Use of site-directed mutagenesis to probe structure-function relationships of alpha-toxin from Clostridium perfringens.

The NH2-terminal domain of the alpha-toxin of Clostridium perfringens is highly homologous to the complete phospholipase C from Bacillus cereus (PC-PLC), for which a high-resolution crystal structure is available. This structural information was used as the basis of a site-directed mutagenesis strategy in which critical amino acid residues of alpha-toxin involved in zinc binding, interaction with substrate, or catalysis were replaced. Biochemical studies with the corresponding toxin variants indicate that there is probably a single active site endowed with lecithinase, sphingomyelinase, and hemolytic activities. By using a highly purified variant in which the catalytic aspartate residue at position 56 was replaced by asparagine, it was shown that phospholipase activity was essential for lethality in vivo and for mediating platelet aggregation in vitro.     

Role of alpha-toxin in Clostridium perfringens infection determined by using recombinants of C. perfringens and Bacillus subtilis.

Clostridium perfringens type A strains which differed in alpha-toxin (phospholipase C [PLC]) productivity were inoculated intraperitoneally or intravenously into mice, and then their 50% mouse lethal doses (LD50) were determined. Strain NCTC 8237 produced ninefold higher PLC activity than strain 13. The mean LD50 for the former was 1 log unit lower than that for the latter. Two isogenic strains were constructed from strain 13: strain 13(pJIR418 alpha) (pJIR418 alpha contains the plc gene), which produced ninefold higher PLC activity than strain 13; and strain 13 PLC-, which showed no PLC productivity at all because of transformation-mediated gene disruption. The mean LD50 for strain 13(pJIR418 alpha) was 1 log unit lower than those for strain 13 PLC- and strain 13. These results indicate that PLC functions as a virulence-determining factor when it is produced in a sufficient amount. Such a difference in LD50 was also observed between Bacillus subtilis with and without the cloned plc gene. Inoculation of B. subtilis PLC+ intravenously into mice caused marked thrombocytopenia and leukocytosis. Mice inoculated with B. subtilis at 2 LD50 died because of circulatory collapse. Histological examination revealed that intravascular coagulation and vascular congestion occurred most prominently in the lungs. These results suggest that PLC plays a key role in the systemic intoxication of clostridial myonecrosis, probably by affecting the functions of platelets and phagocytes.     

Clostridium sordellii phospholipase C: gene cloning and comparison of enzymatic and biological activities with those of Clostridium perfringens and Clostridium bifermentans phospholipase C

The gene encoding Clostridium sordellii phospholipase C (Csp) was cloned and expressed as a histidine-tagged (His-tag) protein, and the protein was purified to compare its enzymatic and biological activities with those of Clostridium perfringens phospholipase C (Cpa) and Clostridium bifermentans phospholipase C (Cbp). Csp was found to consist of 371 amino acid residues in the mature form and to be more homologous to Cbp than to Cpa. The egg yolk phospholipid hydrolysis activity of the His-tag Csp was about one-third of that of His-tag Cpa, but the hemolytic activity was less than 1% of that of His-tag Cpa. His-tag Csp was nontoxic to mice. Immunization of mice with His-tag Cbp or His-tag Csp did not provide effective protection against the lethal activity of His-tag Cpa. These results indicate that Csp possesses similar molecular properties to Cbp and suggest that comparative analysis of toxic and nontoxic clostridial phospholipases is helpful for characterization of the toxic properties of clostridial phospholipases.     

Site-specific mutagenesis of the catalytic subunit of cholera toxin: substituting lysine for arginine 7 causes loss of activity.

Cholera and pertussis toxins each contain a subunit with ADP-ribosyltransferase activity, sharing a region of nearly identical amino acid sequence near the NH2 terminus. Previous investigations have shown that substitution of a lysine residue for Arg-9 in the catalytic A subunit of pertussis toxin substantially eliminates its enzyme activity. We now report that substitution of lysine for the position-equivalent Arg-7 of cholera toxin subunit A leads to a similar loss of catalytic activity. This result suggests a correlation of function with structure between the sequence-related cholera and pertussis toxin A subunits and may contribute to the design of a vaccine containing an enzymatically inert analog of cholera toxin.     

Molecular cloning and nucleotide sequence of the alpha-toxin (phospholipase C) of Clostridium perfringens.

