Abstract
Cell-cell fusion is an indispensable process in the conception, development and physiology of multicellular organisms. Here we demonstrate a direct and noncanonical role for dynamin, best known as a fission GTPase in endocytosis, in cell-cell fusion. Our genetic and cell biological analyses show that dynamin colocalizes within the F-actin-enriched podosome-like structures at the fusogenic synapse, which is required for generating invasive membrane protrusions and myoblast fusion in vivo, in an endocytosis-independent manner. Biochemical, negative stain EM and cryo-electron tomography (cryo-ET) analyses revealed that dynamin forms helices that directly bundles actin filaments by capturing multiple actin filaments at their outer rim via interactions with dynamin’s proline-rich domain. GTP hydrolysis by dynamin triggers disassembly of the dynamin helix, exposes the sides of the actin filaments, promotes dynamic Arp2/3-mediated branched actin polymerization, and generates a mechanically stiff actin network. Thus, dynamin functions as a unique actin-bundling protein that enhances mechanical force generation by the F-actin network in a GTPase-dependent manner. Our findings have universal implications for understanding dynamin-actin interactions in various cellular processes beyond cell-cell fusion.
Cell-cell fusion is essential for the conception, development, regeneration and physiology of multicellular organisms1-3. A common theme underlying cell-cell fusion events ranging from insects to mammals is the presence of actin-propelled invasive protrusions at the site of fusion, known as the fusogenic synapse4-8. Such invasive protrusions were first identified as part of an F-actin-enriched podosome-like structure (PLS) in Drosophila myoblast fusion7. The PLS is generated by an “attacking” fusion partner, the fusion-competent myoblast (FCM), which drills multiple invasive protrusions into the “receiving” fusion partner, the muscle founder cell, to promote cell membrane juxtaposition and fusion pore formation7,9,10. Although each protrusion has a diameter similar to that of a filopodium, these protrusions are mechanically stiffer and are capable of triggering mechanosensitive responses in the receiving cell11. Mechanoresponsive accumulations of myosin II and spectrin in the receiving cell, in turn, increase the cortical tension of the receiving cell, thus generating resisting forces against the invasive protrusions and bringing the two cell membranes to close proximity to promote cell fusion11,12. The invasive protrusions from the attacking cells are propelled by Arp2/3-mediated branched actin polymerization, which generates short actin filaments more suitable for mechanical work than long and linear filaments9. However, how the branched actin filaments are organized to generate mechanically stiff membrane protrusions during fusion is unknown.
Dynamin (Dyn) is a large GTPase best known for its role in endocytosis. Dynamin catalyzes membrane fission by forming rings/helices around necks of a budding endocytic vesicles13-15. The dynamin protein consists of an amino-terminal G domain that catalyzes GTP hydrolysis, a three helix bundle-signaling elements (BSEs), a stalk critical for oligomerization, a pleckstrin homology (PH) domain that interacts with phosphatidyl-inositol-4,5-biphosphate (PIP2) in the plasma membrane, and a carboxy-terminal proline-rich domain (PRD) for SH3 domain interactions that support proper dynamin localization13-15. Crystal structures of dynamin revealed a hairpin-like folding of each monomeric protein mediated by the stalk regions and the BSEs16,17. When assembled into rings or helices, the PH domain at the vertex of the hairpin is positioned inside the dynamin ring, directly interacting with the enclosed membrane tubule that is wrapped inside, whereas the G domain is localized at the outer rim of the dynamin helix. The stalk regions form the backbone of the helix in between the PH and the G domains. The unstructured PRD has not been observed in high-resolution structure, and is predicted to be neighboring the G domain and extend outward beyond the G domain13,18.
In addition to endocytosis, dynamin regulates the formation and/or function of F-actin-enriched cellular structures, such as podosomes19,20, invadopodia21, filopodia22, lamellipodia23, actin comet tails24,25, phagocytic cups26, and stress fibers27,28. While mechanisms underlying dynamin-mediated endocytosis have been extensively studied, how dynamin interacts with actin remains a widely-documented but poorly understood phenomenon. Studies suggest that dynamin bundles actin either directly29,30 or indirectly in a cortactin-dependent manner22,31,32. The binding of actin filaments directly by dynamin was proposed to antagonize gelsolin-mediated filament capping thus promoting actin polymerization29, whereas the cortactin-mediated dynamin-actin interaction was thought to stabilize F-actin filaments22,31,32. In either model, the precise mode and functions of dynamin-actin interactions remain unclear.
Results
Dynamin is required for myoblast fusion in vivo and colocalizes with the F-actin foci at the fusogenic synapse
The presence of a podosome-like structure at the fusogenic synapse7 prompted us to examine whether dynamin, which was implicated in podosome formation19, plays a role in Drosophila myoblast fusion. The Drosophila genome encodes one dynamin, named shibire (shi), which shares 67% protein sequence identity with human dynamins. The availability of temperature-sensitive (ts) alleles of shi allowed us to eliminate both maternal and zygotic Shi function in a temporally controlled manner. We incubated the shits embryos at permissive temperature (18°C) for 14.5 hours just prior to the start of muscle development, before shifting to restrictive temperature (34°C) for 4 hours and analyzing the resulting muscle phenotype. Both shits1 and shits2, which contain single point mutations in the G domain, exhibited a severe myoblast fusion defect at 34°C characterized by the presence of many unfused, mononucleated myoblasts (Fig. 1A, B; Fig. S1A, B, C). These myoblasts expressed tropomyosin and muscle myosin heavy chain (MHC), indicating that they had properly differentiated (Fig. 1A; Fig. S1A, B). Expressing wild-type Shi-GFP in shits2 mutant embryos with an all muscle cells twi-GAL4 driver or specifically in FCMs with an sns-GAL4 driver significantly rescued myoblast fusion at 34°C (Fig. 1B; Fig. S1D). In contrast, expressing Shi-GFP specifically in founder cells with the rP298-GAL4 driver did not rescue the fusion defect. These data establish that the fusion defect was caused specifically by the loss of Shi, especially in the FCMs.
Consistent with Shi’s role in FCMs, Shi colocalized within the FCM-specific F-actin foci at the fusogenic synapse, visualized by both anti-Shi antibody and muscle-specific expression of Shi-GFP (Fig. 1C; Fig. S2A). Live imaging of embryos coexpressing Shi-GFP and actin-mRFP in muscle cells revealed that Shi colocalized with the F-actin foci during their entire life span and dissolved with the F-actin foci when fusion was completed (Fig. 1C; Movie 1). Recruitment of Shi to the fusogenic synapse dependent on the FCM-specific cell adhesion molecule Sns, but not on the actin nucleation-promoting factors (NPFs) for the Arp2/3 complex, including WASP, the WASP-interacting protein (WIP) or Scar (Fig. S2B-E).
A general role for Shi in cell-cell fusion was supported by experiments in Drosophila S2R+ cells that were induced to fuse by co-expressing Sns and a C. elegans fusogen Eff-18. Knocking down (KD) Shi using dsRNAs against the 5’ untranslated region (UTR) led to a fusion defect in S2R+ cell fusion, which was significantly rescued by wild-type Shi expression (Fig. 1D, E). As in Drosophila embryos, Shi colocalized with F-actin foci at the fusogenic synapse in S2R+ cells (Fig. 1F), consistent with a specific function of Shi at the fusogenic synapse. Although a role for dynamin 2 (Dyn2) in the fusion of cultured mouse myoblasts and osteoblasts has been reported in previous studies, the underlying mechanisms have been attributed to either dynamin-mediated endocytosis or fusion pore expansion5,33. The nature of any direct, physical associations between dynamin and the actin-enriched structure at the fusogenic synapse have not been investigated.
Dynamin is required for generating invasive protrusions at the fusogenic synapse independent of endocytosis
In wild-type embryos, founder cells are invaded by the FCM-specific PLS with an average depth of invasion of 2.2 ± 0.7 μm (n=102) (Fig. 1G, left panels; Fig. 1H). However, the PLS in shits2 mutant were broadened and had a decreased invasive depth of 1.2 ± 0.5 μm (n=76) at the restrictive temperature (Fig. 1G, right panels; Fig. 1H). Consistent with the observations using confocal microscopy, ultrastructural studies using transmission electron microscopy (TEM) revealed stubby protrusions originating from a broad base of actin enrichment in shits2 mutant embryos, as compared to the long finger-like protrusions projecting from a dense actin-enriched area in wild-type embryos (Fig. 1I). These data suggest that Shi is involved in promoting the stiff, elongated finger-like protrusions in wild-type embryos.
