Actin ADP-ribosylation at Threonine148 by Photorhabdus luminescens toxin TccC3 induces aggregation of intracellular F-actin
2016; Wiley; Volume: 19; Issue: 1 Linguagem: Inglês
10.1111/cmi.12636
ISSN1462-5822
AutoresAlexander E. Lang, Zheng Qu, Carsten Schwan, Unai Silván, Andreas Unger, Cora‐Ann Schoenenberger, Klaus Aktories, Hans Georg Mannherz,
Tópico(s)Heat shock proteins research
ResumoCellular MicrobiologyVolume 19, Issue 1 e12636 Original articleFree Access Actin ADP-ribosylation at Threonine148 by Photorhabdus luminescens toxin TccC3 induces aggregation of intracellular F-actin Alexander E. Lang, Alexander E. Lang Institut für Experimentelle und Klinische Pharmakologie und Toxikologie, Albert-Ludwigs-Universität Freiburg, Freiburg, GermanySearch for more papers by this authorZheng Qu, Zheng Qu Abteilung für Anatomie und Molekulare Embryologie, Ruhr-Universität Bochum, Bochum, GermanySearch for more papers by this authorCarsten Schwan, Carsten Schwan Institut für Experimentelle und Klinische Pharmakologie und Toxikologie, Albert-Ludwigs-Universität Freiburg, Freiburg, GermanySearch for more papers by this authorUnai Silvan, Unai Silvan ETH Zürich, Institute for Biomechanics, University of Zürich, Balgrist Campus, Zürich, SwitzerlandSearch for more papers by this authorAndreas Unger, Andreas Unger Department of Cardiovascular Physiology, Ruhr-University Bochum, Bochum, GermanySearch for more papers by this authorCora-Ann Schoenenberger, Cora-Ann Schoenenberger Focal Area Structural Biology and Biophysics, Biozentrum, University of Basel, SwitzerlandSearch for more papers by this authorKlaus Aktories, Klaus Aktories Institut für Experimentelle und Klinische Pharmakologie und Toxikologie, Albert-Ludwigs-Universität Freiburg, Freiburg, Germany Centre for Biological Signalling Studies (BIOSS), University of Freiburg, Freiburg, Germany Freiburg Institute of Advanced Studies (FRIAS), University of Freiburg, Freiburg, GermanySearch for more papers by this authorHans Georg Mannherz, Corresponding Author Hans Georg Mannherz hans.g.mannherz@rub.de Abteilung für Anatomie und Molekulare Embryologie, Ruhr-Universität Bochum, Bochum, GermanyFor correspondence. E-mail hans.g.mannherz@rub.de; Tel. +49 234 3224553; Fax +49 234 3214474.Search for more papers by this author Alexander E. Lang, Alexander E. Lang Institut für Experimentelle und Klinische Pharmakologie und Toxikologie, Albert-Ludwigs-Universität Freiburg, Freiburg, GermanySearch for more papers by this authorZheng Qu, Zheng Qu Abteilung für Anatomie und Molekulare Embryologie, Ruhr-Universität Bochum, Bochum, GermanySearch for more papers by this authorCarsten Schwan, Carsten Schwan Institut für Experimentelle und Klinische Pharmakologie und Toxikologie, Albert-Ludwigs-Universität Freiburg, Freiburg, GermanySearch for more papers by this authorUnai Silvan, Unai Silvan ETH Zürich, Institute for Biomechanics, University of Zürich, Balgrist Campus, Zürich, SwitzerlandSearch for more papers by this authorAndreas Unger, Andreas Unger Department of Cardiovascular Physiology, Ruhr-University Bochum, Bochum, GermanySearch for more papers by this authorCora-Ann Schoenenberger, Cora-Ann Schoenenberger Focal Area Structural Biology and Biophysics, Biozentrum, University of Basel, SwitzerlandSearch for more papers by this authorKlaus Aktories, Klaus Aktories Institut für Experimentelle und Klinische Pharmakologie und Toxikologie, Albert-Ludwigs-Universität Freiburg, Freiburg, Germany Centre for Biological Signalling Studies (BIOSS), University of Freiburg, Freiburg, Germany Freiburg Institute of Advanced Studies (FRIAS), University of Freiburg, Freiburg, GermanySearch for more papers by this authorHans Georg Mannherz, Corresponding Author Hans Georg Mannherz hans.g.mannherz@rub.de Abteilung für Anatomie und Molekulare Embryologie, Ruhr-Universität Bochum, Bochum, GermanyFor correspondence. E-mail hans.g.mannherz@rub.de; Tel. +49 234 3224553; Fax +49 234 3214474.Search for more papers by this author First published: 24 June 2016 https://doi.org/10.1111/cmi.12636Citations: 19AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Summary Intoxication of eukaryotic cells by Photorhabdus luminescens toxin TccC3 induces cell rounding and detachment from