Effect of ammodytin L from Vipera ammodytes on L-6 cells from rat skeletal muscle

Lys49 myotoxins, secreted phospholipase A2-like proteins of viperid venoms: A comprehensive review

Muscle necrosis is a potential clinical complication of snakebite envenomings, which in severe cases can lead to functional or physical sequelae such as disability or amputation. Snake venom proteins with the ability to directly damage skeletal muscle fibers are collectively referred to as myotoxins, and include three main types: cytolysins of the "three-finger toxin" protein family expressed in many elapid venoms, the so-called "small" myotoxins found in a number of rattlesnake venoms, and the widespread secreted phospholipase A₂ (sPLA₂) molecules. Among the latter, protein variants that conserve the sPLA₂ structure, but lack such enzymatic activity, have been increasingly found in the venoms of many viperid species. Intriguingly, these sPLA₂-like proteins are able to induce muscle necrosis by a mechanism independent of phospholipid hydrolysis. They are commonly referred to as "Lys49 myotoxins" since they most often present, among other substitutions, the replacement of the otherwise invariant residue Asp49 of sPLA₂s by Lys. This work comprehensively reviews the historical developments and current knowledge towards deciphering the mechanism of action of Lys49 sPLA₂-like myotoxins, and points out main gaps to be filled for a better understanding of these multifaceted snake venom proteins, to hopefully lead to improved treatments for snakebites.

Cytotoxicity of snake venom Lys49 PLA2-like myotoxin on rat cardiomyocytes ex vivo does not involve a direct action on the contractile apparatus

Viperid snake venoms contain a unique family of cytotoxic proteins, the Lys49 PLA2 homologs, which are devoid of enzymatic activity but disrupt the integrity of cell membranes. They are known to induce skeletal muscle damage and are therefore named 'myotoxins'. Single intact and skinned (devoid of membranes and cytoplasm but with intact sarcomeric proteins) rat cardiomyocytes were used to analyze the cytotoxic action of a myotoxin, from the venom of Bothrops asper. The toxin induced rapid hypercontraction of intact cardiomyocytes, associated with an increase in the cytosolic concentration of calcium and with cell membrane disruption. Hypercontraction of intact cardiomyocytes was abrogated by the myosin inhibitor para-aminoblebbistatin (AmBleb). No toxin-induced changes of key parameters of force development were observed in skinned cardiomyocytes. Thus, although myosin is a key effector of the observed hypercontraction, a direct effect of the toxin on the sarcomeric proteins -including the actomyosin complex- is not part of the mechanism of cytotoxicity. Owing to the sensitivity of intact cardiomyocytes to the cytotoxic action of myotoxin, this ex vivo model is a valuable tool to explore in further detail the mechanism of action of this group of snake venom toxins.

Inflammatory effect of Bothropstoxin-I from Bothrops jararacussu venom mediated by NLRP3 inflammasome involves ATP and P2X7 receptor

Muscle tissue damage is one of the local effects described in bothropic envenomations. Bothropstoxin-I (BthTX-I), from Bothrops jararacussu venom, is a K49-phospholipase A2 (PLA2) that induces a massive muscle tissue injury, and, consequently, local inflammatory reaction. The NLRP3 inflammasome is a sensor that triggers inflammation by activating caspase 1 and releasing interleukin (IL)-1β and/or inducing pyroptotic cell death in response to tissue damage. We, therefore, aimed to address activation of NLRP3 inflammasome by BthTX-I-associated injury and the mechanism involved in this process. Intramuscular injection of BthTX-I results in infiltration of neutrophils and macrophages in gastrocnemius muscle, which is reduced in NLRP3- and Caspase-1-deficient mice. The in vitro IL-1β production induced by BthTX-I in peritoneal macrophages (PMs) requires caspase 1/11, ASC and NLRP3 and is dependent on adenosine 5'-triphosphate (ATP)-induced K+ efflux and P2X7 receptor (P2X7R). BthTX-I induces a dramatic release of ATP from C2C12 myotubes, therefore representing the major mechanism for P2X7R-dependent inflammasome activation in macrophages. A similar result was obtained when human monocyte-derived macrophages (HMDMs) were treated with BthTX-I. These findings demonstrated the inflammatory effect of BthTX-I on muscle tissue, pointing out a role for the ATP released by damaged cells for the NLRP3 activation on macrophages, contributing to the understanding of the microenvironment of the tissue damage of the Bothrops envenomation.

