphiphysin 1 on the GTPase activity of dynamin in the presence of liposomes, which is dependent on the size of liposomes. Results Amphiphysin potentiates vesicle formation by dynamin Dynamin-dependent vesicle formation can be reconstituted in vitro by incubation of large unilamellar liposomes with brain cytosol in the presence of nucleotides, and monitored quantitatively by dynamic light scattering (DLS) (Kinuta et al, 2002). To establish the importance of amphiphysin in vesicle formation, we compared brain cytosol from wild-type and amphiphysin 1 knockout mice in this assay. The brain cytosol of the knockout mice is also nearly completely devoid of amphiphysin 2. When incubated with large unilamellar liposomes (1779.5±461.7 nm in diameter) (Figure 1A), wild-type cytosol induced a massive formation of small vesicles (112.3±19.7 nm in diameter) (Figure 1B), while knockout mice brain cytosol was very inefficient in this process (Figure 1C). At the end of the incubation, small vesicles represented 82.9±11.7% of the total vesicle number in the wild-type cytosol sample, but only 6.8±5.5% in the knockout cytosol sample (Figure 1B and C). Addition of purified amphiphysin 1 (0.5 μg/ml) to the knockout cytosol rescued vesicle formation, with small vesicles (176.8±33.9 nm in diameter) representing 82.1±4.3% of the total vesicles at the end of the reaction (Figure 1E). No vesicle formation was observed when liposomes were incubated with amphiphysin 1 alone (Figure 1D). Results of these assays are summarized in Figure 1F. Figure 1.Requirement of amphiphysin 1 for vesicle formation by brain cytosol. (A–E) Vesicle formation quantified by the DLS assay. Data represented are the relative distribution in numbers of formed vesicles. Large unilamellar liposomes were incubated in the presence of 200 μM GTP and 2 mM ATP under the conditions indicated. (F) Percentage of small vesicles formed in each incubation condition. Vesicles smaller than 200 nm in diameter were defined as small vesicles. Values represent the mean±s.d. for more than three independent measurements. Amph=amphiphysin 1; Amph KO cyto.=amphiphysin knockout mice brain cytosol; WT cyto.=wild-type mice brain cytosol. Download figure Download PowerPoint We next examined the effect of purified amphiphysin 1 on vesicle formation by purified dynamin. Large unilamellar liposomes were incubated with dynamin alone or a mixture of dynamin plus amphiphysin 1 in the presence of GTP. At the end of the reaction, small vesicles represented 15.7±12.7% of the total vesicles (Figure 2B and D) in the sample containing dynamin 1 alone, and 71.1±4.5% of the total in the samples containing both proteins (Figure 2C and D), thus indicating a potent action of amphiphysin 1 on the vesicle-generating properties of dynamin. Massive fragmentation of liposomes into small vesicles following incubations with dynamin and amphiphysin 1 was confirmed by negative staining electron microscopy (EM) (Figure 2E). Figure 2.Quantitative analysis of vesicle formation from liposomes by dynamin and amphiphysin 1. (A–C) Vesicle formation by purified proteins was quantified by DLS assay, and the relative number distribution was represented. Large unilamellar liposomes were incubated in the presence of 1 mM GTP under the conditions indicated. The ratio of dynamin (Dyn) to amphiphysin 1 (Amph) was 1:2.67 (mol/mol), a ratio of 1:2 (w/w). (D) Percentage of small vesicles formed in each incubation condition. Values represent the mean±s.d. for more than three measurements. (E) Negative-staining EM of large unilamellar liposomes before incubation (left), and vesicles formed by incubation with dynamin and amphiphysin 1 (right). Bar=100 nm. Download figure Download PowerPoint Stimulation of dynamin GTPase activity by amphiphysin is affected by liposome size To gain some insight into the mechanisms through which amphiphysin 1 stimulates dynamin-dependent vesicle formation, we examined whether amphiphysin 1 enhanced dynamin GTPase activity. Addition of increasing amount of amphiphysin 1 to a reaction mixture containing dynamin, [γ-32P]GTP and large unilamellar liposomes (1779.5±461.7 nm in diameter) composed of a brain lipid extract (Folch fraction type 1, FF) (74%, w/w), cholesterol (20%) and PtdIns(4,5)P2 (6%) produced a prominent increase in the GTPase activity. The activity was maximal when the molar ratio of dynamin to amphiphysin 1 was in the 1:1–2 range (Figure 3A). Such effect may be explained in part by enhanced recruitment of dynamin, because amphiphysin also increased the recruitment of dynamin to large liposomes (approximately two-fold) (Figure 3D). Figure 3.Stimulation of dynamin GTPase activity by amphiphysin 1. Dynamin (0.2 μM) and increasing amount of amphiphysin 1 were incubated either with 2 μg of large unilamellar liposomes (A) or with 2 μg of sonicated small liposomes (B), in the presence of 1 mM GTP containing [γ-32P]GTP (0.8 μCi). After incubation, 32Pi released from the radiolabeled GTP was measured. Data represent the mean value of triplicate measurements. (C) GTPase activity assayed using extruded small liposomes. (D) Recruitment of dynamin to large unilamellar liposomes (left two lanes) or to sonicated small liposomes (right two lanes) in the presence or absence of amphiphysin 1. Results shown are the average of three assays±s.d. Download figure Download PowerPoint Since small liposomes were often utilized in previous studies on dynamin GTPase activity (Tuma et al, 1993; Barylko et al, 1998), we also tested small liposomes prepared by sonication (80.7±10.8 nm in diameter) in the GTPase assay. In the absence of amphiphysin, the GTPase activity of dynamin was much higher with small liposomes than with large liposomes, possibly due to a more efficient oligomerization of dynamin on small liposomes. Surprisingly, addition of increasing amount of amphiphysin 1 led to a drastic decrease in the GTPase activity in the presence of the small liposomes. The molar ratio that generated maximal stimulation in the presence of large liposomes produced a 75% inhibition under these conditions (Figure 3B). A similar inhibitory effect was observed also with small liposomes prepared by extrusion (143.8±31.7 nm in diameter) (Figure 3C), ruling out a possible effect of sonication on the properties of the bilayer. Thus, the effect of amphiphysin 1 on dynamin GTPase activity is strongly affected by the size of the liposomes. An attractive possibility is that the stimulation of dynamin GTPase activity is related to the property of amphiphysin and dynamin to co-oligomerize into ring structures around membrane tubules. While large liposomes allow massive tubulation (Takei et al, 1999), the size of the small liposomes may be incompatible with tubulation and constriction, explaining the different results obtained with large and small liposomes. Stimulatory action of amphiphysin is affected by lipid component If the effect of amphiphysin on dynamin GTPase activity is linked to its property to co-oligomerize with dynamin on tubular membranes, the effect should be strongly dependent on the presence of liposomes. Therefore, we examined whether amphiphysin stimulates dynamin GTPase activity in the presence or absence of liposomes. In the absence of liposomes, the GTPase activity of dynamin was low, and amphiphysin 1 stimulated this activity only slightly, approximately 1.2-fold (compare lanes 1 and 2 in Figure 4A). The presence of large unilamellar liposomes (FF (74%, w/w), cholesterol (20%) and PtdIns(4,5)P2 (6%)) produced a 2.6-fold increase in dynamin GTPase activity (compare lanes 1 and 3 in Figure 4A), and the additional presence of amphiphysin 1 produced a further 3.1-fold stimulation (compare lanes 3 and 4 in Figure 4A). Replacement of PtdIns(4,5)P2 with L-α-phosphatidylcholine (PC) in the liposomes reduced the stimulation produced by the liposomes (compare lanes 3 and 5 in Figure 4A), consistent with the known role of PtdIns(4,5)P2 in dynamin recruitment and activation. However, the stimulatory effect of amphiphysin 1 on the GTPase activity of dynamin was nearly unaffected (approximately 3.4-fold) (compare lanes 5 and 6 in Figure 4A). Figure 4.Effect of the lipid composition on stimulation of dynamin GTPase activity by amphiphysin 1. (A) Dynamin GTPase activity measured with or without amphiphysin 1, and in the absence or presence of liposomes. The lipid composition of FF+PtdIns(4,5)P2 and FF+PC liposomes are cholesterol/FF/PtdIns(4,5)P2=20:74:6% (w/w) and cholesterol/FF/PC=20:74:6% (w/w), respectively. Results