the substratum within a few hours and compromises a number of cell functions like phagocytosis. Here, we used morphological and biochemical procedures to analyse the mechanism of TccC3 intoxication. Life imaging of TccC3-intoxicated HeLa cells transfected with AcGFP-actin shows condensation of F-actin into large aggregates. Life cell total internal reflection fluorescence (TIRF) microscopy of identically treated HeLa cells confirmed the formation of actin aggregates but also disassembly of F-actin stress fibres. Recombinant TccC3 toxin ADP-ribosylates purified skeletal and non-muscle actin at threonine148 leading to a strong propensity to polymerize and F-actin bundle formation as shown by TIRF and electron microscopy. Native gel electrophoresis shows strongly reduced binding of Thr148-ADP-ribosylated actin to the severing proteins gelsolin and its fragments G1 and G1–3, and to ADF/cofilin. Complexation of actin with these proteins inhibits its ADP-ribosylation. TIRF microscopy demonstrates rapid polymerization of Thr148-ADP-ribosylated actin to curled F-actin bundles even in the presence of thymosin β4, gelsolin or G1–3. Thr148-ADP-ribosylated F-actin cannot be depolymerized by gelsolin or G1–3 as verified by TIRF, co-sedimentation and electron microscopy and shows reduced treadmilling as indicated by a lack of stimulation of its ATPase activity after addition of cofilin-1. Introduction Many bacterial toxins excert their toxic effects by targeting the cellular cytoskeleton, in particular the actin cytoskeleton (for reviews see Aktories et al., 2011; Carlier et al., 2015). The bacterium Photorhabdus luminescens is mutualistically associated with entomopathogenic nematodes. The nematodes, carrying the bacteria, invade insect larvae where the released bacteria kill the host through the action of toxin complexes (Tc's) (Ciche et al., 2008; Waterfield et al., 2009). Recently, we have shown that TccC3 and TccC5 are the biologically active TcC components of the tripartite P. luminescens toxin complex (~1.7 megadalton), which consists of TcA, TcB and TcC components. TcA is the pentameric binding and membrane translocation component, whereas TcB is a functional linker between TcC and TcA and also involved in a novel mode of protein unfolding and translocation of the TcC components through the endosomal membrane into the cytoplasm (Lang et al., 2010). While TccC3 ADP-ribosylates actin at threonine148, TccC5 modifies Rho proteins at glutamine61/63 (Lang et al., 2010). Both modifications result in major alterations of the dynamic behaviour of the actin cytoskeleton characterized by increased polymerization and clustering of F-actin leading to an immediate inhibition of cellular activities like for instance of phagocytosis. The freezing of the actin cytoskeleton in the polymerized state probably leads to cell death as indicated by in vivo data (Lang et al., 2010; Sheets et al., 2011). In order to clarify the functional consequences of actin ADP-ribosylation at Thr148 by TccC3, we analysed the effects of this modification on its polymerization behaviour and interaction with a number of actin binding proteins. The intracellular actin cytoskeleton is constantly reorganized according to the functional activities of the cell. Existing actin filaments are disassembled to replenish the pool of polymerization-competent monomeric actin, from which new filaments are formed or existing ones elongated. This leads to a constant cycling of a large fraction of actin subunits between monomeric and filamentous states. Actin filaments are polar dynamic structures in that actin monomers are added to the fast growing plus ends, whereas at the same time an equal number of actin subunits dissociate from the minus ends. This process termed treadmilling (Wegner, 1976; Carlier et al., 2015) is particularly evident in plasma membrane extensions of migrating cells like lamellipodia and filopodia (Pollard and Borisy, 2003; Lai et al., 2008). Within these membrane organelles the plus ends of the actin filaments are oriented towards the plasma membrane and the constant incorporation of new actin molecules at the plus ends exerts the protrusive force for their forward extension (Pollard and Borisy, 2003; Lai et al., 2008). Actin