Differential Macrophage Subsets in Muscle Damage Induced by a K49-PLA2 from Bothrops jararacussu Venom Modulate the Time Course of the Regeneration Process

Bothrops snakes cause around 80% of snakebites in Brazil, with muscle tissue damage as an important consequence, which may cause dysfunction on the affected limb. Bothropstoxin-I (BthTX-I) from Bothrops jararacussu is a K49-phospholipase A₂, involved in the injury and envenomation's inflammatory response. Immune system components act in the resolution of tissue damage and regeneration. Thus, macrophages exert a crucial role in the elimination of dead tissue and muscle repair. Here, we studied the cellular influx and presence of classical and alternative macrophages (M1 and M2) during muscle injury induced by BthTX-I and the regeneration process. BthTX-I elicited intense inflammatory response characterized by neutrophil migration, then increased influx of M1 macrophages followed by M2 population that declined, resulting in tissue regeneration. The high expressions of TNF-α and IL6 were changed by increased TGF-β expression after BthTX-I injection, coinciding with the iNOs and arginase expression and the peaks of M1 and M2 macrophages in muscle tissue. A coordinated sequence of PAX7, MyoD, and myogenin expression involved in muscle regenerative process appeared after BthTX-I injection. Together, these results demonstrate a direct correlation between the macrophage subsets, cytokine microenvironment, and the myogenesis process. This information may be useful for new envenomation and muscular dysfunction therapies.

Screening of cytotoxic, anti-angiogenic, anti-tumorogenic and antimicrobial activities of Anatolian Vipera ammodytes (Nose-horned viper) venom

Objective:

In the present study, we aimed to screen the cytotoxic, antimicrobial, anti-angiogenic and anti-tumorogenic activities of Anatolian

Material and methods:

The cytotoxicity was screened against PC3, HeLa, CaCo-2, U-87MG, MCF-7 and Vero cells by using MTT assay. The antimicrobial activity on

Results:

The IC

Conclusion:

The results of the present study contributed to the knowledge of the biological activities of Anatolian

Cytotoxicity of Lachesis muta muta snake (bushmaster) venom and its purified basic phospholipase A2 (LmTX-I) in cultured cells

Human envenoming by Lachesis muta muta venom, although infrequent, is rather severe, being characterized by pronounced local tissue damage and systemic dysfunctions. Studies on the pharmacological actions of L. m. muta venom are relatively scant and the direct actions of the crude venom and its purified phospholipase A(2) (PLA(2)) have not been addressed using in vitro models. In this work, we investigated the cytotoxicity of L. m. muta venom and its purified PLA(2) isoform LmTX-I in cultured Madin-Darby canine kidney (MDCK) and in a skeletal muscle (C2C12) cell lines. As revealed by neutral red dye uptake assay, the crude venom (10 or 100 microg/ml) induced a significant decrease in cell viability of MDCK cells. LmTX-I at the concentrations tested (70-270 microg/ml or 5-20 microM) displayed no cytotoxicity in both MDCK and C2C12 cell lines. Morphometric analysis of Feulgen nuclear reaction revealed a significant increase in chromatin condensation (pyknosis), apparent reduction in the number of mitotic nuclei and nuclear fragmentation of some MDCK cells after incubation with L. m. muta venom. Monolayer exposure to crude venom resulted in morphological changes as assessed by scanning electron microscopy. The staining with TRITC-labelled phalloidin showed a marked disarray of the actin stress fiber following L. m. muta venom exposure. In contrast, LmTX-I had no effect on nucleus and cell morphologies as well as on stress fiber organization. These results indicate that L. m. muta venom exerts toxic effects on cultured MDCK cells. The LmTX-I probably does not contribute per se to the direct venom cytotoxicity, these effects are mediated by metalloproteinases/disintegrins and other components of the venom.

