ATP-dependent membrane assembly of F-actin facilitates membrane fusion.

We recently established an in vitro assay that monitors the fusion between latex-bead phagosomes and endocytic organelles in the presence of J774 macrophage cytosol (). Here, we show that different reagents affecting the actin cytoskeleton can either inhibit or stimulate this fusion process. Because the membranes of purified phagosomes can assemble F-actin de novo from pure actin with ATP (), we focused here on the ability of membranes to nucleate actin in the presence of J774 cytosolic extracts. For this, we used F-actin sedimentation, pyrene actin assays, and torsional rheometry, a biophysical approach that could provide kinetic information on actin polymerization and gel formation. We make two major conclusions. First, under our standard in vitro conditions (4 mg/ml cytosol and 1 mM ATP), the presence of membranes actively catalyzed the assembly of cytosolic F-actin, which assembled into highly viscoelastic gels. A model is discussed that links these results to how the actin may facilitate fusion. Second, cytosolic actin paradoxically polymerized more under ATP depletion than under high-ATP conditions, even in the absence of membranes; we discuss these data in the context of the well described, large increases in F-actin seen in many cells during ischemia.

Characterization of a transporting system in rat hepatocytes. Studies with competitive and non-competitive inhibitors of phalloidin transport.

Primary cultures of rat hepatocytes were used for assaying several drugs not previously known for inhibiting the transport of phalloidin. In order to have 50% inhibition (IC50) of the entrance of a tritiated phallotoxin derivative ([3H]demethylphalloin, 1 microM) from the medium into the cells the following concentrations (microM) of the various inhibitors were determined: cyclolinopeptide (0.5), Nocloprost (5.0), Nileprost (7.0), beta-estradiol (42), Verapamil (70). For comparison, the corresponding IC50 values of some known antagonists of phalloidin toxicity were determined by the same method. Moreover, we studied several natural and synthetic phallotoxins and alpha-amanitin for their ability to displace [3H]demethylphalloin from the transporting system. Lineweaver-Burk plots made it obvious that two groups of inhibitors exist. Competitive inhibitors are, for example, antamanide, beta-estradiol, silybin, Nileprost, taurocholate, and the cyclic somatostatin analog cyclo[Phe-Thr-Lys-Trp-Phe-D-Pro], whereas Verapamil and monensin inhibit phallotoxin uptake in a non-competitive way. Considering the very different chemical features of the competitive inhibitors, we tentatively conclude that the phallotoxin transport system selects compounds not on the basis of their chemical features, but rather their physical properties. The physical properties of a typical substrate are low molecular mass, lipophilic nature, and, possibly the presence of rigid ring structures. Negative charges accelerate the transport of a substrate, while positive charges have the opposite effect. The phalloidin-transporting system may represent part of a hepatic equipment which clears portal blood from, for example, bile acids, lipophilic hormones, or xenobiotics. By chance, the transporting system incorporates phallotoxins into the hepatocytes leading to the death of these cells.

Membrane damage of liposomes by the mushroom toxin phallolysin.

We have investigated the membrane-damaging effect of phallolysin on liposomes varying in phospholipid composition, net charge and physical constitution. Liposomes were prepared from lipids extracted from bovine or human erythrocyte ghosts. The liposomes composed of bovine lipids (the intact cell showing little sensitivity to phallolysin) were found comparably sensitive to those prepared from lipids of human red cells (these cells being of high sensitivity). In addition, artificial mixtures of lipids were used for the preparation of liposomes, consisting of (a) negatively charged phospholipids such as dicetyl phosphate or phosphatidylserine, (b) cholesterol, and (c) either sphingomyelin (as the major component of erythrocytes from ruminants) or phosphatidylcholine (as the major component of erythrocytes from non-ruminants). Again, we found only little difference in the susceptibilities of sphingomyelin- and phosphatidylcholine-containing liposomes. On the other hand, the susceptibility depended on the presence of phospholipids with negative net charges. Omittance of phosphatidylcholine or dicetyl phosphate, or replacement by the positively charged stearylamine, decreased the susceptibility by a factor of more than 20. Finally, we prepared liposomes from dicetyl phosphate, cholesterol and phosphatidylcholine in two physical states: large unilamellar and smaller multilamellar liposomes. The unilamellar liposomes were about 10-times more sensitive to phallolysin. We conclude: (1) Phallolysin damages phospholipid-membranes in the absence of receptor proteins, but high concentrations of the toxin are required. (2) Membrane damage takes place with liposomes containing phosphatidylcholine as well as those containing sphingomyelin. (3) Phallolysin damages only liposomes containing phospholipids with a negative net charge.

