Functional modification of a 21-kilodalton G protein when ADP-ribosylated by exoenzyme C3 of Clostridium botulinum

ADP-ribosylation and de-ADP-ribosylation of the rho protein by Clostridium botulinum exoenzyme C3. Regulation by EDTA, guanine nucleotides and pH

Pretreatment of rho protein purified from pig brain cytosol with EDTA (3 mM) for 10 min at 30 degrees C inhibited its ADP-ribosylation by Clostridium botulinum C3 ADP-ribosyltransferase by more than 90%. The EDTA effect was not caused by alteration of C3. GDP or GDP beta S present during the pretreatment period completely prevented the decrease in ADP-ribosylation with half-maximal and maximal effects at 3 and 300 microM, respectively. GTP or GTP gamma S were less efficacious in preventing the decrease in ADP-ribosylation, but were more potent (half-maximal and maximal effects at 0.1 and 3 microM, respectively). [32P]ADP-ribose incorporated in pig brain rho by C3 was de-ADP-ribosylated by the enzyme in the presence of nicotinamide and at low pH. Concomitantly, [32P]NAD was formed. The pH optima for ADP-ribosylation and de-ADP-ribosylation were pH 7.5 and 5.5, respectively. De-ADP-ribosylation was most efficient with nicotinamide, less effective with 3-acetylpyridine and not observed with 3-aminopyridine, 4-aminopyridine, 4-acetylpyridine and isonicotinic acid. As observed for the ADP-ribosylation, the de-ADP-ribosylation by C3 was maximal with the GDP-bound form of rho and blocked after EDTA treatment.

A geranylgeranyltransferase for rhoA p21 distinct from the farnesyltransferase for ras p21S

We have clarified that rhoA p21 purified from bovine aortic smooth muscle is geranylgeranylated at the cysteine residue in the C-terminal CAAX motif (A is an aliphatic amino acid and X is any amino acid). In this paper, a geranylgeranyltransferase for rhoA p21 (rhoA p21 GGT) was partially purified from bovine brain cytosol. This enzyme transferred a geranylgeranyl moiety from geranylgeranyl pyrophosphate to rhoA p21 having the CAAX motif (rhoA p21-CAAX) but not to rhoA p21 lacking the AAX portion. rhoA p21 GGT was separated from the previously reported farnesyltransferase for ras p21s (ras p21 FT) by column chromatographies and did not geranylgeranylate or farnesylate c-Ha-ras p21-CAAX. ras p21 FT did not geranylgeranylate or farnesylate rhoA p21-CAAX. These results indicate that rhoA p21 GGT distinct from ras p21 FT is present in bovine brain cytosol.

Tissue and subcellular distributions of an inhibitory GDP/GTP exchange protein (GDI) for the rho proteins by use of its specific antibody

We have recently purified from bovine brain cytosol to near homogeneity a GDP/GTP exchange protein for the rho proteins, named rho GDI, that inhibits the dissociation of GDP from and the subsequent binding of GTP to the rho proteins. In the present study, we made a monoclonal antibody against rho GDI and studied its tissue distribution in rat and its subcellular distribution in rat cerebrum by use of this antibody. rho GDI was found in most rat tissues as described for the rho proteins. In rat cerebrum, rho GDI was mostly found in the cytosol of neuron body and synaptosome. In synaptosome, it was mainly found in the synaptic cytosol.

Molecular cloning and sequencing of the epidermal cell differentiation inhibitor gene from Staphylococcus aureus

We recently purified to homogeneity a protein inhibiting differentiation of cultured keratinocytes from extracellular products of Staphylococcus aureus, and named it epidermal cell differentiation inhibitor (EDIN). In the present study, we isolated and sequenced the structural gene coding for EDIN from Staphylococcus aureus E-1 using oligonucleotide probes on the basis of the partial amino acid sequence of the purified EDIN. DNA sequencing of the cloned DNA revealed an open reading frame encoding 247 amino acids as a precursor of EDIN, which included an NH2-terminal signal sequence of 35 amino acid residues. Processing of this precursor produces a mature EDIN protein composed of 212 amino acids with a calculated Mr of 23,782. The EDIN shared 35% amino acid homology with the ADP-ribosyltransferase C3 of Clostridium botulinum. These results with biological properties of EDIN described previously indicate that EDIN is a novel protein.