If the stubby membrane protrusions in shits2 mutant were caused by a defect in endocytosis, one would expect an accumulation of endocytic pits along the cell membranes at the fusogenic synapse at restrictive temperature, as observed at the neuromuscular junction in the shits1 mutant34. However, no endocytic pits were observed at the fusogenic synapse under the restrictive temperature (Fig. 1I). In addition, neither clathrin light chain nor the α subunit of the AP-2 clathrin adaptor complex was enriched at the fusogenic synapse in wild type embryos (Fig. S2F). Thus, clathrin-mediated endocytosis appears not to be associated with the enrichment of Shi at the fusogenic synapse.
Dynamin bundles actin filaments by forming regularly spaced ring/helices
The colocalization of Shi with the F-actin foci and its role in promoting invasive protrusions prompted us to investigate potential interactions between Shi and actin. Low speed F-actin co-sedimentation assays revealed that purified Shi bundled rabbit skeletal muscle actin in a concentration-dependent manner under low ionic strength solution conditions (50 mM KCl). which is known to allow the formation of higher order structures of dynamin, such as rings and helices35,36 (Fig. 2A; Fig. S3A). In addition, high speed co-sedimentation assays showed that Shi bound F-actin (Fig. S3B), and the apparent Kd of Shi-actin interaction was 0.19 ± 0.03 μM (Fig. S3C, D), similar to that of human Dyn1-actin interaction (∼ 0.4 μM)29.
Using total internal reflection fluorescence (TIRF) microscopy, we demonstrated that actin filaments converge together to form thick actin bundles in the presence of Shi in all three conditions tested: on coverslips coated with N-ethylmaleimide (NEM)-modified myosin fragments (Fig. 2B), on supported lipid bilayers (Fig. 2D; Movie 2), and on coverslips coated with mPEG-Silane (Fig. 2E; Movie 3). The thick actin bundles were significantly straighter compared to single actin filaments, indicating that the bundles were mechanically stiffer (Fig. 2C). When single filaments were imaged in a floe chamber, we observed that once an initial filament was generated, a second filament appeared to grow alongside it to form a bi-filament “mini-bundle”, which was then joined by a third filament growing alongside the bi-filament bundle (Fig. 2E, F; Movie 3).
Negative stain EM revealed that when mixed at equal concentrations (1 μM) Shi formed regularly spaced, lighter staining “bands” along the actin bundles (Fig. 2Ga, b). The average distance between the Shi bands was 15.1 ± 1.8 nm (n= 401), comparable to the average pitch of Shi helices around lipid tubules (15.5 ± 1.6 nm; n= 331) (Fig. 2I, J). We refer to the Shi “bands” as Shi rings, since at this resolution it was difficult to determine whether Shi formed individual rings or helices. Interestingly, at low Shi:actin (0.125 μM Shi:1 μM actin), Shi rings appeared to be arranged in “doublets” along the actin bundle and there were stretches of the actin bundle without Shi rings (Fig. 2H). The average distance between the two Shi rings within a doublet was 16.3 ± 2.5 nm (n= 315), and that between two doublets was 20.2 ± 1.6 nm (n= 189) (Fig. 2I), with the sum of the two distances (∼ 36.5 nm) largely coinciding with a half helical pitch of actin (36 nm). Such characteristic spacing of Shi rings at a low Shi concentration suggests that Shi may have high affinity interactions with specific sites on the actin filaments. In addition to single actin bundles, we also observed “super-bundles” in which two, three or more actin bundles aligned in parallel, each of which could be distinguished by the slightly misaligned Shi rings (Fig. 2Gc, d). Occasionally, the Shi rings appeared more aligned across multiple actin bundles (Fig. 2Ge).
Interestingly, the average outer diameter of an actin bundle was 32.0 ± 3.6 nm (n= 259), which is similar to that of Shi rings alone (31.3 ± 2.9 nm; n= 282) (Fig. 2G, K, L), indicating that bundling actin did not significantly change the size of the Shi ring. This is in contrast to the increased diameter of Shi helices surrounding membrane nanotubes (50.7 ± 3.3 nm; n=252) (Fig. 2J, L). Like Drosophila Shi, human Dyn1 and 2 also formed regularly spaced rings that directly bundled actin shown by negative stain EM analysis of negative stained samples (Fig. S4Aa-c and e-f). The average distances between the hDyn1 and hDyn2 rings along actin bundle were 15.6 ± 1.4 nm (n=307) and 15.4 ± 1.4 nm (n=245), respectively (Fig. S4A, B), similar to the average pitch of hDyn1 helices around nanotubes, 15.4 ± 1.7 nm (n=252) (Fig. S4B, E). Moreover, the average outer diameter of the hDyn1 rings along actin bundles (44.7 ± 3.0 nm; n= 220) (Fig. S4A, C) was significantly larger than that of Drosophila Shi (32.0 ± 3.6 nm; n= 259) (Fig. 2G, L), but comparable to the outer diameter of hDyn1 rings alone (40.2 ± 4.3 nm; n= 132), and smaller than that of hDyn1 around nanotubes (51.3 ± 1.8 nm; n=226) and constricted liposomes (∼50 nm)37 (Fig. S4C, D, E). Thus, while there are similarities in the spacing between dynamin rings on lipid substrates and those associated with F-actin filaments, the differences in ring diameter suggest structural differences in the manner by which dynamin rings interact with actin filaments vs. membrane tubules.
To visualize Shi along actin bundles using light microscopy, we generated C-terminally SNAP-tagged Shi, which showed normal activity in assembling around nanotubes and bundling actin (Fig. S3E). Super resolution microscopy revealed Shi-SNAP along the actin bundles (Fig. 2M). At sub-stoichiometric concentrations (0.125 μM Shi-SNAP:1 μM actin), Shi-SNAP appeared in thin and discontinuous stretches along actin bundles (Fig. 2Ma), consistent with the EM observation in Fig. 2H. At stoichiometric concentrations (1 μM each), Shi-SNAP was more densely distributed along the super-bundle and formed periodically repeated “bands” across the entire width of the bundle (Fig. 2Mb), likely corresponding to the aligned trans-interacting Shi-SNAP stretches observed by EM (Fig. 2Ge). At 2 μM, Shi-SNAP uniformly covered the entire super-bundle (Fig. 2Mc).
GTP hydrolysis of dynamin disassembles the dynamin ring and loosens the actin bundles
GTP addition significantly decreased the amount of Shi-SNAP associated with the actin super-bundles, presumably by disassembling the Shi rings, leaving sparsely localized Shi-SNAP punctae (Fig. 3A). Under EM, the actin filaments no longer associated with Shi rings and were loosely scattered by the addition of GTP (Fig. 3B). Consistent with a role for GTP hydrolysis in disassembling Shi ring and actin bundles, point mutations in the Shi G domain that disable GTP hydrolysis blocked GTP-induced Shi ring disassembly and actin bundle dissociation. Specifically, Shi-Q35E (corresponding to human Dyn1-Q40E) (Fig. 3B), Shi-G141S (Drosophila ts2 mutant; Dyn1-G146S) (Fig. 3C) and Shi-K137A (Dyn1 K142A) (Fig. 3B) all bundled actin filaments as wild-type Shi in the absence of GTP, but they either completely or partially inhibited ring disassembly and actin bundle dissociation upon GTP addition (at the restrictive temperature for Shi-G141S).
Fluorescence recovery after photobleaching (FRAP) analyses revealed that actin bundling by Shi in the presence of GTP is a dynamic process. Thus, while little fluorescence recovery was detected after photobleaching a segment of actin-associated Shi-SNAP in the absence of GTP (7.7% ± 8.9%), there was a significant fluorescence recovery (52.0% ± 16.9%) in the presence of GTP. This difference suggests that Shi equilibrates between ring assembly and disassembly when GTP is present (Fig. 3D, E; Movie 4). Similar to Shi, hDyn1 and 2 rings along actin bundles were also disassembled by GTP addition, resulting in loosened actin bundles (Fig. S4Ad, g). We further note that the GTPase activity of Shi was stimulated ∼35 fold in the presence of actin filaments, compared to Shi’s basal activity, consistent with actin-induced, assembly-stimulated hydrolysis (Fig. S6).
Dynamin bundles actin filaments in vivo
If Shi is involved in bundling actin filaments at the fusogenic synapse, one would expect Shi to colocalize with the actin bundles. However, due to the presence of an average of ∼0.5 μM GTP in cells38, one would expect a punctate Shi localization pattern, similar to that of Shi-SNAP on actin super-bundles in the presence of GTP (Fig. 3A). Indeed, super resolution microscopy (SIM and STED) revealed Shi-GFP and Shi-HA punctae decorating the actin bundles at the fusogenic synapse in Sns-Eff-1-expressing S2R+ cells (Fig. 3F; Fig. S5A). To ask whether dynamin exhibits a similar pattern of localization on other actin-enriched structures, such as actin comet tails induced by the bacteria Listeria monocytogenes, we examined hDyn2 on comet tails by super resolution microscopy. Indeed, we observed similar punctate hDyn2 localization on actin comet tails in Listeria-infected Hela cells (Fig. 3G; Fig. S5B). These data suggest that dynamin is involved in bundling actin filaments in vivo and that GTP hydrolysis allows dynamic assembly and disassembly of dynamin rings along the actin bundles.