Glutathionyl(cysteine-374) actin forms filaments of low mechanical stability.

Rabbit muscle actin reacts with 2,4-dinitrophenylglutathionyldisulfide, forming a mixed disulfide in position 374. the product S-(cysteine-374)glutathionyl actin forms filaments which are easily disrupted under shearing stress. Even weak mechanical strain, as exerted, for example, during capillary viscometry or heating the solution to 37 degrees C, leads to considerable breakage of these filaments. Because of spontaneous repair which consumes ATP, the mechanically broken filaments exhibit an approx. 6-fold enhanced steady-state ATPase activity as compared to normal F-actin. Monomers of glutathionyl actin have a reduced affinity for their bound nucleotide and a slightly increased critical concentration. Disruption of the filaments and enhanced ATPase activity are reversed by the addition of KCl or the mushroom toxin phalloidin. By the large stabilizing effects of KCl and phalloidin on glutathionyl actin filaments we propose glutathionyl actin as a tool for detecting filament-stabilizing agents and for studying the different mechanisms of filament stabilization.

Evaluating atomic models of F-actin with an undecagold-tagged phalloidin derivative.

We have prepared an undecagold-tagged phalloidin derivative to determine this mushroom toxin's binding site and orientation within the F-actin filament by scanning transmission electron microscopy (STEM) and 3-D helical reconstruction. Remarkably, when stoichiometrically bound to F-actin, the undecagold moiety of the derivative could be directly visualized by STEM along the two half-staggered long-pitch helical strands of single filaments. Most importantly, the structural data obtained when combined with various biochemical constraints enabled us to critically evaluate two distinct atomic models of the F-actin filament (i.e. the Holmes-Lorenz versus the Schutt-Lindberg model). Taken together, our data are in excellent agreement with the Holmes-Lorenz model.

A beta-turn in alpha-amanitin is the most important structural feature for binding to RNA polymerase II and three monoclonal antibodies.

Four amatoxin-binding proteins with KD values in the nanomolar range, 3 monoclonal antibodies and RNA polymerase II, were studied with respect to their affinities to 24 alpha-amanitin derivatives with modified side chains. From KD values we estimated the amounts of binding energy that single side chains of the amatoxins contribute to complex formation. Ile6, previously identified by X-ray analysis to be part of a beta-turn (Kostansek EC, Lipscomb WN, Yocum RR, Thiessen WE, 1978, Biochemistry 17:3790-3795) proved to be of outstanding importance in all complexes. Replacement of the isoleucine with alanine reduced the affinity to all binding proteins to < 1%, suggesting a strong hydrophobic interaction. A strong effect was also seen when Gly5 was replaced with alanine, suggesting that the absence of a side chain in proximity to the beta-turn is likewise important. In addition to the beta-turn, each of the proteins showed at least 2 other points of strong contact formed by hydrogen bonds. Donors are the indole NH of 6'-hydroxy-Trp4 and OH of hydroxy-Pro2 and dihydroxy-Ile3. All the antibodies, but not RNA polymerase II, recognized the indole nucleus of 6'-hydroxy-Trp4. The geometric arrangement of the 4 strongest contact points suggests that the amatoxin binding site is different in each of the 4 proteins, except for the 2 antibodies raised in the same animal. Here, most of the contact points were identical but differed in strength of interaction. The method of structural analysis presented in this study is useful for identifying contact sites in complexes of proteins with peptides of rigid conformation. Furthermore, the method complements X-ray data by providing information on the amount of binding energy contributed by single structural elements.