treadmilling and filament assembly or disassembly are regulated by a large number of actin binding proteins (AbP), which either support filament formation (like the formins) or filament disassembly like gelsolin or the proteins of the ADF (actin depolymerization factor)/cofilin family (Pollard and Cooper, 1986; Andrianantoandro and Pollard, 2006; Goode and Eck, 2007; Bernstein and Bamburg, 2010). In a previous communication it was shown that ADP-ribosylation of actin at Thr148 by TccC3 strongly reduces its ability to interact with thymosin β4 (Lang et al., 2010). Thymosin β4 (Tβ4) is a small peptide (5 kDa) and by binding to monomeric actin blocks its ability to polymerize to filaments. Normally Tβ4 sequesters monomeric actin, i.e. removes it from the polymerization competent fraction (for reviews see Carlier and Pantaloni, 1994; Mannherz and Hannappel, 2009). However, an inhibition of the interaction of actin and Tβ4 should lead to an increase in the concentration of polymerizable actin (Mannherz and Hannappel, 2009; Al Haj et al., 2014). Here, we show by life cell imaging that within TccC3-intoxicated HeLa cells the actin filaments transiently form peripheral F-actin bundles, which are subsequently disassembled to aggregates of varying sizes. Therefore, we assumed that the cell lost the ability for an ordered regulation of actin cycling or teadmilling after its Thr148-ADP-ribosylation. Consequently, we analysed the interaction of AbPs known to disassemble F-actin or to sequester monomeric actin with Thr148-ADP-ribosylated actin. As typical representatives of proteins with F-actin depolymerization activity, we chose deoxyribonuclease I (DNase I) (Mannherz et al., 1980), gelsolin and some of its functional segments (Way et al., 1989; Nag et al., 2013), the proteins of the ADF/cofilin family (Yeoh et al., 2002), and thymosin β4 (Mannherz and Hannappel, 2009; Al Haj et al., 2014). Our data show that the interaction of Thr148-ADP-ribosylated actin with these proteins is strongly reduced and, furthermore, that polymerized Thr148-ADP-ribosylated actin is almost resistant to their F-actin depolymerizing activity. Results ADP-ribosylation of actin by TccC3 toxin from P. luminescens leads to intracellular F-actin aggregation Previously it was observed that HeLa cells round up and detach from the substratum within a few hours after intoxication with PTC3 (TcdA1 plus TcdB2-TccC3) toxin (Lang et al., 2010; Sheets et al., 2011). We have extended these observations by visualization the F-actin cytoskeleton of HeLa cells transfected with AcGFP-tagged cytoplasmic β-actin. Life imaging for 4 h revealed the alterations of the actin cytoskeleton occurring after TccC3 intoxication. Figure 1 (and Movie S1) shows a decrease of F-actin stress fibres and the appearance of prominent aggegates already after 60 min coexisting with stress fibres (early time points up to 60 min of stress fibre disassembly and aggregate formation are given in Fig. S1A). These aggregates enlarge and increase in number with ongoing time parallel to a further disassembly of the stress fibres first in the cell centre, followed by their complete degradation after about 2 h concomitant to cell shrinkage and rounding, starting at about 3.5 h (Movie S1). In addition, we observed after about 1 h a compaction of the F-actin fluorescence in the cell periphery and newly formed long and thick filopodia-like extensions suggesting F-actin bundling (at about 80 min see Movie S1) that were, however, subsequently also disassembled and most probably integrated into the enlarging aggregates (legend to Movie S1 also gives a semi-quantitative evaluation of the aggregate formation). Similar actin aggregate formation had been observed previously by TRITC-phalloidin staining of identically intoxicated HeLa cells suggesting that these aggregates consist of short or distorted actin filaments (Lang et al., 2010). In contrast, control cells showed the maintenance of intact stress fibres during the observation period. Life imaging of control cells indicated a slight membrane activity probably affected by filopodia and cell movement (see Movie S1). Figure 1Open in figure viewerPowerPoint Life cell imaging of HeLa cells after intoxication with PTC3. HeLa