Dynamin ring disassembly is required for Arp2/3-mediated actin polymerization and cell-cell fusion
Given the colocalization of dynamin and the branched actin network at fusogenic synapses and comet tails, we asked how dynamin-actin interaction might regulate Arp2/3-mediated actin polymerization. We reasoned that dynamin rings along the actin bundles could block Arp2/3-actin interaction therefore preventing branched actin polymerization. Indeed, in the absence of GTP, Shi-bound actin bundles failed to initiate Arp2/3-mediated branched actin polymerization (Fig. 3H; Fig. S7A; Movie 5, 6). However, upon GTP addition, the disassembly of Shi rings resulted in the exposure of actin bundle sides, thus allowing robust branched actin polymerization from their sides (Fig. 3H; Fig. S7A; Movie 5, 6). Similarly, disassembly of human Dyn1 rings upon GTP addition also led to robust Arp2/3-mediated branched actin polymerization (Fig. S7B; Movie 7). Together, our data suggest that GTP hydrolysis driven disassembly of dynamin rings exposes the sides of the actin bundles to allow the generation of new short and branched actin filaments suitable for mechanical work. These branched filaments can be bundled by newly assembled dynamin rings to further increase the mechanical strength of the branched actin network to promote membrane protrusions.
To test whether GTP-induced disassembly of the Shi rings is required for cell-cell fusion, we performed rescue experiments using the G domain mutants in S2R+ cells co-expressing Sns, Eff-1 and dsRNA against the 5’ UTR of shi. Shi-Q35E, Shi-K137A and the ts2 mutant (at restrictive temperature) failed to rescue the cell fusion defect in Shi KD cells (Fig. 3I, J). In addition, overexpression of Shi-Q35E in wild-type S2R+ cells coexpressing Sns and Eff-1 caused strong dominant-negative effect on cell-cell fusion, whereas Shi-K137A had a weaker dominant-negative effect (Fig. 3I), consistent with the partial disassembly of Shi-K137A ring in the presence of GTP (Fig. 3B). Taken together, these results support a role for GTP hydrolysis-mediated dynamin ring disassembly in promoting cell-cell fusion.
The stalk region and the PH domain are not specifically involved in dynamin-actin interactions
To further probe the mechanism of dynamin-actin bundling, we set out to map the specific actin-binding sites in dynamin. A previous study mapped the actin-binding domain of human Dyn1 to a region within the stalk, and showed that mutating several positively charged amino acids (Dyn1-K414/415/419/421/426E; known as Dyn1-AKE) significantly impaired Dyn1-actin interaction29. We made the corresponding Shi quadruple mutant, Shi-K410E, R411E, K417E, K422E (Shi-AKE), but found that the expression level of Shi-AKE was significantly lower (20.2% ± 15.7%) than wild-type Shi and most of the mutant protein was unstable during purification (Fig. S8A, B). Shi-AKE expression neither rescued the cell fusion nor endocytosis defect in Shi KD cells (Fig. S8C, D, E), and failed to cause a dominant-negative effect in cell fusion or endocytosis in wild-type cells (Fig. S8C, E). These results suggest that the Dyn1-AKE mutant likely affects Shi protein stability and assembly, rather than specifically blocking Shi-actin interactions.
The ability of dynamin to bind and assemble onto lipid nanotubes requires interactions with the PH domain15. Thus, we made a Shi mutant carrying a deletion of the PH domain (ShiΔPH) and tested its interaction with actin (Fig. 4A). High-speed and low-speed F-actin co-sedimentation assays and EM of negative stained samples revealed that ShiΔPH bound and bundled actin as effectively as wild-type Shi (Fig. 4B,C; Fig. S9A), but failed to form helices around lipid tubules as predicted (Fig. S9B). Similarly, a point mutation in the PH domain (Shi-I527A, corresponding to Dyn1-I533A), which is required to interact with PIP2, did not affect actin bundling (Fig. 3B). However, and unexpectedly, Shi-I527A failed to disassemble in the presence of GTP (Fig. 3B), suggesting again that the nature of dynamin-assembly with actin was structurally distinct from dynamin’s assembly around lipid templates. Consistent with a defect in disassembly, Shi-I527A failed to rescue the cell-cell fusion defect in Shi KD cells, and caused a dominant negative effect on cell-cell fusion in wild-type cells (Fig. 3I). These results suggest that actin filaments are not bundled in the center of the Shi rings, and that the PH domain is involved in conformational changes required to disassemble actin-associated Shi rings upon GTP hydrolysis.
The PRD of dynamin plays a primary role in mediating dynamin-actin interactions
Deleting the PRD domain (ShiΔPRD) significantly decreased Shi’s actin binding and bundling activities, shown by high-speed and low-speed co-sedimentation assays (Fig. 4B; Fig. S9A). Consistent with this, negative-stain EM revealed that ShiΔPRD failed to organize actin filaments in tight bundles and failed to form regularly spaced rings (Fig. 4C), despite forming normal helices around lipid tubules (Fig. S9B). Since PRD is best known for interacting with SH3 domains, we tested whether isolated SH3 domains would interfere with PRD-actin interaction. As shown in Fig. S9C, although SH3 domains compromised Shi-actin interaction, a significant portion of Shi remained in the pellet in the high-speed sedimentation assay, suggesting that the PRD can accommodate interactions with both SH3 domains and actin simultaneously (Fig. S9C). Indeed, most of the SH3-interacting PxxP motifs reside in the N-terminal half of the Shi PRD, whereas most of the conserved and positively charged amino acids reside in the C-terminal half of the PRD (Fig. S9D). The critical function of the PRD in actin binding is not specific for Drosophila Shi, since hDyn2ΔPRD also exhibited significantly decreased activities in actin binding and bundling (Fig. S4F, G). Interestingly, deleting the G domain (ShiΔG) also decreased Shi’s actin bundling activity, albeit to a lesser extent than deleting the PRD (Fig. 4B,C; Fig. S9A). Negative stain EM revealed that ShiΔG could bundle actin to some extent (likely due to the presence of the PRD) without forming regularly spaced rings (Fig. 4C). Taken together, these results suggest that the PRD is the primary actin-binding domain in Shi and that the G domain may also play a minor role in facilitating PRD in bundling actin and/or in enhancing ring assembly so as to position the PRDs for efficient actin binding.
To test whether the PRD itself has actin-binding activity, we purified the Shi PRD alone, as well as two PRDs linked in a tandem repeat (2xPRD). PRD alone exhibited a low actin-binding activity detected by high-speed F-actin co-sedimentation assay (Fig. S9E), and the PRD did not bundle actin as probed by low-speed F-actin co-sedimentation assay and negative-staining EM (Fig. 4D,E). However, 2xPRD showed a higher actin-binding activity (Fig. S9E) and organized tight actin bundles despite not forming regularly spaced rings like Shi (Fig. 4D, E). Therefore, the PRD binds actin and the tandem PRD serves as an artificial two-filament actin crosslinker. To further test the ability of PRD in actin bundling, we fused the Shi PRD domain to the C-terminus of Drosophila dynamin-related protein 1 (Drp1), which does not have a PRD and is normally required for mitochondria fission. Although the human Drp1 (hDrp1) has been shown to bundle actin by light microsocpy39, our negative stain EM revealed relatively loose actin bundles organized by either Drosophila Drp1 or hDrp1 without regularly spaced Drp1 bands (Fig. 4F; Fig. S10). Strikingly, however, the Drp1-ShiPRD fusion protein organized tight actin bundles with regularly spaced rings as Shi (Fig. 4F; Fig. S10), demonstrating a critical function of PRD in mediating the formation of tight actin bundles. Consistent with this, Drp1-ShiPRD, but not Drp1, partially rescued the fusion defect in Shi KD cells expressing Sns and Eff-1 (Fig. 4G). The partial rescue in vivo may be due to the requirement of a higher concentration of Drp1-ShiPRD than that of Shi to organize tight actin bundles (Fig. S10). Consistent with this, overexpressing (OE) Drp1-ShiPRD, which was enriched at the fusogenic synapse (Fig. 4H), exhibited a dominant negative effect in the fusion of wild-type cells, presumably by interfering endogenous Shi function as a less efficient actin-bundling protein (Fig. 4G). Taken together, these results demonstrate that the PRD is the primary actin-binding domain in Shi.