The mechanism of cytolysis of erythrocytes by the mushroom toxin phallolysin. Morphological and biochemical evidence for sodium influx and swelling.

Phallolysin, a basic protein of multiple molecular forms, was isolated from the mushroom Amanita phalloides. The toxic protein causes cytolysis in most mammalian cells at concentrations as low as 10(-8) M. Different from the cytolytic activities of detergents or sea-anemone toxin, cytolysis by low concentrations of phallolysin occurs only after a lag time of 2 to 3 minutes. In erythrocytes it was shown that phallolysin causes a rapid influx of sodium ions accompanied by a loss of potassium ions. Simultaneously, the cells swell up to their critical volume until they burst. In morphological studies the disruption of red blood cells following phallolysin treatment was found to be indistinguishable from that caused by incubation in hypotonic medium. Freeze-etching studies of phallolysin-treated erythrocytes showed changes in the membrane structure and a decrease of intramembranous particles. It is suggested that phallolysin either forms or opens preformed cation channels in the plasma membrane resulting in cation leakage. Concomitantly with the ion fluxes, water penetrates and causes the cells to swell until lysis occurs.

Phalloidin depletes the mitochondrial Ca2+ compartment of hepatocytes.

In 3 h of incubation, primary cultures of rat hepatocytes attach to the substratum and exchange about 2.2 nmol 45Ca2+ per mg protein. In the presence of 1 microM phalloidin, the exchanged amount of 45Ca2+ was found to be decreased by about 30%. Using the uncoupling agent FCCP and the ionophore A23187 for further characterisation we determined that the 45Ca2+ deficit caused by phalloidin occurs in the FCCP-sensitive compartment, i.e., the mitochondria.

Interchain and intrachain crosslinking of actin thiols by a bifunctional thiol reagent.

From 1,9-nonylenedithiol and Ellman's Reagent the bifunctional asymmetric disulfide n-nonylene-1,9-bis-[5-dithio-(2-nitrobenzoic acid)] (NBDN) was prepared. By monovalent reaction with cysteine-374 the crosslinker could be introduced into monomeric actin, with release of one equivalent of yellow 2-nitro-5-thiobenzoate (NTB). From the monovalent actin derivative we prepared a crosslinked actin dimer (Cys-374-Cys-374') as well as a monomer with a crosslink between Cys-374 and Cys-10. Neither crosslinked actin species was able to polymerize the crosslinked monomer even in the presence of phalloidin. The crosslinked monomer polymerized on the addition of dithiothreitol, thus providing the first unpolymerizable actin species whose polymerizability can be restored under mild conditions. We suggest the use of NBDN as a thiol-specific crosslinker that reacts under spectrophotometric control and can be removed by the addition of thiols.

Probing the phalloidin binding site of actin.

Phallotoxins form tight complexes with filamentous actin and stabilize the polymer against shearing stress. In the present study a phalloidin derivative containing a thiol-capturing moiety was prepared and reacted with single thiol groups of monomeric muscle actin. Sites of attachment in the protein were Cys-374 next to the C-terminus and Cys-10, close to the N-terminus; the latter was recently shown to be uncovered during a slow but reversible conformational transition occurring in ADP-G-actin. Phalloidin bound to Cys-374 stabilizes filaments against shearing stress almost as effectively as free phalloidin, indicating that the phalloidin binding site cannot be far from the C-terminus of actin. Stabilization was also achieved when the phalloidin reagent was added to F-actin, however, the subsequent formation of a covalent linkage with Cys-374 was not observed, most likely due to a restricted mobility of the reactants. In contrast to the efficient stabilization of filaments by phalloidin linked to Cys-374 a destabilizing effect was observed when phalloidin was attached to Cys-10. It appears that phalloidin located close to the N-terminus is unable to bind to the normal binding site in its own filament. Pronounced gelification of this actin derivative suggests that the toxin is able to mediate crosslinking with neighbouring filaments. From these results we conclude that the phalloidin binding site of actin is distant from the N-terminus, but close to the C-terminus. Furthermore, the data provide evidence that binding of phalloidin reduces the mobility of the C-terminus.