were normalized to the average of control value (1.06 nmol Pi/min/nmol Dyn). (B) Negative-staining electron micrographs of liposomes incubated with amphiphysin 1. Left, a single liposome containing 40% PS (cholesterol/PS/PC=20:40:40%, w/w) generated irregular surface, but little tubulation was observed. Right, a single liposome containing 80% PS (cholesterol/PS=20:80%, w/w) formed massive tubules. Bar=500 nm. (C) Dynamin GTPase activity measured with (closed circles) or without (open circles) amphiphysin 1, in the presence of liposomes containing various amounts of PS, PC and 20% cholesterol. Data represent the mean value of triplicate measurements. Download figure Download PowerPoint Another acidic phospholipid was tested. L-α-Phosphatidylserine (PS) was reported to support the binding and polymerization of both dynamin and amphiphysin 1 and to stimulate the GTPase activity of dynamin (Figure 4B, Tuma et al, 1993; Sweitzer and Hinshaw, 1998; Takei et al, 1999). Thus, liposomes containing variable amounts of PS (from 20 to 80%) were used in the tubulation and GTPase assays. Amphiphysin 1 tubulated very efficiently liposomes containing the higher concentrations of PS (Figure 4B, Takei et al, 1999). The GTPase activity of dynamin in the absence of amphiphysin 1 increased linearly in parallel with the concentration of PS (Figure 4C, open circles). The presence of amphiphysin 1 produced an additional powerful stimulation of the GTPase activity, which was highly synergistic as the amount of PS was increased. The sigmoidal shape of the stimulation curve indicated a strong positive cooperativity when the amount of PS was between 40 and 60% (Figure 4C, closed circles), possibly reflecting gathering of PS and preferential assembly of dynamin–amphiphysin polymers on PS-rich microdomains. Both BAR domain and SH3 domain of amphiphysin 1 are required for the stimulation of dynamin GTPase activity The effect of truncated amphiphysin 1 molecules (Figure 5A) was investigated to identify the domains of the protein responsible for its stimulatory action on the catalytic activity of dynamin. These fragments were tested in the presence of large liposomes composed of FF (74%), cholesterol (20%) and PtdIns(4,5)P2 (6%). Under this assay condition, full-length amphiphysin 1 stimulated the GTPase activity about three-fold (Figure 5B, see also lanes 3 and 4 in Figure 4A). A roughly similar stimulation was produced by Amph 1–226, which comprises the BAR domain only, and by Amph 1–306, which contains an additional flanking region at the C-terminal side of this domain. Strikingly, Amph Δ248–601, which contains both BAR and SH3 domains, but lacks the central region, produced a 26-fold stimulation of the GTPase activity. As the proline-rich stretch binds intramolecularly to the SH3 domain and regulates negatively the interaction of amphiphysin with dynamin (Farsad et al, 2003), we also tested Amph Δ248–315, a mutant lacking the proline-rich stretch. This mutant stimulated the GTPase activity about three-fold over the value obtained with full-length amphiphysin 1. Amph 545–695 alone, corresponding to the C-terminal SH3 domain, had no effect (Figure 5B). Furthermore, stimulation of the GTPase activity by the BAR domain was not enhanced by the presence of the free SH3 domain (data not shown). These results suggested that presence of both the BAR domain and the SH3 domain in a single molecule was critical for the stimulation of dynamin GTPase activity. Figure 5.Amphiphysin 1 domains required for the stimulation of dynamin GTPase activity. (A) Domain structure of amphiphysin 1 constructs used for assays. Numbers indicate amino-acid residues of full-length amphiphysin 1. (B) Dynamin GTPase activity in the presence of each construct and of large unilamellar liposomes containing 74% FF, 20% cholesterol and 6% PtdIns(4,5)P2. Dyn:Amph=1:2 (mol/mol). Data represent the mean value of triplicate measurements. Results were normalized to the average of control value (2.86 nmol Pi/min/nmol Dyn). (C) Dynamin binding to large liposomes in the presence of amphiphysin or truncated mutants was checked by the sedimentation assay. BAR=BIN/amphiphysin/Rvs; PRS=proline-rich stretch; CLAP=clathrin-AP-2-binding domain; SH3=src-homology 3. Download