Point mutations in the PRD differentially inhibit actin bundling and endocytosis
To gain further insight into the PRD-actin interaction, we made point mutations in the four highly conserved, positively charged residues in the Shi PRD, R804D, R829D, R846D, and R853D (Shi-4RD) (Fig. S9D), and tested the interactions between the dynamin mutants with negatively charged actin. While R804D resides in a PxxP motif and R829 is two amino acids away from a PxxP motif, the two other mutations, R846D and R853D, did not overlap with any PxxP motifs (Fig. S9D). The single mutants (Shi-R804D and Shi-R829D) and the double mutant (Shi-R846/853D) did not affect Shi-actin interaction shown by low/high-speed F-actin co-sedimentation assays and negative-staining EM (Fig. 5A, B, C). However, the quadruple mutant (Shi-R804/829/846/853D or Shi-4RD) exhibited a significant decrease in actin binding and bundling, demonstrating the specificity of the PRD-actin interaction (Fig. 5A, B, C). Furthermore, Shi-4RD, but not the single or double mutants, failed to rescue the fusion defect in Sns/Eff-1-expressing Shi KD cells (Fig. 5D, E), despite their normal colocalization with the F-actin-enriched foci at the fusogenic synapse, likely mediated by PxxP-SH3 domain interactions (Fig. 5F; Fig. S11A). Although Shi-4RD was unable to rescue fusion, this mutant could significantly, albeit partially, rescue the endocytosis defect in Shi KD cells (Fig. 5G, H; Fig. S11B). These data, together with our inability to detect endocytic structures at the fusion synapse even when fission is blocked (Fig. 1I), suggest that dynamin’s function in actin bundling is uncoupled from its role in endocytosis. The incomplete rescue of endocytosis by Shi-4RD supports a potential involvement of dynamin-mediated actin bundling in endocytosis.
Multiple actin filaments are locked to the outer rim of the dynamin helix in each actin bundle
That the externally oriented PRD is the primary site for actin binding suggests that actin filaments are likely to be attached to the outer rim of the dynamin ring. We next applied negative stain and cryo-ET to further explore the nature of dynamin-actin interactions. In the negative stain EM experiments with a low concentration of Shi (0.125 μM), actin filaments were occasionally not bundled together by dynamin at the ends. Such “split ends” allowed us to discern ∼8 actin filaments within an actin bundle (Fig. 6A, top panel). In addition, in some middle segments of an actin bundle where Shi rings were absent, we detected actin filaments running along the outer rim of the Shi rings (Fig. 6A, bottom panel). These data strongly support our model that the actin filaments bind to the outer edges of the Shi rings.
To achieve a higher resolution view of dynamin-mediated actin bundles, we performed cryo-electron tomography (cryo-ET) using hDyn1 (or Shi) with actin. Cryo-ET revealed that hDyn1 assembled as a helix surrounded by actin filaments (Fig. 6B-E; Fig. S12A, B; Movie 8-11). The hDyn1 helices were evident by sectioning through the tomogram and observing that the hDyn1 bands on the top of the bundle had a right-handed tilt, whereas the corresponding hDyn1 bands on the bottom of the bundle exhibited a left-handed tilt (Fig. 6C; Fig. S12A; Movie 9). The middle section of the bundle contained T-shaped structures, similar to images of dynamin decorated lipid tubes13, with the top of the T near the actin filaments that were positioned at the outer rim of the dynamin helix (Fig. 6C; Fig. S12A; Movie 9). Based on the outer diameter (42 nm) and the pitch (16.8 nm) of the hDyn1 helix observed by cryo-ET, we made a model of the hDyn1 helix and positioned actin filaments based on the tomogram (Fig. 6D, E; Fig. S12B). Since the length per turn (L) of the hDyn1 helix was 133 nm, [L = (C^2 + H^2) ^(0.5); C = circumference, H = height or pitch], and the distance between hDyn1 dimers is 8.3 nm based on the high-resolution cryo-EM map of hDyn140, we determined that there were ∼16 hDyn1 dimers per turn (133 nm divided by 8.3 nm) in the actin bundle, which correlated with the helical parameters of 22.5-degree twist and 10.5 Å rise between dimers. Segmentation of a hDyn1 helix and its associated actin filaments revealed 11 actin filaments associated with the outer rim of this hDyn1 helix (Fig. 6D, E; Fig. S12B; Movie 9-11), corresponding to a partially occupied hDyn1 helix in which 11 of the 16 PRDs per turn engaged with actin filaments. Cryo-ET also revealed that Drosophila Shi assembled as a helix surrounded by actin filaments (Fig. S12C, D; Movie 12). Based on the outer diameter (32 nm) and the pitch (15.1 nm) of Shi, we estimated that there are ∼12 Shi dimers per turn in the actin bundle. The 8-filament bundle shown in Fig. 6A therefore is a partially occupied Shi helix. We conclude from these results that each dynamin-mediated actin bundle consists of a centrally localized dynamin helix and multiple actin filaments attached to the outer rim of a dynamin helix.
Discussion
Our data demonstrate that dynamin oligomerizes to regulate actin bundling as it does in regulating endocytosis. However, the mechanisms by which dynamin helices and rings interact with different substrates, such as lipid tubules vs. actin filaments, are distinct. While the lipid tubule is wrapped in the center of a dynamin helix via its interaction with the PH domain, the actin filaments are locked to the outer rim of dynamin rings via their specific interactions primarily with the PRD (Fig. S13A). Our data demonstrate that the PRD itself can directly bind and bundle actin without other linker proteins, in contrast to previous studies suggesting that the PRD interacts with actin filaments through a linker protein, cortactin22,31,32. In vivo studies have shown that the homozygous Drosophila cortactin mutant is viable and phenotypically normal41, and that double knockout of cortactin and its related protein HS1 in mouse platelets does not affect the formation of podosomes or platelet function42. In addition, pharmacological inhibition or genetic depletion of cortactin did not affect actin comet tail formation or Listeria motility in Listeria-infected mammalian cells43,44, whereas dynamin depletion significantly reduces both24,25,44. Thus, cortactin does not appear to have an essential role in dynamin-mediated actin bundling. Although the PRD is best known for its role in dynamin localization via its interaction with the SH3 domain-containing proteins, the PRD can accommodate SH3 and actin interactions simultaneously, likely due to the largely separated SH3-interacting PxxP motifs and the positively charged amino acids involved in actin binding. The identification of the positively charged amino acids in the PRD allows us to uncouple dynamin’s function in actin bundling vs. fission of lipid tubules. In this regard, Shi-4RD abolishes dynamin’s activity to bundle actin and promote cell fusion, but only partially affected its activity in supporting endocytosis. Thus, the dynamin-actin interactions are essential for actin bundling, and only facilitates, but is not required for, endocytosis.
Our study establishes dynamin as a unique F-actin-bundling protein. Most actin bundling factors, such as α-actinin, fascin, fimbrin, filamin, myosin II, and spectrin are two-filament crosslinkers45-47. They form linear or V-shaped structures (can be monomer, dimer or heterotetramer), each of which has two actin-binding sites located at the two opposite ends of the molecule and thus crosslinking two actin filaments like a bridge. Dynamin, on the other hand, appears to be a multi-filament bundler. Dynamin does this by forming a helical structure composed of either 16 dimers (hDyn1) or 12 (Shi) dimers per helical turn, allowing the binding of a maximal number of 16 (hDyn1) or 12 (Shi) actin filaments to a single helix (Fig. S13A). Due to the tilt of the dynamin dimers along the long axis of the actin bundle, each actin filament is bound by two PRDs from two different dimers in each turn of a dynamin helix. By interacting with additional PRDs along the dynamin helix, the multi-filament actin bundle is further stabilized. Moreover, the multi-filament single bundles can align in parallel to form super bundles, likely mediated by trans-bundle PRD-actin interactions. The ability to mediate the formation of multi-filament single bundles, as well as super bundles, makes dynamin a unique and efficient actin-bundling protein.
Our data strongly support a critical function of dynamin in mechanical force generation. By forming single bundles and super bundles, dynamin increases the mechanical strength of the actin filaments. Moreover, the actin-stimulated GTP hydrolysis allows dynamic disassembly of dynamin helices, freeing the sides of the actin bundles and promoting Arp2/3-meidated branched actin polymerization. The short, branched actin filaments may join the pre-existing, partially occupied actin bundles or form new bundles de novo via actin-PRD interactions, thus further increasing the mechanical strength of the actin filaments (Fig. 13B). Thus, by equilibrating between helix assembly and disassembly as a function of GTP hydrolysis, dynamin gradually increases the mechanical strength of the actin network. As a consequence, the podosome-like structure at the fusogenic synapse is able to propel membrane protrusions (Fig. 13C) and the comet tails are able to propel bacteria movement. We speculate that the mechanisms underlying dynamin-actin interaction uncovered in this study will be generally applicable to other cellular processes involving dynamin-mediated actin cytoskeletal rearrangements.