The molecular mechanism of interaction of Et3Pb+ with tubulin.

Triethyllead ion (Et3Pb+) was found to interact with 2 out of 18 thiol groups present in tubulin dimers. Specificity of the interaction was shown by the high affinity of Et3Pb+ to tubulin, by the fact that the 16 residual thiol groups in tubulin remained unaffected, and by the observation that other proteins with exposed thiol groups, e.g., actin, did not react with Et3Pb+. After complexation of the two thiol groups, tubulin in vitro had lost its capability for microtubule assembly. Likewise, polymerized tubulin disassembled on addition of the lead compound.

Polymerization of actin from the thymosin beta 4 complex initiated by the addition of actin nuclei, nuclei stabilizing agents or myosin S1.

Thymosin beta 4 forms a 1:1 complex with actin and thereby prevents polymerization. Rapid formation of filaments from this complex was observed, however, when actin trimers were added. Polymerization can likewise be initiated by the addition of one equivalent of phalloidin or, less effectively, cytochalasin B. Since both toxins, which reportedly support nucleation, have similar effects as the covalently linked actin trimers, it appears that the formation of filaments from the actin-thymosin beta 4 complex depends on the availability of stable actin nuclei. Remarkably, rapid polymerization was also observed if small amounts of myosin S1 were added, suggesting that also myosin, a protein functionally connected with polymeric actin, can serve as a nucleation center. Considering the existence of thymosin beta 4 and related peptides in numerous mammalian tissues, our data suggest that spontaneous formation of microfilaments in non-muscle cells may be regulated at the level of nucleation. Uncontrolled polymerization induced by the formation of phalloidin-stabilized nuclei may explain the acute toxic effects of phalloidin in hepatocytes.

Use of bimanyl actin derivative (TMB-actin) for studying complexation of beta-thymosins. Inhibition of actin polymerization by thymosin beta 9.

By reacting trimethylammoniobromobimane bromide (TMB bromide) with rabbit muscle actin, a fluorescent reporter group was linked to cysteine at position 374. Fluorescence of TMB-actin decreased significantly on addition of thymosin beta 4 (T beta 4), a peptide of 43 amino acid residues reported to bind to monomeric actin and to prevent filament formation. Based on this effect, we determined the KD value of the thymosin beta 4 complex as 0.8 microM, a value that is in agreement with previous determinations. In addition to the main compound thymosin beta 4, bovine tissue contains a related peptide, thymosin beta 9 (T beta 9), which has 41 amino acid residues and ca. 75% sequence homology. In the present study we show for the first time that T beta 9, similar to T beta 4, forms a 1:1 complex with monomeric actin, and hereby inhibits actin polymerization. With a KD value of 1.1 microM the affinity of T beta 9 is in the same range as that of T beta 4, suggesting that T beta 9, like T beta 4, contributes to maintaining the pool of monomeric actin in bovine non-muscle cells. Further proof of the interaction of T beta 9 with actin was provided by native PAGE, where the complex showed the reported higher mobility, as well as by crosslinking experiments. Using different crosslinking reagents, like water-soluble carbodiimide (EDC), m-maleimidobenzoyl-N-hydroxysuccinimidate (MBS), and disuccinimidylsuberate (DSS), we were able to produce conjugates of 47 kDa. In one of these (from MBS) both actin and T beta 9 could be identified by immunoblotting. When, in the MBS crosslinking experiments, native actin was replaced with (374-NEM)-actin, the 47 kDa band was not seen, indicating that Cys-374 takes part in the thiol-specific crosslinking reaction. This suggests that part of the binding site of T beta 9 must be located close to the carboxy-terminus.