A fragment of DNA containing the gene coding for the phospholipase C (alpha-toxin) of Clostridium perfringens was cloned into Escherichia coli. The cloned DNA appeared to code only for the alpha-toxin and contained both the coding region and its associated gene promoter. The nucleotide sequence of the cloned DNA was determined, and an open reading frame was identified which encoded a protein with a molecular weight of 42,528. By comparison of the gene sequence with the N-terminal amino acid sequence of the protein, a 28-amino-acid signal sequence was identified. The gene promoter showed considerable homology with the E. coli sigma 55 consensus promoter sequences, and this may explain why the gene was expressed by E. coli. The cloned gene product appeared to be virtually identical to the native protein. A 77-amino-acid stretch that was close to the N terminus of the alpha-toxin showed considerable homology with similarly located regions of the Bacillus cereus phosphatidylcholine, preferring phospholipase C and weaker homology with the phospholipase C from Pseudomonas aeruginosa.     

Effect of zinc and calcium ions on the production of alpha-toxin and proteases by Clostridium perfringens.

Clostridium perfringens produced at least three distinct proteases in a synthetic medium containing calcium. Two of them, thiol and ethylenediaminetetraacetic acid disodium salt-sensitive proteases, appeared at an early stage of growth, but the other one, perhaps being identical to the one produced in a calcium-deficient medium, appeared at a late stage. The production of these proteases depended on Ca2+ but not on Zn2+ in the medium. Alpha-toxin, perhaps being a zinc-containing metalloenzyme, was rather resistant to the proteases, but toxin, produced in a zinc-deficient medium or deprived of zinc with ethylenediaminetetraacetic acid disodium salt, was very sensitive. By adding Zn2+, the toxin lacking zinc may have been converted to the zinc-containing metalloprotein that is resistant to proteases. This may explain why alpha-toxin activity increased progressively in a zinc-containing medium in spite of simultaneous production of potent proteases and why it disappeared rapidly in a zinc-deficient medium.     

Molecular genetic analysis of beta-toxin of Clostridium perfringens reveals sequence homology with alpha-toxin, gamma-toxin, and leukocidin of Staphylococcus aureus.

Oligonucleotide probes designed on the basis of the N-terminal sequence of Clostridium perfringens beta-toxin were used to isolate the encoding gene (cpb). The nucleotide sequence of cpb was determined, and on the basis of DNA hybridization experiments it was shown that the gene is found only in type B and C strains of C. perfringens. The deduced amino acid sequence of the beta-toxin revealed homology with the alpha-toxin, gamma-toxin, and leukocidin of Staphylococcus aureus. The beta-toxin purified from C. perfringens appeared to exist in monomeric and multimeric forms. Recombinant beta-toxin, produced in Escherichia coli, appeared to be mainly in the multimeric form.     

Measurement of biological activities of purified and crude enterotoxin of Clostridium perfringens.

Enterotoxin of Clostridium perfringens was assayed and compared with toxicity in mice and erythemal activity in guinea pigs. Conversion factors were used to express these biological activities of crude enterotoxin in terms of weight of pure enterotoxin protein. One microgram of enterotoxin was equivalent to 3.41 erythema units and to 0.68 mouse median lethal dose.     

Hemolytic and sphingomyelinase activities of Clostridium perfringens alpha-toxin are dependent on a domain homologous to that of an enzyme from the human arachidonic acid pathway.

The N-terminal domain of Clostridium perfringens alpha-toxin, homologous with the nontoxic phospholipase C of Bacillus cereus, was expressed in Escherichia coli and shown to retain all of the phosphatidylcholine hydrolyzing activity of the alpha-toxin, but not the sphingomyelinase, hemolytic, or lethal activities. The C-terminal domain of alpha-toxin showed sequence and predicted structural homologies with the N-terminal region of arachidonate 5-lipoxygenase, an enzyme from the human arachidonic acid pathway which plays a role in inflammatory and cardiovascular diseases in humans.     