Materials and Methods
Fly stocks and genetics
The following stocks were obtained from the Bloomington Drosophila Stock Center: w1118 (wild type; BL#3605), shits1 (BL#7068), shits2 (BL#2248), ketteJ4-48 (BL#8753), twi-GAL4 (BL#914), UAS-Actin5CmRFP (BL#24778), UAS-Clc-GFP (BL#7107), and UAS-AP-2α-GFP (AP-2αMI06502-GFSTF.1) (BL#59834). Other stocks used were: sns40-49/CyO48, wasp3/TM349, sltrS1946/CyO50, rP298-GAL451, and sns-GAL452. Transgenic flies carrying UAS-Shi-GFP were generated by P-element-mediated germline transformation. To express Shi-GFP in all muscle cells, flies carrying the UAS-Shi-GFP transgene were crossed with twi-GAL4. For transgenic rescue experiments, shits2; UAS-shi-GFP flies were crossed with shits2; twi-GAL4 (all muscle cells), or shits2, rP298-GAL4 (founder cells), or shits2; sns-GAL4 (FCMs) flies. For phenotypic analyses of embryos carrying shits1 or shits2 mutations, embryos were collected for 1 or 2 hrs at 20 °C and incubated for 16 or 15 hours at 18 °C prior to temperature shift. The embryos were then shifted to the restrictive temperature of 32 (or 34) °C for either 2 hours (or 4 hours), followed by fixation and immunostaining experiments to assess the F-actin foci phenotype (or myoblast fusion phenotype). Washes and dechorionation prior to fixation were carried out with buffers pre-warmed to 32 °C.
Cell culture
S2R+ cells were cultured in Schneider’s medium (Gibco) supplemented with 10% fetal bovine serum FBS (Gibco) and penicillin/streptomycin (Sigma). S2R+ cells were induced to fuse as previously described8. Briefly, 1.2×106 cells were plated into each well of a 6-well plate and transfected with Sns and Eff-1 constructs (200ng each) using Effectene (Qiagen) per the manufacturer’s instructions. Transfected cells were then fixed after 48 or 72 hrs for phenotypic analyses. For Shi RNAi knockdown (KD) experiments, cells were incubated with 3 µg/ml of dsRNA for four days, transfected with appropriate DNA constructs, and continued to be incubated with 3 µg/ml of dsRNA for 48-72 hrs. For rescue experiments, a Shi construct (100 ng) was co-transfected with Sns and Eff-1 in the Shi KD cells. For drug treatment, cells were incubated with 2 μM, 3 μM or 4 μM of MiTMAB (Abcam, ab120466) at 24 hrs post-transfection, and incubated for another 24 hrs before fixation.
Sf9 cells were cultured in suspension medium of Sf-900 III SFM (Gibco) with L-Glutamine (Gibco) and Antibiotic-Antimyotic (Gibco). The cells were grown at a density of 0.5 × 106 − 6 × 106 cells/ml with 150 rpm shaker at 27 °C. For protein expression, cells were diluted to 0.5 × 106 cells/ml with fresh medium and cultured for 24 hrs, followed by transfection with corresponding DNA constructs using Transfection Reagent I (Avanti) per manufacturer’s instructions and continued culturing for 48 hrs.
HeLa cells (ATCC CCL-2) were cultured in DMEM (Gibco) supplemented with 1X non-essential amino acids (Gibco) and 10% FBS (Gibco).
Immunofluorescence staining and confocal microscopy
For Drosophila embryo staining, embryos were dechorionated, rinsed, and fixed in 1:1 mix of heptane and 4% formaldehyde in PBS for 20 min, devitellinized in 1:1 heptane/methanol by vigorous shaking, washed in methanol, and incubated in PBSBT (0.2% BSA and 0.1%Triton in PBS) for 2 hrs, unless otherwise noted. Samples were incubated with primary antibodies overnight at 4 °C, followed by PBSBT washes and incubation with secondary antibodies. The following primary antibodies were used: rabbit anti-Shi (1:500)53, rabbit anti-MHC (1:2000; gift from Bruce Paterson), rat anti-Tropomyosin (1:1000; Abcam, ab50567), guinea pig anti-Duf (1:500)7, rabbit anti-Ants (1:2000)54; rabbit anti-Eve (1:30; Developmental Studies Hybridoma Bank (DSHB), 3C10), rabbit anti-GFP (1:500; Invitrogen, A11122), chicken anti-GFP (1:5000; Invitrogen, A10262). The following secondary antibodies were used at 1:200: Alexa488-, 568-, and 647-conjugated (Invitrogen) and biotinylated antibodies (Vector Laboratories) made in goat. Vectastain ABC kit (Vector Laboratories, PK-4000) and the TSA system (Perkin Elmer, SAT701001EA) were used to amplify weak fluorescent signals for anti-Eve staining. Nuclei were labeled with Hoechst 33342 (1:1000; Invitrogen, H3570). For Shi antibody staining, embryos were fixed in 1:1 mix of heptane and 37% formaldehyde for 5 min. For F-actin labeling, embryos were fixed in formaldehyde-saturated heptane for 1 hour at room temperature, hand-devitellinized, and incubated with Alexa488-, Alexa568- or Alexa647-conjugated phalloidin (1mg/ml) (Sigma) at 1:200 for 2 hrs at room temperature.
For S2R+ cell staining, cultured cells were fixed in 3.7% formaldehyde for 10 min, unless otherwise noted. Samples were incubated with primary antibodies for 1 hr at room temperature, followed by PBSBT washes and incubation with secondary antibodies for 1 hr at room temperature. The following primary antibodies were used: mouse anti-hTfR (1:600)55, rabbit anti-HA (1:600; Sigma, H6908), and hDyn2 (1:50; Santa Cruz Biotechnology, G4). Nuclei were labeled with Hoechst 33342 (1:1000; Abcam, 27671). The following secondary antibodies were used at 1:200: Alexa488-, 568-, and 647-conjugated (Invitrogen) and biotinylated antibodies (Vector Laboratories) made in goat. F-actin was labeled with Alexa488-, Alexa568- or Alexa647-conjugated phalloidin (Sigma) at 1:900.
Labeled samples were mounted in Vectashield (Vector Laboratories) or Aqua-Poly/Mount (Polysciences). Confocal microscopy was performed using a Zeiss LSM 700 microscope equipped with EC Plan Neofluar DIC M27 40x/1.3 and Plan Apochromat DIC 63x/1.4 oil objectives, a Nikon A1R confocal microscope (Ti-E; Nikon) equipped with CFI Plan Fluor 40x/1.3 and CFI Plan Apo VC 60x/1.4 oil objectives, and Leica TCS SPE confocal microscope (DMi8; Leica) equipped with an HC PL APO 20x/0.8 dry objective. The images were processed using Adobe Photoshop CS6 and ImageJ (NIH) softwares.
Molecular biology
All molecular reagents generated and used in this study are described in Table S1.
Electron microscopy
The high-pressure freezing and freeze substitution (HPF/FS) method was used to fix fly embryos as described7,56. Briefly, fly embryos at stage 12-14 were frozen with a Bal-Tec device, followed by freeze-substitution with 1% osmium tetroxide, 0.1% uranyl acetate in 98% acetone and 2% methanol on dry ice. Embryos were then embedded in Epon (Sigma), cut in thin sections (70 nm) with an ultramicrotome (Leica EM UC7), mounted on copper grids, and post-stained with 2% uranyl acetate for 10 min and Sato’s lead solution57 for 1 min for contrast enhancement. Images were acquired on a transmission electron microscope (CM120; Philips).
Proteins purification
Most recombinant proteins generated in this study (Drosophila Shi, ts2, Q35E, K137A, I527A, Shi-SNAP, Shi-ΔG, Shi-ΔPRD, Shi-ΔPH, PRD, 2×PRD, Drp1, Drp1-ShiPRD, Shi-4RD, R804D, R829D, and R846/853D; human Dyn1, Dyn2 and Dyn2-ΔPRD) were expressed in inset Sf9 cells transiently transfected with the corresponding constructs unless otherwise noted. These N-terminal His-tagged fusion proteins were purified with nickel-nitrilotriacetic acid resin (Thermo) according to the manufacturer’s instructions. PRD and 2×PRD were further purified with gel filtration. All purified proteins were dialyzed against 20 mM HEPES, pH 7.3, 150 mM KCl, 1 mM EGTA and 1 mM DTT, flash frozen in liquid nitrogen, and stored at −80°C.