The ras superfamily of small GTP-binding proteins

Considerable advances have recently been made in understanding the structure and function of the proteins encoded by the ras proto-oncogenes. In addition, a large number of ras-related small GTP-binding proteins with very diverse activities have now been identified. This review explores developments in this rapidly expanding field.

Morphological alterations of Xenopus oocytes induced by valine-14 p21rho depend on isoprenylation and are inhibited by Clostridium botulinum C3 ADP-ribosyltransferase

Microinjection of the constitutively active recombinant Val-14 p21rho A into Xenopus oocytes induced dramatic morphological changes with redistribution of pigments from the animal pole resulting in spotted oocytes. The effects induced by Val-14 p21rho A were regulated by progesterone in a dose-dependent manner whereas prior ADP-ribosylation of the rho protein blocked its activity. About 30 min after microinjection, p21 rho was associated with the plasma membrane. The membrane association of p21rho and its biological activity were inhibited by lovastatin, an inhibitor of the 3-hydroxy-3-methylglutaryl coenzyme A reductase. The findings suggest that membrane attachment and biological activity of p21rho depend on isoprenylation of the GTP-binding protein.

Interaction of recombinant rho A GTP-binding proteins with photoexcited rhodopsin

The small molecular mass GTP-binding proteins rho A, B and C are targets for ADP-ribosyltransferase activity of the botulinum exoenzyme C3. The possible interaction of recombinant rho A proteins expressed in E. coli with photoexcited rhodopsin was studied by reconstitution with bovine rod outer segment (ROS) membranes depleted of endogenous GTP-binding proteins by treatment with urea. As reported for C3 substrates present in untreated ROS membranes, ADP-ribosylation of recombinant rho A proteins, both normal and Val-14 mutant, by C3 was inhibited when reconstituted with illuminated compared to dark-adapted ROS membranes pretreated with urea. GDP reduced the light-induced inhibition, while GTP[S] and light inhibited ADP-ribosylation of rho A proteins in a synergistic manner.

Isoprenylation of the low molecular mass GTP-binding proteins rac 1 and rac 2: possible role in membrane localization

Ras proteins can be modified at their COOH-terminal cysteine in the motif Cys-Ali-Ali-Xaa by a farnesyl isoprenoid. This modification is essential for membrane association and biological activity of ras proteins. A similar COOH-terminal amino acid sequence, Cys-Xaa-Ali-Xaa, exists in the ras-related GTP-binding proteins rac 1 and rac 2. To determine whether these proteins were similarly modified, COS cells were transfected with rac 1 and rac 2 cDNA and expressed proteins were labeled with [3H]mevalonic acid. We report here that both rac 1 and rac 2 are post-translationally modified by addition of an isoprenoid group, the likely site of which is the COOH-terminal cysteine. Isoprenylation was found only in racs associated with particulate cell fractions, suggesting that this modification may be associated with membrane localization of the proteins. These data specifically identify mammalian low molecular mass GTP-binding proteins other than ras that undergo post-translational modification and further define the COOH-terminal consensus sequence, Cys-Ali-Ali-Xaa, as an isoprenylation signal. This sequence may identify a larger family of low molecular mass GTP-binding proteins which are isoprenylated.

Identification of a major GTP-binding protein in bovine aortic smooth muscle cytosol as the rhoA gene product

In bovine aortic smooth muscle, about 50% of total GTP-binding activity was present in the cytosol fraction. A major GTP-binding protein (G protein) with a Mr value of about 21,000 (21K G) in this fraction was purified to near homogeneity and characterized. 21K G bound maximally about 0.8 mol of [35S]guanosine 5'-(3-O-thio)triphosphate/mol of protein with a Kd value of about 20 nM. 21K G showed GTPase activity with a turnover number of about 0.007 min-1. 21K G was ADP-ribosylated by botulinum ADP-ribosyltransferase and about 0.4 mol of ADP-ribose was maximally incorporated into 1 mol of 21K G. 21K G and the bovine brain rhoA gene product (rhoA p21) were eluted at the same retention time on C4 reversed-phase high performance liquid chromatography and migrated at the same positions on two-dimensional gel electrophoresis. These results indicate that the major G protein in bovine aortic smooth muscle cytosol is rhoA p21.