cells transfected with pAcGFP1-actin were treated with or without the toxin complex PTC3 (TcdA1 + TcdB2-TccC3). Changes of the actin cytoskeleton were visualized by confocal time-lapse microscopy. Scale bar, 10 µm. In addition, we examined HeLa cells transfected with AcGFP-tagged cytoplasmic β-actin by life cell TIRF microscopy after intoxication with PTC3 that allowed to visualize alterations of the actin cytoskeleton in close proximity to regions adherent to the substratum (Fig. S1B). Again actin aggregate formation was clearly visible after 8 min (arrows in Fig. S1B). Furthermore, the data showed that about 90 min after intoxication shortening of F-actin filaments occurred indicating their disassembly (see arrows in Fig. S1B). These data suggested that TccC3 intoxication leads to stress fibre degradation and finally to aggregation of the intracellular actin. The critical concentrations of polymerization of native and Thr148-ADP-ribosylated actin are similar In view of the strong propensity of Thr148-ADP-ribosylated actin to form intracellular aggregates, we first tested whether thus modified purified actin retains its native state. Purified skeletal muscle actin was Thr148-ADP-ribosylated by recombinant TccC3 toxin and tested for its DNase I inhibitory capacity as an indicator for its native state (Mannherz et al., 1980). The comparison to native actin gave an identical inhibitory capacity of Thr148-ADP-ribosylated actin (Fig. 2A), indicating retention of the native state after this modification. Similarly, its ability to bind ATP was preserved as tested by its exchange for fluorescent etheno-ATP (see later). Figure 2Open in figure viewerPowerPoint Determination of DNase I inhibition and the critical concentrations of polymerization of native or TccC3-ADP-ribosylated actin in the presence of Tβ4. A. Inhibition of the DNase I activity by native and Thr148-ADP-ribosylated actin. DNase I at 5.4 μM was preincubated with the actins, set to the concentrations indicated for 30 min at room temperature. Then, the DNase I activity of a 10 µl aliquot was determined by using the hyperchromicity test (14). DNase I activity is traditionally expressed in Kunitz units (KU): 1 KU = ΔOD260 nm of 0.001/min (Kunitz, 1950). B. Determination of the critical concentrations of polymerization of native and Thr148-ADP-ribosylated actin in the absence or C. presence of a constant amount of Tβ4 (10 μM) using pyrene-labelled actin as described (Al Haj et al., 2014). Because we suspected that Thr148-ADP-ribosylated actin possesses a different critical concentration of polymerization (Cc), we determined the Cc by using pyrene-labelled actin as fluorescent indicator for polymerization. The results obtained indicated a Cc for Thr148-ADP-ribosylated actin almost identical to native actin (0.35 μM) (Fig. 2B) as shown before (Lang et al., 2010). In the presence of 5 μM thymosin β4, the Cc for native actin increased to about 1 μM (Ballweber et al., 1997), whereas it remained unchanged for the modified actin (Fig. 2C), indicating in agreement with previous data a greatly reduced affinity of Tβ4 to Thr148-ADP-ribosylated actin (Lang et al., 2010). These experiments have been repeated many times (> 10×) with different actin preparations and always showed that Thr148-ADP-ribosylation of actin mainly affected its interaction with particular actin binding proteins but not its general polymerization behaviour. Thr148-ADP-ribosylation inhibits actin interaction with a number of binding proteins Intracellularly existing actin filaments are rapidly depolymerized by the activity of F-actin severing proteins replenishing the monomeric actin pool. Because the life imaging experiments suggested that TccC3 intoxication leads to F-actin bundling and actin aggregation, we examined the interaction of Thr148-ADP-ribosylated actin with a number of AbPs with F-actin depolymerizing or severing activity like ADF/cofilin and gelsolin and its N-terminal segment G1 and N-terminal half G1–3. Effective Thr148-ADP-ribosylation of actin was verified by native gel electrophoresis (NGE) and subsequently, we used NGE to compare the binding of these AbPs to Thr148-ADP-ribosylated and control actin. The results clearly show a strongly reduced ability of