The ternary complex of DNase I, actin and thymosin beta4.

We have recently described a method for identifying contact sites between actin and thymosin beta4 (Tbeta4) by following spectrophotometrically the extent and kinetics of distinct, thiol-specific crosslinking reactions between appropriate derivatives of the two proteins [Reichert et a]. (1996) J. Biol. Chem. 271, 1301-1308]. In the present study this method was used to show that such crosslinking, which is indicative of complex formation, occurs to the same extent with the actin-DNase I complex as with pure actin, although at a somewhat lower rate. Further evidence for the formation of the ternary complex was given by gel electrophoresis. From fluorescence spectroscopy the KD value of Tbeta4 from the actin-DNase I complex was found to be identical to that from pure actin. In line with these data, the capacity of actin for inhibiting DNase I was not affected by the addition of Tbeta4. In conclusion, DNase I and Tbeta4 are independent of each other in their interaction with actin, suggesting that the binding sites of thymosin beta4 and DNase I on actin do not overlap. A ternary complex of DNase I, actin and Tbeta4, if obtained in crystalline form, could thus provide an approach for studying the interface of Tbeta4 and actin by X-ray analysis.

Galactosylated poly(L-lysine) as a hepatotropic carrier of 9-beta-D-arabinofuranosyladenine 5'-monophosphate.

D-Galactopyranosyl residues were coupled to poly(L-lysine) and the antiviral agents arabinofuranosyladenine 5'-monophosphate (ara-AMP) and acyclovir were conjugated with this glycosylated polymer. In mice the ara-AMP conjugate accomplished a selective drug delivery to liver cells.

Biotinylphallotoxins: preparation and use as actin probes.

We describe the synthesis of four phalloidin derivatives conjugated with biotin. An aminomethyldithiolane derivative of ketophalloidin was used as a reactive starter compound, and biotin residues were coupled to this molecule either directly, separated by spacer chains comprised of one or two glycyl residues, or of a 12-atom long chain constructed from succinic acid and hexamethylendiamine. Although all products still displayed a high affinity for F-actin, as seen in competition experiments with [3H]-demethylphalloidin, only the one with the longest spacer (BHPP) showed specific and high-affinity decoration of actin filaments in permeabilized cells, in conjunction with FITC-coupled avidin and fluorescence microscopy. Combined with gold-streptavidin, BHPP decorated the actin filament system at the light and electron microscopic level faithfully and with satisfactory density. Actin filaments polymerized in vitro from purified protein were not as densely labeled as had been expected. However, in all these experiments the new phalloidin probe, when combined with avidin or streptavidin, yielded clear and highly specific labeling of F-actin. Therefore, this system is useful to identify and localize actin unambiguously in microfilaments, independent of actin antibodies, and should facilitate double-label experiments on cytoskeletal components at the ultrastructural level.

Lack of intestinal transport of [3H]-demethylphalloin: comparative studies with phallotoxins and bile acids on isolated small intestinal cells and ileal brush border membrane vesicles.