Identification and localization of low-molecular-mass GTP-binding proteins associated with synaptic vesicles and other membranes

GTP-binding proteins were studied in synaptic vesicles prepared from bovine brain by differential centrifugation and separated further from plasma membranes using gel permeation chromatography. Following separation by SDS-PAGE of proteins from the different fractions, and transfer to nitrocellulose sheets, the presence and localization of low-molecular-mass GTP-binding proteins were assessed by [alpha-32 P]GTP binding. The vesicle-membrane fraction (SV) was enriched in synaptophysin (p38, a synaptic vesicle marker) and contained low-molecular-mass GTP-binding proteins; these consisted of a major 27 kDa protein and minor components (Mr 26 and 24 kDa) which were trypsin-sensitive and immunologically distinguishable from ras p21 protein. GTP-binding proteins of low molecular mass, but displaying less sensitivity to trypsin, were also found in the plasma membrane fraction (PM; enriched in Na+/K(+)-ATPase). In addition, the PM fraction contained GTP-binding proteins with higher Mr (Gi alpha and G0 alpha), together with another GTP-binding protein, ras p21. Putative function(s) of these GTP-binding proteins with low mass are discussed.

A GTPase-activating protein for rhoB p20, a ras p21-like GTP-binding protein--partial purification, characterization and subcellular distribution in rat brain

A protein stimulating the GTPase activity of rhoB p20, a ras p21-like GTP-binding protein (G protein), was partially purified from the cytosol fraction of bovine brain. This protein, designated as rhoB p20 GTPase-activating protein (GAP), did not stimulate the GTPase activity of other ras p21/ras p21-like G proteins including c-Ha-ras p21, smg p21 and smg p25A. The activities of c-Ha-ras p21 GAP and smg p21 GAP were also detected in the cytosol fraction of bovine brain and rhoB p20 GAP was separated from these GAPs. The activity of rhoB p20 GAP was eliminated by tryptic digestion or boiling. The Mr value of rhoB p20 GAP was estimated to be 150-200 x 10(3) and 37 x 10(3) by gel filtration and sucrose density gradient ultracentrifugation, respectively. These results indicate that there is rhoB p20 GAP in addition to c-Ha-ras p21 GAP and smg p21 GAP in bovine brain. In rat brain, about 50% of rhoB p20 GAP was found with the highest specific activity in the P2 fraction containing myelin, synaptosomes and mitochondria. In the P2 fraction, about 30% of rhoB p20 GAP was found in the P2C fraction containing mainly synaptosomes. rhoB p20 GAP was detected in the cytosol and particulate fractions of not only rat brain but also other rat tissues.

Partial purification and characterization of GDP dissociation stimulator (GDS) for the rho proteins from bovine brain cytosol

A novel type of regulatory proteins for the rho proteins (rhoA p21 and rhoB p20), ras p21-like small GTP-binding proteins (G proteins), are partially purified from bovine brain cytosol. These regulatory proteins, named rho GDP dissociation stimulator (GDS) 1 and -2, stimulate the dissociation of GDP from rhoA p21 and rhoB p20. rho GDS1 and -2 are inactive for other ras p21/ras p21-like small G proteins including c-Ha-ras p21, smg p21B, and smg p25A. Since we have previously shown that the rate limiting step for the GDP/GTP exchange reaction of the rho proteins is the dissociation of GDP from these proteins, the present results suggest that rho GDS1 and -2 stimulate the GDP/GTP exchange reaction of the rho proteins. rho GDS1 and -2 are distinct from the GAP- and GDI-types of regulatory proteins for the rho proteins previously purified from bovine brain cytosol. rho GAP stimulates the GTPase activity of the rho proteins and rho GDI inhibits the GDP/GTP exchange reaction of the rho proteins. The present results together with these earlier observations indicate that the rho proteins are regulated by at least three different types of regulatory proteins, GDS, GDI, and GAP.

ADP-ribosylation of the rho/rac proteins induces growth inhibition, neurite outgrowth and acetylcholine esterase in cultured PC-12 cells

Botulinum ADP-ribosyltransferase C3 (C3 exoenzyme) was purified to homogeneity and added to cultured rat pheochromocytoma PC-12 cells. Incubation with this exoenzyme caused inhibition of cell growth and induced neurites as well as acetylcholine esterase in these cells. These changes were dependent on the amount of the enzyme added to the culture, which correlated with the in situ ADP-ribosylation of the rho/rac proteins in the cells. Preincubation with a specific anti-C3 exoenzyme monoclonal antibody inhibited both the ADP-ribosyltransferase activity and the neurite-inducing activity of the enzyme preparation. These results suggest that C3 exoenzyme affected the cellular function of the rho/rac proteins by ADP-ribosylation to induce these changes in the cells.