figure Download PowerPoint The observed effects of the amphiphysin deletion constructs on the GTPase activity of dynamin correlated qualitatively with their property to enhance dynamin recruitment to large liposomes, with Amph Δ248–601 producing a more than three-fold increase in recruitment (Figure 5C). Both BAR domain and SH3 domain of amphiphysin 1 are sufficient for the formation of rings The incubation of dynamin and amphiphysin 1 in physiological buffer results in the formation of ring structures, similar in size to dynamin-containing collars of endocytic pits (Takei et al, 1999). These collars are thought to represent a transient intermediate in the fission reaction. We used negative staining EM to determine whether these rings were present following incubation of dynamin with the truncated amphiphysins. Rings were observed after incubation of dynamin with full-length amphiphysin 1, with Amph Δ248–315, or with Amph Δ248–601, but not with the other amphiphysin 1 fragment (Figure 6). The diameter of rings with Amph Δ248–601 (41.2±3.0 nm of outer diameter and 22.3±2.6 nm of inner diameter) was slightly shorter than that of rings with amphiphysin 1 (outer diameter 51.7±4.3 nm and inner diameter 27.5±3.3 nm). Figure 6.Negative staining EM of reaction mixtures containing dynamin and each construct indicated. Note the massive ring formation with full-length amphiphysin 1, Amph Δ248–315 or with Amph Δ248–601. Bar=100 nm. Download figure Download PowerPoint Discussion We provide strong evidence for the physiological role of amphiphysin in the regulation of dynamin function and new insights concerning the mechanisms through which amphiphysin achieves this effect. Several previous reports, including studies in living cells and cell-free studies, had suggested an important partnership of amphiphysin and dynamin in the fission reaction of endocytic vesicles, in particular in the fission of recycling synaptic vesicles at synapses (Shupliakov et al, 1997; Takei et al, 1999). Yet, the importance of this interaction has been challenged by studies in Drosophila, where amphiphysin is expressed primarily outside the nervous system (Leventis et al, 2001; Razzaq et al, 2001; Zelhof et al, 2001), and by the relatively mild neurotransmission defects observed in amphiphysin 1 knockout mice (Di Paolo et al, 2002), which lack both amphiphysin 1 and 2 in the neuronal cytosplasm. In contrast with these findings, we have now found that the property of brain cytosol to vesiculate large liposomes is drastically impaired in the cytosol of amphiphysin 1 knockout brain (Figure 1). Since this property was previously shown to be mediated mainly by dynamin (Takei et al, 1998; Kinuta et al, 2002), our findings point to a role of amphiphysin in the regulation of dynamin. Addition of amphiphysin 1 is sufficient to rescue the defect of the knockout brain cytosol, indicating that in this assay the amphiphysin 1 homodimer and the amphiphysin 1/2 heterodimer have similar properties. The discrepancy between these in vitro results and in vivo data may reflect the compensatory action in vivo of other BAR domain-containing proteins (Peter et al, 2004), such as endophilin and syndapin/pacsin, that have a domain structure similar to amphiphysin (Qualmann and Kelly, 2000; Habermann, 2004; Peter et al, 2004) and are also present at synapses. The importance of amphiphysin relative to other proteins with overlapping function may be greater in in vitro assays than in situ because of differential compartmentalization and/or regulation in living nerve terminals. Furthermore, yet unidentified molecules present in cellular membranes, and absent in our cell-free assays, may compensate for the absence of amphiphysin. We have found that the property of amphiphysin to stimulate the GTPase activity of dynamin is critically influenced by the presence of liposomes and that this effect, in turn, is strongly affected by the composition and size of liposomes (Figures 3 and 4). The presence of acidic phospholipids is required, as expected. Amphiphysin stimulated the GTPase activity of dynamin and its binding on large unilamellar liposomes. Large liposomes, which can be evaginated to narrow tubules, may allow dynamin and amphiphysin to co-oligomerize
Referência(s)