Effects of Clostridium perfringens alpha-toxin (PLC) and perfringolysin O (PFO) on cytotoxicity to macrophages, on escape from the phagosomes of macrophages, and on persistence of C. perfringens in host tissues

Clostridium perfringens is the most common cause of clostridial myonecrosis (gas gangrene). Polymorphonuclear cells (PMNs) appear to play only a minor role in preventing the onset of myonecrosis in a mouse animal model of the disease (unpublished results). However, the importance of macrophages in the host defense against C. perfringens infections is still unknown. Two membrane-active toxins produced by the anaerobic C. perfringens, alpha-toxin (PLC) and perfringolysin O (PFO), are thought to be important in the pathogenesis of gas gangrene and the lack of phagocytic cells at the site of infection. Therefore, C. perfringens mutants lacking PFO and PLC were examined for their relative cytotoxic effects on macrophages, their ability to escape the phagosome of macrophages, and their persistence in mouse tissues. C. perfringens survival in the presence of mouse peritoneal macrophages was dependent on both PFO and PLC. PFO was shown to be the primary mediator of C. perfringens-dependent cytotoxicity to macrophages. Escape of C. perfringens cells from phagosomes of macrophage-like J774-33 cells and mouse peritoneal macrophages was mediated by either PFO or PLC, although PFO seemed to play a more important role in escape from the phagosome in peritoneal macrophages. At lethal doses (10(9)) of bacteria only PLC was necessary for the onset of myonecrosis, while at sublethal doses (10(6)) both PFO and PLC were necessary for survival of C. perfringens in mouse muscle tissue. These results suggest PFO-mediated cytotoxicity toward macrophages and the ability to escape macrophage phagosomes may be important factors in the ability of C. perfringens to survive in host tissues when bacterial numbers are low relative to those of phagocytic cells, e.g., early in an infection.     

Stable L-Forms of Clostridium perfringens: Growth, Toxin Production, and Pathogenicity

Growth and toxin production of stable L-forms of Clostridium perfringens grown in a mini-fermentor were monitored. A gradual but steady increment in viable count occurred over a 7-h period, followed by death. The peak of viability preceded the optical density peak by 3 h. Theta, alpha, kappa, and lambda toxins were measured, with theta toxin appearing first in the culture supernate. Growth of the parent bacillus form of C. perfringens was compared under similar conditions. Toxin levels achieved by the bacillus culture exceeded those of the L-form culture four- to eightfold; however, based upon viable count, the L-form organism produced 8 to 16 times as much toxin as did the bacillus. The amounts of extracellular toxin produced by both forms were similar when related to cell protein rather than cell number. Guinea pig inoculation showed that the L-form of C. perfringens did not produce gas gangrene, although it was not entirely without effect. Both guinea pig and human sera were inhibitory to these L-forms, a fact attributable to a heat-liable component in the sera, most likely complement.     

Identification and antimicrobial resistance patterns of clinical isolates of Clostridium clostridioforme, Clostridium innocuum, and Clostridium ramosum compared with those of clinical isolates of Clostridium perfringens.

Clostridium ramosum, C. innocuum, and C. clostridioforme are frequently isolated from clinical specimens including blood. Because of Gram stain variability, a lack of spores, and atypical colonial morphology, identification of these species is often difficult. Three anaerobe identification kits were evaluated for their abilities to identify these species. For comparison, 11 strains of C. perfringens were evaluated in parallel. By using profile numbers and codebooks, the correct genus and species were identified, as follows: with the RapID ANA II kit, 100% (20 of 20) of C. ramosum isolates, 24% (5 of 21) of C. innocuum isolates, and 50% (10 of 20) of C. clostridioforme isolates; with the AnIDent kit, 60% (12 of 20) of C. ramosum isolates, 28% (6 of 21) of C. innocuum isolates, and 90% (18 of 20) of C. clostridioforme isolates; with the ATB32A kit, 70% (14 of 20) of C. ramosum isolates, 0% (0 of 21) of C. innocuum isolates, and 40% (8 of 20) of C. clostridioforme isolates. Profile numbers that overlapped several species were obtained as follows: with the RapID ANA II kit, 0% of C. ramosum isolates, 76% of C. innocuum isolates, and 40% of C. clostridioforme isolates; with the AnIDent kit 40% of C. ramosum isolates, 62% of C. innocuum isolates, and 5% of C. clostridioforme isolates; with the ATB32A kit, 15% of C. ramosum isolates, 52% of C. innocuum isolates, and 25% of C. clostridioforme isolates. One strain of C. innocuum was misidentified by the AnIDent kit, and the remainder yielded profile numbers that were not listed in the codebooks. The MICs of 11 antimicrobial agents including penicillin G, metronidazole, clindamycin, cefoxitin, cefotetan, imipenem, meropenem, amoxicillin-clavulanate, ampicillin-sulbactam, piperacillin-tazobactam, and vancomycin were determined by the agar dilution method. All C. perfringens strains were susceptible to all antimicrobial agents tested. Various levels of resistance to cefoxitin, cefotetan, and penicillin G were noted with C. ramosum, C. clostridioforme, and C. innocuum. In addition, resistance to clindamycin was noted with C. ramosum (5%) and C. innocuum (10%). Most strains of C. innocuum were only moderately susceptible to vancomycin (MIC at which 90% of strains are inhibited, 4 micrograms/ml).     