Multiple small molecular weight guanine nucleotide-binding proteins in human erythrocyte membranes

Native membranes from human erythrocytes contain the following G proteins which are ADP-ribosylated by a number of bacterial toxins: Gi alpha and Go alpha (pertussis toxin), Gs alpha (cholera toxin), and three proteins of 27, 26 and 22 kDa (exoenzyme C3 from Clostridium botulinum). Three additional C3 substrates (18.5, 16.5 and 14.5 kDa) appeared in conditions of unrestrained proteolysis during hemolysis. SDS-PAGE separation of erythrocyte membrane proteins followed by electroblotting and incubation of nitrocellulose sheets with radiolabeled GTP revealed consistently four GTP-binding proteins with Mr values of 27, 26, 22 and 21 kDa. Although a 22 kDa protein was immunochemically identified as ras p21, the C3 substrate of 22 kDa is a different protein probably identifiable with a rho gene product. Accordingly, at least five distinct small molecular weight guanine nucleotide-binding proteins, whose functions are so far undetermined, are present in native human erythrocyte membranes.

Purification and characterization from bovine brain membranes of a GTP-binding protein with a Mr of 21,000, ADP-ribosylated by an ADP-ribosyltransferase contaminated in botulinum toxin type C1--identification as the rhoA gene product

We have previously shown that there are multiple GTP-binding proteins (G proteins) with Mr values of about 20,000 in bovine brain membranes and identified one G protein with a Mr of 20,000 as the rho gene product. We have also shown that this rho gene product is ADP-ribosylated by an ADP-ribosyltransferase contaminated in botulinum toxin type C1. In the present studies, we have purified another G protein with a Mr of about 21,000 to near homogeneity from bovine brain membranes by several column chromatographies and identified it as the rhoA gene product. Further analysis of the amino acid sequence of the G protein, which we have purified and identified as the rho gene product previously, has revealed that this G protein is the rhoB gene product. The rhoA gene product binds maximally about 0.9 mol of [35S]guanosine 5'-(3-O-thio) triphosphate (GTP gamma S)/mol of protein with a K d value of about 20 nM. [35S]GTP gamma S-binding to the rhoA gene product is inhibited by pretreatment with N-ethylmaleimide. The rhoA gene product hydrolyzes GTP to liberate Pi with a turnover number of about 0.01 min-1. Moreover, the rhoA gene product is ADP-ribosylated by an ADP-ribosyltransferase contaminated in botulinum toxin type Cl. About 0.3 mol of ADP-ribose is maximally incorporated into 1 mol of the rhoA gene product. The ADP ribosylation of the rhoA gene product does not affect its GTP gamma S-binding or GTPase activity. These properties of the rhoA gene product are similar those of the rhoB gene product described previously. These results together with the earlier observations indicate that there are at least two rho gene products (rhoA, B) among three members of the rho gene family (rhoA, B, C) in bovine brain membranes and that both of them are ADP-ribosylated by an ADP-ribosyltransferase contaminated in botulinum toxin type C1.

Botulinum ADP-ribosyltransferase C3: a new tool to study low molecular weight GTP-binding proteins

It is well known that certain bacterial toxins, e.g. cholera and pertussis toxins, ADP-ribosylate eukaryotic regulatory proteins. They have become invaluable tools in the study of G protein-linked receptors. Less well appreciated is the fact that certain strains of Clostridium botulinum types C and D produce an ADP-ribosyltransferase, termed C3. This enzyme is structurally and functionally distinct from botulinum neurotoxins C1 and D. Its substrate is the 21 kDa GTP-binding protein rho. Klaus Aktories and Alan Hall explain why C3 is now an important tool in analysing the regulatory function of the ras-related protein rho.

Rabbit intestine contains a protein that inhibits the dissociation of GDP from and the subsequent binding of GTP to rhoB p20, a ras p21-like GTP-binding protein

A novel regulatory protein for rhoB p20, a ras p21-like GTP-binding protein (G protein), was partially purified from the cytosol fraction of rabbit intestine. This protein, designated as rhoB p20 GDP dissociation inhibitor (GDI), inhibited the dissociation of GDP from rhoB p20. rhoB p20 GDI also inhibited the binding of guanosine 5'-(3-O-thio)triphosphate (GTP gamma S) to the GDP-bound form of rhoB p20 but not of that to the guanine nucleotide-free form. GDI did not affect the GTPase activity of rhoB p20 and by itself showed no GTP gamma S-binding activity. GDI was inactive for other ras p21/ras p21-like G proteins including c-Ha-ras p21, smg p21 and smg p25A. The Mr value of GDI was estimated to be about 27,000 from the S value. These results indicate that rabbit intestine contains a novel regulatory protein that inhibits the dissociation of GDP from and thereby the subsequent binding of GTP to rhoB p20.