Non-tagged hDyn1 used in this study (Fig. S4A-E) was purified using GST-Amphiphysin II-SH3 domain as described 58. Non-tagged hDyn1 was used to compare with N-terminal His-tagged hDyn1 in the activity of actin bundling and no differences were found revealed by negative stain EM assay. Human Drp1 was purified using bacterial BL21 DE3 RIPL cells as described previously59.
All the SH3 domains used in this study (Drosophila Drk-SH3.1, Drk-SH3.2, Dock-SH3.2, Dock-SH3.3, SH3PX1-SH3, and amphiphysin II-SH3) were expressed in Escherichia coli BL21 (DE3). GST-tagged fusion proteins were affinity purified using Glutathione superflow agarose (Thermo) according to the manufacturer’s instructions. The GST-tags were removed by thrombin. All purified proteins were dialyzed against 10 mM Tris-HCl, pH 8.0, flash frozen in liquid nitrogen, and stored at −80°C.
Co-sedimentation assay
High-speed and low-speed F-actin co-sedimentation assays were performed to determine the F-actin binding and bundling activity of Shi (and its mutants), respectively. 3 μM preassembled F-actin was incubated with the indicated concentrations of Shi or a mutant in 5 mM HEPES, pH 7.3, 50 mM KCl, 2 mM MgCl2, 0.5 mM EGTA, 3 mM imidazole, 60 nM ATP, 60 nM CaCl2, 0.4 mM DTT and 0.003% NaN3 for 30 min at room temperature. The samples were then spun for 30 min at 4 °C at 50,000g for high-speed and 13,600g for low-speed co-sedimentation, respectively. The supernatants and pellets resulting from centrifugation were resolved by 10% SDS-PAGE and stained with InstantBlue (expedeon).
Total internal reflection fluorescence microscopy
To visualize Shi-actin bundles, 2 μM rabbit skeletal muscle actin (90% unlabeled and 10% rhodomine-actin) (Cytoskeleton Inc.), 1 μM Shi were incubated in 6.5 mM HEPES, pH 7.3, 50 mM KCl, 1 mM MgCl2, 0.3 mM EGTA, 80 nM ATP, 80 nM CaCl2, 0.3 mM DTT and 0.004% NaN3 for 2 hrs. 10 μl of the reaction mix was mixed with 2× TIRF buffer (1x TIRF buffer: 50 mM KCl, 1 mM MgCl2, 1 mM EGTA and 10 mM imidazole, 100 mM DTT, 0.2 mM ATP, 15 mM glucose, 20 μg/ml catalase, 100 μg/ml glucose oxidase, 0.2% BSA, 0.5% methylcellulose, pH 7.0), and loaded onto flow chambers prepared as follows. Flow chambers were coated with 10 nM NEM-myosin in HS-TBS buffer (50 mM Tris-HCl, pH 7.5 and 600 mM NaCl) for 1 min, washed with HS-BSA (1% BSA, 50 mM Tris-HCl, pH 7.5 and 600 mM NaCl) twice, washed with LS-BSA (1% BSA, 50 mM Tris-HCl, pH 7.5 and 150 mM NaCl) twice and incubated for 5 min, and washed with TIRF buffer. The loaded samples on the flow chambers were incubated for 5 min and subjected to TIRF imaging. Nikon Eclipse Ti microscopy equipped with an Apochromat DIC N2 TIRF 100x/ 1.49 oil objective lens and an Andor DU-897 X-8654 camera. Convolutedness is calculated as the ratio of traced filament length to the length of the longest side of a bounding rectangle encompassing the same filament60.
To visualize branched actin polymerization mediated by Arp2/3 on dynamin-actin bundles, both Shi-actin and human Dyn1-actin bundles were examined. Shi-actin bundles were preassembled with 2 μM Shi, 1 μM G-actin and 1 μM Alexa 488-conjugated phalloidin. The preassembled samples were loaded onto flow chambers, incubated for 5 min, washed and incubated with TIRF buffer with or without 1 mM GTP for an additional 5 min. Subsequently, 50 nM Arp2/3 (Cytoskeleton Inc.), 1 μM VCA (Cytoskeleton Inc.) and 250 nM rhodamine-actin were loaded onto flow chambers with or without 1mM GTP and subjected to TIRF imaging immediately. To visualize Shi in the actin polymerization process, Shi-SNAP was labeled fluorescently with SNAP-Surface 488 (New England Biolabs), and Shi-SNAP-Surface 488-mediated actin bundles were visualized by TIRF imaging. hDyn1-actin bundles were preassembled with 2 μM hDyn1, 1 μM G-actin and 1 μM Alexa 488-conjugated phalloidin. The preassembled samples were loaded onto flow chambers, incubated for 5 min, washed and incubated with TIRF buffer with or without 0.01 mM GTP for an additional 1 min. After another brief wash with TIRF buffer, 50 nM Arp2/3, 1 μM VCA and 250 nM rhodamine-actin were loaded into flow chambers without GTP and subjected to TIRF imaging immediately.
To monitor the process of Shi bundling of preassembled actin filaments, experiments were performed on His-tagged Ezrin-bound supported lipid bilayers (SLBs). SLBs were prepared as previously described 61. Briefly, 96-well glass-bottomed plate (Matrical) was pretreated with Hellmanex III (Hëlma Analytics) overnight, thoroughly rinsed with MilliQ H2O, washed with 300 μl of 6 M NaOH for 30 min at 45°C and repeated once, rinsed with 500 μl of MilliQ H2O for three times, and washed with 500 μl of HEPES buffer (50 mM HEPES, pH 7.3, 150 mM NaCl and 1 mM TCEP) for three times to equilibrate wells and finally 200 μl of HEPES buffer was kept in each well. 10 μl of SUVs (98% POPC, 2% DOGS-NTA and 0.1% PEG-5000PE) were added into wells and incubated for 1 hr at 40 °C to form SLBs, followed by three washes with 500 μl of HEPES buffer. Next, 100 μl of BSA buffer (50 mM HEPES, pH 7.3, 50 mM KCl, 1 mM TCEP, 1 mM MgCl2 and 1 mg/ml BSA) was added into each well and incubated for 30 min at room temperature. Ultimately, His-tagged Ezrin at 40 nM in BSA buffer was added into each well and incubated for 1 hr at room temperature, followed by five washes with 500 μl of BSA buffer to remove unbound His-Ezrin. Preassembled actin filaments (90% unlabeled actin and 10% rhodamine-actin) was added into wells and incubated for 5 min at room temperature followed by 100 μl of image buffer addition (0.02 mg/ml catalase, 0.1 mg/ml glucose oxidase and 15 mM glucose in BSA buffer) to adjust the actin concentration to 100 nM. Immediately before imaging, Shi was added into a well to a final concentration of 10 nM and imaged by a TIRF/iLAS2 TIRF/FRAP module (Biovision) mounted on a Leica DMI6000 microscope base equipped with a 100x/1.49 oil objective lens and a Hamamatsu ImagEMX2 EM-CCD camera.
To monitor Shi-mediated actin bundling in the process of actin polymerization, experiments were performed in flow chambers prepared as previously described 62 except that piranha cleaning was replaced with UVO treatment. Briefly, microscope slides and coverslips (#1.5; Fisher Scientific) were washed for 30 min with acetone and for 10 min with 95% ethanol, sonicated for 2 hrs with Helmanex III detergent (Hellma Analytics), washed extensively with milliQ water and treated with UVO cleaner (Jetlight Company) for 10 min. Slides were then immediately incubated for 18 hrs with 1 mg/ml mPeg-Silane (5,000 MW) in 95% ethanol (pH 2.0). Parallel strips of double-sided tape were placed on a coverslip to create multiple flow chambers. To initiate the experiment, 1.5 μM Mg-ATP-actin (10% tetramethylrhodamine (TMR)-labeled) and 1 μM Shi was mixed with 2X TIRF buffer [1X TIRF buffer: 10mM imidazole (pH 7.0), 50 mM KCl, 1 mM MgCl2, 1 mM EGTA, 50 mM DTT, 0.2 mM ATP, 50 μM CaCl2, 15 mM glucose, 20 μg/ml catalase, 100 μg/ml glucose oxidase, and 0.2% (400 centipoise) methylcellulose] and transferred to a flow chamber for imaging. TIRF microscopy images were collected at 10-sec intervals using a Nikon Inverted Ti-E microscope base equipped with a 100x/1.49 oil objective lens and an Evolve EMCCD camera (Photometrics) with through-the-objective TIRF illumination.