Differential susceptibility of C2C12 myoblasts and myotubes to group II phospholipase A2 myotoxins from crotalid snake venoms

Group II phospholipase A(2) (PLA(2)) myotoxins isolated from Viperidae/Crotalidae snake venoms induce a rapid cytolytic effect upon diverse cell types in vitro. Previous studies suggested that this effect could be more pronounced on skeletal muscle myotubes than on other cell types, including undifferentiated myoblasts. This study utilized the murine skeletal muscle C2C12 cell line to investigate whether differentiated myotubes are more susceptible than myoblasts, and if this characteristic is specific for the group II myotoxic PLA(2)s. The release of lactic dehydrogenase was quantified as a measure of cytolysis, 3 h after cell exposure to different group II PLA(2)s purified from Bothrops asper, Atropoides nummifer, Cerrophidion godmani, and Bothriechis schlegelii venoms. In addition, susceptibility to lysis induced by synthetic melittin and group III PLA(2) from bee (Apis mellifera) venom, as well as by anionic, cationic, and neutral detergents, was comparatively evaluated on the two cultures. Myotubes were significantly more susceptible to group II PLA(2) myotoxins, but not to the other agents tested, under the same conditions. Moreover, the increased susceptibility of myotubes over myoblasts was also demonstrated with two cytolytic synthetic peptides, derived from the C-terminal region of Lys49 PLA(2) myotoxins, that reproduce the action of their parent proteins. These results indicate that fusion and differentiation of myoblasts into myotubes induce changes that render these cells more susceptible to the toxic mechanism of group II PLA(2) myotoxins, but not to general perturbations of membrane homeostasis. Such changes are likely to involve myotoxin acceptor site(s), which remain(s) to be identified.

Rapid nongenomic effects of 3,5,3'-triiodo-L-thyronine on the intracellular pH of L-6 myoblasts are mediated by intracellular calcium mobilization and kinase pathways

L-T3 and L-T4 activated the Na+/H+ exchanger of L-6 myoblasts, with a fast nongenomic mechanism, both in the steady state and when cells undergo acid loading with ammonium chloride. Monitored with the intracellular pH-sensitive fluorescent probe 2',7'-bis(carboxyethyl)-5(6)-carboxyfluorescein, activation of the exchanger appeared to be initiated at the plasma membrane, because T3-agarose reproduced the effect of L-T3, and triiodothyroacetic acid, a hormone analog previously shown to inhibit membrane actions of thyroid hormone, blocked the action of L-T3 on the exchanger. We show here for the first time that transduction of the hormone signal in this nongenomic response requires tyrosine kinase-dependent phospholipase C activation and two different signaling pathways: 1) mobilization of intracellular calcium, assessed by the fluorescent probe fura-2, through activation of inositol trisphosphate receptors and without contributions from extracellular calcium or ryanodine receptors; and 2) protein phosphorylation involving protein kinase C and MAPK (ERK1/2), as shown by the use of kinase inhibitors and by immunoblotting for activated kinases.

Skeletal muscle degeneration induced by venom phospholipases A2: insights into the mechanisms of local and systemic myotoxicity

Local and systemic skeletal muscle degeneration is a common consequence of envenomations due to snakebites and mass bee attacks. Phospholipases A2 (PLA2) are important myotoxic components in these venoms, inducing a similar pattern of degenerative events in muscle cells. Myotoxic PLA2s bind to acceptors in the plasma membrane, which might be lipids or proteins and which may differ in their affinity for the PLA2s. Upon binding, myotoxic PLA2s disrupt the integrity of the plasma membrane by catalytically dependent or independent mechanisms, provoking a pronounced Ca2+ influx which, in turn, initiates a complex series of degenerative events associated with hypercontraction, activation of calpains and cytosolic Ca(2+)-dependent PLA2s, and mitochondrial Ca2+ overload. Cell culture models of cytotoxicity indicate that some myotoxic PLA2s affect differentiated myotubes in a rather selective fashion, whereas others display a broad cytolytic effect. A model is presented to explain the difference between PLA2s that induce predominantly local myonecrosis and those inducing both local and systemic myotoxicity. The former bind not only to muscle cells, but also to other cell types, thereby precluding a systemic distribution of these PLA2s and their action on distant muscles. In contrast, PLA2s that bind muscle cells in a more selective way are not sequestered by non-specific interactions with other cells and, consequently, are systemically distributed and reach muscle cells in other locations.