Kinetic analysis of the binding of guanine nucleotides to bovine brain rhoB p20, a ras p21-like GTP-binding protein

We have investigated the kinetics of the binding of guanine nucleotides to bovine brain rhoB p20, a ras p21-like GTP-binding protein with GTPase activity. The initial velocities of the binding of guanosine 5'-(3-O-thio)triphosphate (GTP gamma S) to GDP-bound rhoB p20 and the dissociation of GDP from this protein were markedly increased by decreasing Mg2+ concentrations. The initial velocity of the binding of GTP gamma S to GDP-free rhoB p20 was not affected by changing Mg2+ concentrations. These results indicate that the dissociation of GDP from rhoB p20 limits the binding of GTP to this protein, and suggest that there is a factor stimulating the dissociation of GDP from rhoB p20 and thereby stimulating the binding of GTP to this protein in mammalian tissues. Consistently, the factor stimulating the dissociation of GDP, but not of GTP gamma S, from rhoB p20 was detected in bovine brain cytosol.

Signal transduction and the ras gene family: molecular switches of unknown function

The ras family consists of 20 or more genes that encode small GTP/GDP-binding proteins, of 20-26 kDa, the functions of which are unknown. This article discusses possible roles of the ras proteins in signal transduction and the interaction of p21(ras) and other members of the ras family with GTPase-accelerating proteins (GAPS) that may be regulatory elements of the signaling machinery.

Immuno-crossreactivity between botulinum neurotoxin type C1 or D and exoenzyme C3

Botulinum neurotoxin type D and exoenzyme C3 have been separately purified from Clostridium botulinum strain D-1873 to apparent homogeneity. Both ADP-ribosylated a rat liver cytosolic protein of 24 kDa. The N-terminal amino acid sequence of C3 was determined and showed a low degree of homology with those of the light and heavy chains of neurotoxins of various types which have been reported previously. However, a polyclonal antibody raised against C3 cross-reacted with the light chains, but not with the heavy chains, of type C1 and D neurotoxins. Furthermore, a monoclonal antibody recognizing the light chains of type C1 and D neurotoxins interacted with C3. These results suggest that the light chain of type C1 or D neurotoxin and exoenzyme C3 share at least one epitope in common with each other.

Identification of rho as a substrate for botulinum toxin C3-catalyzed ADP-ribosylation

Recombinant Aplysia rho and a GTP-binding protein purified from human neutrophil membranes (G22K) were ADP-ribosylated by botulinum toxin C3 with stoichiometries of 0.8 and 0.6, respectively. Rho and G22K appeared to be different proteins since (i) rho migrated faster on polyacrylamide gels, (ii) unlike G22K, rho did not require the presence of cytosol to be ADP-ribosylated, (iii) G22K was not recognized by an anti-rho antiserum, and (iv) antibody 142-24E05 recognized G22K effectively but only poorly cross reacted with rho. ADP-ribosylation had no effect on the ability of rho to bind or hydrolyse GTP. Therefore, it appears that there are multiple botulinum toxin C3 substrates and that the toxin exerts its effects on cell function by a mechanism other than modulating the GTPase activity of rho.

The rho gene product expressed in E. coli is a substrate of botulinum ADP-ribosyltransferase C3

The ras-related rho A protein expressed in E. coli, was ADP-ribosylated by botulinum ADP-ribosyltransferase C3. C3 also modified the valine-14 mutant rho protein but not the products of H-ras, R-ras, ral, ypt, and rap 1 genes. A ras-rho chimaera consisting of 60 amino acids from the amino terminus of ras fused to 133 amino acids from the carboxy terminus of rho was not modified by C3. Antibodies raised against the porcine brain cytosolic substrate of C3 cross reacted with the rho, valine-14 rho and ras-rho proteins, but not with the gene products of H-ras, R-ras, ral or rap 1. Polyclonal anti-H-ras antibodies cross reacted with H-ras but not with ral, rho, or the C3 substrate purified from porcine brain.