the ADP-ribosylated actin to bind these proteins as shown for G1, G1–3, ADF (Fig. 3A–C) and cofilin-1 (not shown). Figure 3Open in figure viewerPowerPoint Interaction of native or TccC3-ADP-ribosylated actin with actin binding proteins (AbPs). Native gel analysis of native and TccC3-ADP-ribosylated actin incubated with gelsolin domain G1 (A), gelsolin domain G1–3 (B) and ADF (C). Rabbit muscle α-actin was treated with the N-terminal ADP-ribosyltransferase domain of TccC3 (0.16 µg/µg of actin). Then, actin was incubated with the indicated AbPs for 20 min at 21°C and subjected to native gel electrophoresis. Note, ADP-ribosylated actin migrates faster in native gels. (D) Chemical cross-linking of native (lanes 1–3) and Thr148-ADP-ribosylated actin (lanes 4–6) with Tβ4 (lane 2 and 5) and profilin (lane 3 and 6) using EDC. We also included profilin in this study, which is known to bind to the same target area as G1 or ADF (Kabsch et al., 1990; Ballweber et al., 1997; Dominguez and Holmes, 2011). Because profilin does not enter the NGE-gel, it was not possible to obtain clear data of profilin-binding to the actins by NGE. Therefore, we analysed this interaction by chemical cross-linking using EDC and Tβ4 as control. The data (Fig. 3D) show a clear reduction in the amount of Tβ4 cross-linked to Thr148-ADP-ribosylated versus control actin, but no difference in the amount of cross-linked profilin. TccC3-catalysed ADP-ribosylation of actin is inhibited by gelsolin and ADF/cofilin Threonine148 is localized at the entrance of the cleft between subdomains 1 and 3 of actin, which forms a hydrophobic binding area for many actin binding proteins like gelsolin, its fragments G1, G1–3, profilin and ADF/cofilin (Dominguez and Holmes, 2011; see also Fig. S2). Therefore, we tested whether Thr148-ADP-ribosylation of actin occurred irrespective of complexation of actin with these AbPs, because during the dynamic intracellular cycling actin might transiently interact with these AbPs. Most of these experiments were performed with isolated skeletal muscle actin, because initial experiments had demonstrated that both α-skeletal and β-cytoplasmic actin were equally well ADP-ribosylated by TccC3 irrespective of their state of polymerization. Indeed, polymerization to F-actin did not inhibit Thr148-ADP-ribosylation (Fig. 4A). Figure 4Open in figure viewerPowerPoint TccC3-mediated in vitro ADP-ribosylation of actin bound to AbPs. A. To test substrate specificity, 2 μM of monomeric G- or F-actin from each isoform was incubated with 60 nM of TccC3hvr and radioactive [32P]NAD+ (0.5 μCi per sample) for 10 min at 21°C. Proteins were detected by SDS-PAGE and radiolabelled actin was visualized by autoradiography (AR). B. α-Actin (2 μM) was incubated with 4 μM of indicated ABPs for 20 min at 21°C. Then, in vitro ADP-ribosylation was started with the addition of radioactive [32P]NAD+ and 60 nM of TccC3hvr. After indicated time points, reaction was stopped by addition of SDS-containing Laemmli buffer and samples were subjected to SDS-PAGE and autoradiography (AR). C. α-Actin (2 μM) was incubated with indicated concentrations of ABPs for 20 min at 21°C. In vitro-ADP-ribosylation with TccC3hvr was performed as before and reaction was stopped after 10 min. Coomassie-stained SDS-PAGE gels (CO) were then subjected to autoradiography (AR). Subsequently, G-actin was incubated with TccC3 after complexation with intact gelsolin and its deletion constructs G1, G1–2, G1–3 and G2–6, furthermore with ADF. The data shown in Fig. 4B give the time dependence of the covalent actin modification by TccC3 (shown by autoradiography in Fig. 4) and demonstrate a clear inhibition of Thr148-ADP-ribosylation of G-actin, when complexed to intact gelsolin and its deletion fragments, containing the N-terminal segment G1. In contrast, G2–6 devoid of G1 did not inhibit Thr148-ADP-ribosylation (Fig. 4B). ADP-ribosylation was also inhibited by ADF (Fig. 4B), but not by DNase I (not shown), which binds opposite to this general target area, i.e. to the so-called D-loop of subdomain 2 (Kabsch et al., 1990). Of note, profilin was also found not to inhibit Thr148-ADP-ribosylation of actin, probably because it binds to the 'back part' of this cleft (see Fig. S2). In