An overview of lysine-49 phospholipase A2 myotoxins from crotalid snake venoms and their structural determinants of myotoxic action

In 1984, the first venom phospholipase A2 (PLA2) with a lysine substituting for the highly conserved aspartate 49 was discovered, in the North American crotalid snake Agkistrodon p. piscivorus [J. Biol. Chem. 259 (1984) 13839]. Ten years later, the first mapping of a 'toxic region' on a Lys49 PLA2 was reported, in Bothrops asper myotoxin II [J. Biol. Chem. 269 (1994) 29867]. After a further decade of research on the Lys49 PLA2s, a better understanding of their structural determinants of toxicity and mode of action is rapidly emerging, with myotoxic effector sites identified at the C-terminal region in at least four proteins: B. asper myotoxin II, A. p. piscivorus K49 PLA2, A. c. laticinctus ACL myotoxin, and B. jararacussu bothropstoxin I. Although important features still remain to be established, their toxic mode of action has now been understood in its more general concepts, and a consistent working hypothesis can be experimentally supported. It is proposed that all the toxic activities of Lys49 PLA2s are related to their ability to destabilize natural (eukaryotic and prokaryotic) and artificial membranes, using a cationic/hydrophobic effector site located at their C-terminal loop. This review summarizes the general properties of the Lys49 PLA2 myotoxins, emphasizing the development of current concepts and hypotheses concerning the molecular basis of their toxic activities.

Animal experimentation in snake venom research and in vitro alternatives

Current experimental techniques used in snake venom research (with and without the use of animals) are reviewed. The emphasis is on the reduction of the use of animals in the development of antivenoms for the clinical treatment of snakebite. Diagnostic and research techniques for the major pathologies of envenoming are described and those using animals are contrasted with non-sentient methods where possible. In particular, LD50 and ED50 assays using animals (in vivo) and fertilised eggs (in vivo, non-sentient) are compared as well as in vitro procedures (ELISA and haemolytic test) for ED50 estimations. The social context of antivenom production, supply and demand is outlined together with the consequent tension between the benefits derived and the increase in opposition to experiments on animals. Stringent regulations governing the use of animals, limited research funds and public pressure all focus the need for progress towards non-animal, or non-sentient, research methods. Some achievements are noted but success is hampered by lack of detailed knowledge of the many constituents of venom which have to be assessed as a whole rather than individually. The only way to evaluate the net pathological effect of venom is to use a living system, usually a rodent, and similarly, the efficacy of antivenoms is also measured in vivo. The pre-clinical testing of antivenoms in animals is therefore a legal requirement in many countries and is strictly monitored by government authorities. New technologies applied to the characterisation of individual venom proteins should enable novel in vitro assays to be designed thus reducing the number of animals required. In the meantime, the principles of Reduce, Refine and Replace relating to animals in research are increasingly endorsed by those working in the field and the many agencies regulating ethical and research policy.

Membrane-perturbing activity of Viperidae myotoxins: an electrostatic surface potential approach to a puzzling problem

Phospholipase-like myotoxins are a class of proteins present in Viperidae venom. Despite the high level of amino acid and structural homology with soluble phospholipases A(2), myotoxins are devoid of enzymatic activity and share cytolytic activity by means of a totally unknown mechanism involving the lipid bilayer perturbation. The distribution of electrostatic surface potentials of four myotoxins and seven phospholipases A(2) has been compared. The charge distribution is similar in all active non-cytolytic phospholipases with a strongly positive side corresponding to the domain interacting with the micellar substrate and with the opposite side negatively charged. In contrast, all myotoxins examined are positively charged on both sides. Myotoxin III, the only known example of a myotoxin sharing enzymatic activity, displays the same electrostatic surface potential as other related toxins. Using liposomes made with non-hydrolysable phospholipids, we demonstrate that myotoxin III perturbs the lipid bilayer like other myotoxins. Based on these results, a molecular model for myotoxin-membrane perturbing activity is proposed. In this model, potential double-face binding of myotoxic phospholipases A(2) to lipid surfaces could trigger a lipid bilayer destabilization and could generate a stable fusion pore, probably because of the presence of hydrophobic moieties that flank the cationic sites.

PLA(2) stimulation of Na(+)/H(+) antiport and proliferation in rat aortic smooth muscle cells

The proliferative properties and the ability to stimulate the Na(+)/H(+) antiport activity of a secretory phospholipase A(2) were studied in rat aortic smooth muscle cells in culture. The requirement of the enzymatic activity of phospholipase A(2) to elicit mitogenesis was assessed by the use of ammodytin L, a Ser(49) phospholipase A(2) from the venom of Vipera ammodytes, devoid of hydrolytic activity. We propose that the proliferative effect is mediated by the same transduction pathway for both proteins. In particular, 1) both secretory phospholipase A(2) and ammodytin L stimulated thymidine incorporation in a dose-dependent manner; 2) both proteins affected the cell cycle, as assessed by cell growth and fluorescence-activated cell sorting experiments; 3) both phospholipase A(2) and ammodytin L increased intracellular pH, a permissive factor for cell proliferation, through activation of the Na(+)/H(+) antiport; 4) ammodytin L was able to displace the (125)I-labeled phospholipase A(2) from specific binding sites in a concentration range consistent with that capable of eliciting a cellular response; and 5) the inhibition by heparin was similar for both proteins, taking into account the ratio of heparin to protein. In conclusion, the enzymatic activity of phospholipase A(2) is not required for the stimulation of mitogenesis. The inhibitory effect of heparin combined with its therapeutic potential could help to clarify the role of phospholipase A(2) in the pathogenesis of several preinflammatory situations.