Negative stain electron microscopy
Negative stain electron microscopy (EM) was performed to visualize the distribution of dynamin and mutants along actin bundles. 1 μM actin were incubated with the indicated concentrations of appropriate proteins in 6.5 mM HEPES, pH 7.3, 50 mM KCl, 2 mM MgCl2, 0.3 mM EGTA, 20 nM ATP, 20 nM CaCl2, 0.3 mM DTT and 0.001% NaN3 at 4°C overnight. The reaction mixture was loaded onto carbon-coated, glow discharged 400 mesh copper grids and incubated for 5 min, followed by two quick washes in actin polymerization buffer (50 mM KCl, 1 mM MgCl2, 1 mM EGTA and 10 mM imidazole). The grid was stained in two drops of 0.75% uranyl formate for 20 sec and air-dried. Electron micrographs were collected on a JEOL 1200-EX II transmission electron microscope.
To test the effects of GTP on dynamin-actin bundles by negative stain EM, 1 μM actin were incubated with the indicated concentrations of dynamin or mutants in the presence or absence of 1 mM GTP at room temperature for 30 min.
To visualize dynamin helices on lipid nanotubes, lipid tubes were generated with 40% phosphatidylcholine (1,2-dioleoyl-sn-glycero-3-phosphocholine, DOPC), 40% NFA-GalCer, 10% cholesterol and 10% PtdIns(4,5)P2 (to a total lipid concentration of 1 mM) (Avanti). 3 μM dynamin or a mutant was incubated with lipid nanotubes in lipid buffer (20 mM HEPES, pH 7.3, 150 mM KCl, 2 mM MgCl2 and 1 mM DTT) at 4°C overnight, followed by negative stain described above except that the wash buffer was done in the lipid buffer.
To visualize dynamin self-assembled rings, 0.2 mg/ml dynamin was dialyzed into 20 mM HEPES, pH 7.3, 25 mM KCl, 1 mM EGTA and 1 mM DTT at 4°C overnight. Samples were loaded onto grids and incubated for 1 min, followed by two quick washes with the same dialysis buffer and staining in two drops of 0.75% uranyl formate for 1 min.
Super resolution microscopy
To visualize the distribution of Shi along actin bundles generated by purified proteins, Shi-SNAP-Surface 488 was incubated with actin at different molar ratios ([Shi]: [actin] = 1:8, 1:1 or 2:1) and an equimolar concentration of Alexa568-conjugated phalloidin as actin with or without 1 mM GTP, in 6.5 mM HEPES, pH 7.3, 50 mM KCl, 1 mM MgCl2, 0.3 mM EGTA and 0.3 mM DTT for 2 hrs at room temperature. Subsequently, the samples were diluted 20-fold in fluorescence buffer (0.5% methycellulose, 50 mM KCl, 1 mM MgCl2, 1 mM EGTA, 10 mM imidazole, 100 mM DTT, 0.02 mg/ml catalase, 0.1 mg/ml glucose oxidase and 15 mM glucose, 0.1 mM CaCl2, 0.1 mM ATP and 0.005% NaN3). 4 μl of diluted samples were mounted onto a slide, covered with poly-L-lysine coated coverslip, and subjected to SIM imaging.
To visualize Shi localization along actin bundles at the fusogenic synapse, 1.2×106 S2R+ cells were seeded onto each well of a 6-well plate and transfected using Effectene (Qiagen) with either pAc-Eff1-HA (150 ng), pAc-Sns-iBlueberry-T2A-HO1 (150 ng) pUAST-Shi-GFP (50 ng) and pUbi-Gal4 (50 ng) for SIM imaging, or pAc-Eff1-FLAG (150 ng), pAc-Sns-HA (150 ng), pUAST-Shi-GFP (50 ng) and pUbi-Gal4 (50 ng) for STED imaging. Samples were washed twice with PBS supplemented with 1 mM CaCl2 and 0.5 mM MgCl2 (PBS++) at 36-48 hrs post transfection, then fixed with 4% paraformaldehyde diluted in cytoskeleton preservation buffer (80 mM PIPES pH6.8, 5mM EGTA, 2mM MgCl2) for 10 min at 37°C63. Samples were washed in PBS, permeabilized and incubated in PBSBT, followed by immunostaining. For SIM imaging, the following labeling reagents were used: rabbit anti-HA antibody (1:600; ThermoFisher, 715500), Alexa488- or Alexa647-conjugated goat anti-rabbit antibody (1:600; Invitrogen) and Alexa568-conjugated phalloidin (1:200; Sigma). For STED imaging, the following labeling reagents were used: rabbit anti-HA antibody (1:600; ThermoFisher, 715500), mouse anti-GFP antibody (1:600; DSHB, 8H11), CF405M-conjugated goat anti-rabbit antibody (1:600; Biotium), Star580-conjugated goat anti-mouse antibody (1:600; Abberior) and StarRed-conjugated phalloidin (1:200; Abberior).
To visualize hDyn2 localization along actin bundles on comet tails, 4.5×105 HeLa cells were seeded onto 25 mm glass coverslips (Warner Instruments, Cat 64-0715), deposited in 6-well plates the day before infection. GFP-expressing Listeria monocytogenes strain 10403s (provided by Dr. Dan Portnoy, UC Berkeley) was cultured in 3 ml of brain heart infusion (BHI) media (Gibco) at 30°C overnight without shaking until OD600 = 0.8 was reached. The following day, 1 ml Listeria culture was centrifuged at 16,000xg for 1 min, washed with PBS, and resuspended in a final volume of 1 ml PBS. HeLa cells were infected with Listeria at a multiplicity of infection (MOI) of 60, and infection synchronized by centrifugation for 10 min at 800g. The cells were then incubated at 37°C for 90 min, washed with PBS, followed by another incubation in fresh media containing 50 µg/ml gentamicin for 5 hours. HeLa cells were washed twice with PBS supplemented with 1 mM CaCl2 and 0.5 mM MgCl2 (PBS++), then fixed with 4% paraformaldehyde diluted in cytoskeleton preservation buffer (80 mM PIPES pH6.8, 5mM EGTA, 2mM MgCl2) for 10 min at 37°C63. Samples were washed in PBS, permeabilized and incubated in PBSBT, followed by immunostaining. The following labeling reagents were used: mouse anti-hDyn2 (1:50; Santa Cruz Biotechnology, G4), Star580-conjugated goat anti-mouse antibody (1:600; Abberior) and StarRed-conjugated phalloidin (1:200; Abberior).
SIM imaging was performed using a Nikon N-SIM E microscope (Ti-E; Nikon), equipped with a CFI SR Apochromat TIRF 100x/1.49 oil objective lens and an ORCA-Flash 4.0 sCMOS camera (Hamamatsu Photonics K.K.). Excitation of the fluorophores was performed with three lasers (488/561/640nm). STED imaging was performed using an Abberior Instruments Expert Line STED microscope (Abberior Instruments, Göttingen, Germany), equipped with an SLM based easy3D module, an Olympus IX83 microscope body and an UPLSAPO 100x/1.4 oil objective. Excitation of the fluorophores was performed with pulsed excitation lasers at 561 nm and 640 nm. A pulsed 775 nm laser (pulse width, ∼1.2 ns) was used for STED. Both lasers were synchronized and operated at 40 MHz. Time-gated detection of the fluorescence was performed at 605-625nm for the orange channel and 650-720 nm for the deep red channel. The images were processed using Adobe Photoshop CS6 and ImageJ (NIH) software.
Fluorescence recovery after photobleaching
To visualize the dynamics of Shi along actin bundles with or without GTP, actin bundles were prepared by incubating 1 μM Shi-SNAP-Surface 488 and 1 μM actin in 6.5 mM HEPES, pH 7.3, 50 mM KCl, 1 mM MgCl2, 0.3 mM EGTA and 0.3 mM DTT for 2 hrs at room temperature. Before imaging, 0.25 μM Shi-SNAP-Surface 488 in fluorescence buffer with or without 1 mM GTP were mixed with the samples. To perform FRAP assay, two pre-photobleaching images were recorded, followed by photobleaching of the regions of interest (ROI) and quickly bleached to ∼15% of its original intensity. Subsequently, images were acquired every 30 sec by a Nikon A1R confocal microscope (Ti-E; Nikon) equipped with a CFI SR Apochromat TIRF 100x/1.49 oil objective lens. The mean fluorescence intensity of the photobleached region was measured by ImageJ software and unbleached region was used to correct the curve.