contrast, the ADP-ribosylation-inhibiting AbPs bind to the 'front part' of this target zone (Dominguez and Holmes, 2011; see Fig. S2). These data strongly suggest that the AbPs, attaching to the 'front part' of the cleft between subdomains 1 and 3, are able to inhibit Thr148-ADP-ribosylation, i.e. they apparently compete with TccC3 for actin-binding. Testing the concentration dependence of the inhibition by intact gelsolin or its N-terminal segment G1 revealed maximal inhibition of ADP-ribosylation when their concentration of actin binding sites equalled that of actin (Fig. 4C). In addition, we investigated the ability of gelsolin and G1 to inhibit actin ADP-ribosylation by the Clostridium difficile transferase CDTa, another bacterial ADP-ribosyltransferase, which modifies Arg177 of actin (Aktories et al., 1986; Aktories, 1994). CDTa was able to ADP-ribosylate actin even in the presence of equimolar concentration of gelsolin or G1 or a high excess of Tβ4 (supplementary Fig. S3), demonstrating that the inhibition of the TccC3-catalysed ADP-ribosylation by these AbPs is site-specific for Thr148-ADP-ribosylation. TIRF microscopy of the polymerization of Thr148-ADP-ribosylated actin We visualized the polymerization of Thr148-ADP-ribosylated actin by total internal reflection fluorescence (TIRF) microscopy (see Materials and methods). Addition of 2 mM MgCl2 and 50 mM KCl to 5 μM Thr148-ADP-ribosylated actin induced the rapid formation of actin filaments subsequently followed by the appearance of F-actin bundles of varying sizes, which sometimes became suddenly visible probably because of their rapid sedimentation into the excitation (evanescence field) zone. These bundles were often curled or circular and of variable diameter (Fig. 5). This behaviour differed from native actin, which after a lag phase formed single actin filaments, which further elongated and exhibited wave-like movements (Fig. 5). Figure 5Open in figure viewerPowerPoint ADP-ribosylation by TccC3 impairs the effects of ABPs on actin polymerization. Time-lapse TIRF microscopy of the effects of TccC3 on actin polymerization. Native or TccC3-modified α-actin (5 μM) together with 10% Alexa-Fluor 488-labelled α-actin was incubated or not with the AbPs Tβ4 (10 μM), gelsolin G1–3 (5 μM) or with intact gelsolin (3.66 μM). Polymerization was induced by addition of polymerization buffer resulting in 50 mM KCl and 1 mM MgCl2. Images were acquired at indicated time points. Next, we tested by TIRF whether the AbPs, which stabilise or bind monomeric actin (see above), are able to prevent the polymerization of Thr148-ADP-ribosylated G-actin. Previous data had indicated that the polymerization of Thr148-ADP-ribosylated actin cannot be inhibited by Tβ4. This effect was also verified by TIRF microscopy (Fig. 5). The data show that even at a high molar excess over Thr148-ADP-ribosylated actin Tβ4 was unable to block actin polymerization or bundle and curl formation (Fig. 5). Similarly, intact gelsolin and its N-terminal half G1–3 were also unable to inhibit polymerization of TccC3-modified actin. Only DNase I, which binds to a different region of actin, inhibited the Mg2+-ion induced polymerization of both native and Thr148-ADP-ribosylated G-actin (Fig. S4). Thr148 ADP-ribosylated actin filaments cannot be depolymerized by gelsolin Next, we analysed by TIRF microscopy the ability of intact gelsolin or its N-terminal half comprising domains 1–3 (G1–3) to depolymerize Thr148-ADP-ribosylated F-actin. To this end, the polymerization of 1 μM Thr148-ADP-ribosylated actin was first followed by TIRF and after 12 min intact gelsolin or its fragment G1–3 was added. As control, native actin was used under identical conditions, which at these concentrations polymerized with almost equal speed as the modified actin. Figure 6A shows that addition of G1–3 to preformed native F-actin led to its rapid and complete depolymerization. In contrast, addition of G1–3 to preformed filamentous Thr148-ADP-ribosylated actin did not lead to appreciable depolymerization (Fig. 6A). This resistence was not dependent on bundle formation, because most of the TccC3-F-actin under these conditions (1 μM) appeared as single filaments of equal diameter as native F-actin. In