Comparative study of the cytolytic activity of myotoxic phospholipases A2 on mouse endothelial (tEnd) and skeletal muscle (C2C12) cells in vitro

A rapid in vitro cytolytic effect of some myotoxic phospholipases A2 (PLA2s) isolated from the venoms of Viperidae snakes has been previously described. This study was undertaken to investigate if cytolytic activity is a common property of the myotoxic proteins from this group. Murine endothelial cells (tEnd) and skeletal muscle myotubes (C2C12) were utilized as targets. The release of lactic dehydrogenase was quantified as a measure of cell damage, 3 h after exposure of cells to the different PLA2s, including representatives from the genera Bothrops, Agkistrodon, Trimeresurus, Crotalus (family Viperidae), and Notechis (family Elapidae). All of the group II myotoxic PLA2s tested displayed rapid cytolytic activity when tested in the micromolar range of concentrations (8-32 microM). In contrast, the group I myotoxic PLA2 notexin was devoid of this activity. Aspartate-49 and lysine-49 PLA2 group II variants showed a comparable cytolytic effect. Skeletal muscle myotubes, obtained after fusion and differentiation of C2C12 myoblasts, were significantly more susceptible to the cytolytic action of myotoxins than endothelial cells, previously reported to be more susceptible than undifferentiated myoblasts under the same assay conditions. Cytolytic activity appears to be a common characteristic of group II myotoxic PLA2s of the Viperidae. Bee venom PLA2, a group III enzyme of known myotoxicity, also displayed cytotoxic activity on C2C12 myotubes, being devoid of activity on endothelial cells. These results suggest that in vitro differentiated skeletal muscle myotubes may represent a suitable model target for the study of myotoxic PLA2s of the structural group II found in snake venoms.

Insulin stimulation of Na/H antiport in L-6 cells: a different mechanism in myoblasts and myotubes

Insulin modulation of the Na/H antiport of L-6 cells, from rat skeletal muscle was studied in both myoblasts and myotubes using the fluorescent, pH sensitive, intracellular probe 2',7' bis (carboxyethyl)-5(6)-carboxyfluorescein. Insulin stimulated the Na/H antiport activity in L-6 cells, showing a bell-shaped dose response typical of other insulin responses: a maximum at 10 nM (delta pH of 0.132 +/- 0.007 and 0.160 +/- 0.040 over basal value, for myoblasts and myotubes, respectively; means +/- SD, n = 6-8) and smaller effects at higher and lower concentrations. Phorbol 12-myristate 13-acetate (PMA), an activator of protein kinase C, also stimulated the antiport in myoblasts but not in myotubes. Surprisingly the rapid increase in intracellular pH was not observed when insulin and PMA were added simultaneously to myoblasts; apparently these two activators mutually excluded each other. Downregulation of protein kinase C, obtained by preincubation of cells with PMA for 20 hr, totally abolished both hormone and PMA effects in myoblasts, whereas in myotubes insulin stimulation was not affected. Inhibitors of tyrosine kinase activity, such as erbstatin analog and genistein abolished insulin effect on the Na/H antiport, both in myoblasts and in myotubes. Different sensitivity to pertussis toxin in the two cell types suggests that the differentiation process leads to a change in the signal pathways involved in the physiological response to insulin.