Several earlier studies suggested that the uptake of phallotoxins by liver cells is a carrier mediated process using a transport system normally handling bile acids (see Frimmer 1982). In this study we have shown whether ileal cells, well known to transport bile acids too, are able to take up phallotoxins. Isolated epithelial cells prepared from guinea pig ileum accumulated [14C]-cholate, whereas [3H]-demethylphalloin ([3H]-DMP) was not taken up. The same observation was made with isolated jejunal cells but the uptake of [14C]-cholate was much slower. [3H]-DMP, however, was partly bound to intestinal cells. This process was not inhibited by cholate, iodipamide, oligomycin and carbonylcyano-chlorophenylhydrazone (CCCP), compounds known to decrease the uptake of phallotoxins into liver cells. Substituting Na+ for choline+ and also Cl- for SCN- did not influence the binding of [3H]-DMP. Frozen intestinal cells from the guinea pig bound two time more [3H]-DMP after thawing compared with intact cells. Supplementary uptake experiments on isolated brush border membrane vesicles from rat ileum revealed that phalloidin does not inhibit taurocholate uptake and that taurocholate does not interfere with [3H]-DMP binding. The results suggest that [3H]-demethylphalloin is not recognized by the bile acid carrier of the guinea pig and the rat ileum. It is concluded that the transport system for bile acids present in ileal cell is different from that of liver cells.

Reduced phallotoxin uptake by livers of young compared with adult rats.

Using [3H]-demethylphalloin as a tracer the uptake of phallotoxins by the liver of young (6, 12, 16 days old) and adult rats was determined in relation to the dose of toxin administered. The maximum amount taken up by the livers of the young rats was only about 50% of that in adults. Nevertheless, with a dose as high as 55 mg/kg body weight the toxin concentration in the young liver reached more than 30 micrograms/g, being markedly higher than the minimum concentration (approximately 20 micrograms/g) required to cause irreversible damage of the liver in adult rats and death of the animals. This suggests that the tolerance of young rats to phallotoxins cannot solely be explained by the reduced uptake of the toxin.

Purification of amatoxin-specific antibodies from rabbit sera by affinity chromatography, their characterization and use in toxicological studies.

Rabbits were immunized with fetuin-beta-amanitin. They produced amatoxin-specific immunoglobulins of various classes, predominantly IgG. The crude IgG fraction was isolated by gel filtration on Sephacryl S-300. By polyethylene glycol precipitation in the presence of tritiated amatoxin, as well as by a solid phase radioimmunoassay technique, the portion of amatoxin-specific antibodies in the IgG fraction was determined to be 5-13%. The affinity of the amatoxin-specific IgG for a tritiated amatoxin derivative was measured by equilibrium dialysis and calculated according to the method of Scatchard. The dissociation constant was 2.6 X 10(-9) M. The amatoxin-specific immunoglobulins were extracted by their affinity to alpha-amanitin-Sepharose 4B and eluted with excess alpha-amanitin. The complex was isolated in 95% yield with a ratio immunoglobulin:toxin of c. 2:1. Alternatively, the uncomplexed IgG could be eluted from the alpha-amanitin-Sepharose 4B with 5 M guanidine hydrochloride; this treatment, however, decreased the binding capacity for amatoxin by 30% (ratio 1.4:1). The purified amatoxin-specific IgG was assayed for its therapeutic efficiency in mice poisoned with alpha-amanitin, but was unable to neutralize the toxic effects of the mushroom toxin. On the contrary, equimolar amounts of the amatoxin-specific immunoglobulins enhanced the toxicity of alpha-amanitin in the mouse by a factor 2.

Strongly enhanced toxicity of the mushroom toxin alpha-amanitin by an amatoxin-specific Fab or monoclonal antibody.

A monoclonal antibody, with high affinity against the mushroom toxin alpha-amanitin, was prepared. Administration of the Fab fragment of the monoclonal antibody to mice caused a 50-fold increase in alpha-amanitin toxicity. Electron micrographs showed normal appearance of hepatocytes but typical, amanitin-induced lesions in cells of the proximal convoluted tubules of the kidney. The pronounced nephrotoxicity is mainly explained by glomerular filtration and tubular reabsorption of the Fab-amatoxin complex and, to a lesser extent, of the immunoglobulin-amatoxin complex, which is still c. Twice as toxic as free alpha-amanitin. To our knowledge this is the first reported case where immunoglobulins or their fragments enhance rather than decrease the activity of a toxin. Accordingly, immunotherapy of Amanita mushroom poisoning in humans does not appear promising.