fact, a large F-actin bundle became only suddenly visible presumably after its sedimentation into the evanescence field (at time point 16 min; Fig. 6A) possibly because of its retarded formation under these conditions. Figure 6Open in figure viewerPowerPoint Influence of TccC3 on the severing activity of gelsolin. A. Native or TccC3-modified α-actin (1 μM) was polymerized together with 10% Alexa-Fluor 488-labelled α-actin by the addition of MgCl2-containing buffer. Asterisk indicates addition of gelsolin domain G1–3 (1.15 μM). Time-lapse TIRF microscopy pictures were made at indicated time points (please note different time scale than in Fig. 5). B. Native or TccC3-modified α-actin was polymerized to F-actin. Then, equal amount of gelsolin was added followed by ultra-centrifugation at 100 000 g for 60 min. Then, supernatants and pellets were subjected to SDS-PAGE. Note, gelsolin-severed F-actin remains in the supernatant (S), whereas TccC3-F-actin is found in the pellet (P). C. EM after negative staining of native or TccC3-ADP-ribosylated F-actin before and after addition of gelsolin domain G1–3. Scale bars, 100 nm. These results were confirmed by co-sedimentation, in which native and Thr148-ADP-ribosylated F-actin were incubated either without or with intact gelsolin. Analysis of the supernatants and pellets by SDS-PAGE after high-speed sedimentation showed that in the absence of gelsolin most of the native F-actin pelleted, whereas actin preincubated with gelsolin remained in the supernatant together with gelsolin (Fig. 6B; molecular mass of about 84 kDa). In contrast, most of the Thr148-ADP-ribosylated F-actin pelleted under both conditions (Fig. 6B). Thus, these data again demonstrated the inability of gelsolin or G1–3 to sever the Thr148-ADP-ribosylated F-actin probably because of the reduced ability of its N-terminal segment G1 to bind modified F-actin (see also Fig. 3) and to initiate its fragmentation (McLaughlin et al., 1993). These results were further verified by EM, comparing polymerized native and Thr148-ADP-ribosylated actin and the effect of G1–3. Electron microscopy of negatively stained samples verified the presence of bundles of polymerized Thr148-ADP-ribosylated actin after initiation of polymerization by addition of 2 mM MgCl2, which in contrast to native F-actin were not depolymerized by gelsolin or G1–3 after an incubation period of 30 min (shown for G1–3 in Fig. 6C). We cannot fully explain why in the presence of G1–3 we did not observe by EM the short, clearly defined G1–3:TccC3-F-actin oligomers seen by TIRF (Fig. 5). The difference might be because of the different experimental conditions and techniques (EM versus TIRF). In the experiments shown in Fig. 6, G1–3 was added to preformed TccC3-F-actin, whereas in those of Fig. 5, G1–3 was first co-incubated with TccC3-G-actin before initiating polymerization. Thr148-ADP-ribosylated F-actin shows reduced treadmilling which is not stimulated by cofilin-1 Next, we tested whether polymerized Thr148-ADP-ribosylated actin was 'frozen' in a polymerized state unable to treadmilling. During F-actin-treadmilling ATP-containing actin subunits are constantly added to the plus ends, which after ATP-hydrolysis travel through the filament and dissociate from the minus ends (Wegner, 1976). Because of the high ATP concentration in the cytoplasm, the bound ADP of the dissociated actin subunits is rapidly exchanged for ATP, recharging the actin for a new association to plus ends. The rate of the treadmilling process can be followed by measuring the rate of ADP exchange by the use of the fluorescent ADP-analogue etheno-ADP (Waechter and Engel, 1975). To this end, the ATP of monomeric native and TccC3-treated actin was exchanged for etheno-ATP as described (Waechter and Engel, 1975). After polymerization, both actins contain etheno-ADP (Yeoh et al., 2002). Subsequently, the rates of etheno-ATP or etheno-ADP release from 5 μM monomeric or filamentous actins, respectively, were determined by measuring the decrease in fluorescence after addition of a molar excess of ATP (1 mM). The data compiled in Fig. 7 show that etheno-ATP is slightly slower released from TccC3-modified actin than from native G-actin (with observed rat
Referência(s)