Proliferative effect of ammodytin L from the venom of Vipera ammodytes on 208F rat fibroblasts in culture

Ammodytin L, purified from the venom of Vipera ammodytes, triggers a rapid and dramatic lytic process in myotubes in vitro, as well as in differentiated muscle cells in vivo, through a mechanism that is not well understood. Despite its great sequence similarity to phospholipase A2, it is devoid of any enzyme activity. Data on artificial membranes demonstrating a direct interaction between this toxin and the hydrophobic core of the lipid bilayer suggest that the toxin also acts on the lipid microenvironment in cell membranes. Recent experiments on living cells do not confirm this hypothesis, and a more intricate mechanism is proposed. In vitro, ammodytin L has necrotic effects only in well-differentiated myogenic cells, whereas other cell types such as platelets, red blood cells and lymphocytes show neither morphological nor functional alterations. In this work we demonstrate that rat 208F fibroblasts in culture after ammodytin L challenge increase [3H]thymidine incorporation, indicating that this toxin has a myogenic effect. Moreover, ammodytin L increases intracellular Ca2+ by acting on intracellular stores probably by activating a phosphatidylinositol-specific phospholipase C. Preincubation of the cells with ammodytin L did not prevent the massive Ca2+ release evoked by bradykinin, a phenomenon observed when fibroblasts were incubated with both thapsigargin and ionomycin. Heparin, an agent that inhibits the necrotic effect of the myotoxin in myotubes, also reduces the effect of ammodytin L on DNA synthesis. Heparin inhibits only the late sustained increase in intracellular Ca2+ induced by the toxin.

Similarities and differences in mechanisms of cardiotoxins, melittin and other myotoxins

Myonecrosis induced in vivo by cardiotoxin, melittin, and Asp49 and Lys49 phospholipase A2 (PLA2) myotoxins involves rapid lysis of the sarcolemma, myofibril clumping, and hypercontraction of sarcomeres. In contrast, skeletal muscle necrosis induced by crotamine and myotoxin a is much slower, consisting of mitochondrial and sarcoplasmic reticulum swelling, myofibril degeneration, and lack of sarcolemma or transverse tubule damage. The mechanisms contributing to the myonecrosis induced by these peptides were evaluated. Two cardiotoxins and two Lys49 PLA2 myotoxins lysed primary cultures of human skeletal muscle within 24 hr at a concentration of 0.25 microM, while melittin, crotamine, and myotoxin a, and an Asp49 PLA2 myotoxin were non-cytolytic at concentrations up to 5.0 microM, suggesting that cytolysis is not a good measure of myotoxicity. Crotamine and the Lys49 PLA2 myotoxin altered Ca2+ ion flux in human heavy sarcoplasmic reticulum by opening the ryanocine receptor. Whole-cell patch-clamp studies demonstrated that administrating crotamine intracellularly increased Na+ currents. Free fatty acids, liberated by activation of tissue phospholipase C or by the PLA2 activity of the myotoxins, were monitored for crotamine, myotoxin a and a Lys49 PLA2 myotoxin in cell cultures in which the lipids had been radiolabeled. Only the Lys49 myotoxin produced significant amounts of fatty acid in cell cultures, supporting a potential role for fatty acid production only in the mechanism of sarcolemma-destroying myotoxins. These findings, coupled with those in the literature, support a hypothesis in which the myotoxins and/or products of lipase activity (e.g. fatty acids) are acting at a site existing on both the Na+ channel and a protein involved in Ca2+ release and probably serving a modulatory function for ion regulation. Based on the similarities in mechanisms between the toxins and fatty acids, the most likely site would be a fatty acid binding site on the protein (either similar to that on fatty acid binding proteins, or an acylated cysteine residue) or in the membrane.

Effect of ammodytin L from the venom of Vipera ammodytes on Xenopus laevis differentiated muscle fibres and regenerating limbs

Ammodytin L is a non-catalytic, phospholipase-like snake venom toxin from Vipera ammodytes, which shows a cytotoxic activity on differentiated myotubes when tested in vitro. In the range of concentrations in which ammodytin L induced necrosis of myogenic cells in culture, other cell types (erythrocytes, platelets, fibroblasts) did not appear to be affected. To test the in vivo toxicity and the effective cytolytic specificity of ammodytin L we have followed the morphological changes in muscle tissue of Xenopus laevis limbs after intramuscular toxin injection. Only muscular cells were affected by ammodytin L, and the toxin did not induce any morphological change in other cell types. Further evidence of the muscle-specific action of the toxin was obtained from experiments carried out using the Xenopus kidney cell line B3.2 in culture. Ammodytin L was unable to affect parameters of cell viability such as lactate dehydrogenase leakage, [3H]thymidine incorporation, growth curves and morphological changes. Moreover, direct ammodytin L application to cultured regenerative limbs did not provoke alterations in undifferentiated myoblasts. These data suggest that ammodytin L, like other phospholipase-like toxins, exerts its toxicity by selectively damaging differentiated muscle fibres.