Purification of the 22 kDa protein substrate of botulinum ADP-ribosyltransferase C3 from porcine brain cytosol and its characterization as a GTP-binding protein highly homologous to the rho gene product

The 22 kDa protein substrate of botulinum ADP-ribosyltransferase C3 was purified from porcine brain cytosol by acetone precipitation, CM-Sephadex, octyl-Sepharose and TSK phenyl-5PW HPLC chromatography to apparent homogeneity. ADP-ribosylation of the protein was increased by guanine nucleotides (GTP, GDP, GTP gamma S, each 100 microM) but not by GMP, ATP or ATP gamma S. The purified 22 kDa protein bound maximally 0.9 mol [35S]GTP gamma S and hydrolyzed GTP with the rate 0.007 mol per mol protein. Amino acid sequences were obtained from two tryptic peptides, selected from an in situ digestion of Immobilon electrotransferred, gel purified ADP-ribosylated substrate. The two sequences obtained, cover 23 residues from the corresponding sequences in human rho.

Different types of ADP-ribose protein bonds formed by botulinum C2 toxin, botulinum ADP-ribosyltransferase C3 and pertussis toxin

We attempted to characterize ADP-ribose-amino acid bonds formed by various bacterial toxins. The ADP-ribose-arginine bond formed by botulinum C2 toxin in actin was cleaved with a half-life of about 2 h by treatment with hydroxylamine (0.5 M). In contrast, the ADP-ribose-cysteine bond formed by pertussis toxin in transducin and the ADP-ribose-amino acid linkage formed by botulinum ADP-ribosyltransferase C3 in platelet cytosolic proteins were not affected by hydroxylamine. HgCl2 cleaved the ADP-ribose-amino acid bond formed by pertussis toxin in transducin but not those formed by botulinum C2 toxin or botulinum ADP-ribosyltransferase C3 in actin and platelet cytosolic proteins, respectively. NaOH (0.5 M) cleaved the ADP-ribose-amino acid bonds formed by botulinum C2 toxin and pertussis toxin but not the one formed by botulinum ADP-ribosyltransferase C3. The data indicate that the ADP-ribose bond formed by botulinum ADP-ribosyltransferase C3 differs from those formed by the known bacterial ADP-ribosylating toxins.

Botulinum toxin-induced ADP-ribosylation and inhibition of exocytosis are unrelated events

The hypothesis that inhibition of secretion by botulinum neurotoxin type D occurs by an intracellular process involving ADP-ribosylation has been directly tested by measuring both the extent of inhibition of secretion and of ADP-ribosylation in the same cells. Although the inhibitory effect of unpurified toxin closely parallels intracellular ribosylation, the two events are clearly unrelated, as using purified D and C3 toxins together with their antibodies, each of these events can be either stimulated or inhibited independently of each other.

A 22 kDa ras-related G-protein is the substrate for an ADP-ribosyltransferase from Clostridium botulinum

A ribosyltransferase from C. botulinum type D ADP-ribosylated a protein of 22 kDa (p22) in human astrocytoma (1321N1) cells. ADP-ribosylation of membrane-bound p22 was potentiated by 2 mM MgCl2 or guanine nucleotides but was much reduced in the presence of 10 mM Mg2+ plus GTP gamma S. p22 was immunoprecipitated by a monoclonal antibody (142-24E05) raised against a peptide sequence common to the ras gene family but not by other ras or G-protein antibodies. p22 was also ADP-ribosylated in Drosophila but was not detected in Dictyostelium. These data suggest that the 22 kDa botulinum toxin substrate is a GTP-binding protein and a member of the ras protein family.

Clostridium spiroforme toxin is a binary toxin which ADP-ribosylates cellular actin

We have purified from Clostridium spiroforme strain 246 an heterogeneous population of proteins (Sa) ranging from 43 to 47 kilodaltons exhibiting ADP-ribosyl transferase activity as do C. botulinum C2 toxin component I or the ia chain of C. perfringens E iota toxin. C. spiriforme Sa had alone no activity upon injection in mice or inoculated to Vero cells. When spiroforme ADP ribosyl transferase were mixed with a trypsin activated protein (Sb) separated from C. spiroforme bacterial supernatant, a lethal effect in mice and cytotoxicity on Vero cells were recorded. The Sa cross-reacted immunologically with either the light chain of C. perfringens E iota toxin or the ADP-ribosyl transferase from C. difficile 196 strain. No immunological relatedness was observed between Sa and C2 toxin component I. C. spiroforme toxin is thus another binary toxin close to iota.