GTPase activity assay
The GTPase activity of Shi was measured according to previously published methods64. For basal GTPase assay, 1 μM Shi was incubated in GTPase assay buffer (GAB-150; 20 mM HEPES, pH 7.5, 150 mM KCl, 2 mM MgCl2 and 1 mM DTT) with a series of GTP concentrations (25, 50, 100, 200, 400, 800 and 1200 μM) at room temperature. Samples were harvested from the reaction mixes at 5-min interval over a 30 min time period, and reactions were stopped by EDTA addition to a final concentration of 100 mM. Malachite green solution was added to each sample and the absorbance at 650 nm was determined using a microplate reader. Released Pi detected by malachite green was fitted to a linear curve vs. time course, and the GTP hydrolysis rate was calculated for each GTP concentration. Subsequently, the GTP hydrolysis rates were plotted against the initial concentrations of GTP to calculate the Michaelis-Menten constants Kcat and Km.
For actin-stimulated GTPase assay, 0.1 μM Shi was pre-incubated with 0.3 μM actin in GTPase assay buffer-50 (GAB-50; 20 mM HEPES, pH 7.5, 50 mM KCl, 2 mM MgCl2 and 1 mM DTT) for 1 hr to form stable Shi-actin bundles at room temperature. GTP was added to reaction mixes to reach a series of final concentrations of 10, 15, 25, 50, 100, 150 and 200 μM. Samples were harvested from the reaction mixes at 2-min interval over a 12 min time period, and processed as described above.
Western blot
Expression constructs were transfected in S2R+ cells. Cells were harvested, washed in PBS and lysed in lysis buffer (50mM Tris pH 7.5, 200 mM NaCl, 10 mM MgCl2, 1% NP-40, 0.5% Triton X-100) supplemented with 1x protease inhibitor cocktail (Sigma). After centrifugation, cell lysates were analyzed on an SDS-PAGE gel. The following antibodies were used in blotting: rabbit anti-HA (1:1000; ThermoFisher, 715500), mouse anti-tubulin (1:1000; DSHB, E7), chicken anti-GFP (1:5000; ThermoFisher; A10262), anti-rabbit-HRP (1;10,000; Invitrogen), anti-mouse-HRP (1;10,000; Invitrogen), and anti-chicken-HRP (1:20,000; Invitrogen).
Pulse-chase transferrin uptake assay
1.2×106 S2R+ cells were plated in each well of a 6-well plate, transfected with 100 ng of human TfR (and 100ng of rescue construct if applicable) using the Effectene (QIAGEN) according to manufacturer’s instructions. At 36 hrs post-transfection, cells were washed with PBS, incubated in PBS++++ (1mM CaCl2, 1mM MgCl2, 5mM Glucose and 0.2% BSA) for 30 min, and incubated with 5 µg/ml Alexa488- or Alexa568-conjugated transferrin (Tf) (Molecular Probes, T13342 and T23365) for 5 min. Subsequently, cells were placed on ice, washed three times with cold PBS, three times with acetic acid (0.2M acetic acid, 0.2M NaCl at pH 2.5) for 5 min each, three times again with cold PBS, and fixed with 4% formaldehyde, and stained for hTfR and HA. Stained cells were mounted and subjected to confocal imaging.
For image analysis, a custom-written ImageJ (Fiji) macro was used. In brief, for each image analyzed, transfected cells were identified and manually outlined, yielding the total surface area (TSA). Then, within these TSAs, fluorescent signals of interest were automatically determined using the triangular threshold method, and particles were analyzed though the build-in ImageJ algorithm. Relative Tf uptake was determined by Tf particle count divided by the TSA, followed by normalization to the relevant control within the dataset.
Cryo-electron tomography
Cryo-ET was used to determine the 3-dimensional organization of actin filaments bundled by dynamin. 5 μM actin was incubated with 0.625 μM of either hDyn1 or Shi in 20 mM HEPES, pH 7.3, 50 mM KCl, 1 mM EGTA, 80 nM ATP, 80 nM CaCl2, 1 mM DTT and 0.004% NaN3 at 4°C overnight. Aliquots of 3.5 μl of the overnight actin/dynamin reaction mixture with 10 nm gold fiducials was loaded onto glow-discharged gold C-flat grids (Electron Microscopy Sciences, CF-1.2/1.-4Au-50) for 20 sec, back-blotted on the sample-free side with filter paper for 2 sec (22 °C, 90% humidity) and then plunged into liquid ethane using the EM Grid Plunger (Leica Microsystems). Electron micrographs for the cryo-ET study were collected with a 200 kV TF20 transmission electron microscope (FEI) at 29,000x magnification, with a defocus of −3.0 μm using a K2 summit camera (Gatan). Tilt series was collected using SeriaEM at angles between ∼−60° and +60° for each data set. Tomograms were generated from the tilt series with the IMOD package65. Modelling of dynamin (EMDB code: EMDB-7957) and actin (provided by Dr. Krishna Chinthalapudi using the PDB entry 5ONV) into tomograms was done in Chimera66, and was also guided by segmented tomograms obtained using the EMAN2 semi-automated segmentation package67. The helical parameters of dynamin were calculated based on the diameter, pitch and known distance between dynamin subunits along the helical path.
Author contributions
S.K. initiated the project. R.Z., N.G., D.M.L., S.K and E.H.C. design the project, performed experiments and discussed the data. R.Z. carried out biochemistry, negative stain EM, and in vitro imaging experiments in Fig. 2 (except for 2E, F), Fig. 3 (except for 3F, G), Fig. 4 (except for 4G, H), Fig. 5A-C, Fig. 6A, Fig. S3, 4, 6, 7, 9 and 10. N.G. carried out fly genetics and cell culture experiments in Fig. 1A, B, G and H, Fig. 5G and H, Fig. S1, Fig. S2F, Fig. S8. D.M.L carried out cell culture and super resolution imaging experiments in Fig. 1D-F, Fig. 3F and G, Fig. 4G and H, Fig. 5D-F, Fig. S5, Fig. S11. S.K performed live imaging and immunostaining experiments in Fig. 1C and Fig. S2A-E. R.Z. and E.H.C. collaborated with J.J. and J.H on the cryo-ET experiments in Fig. 6 and Fig. S12, with J.W. and M.G. on TIRF imaging experiment in Fig. 2E and F, with R.K. and A.F. on experiments with human Drp1 in Fig. S10, and with J.D. and M.R. on live imaging of actin bundling in Fig. 2D. N.G., D.M.L. and E.H.C. collaborated with M.M. and S.S. on the endocytosis experiments in Fig. 5G, Fig. S8D and Fig. S11B. D.M.L. and E.H.C. collaborated with M.A. and N.A. on the actin comet tail experiments in Fig. 3G and Fig. S5B. D.L. and E.H.C. collaborated with G.Z. on the EM experiments in Fig. 1I. P.K. worked with D.M.L. in generating constructs used in Fig. 1F, Fig. 5D-F and Fig. S11B. P.P. worked with R.Z. on the negative stain EM experiments in Fig. 2–6, Fig. S3, S4 and S10. B.R. worked with R. Z. on the lipid tubule experiments in Fig. 2J, Fig. S4E and Fig. S9B. J.K. worked with N.G. on foci invasion measurement in Fig. 1G. R.Z., N.G., D.M.L. and E.H.C. made the figures. R.Z., D.M.L. and E.H.C. wrote the paper. All authors commented on the manuscript.
Data availability
The main data supporting the findings of this study are available within the article and its Supplementary Information files. All other data supporting the findings of this study are available from the corresponding author upon reasonable request.
Author information
The authors declare no competing financial interests.
Acknowledgements
We thank the Bloomington stock center for fly stocks, Dr. Bruce Paterson for the MHC antibody, Dr. Melissa Mikolaj for help with generating the model in iMOD, Drs. Haifeng He and Huaibin Wang for help with generating the cryo-tomograms using SerialEM, and Dr. Kishna Chinthalapudi for providing the actin filament for modeling in Chimera. This work was supported by NIH grants (R01 AR053173 and R01 GM098816), American Heart Association Established Investigator Award, and HHMI Faculty Scholar Award to E.H.C.; NIH grant (R01 AI083359), HHMI and Simons Foundation, and Welch Foundation (I-1704) to N.M.A.; NIH grant (R01 GM127673), Chan Zuckerberg Biohub Investigator Award, and HHMI Faculty Scholar Award to A. F.; NIH grant (R01 GM42455) and Welch Foundation (I-1823) to S.L.S. M.K.R is an HHMI investigator. R. Z. was supported by an American Heart Association postdoctoral fellowship; N. G. by an American Heart Association predoctoral fellowship; D.M.L. by a Canadian Institute of Health Research postdoctoral fellowship; and J.A.D by a National Research Service Award from NIDDK (F32 DK101188).