Purification, characterization, and primary structure of Clostridium perfringens lambda-toxin, a thermolysin-like metalloprotease.

The lambda-toxin of Clostridium perfringens type B NCIB10691 was purified by ammonium sulfate precipitation, followed by size exclusion, anion-exchange, and hydrophobic interaction chromatography. The purified toxin had an apparent molecular mass of 36 kDa, as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The toxin possessed casein-hydrolyzing activity, which was inhibited specifically by metal chelators, indicating that the toxin is a metalloprotease. The gene encoding the lambda-toxin (lam), which was shown by Southern analysis to be located on a 70-kb plasmid, was cloned into Escherichia coli cells. Nucleotide and N-terminal amino acid sequencing revealed that the lam gene encodes a 553-amino-acid protein, which is processed into a mature form, the molecular mass of which was calculated to be 35,722 Da. The deduced amino acid sequence of the mature enzyme contains an HEXXH motif characteristic of zinc metalloproteases and is homologous to other known enzymes belonging to the thermolysin family. The purified toxin degraded various biologically important substances, such as collagen, fibronectin, fibrinogen, immunoglobulin A, and the complement C3 component. It caused an increase in vascular permeability and hemorrhagic edema on injection into the dorsal skin of mice. These results suggest that the toxin contributes to the pathogenesis of histolytic infection by lambda-toxin-producing C. perfringens.     

Cloning, nucleotide sequencing, and expression of the Clostridium perfringens enterotoxin gene in Escherichia coli.

A complete copy of the gene (cpe) encoding Clostridium perfringens enterotoxin (CPE), an important virulence factor involved in C. perfringens food poisoning and other gastrointestinal illnesses, has been cloned, sequenced, and expressed in Escherichia coli. The cpe gene was shown to encode a 319-amino-acid polypeptide with a deduced molecular weight of 35,317. There was no consensus sequence for a typical signal peptide present in the 5' region of cpe. Cell lysates from recombinant cpe-positive E. coli were shown by quantitative immunoblot analysis to contain moderate amounts of CPE, and this recombinant CPE was equal to native CPE in cytotoxicity for mammalian Vero cells. CPE expression in recombinant E. coli appeared to be largely driven from a clostridial promoter. Immunoblot analysis also demonstrated very low levels of CPE in vegetative cell lysates of enterotoxin-positive C. perfringens. However, when the same C. perfringens strain was induced to sporulate, much stronger CPE expression was detected in these sporulating cells than in either vegetative C. perfringens cells or recombinant E. coli. Collectively, these results strongly suggest that sporulation is not essential for cpe expression, but sporulation does facilitate high-level cpe expression.     

Synergistic haemolysis test for presumptive identification and differentiation of Clostridium perfringens, C. bifermentans, C. sordellii, and C. paraperfringens.

A new test for the presumptive identification of Clostridium perfringens, C. bifermentans, C. sordellii, and C. paraperfringens is described. The test is based on the synergistic haemolysis shown by the clostridia and group B streptococci on sheep and human and CaCl2-supplemented human blood agar. C. perfringens gave crescent-shaped synergistic lytic zones (7 to over 20 mm in length), and C. paraperfringens usually small-sized (3 mm), bullet-shaped reactions on all three types of media. C. bifermentans showed a horseshoe-shaped synergistic reaction only on human blood containing media, and C. sordellii only on CaCl2-supplemented human blood agar. C. perfringens type A antiserum inhibited synergistic lytic activities of the four species. The test provided a reliable method for presumptive identification and differentiation of the four clostridial species and may obviate the need for the Nagler test.     

Evaluation of ELISA, RPLA, and Vero cell assays for detecting Clostridium perfringens enterotoxin in faecal specimens.

Three hundred and ninety two faecal specimens from 70 separate outbreaks of suspected Clostridium perfringens food poisoning were examined by enzyme linked immunosorbent assay (ELISA), reversed passive latex agglutination (RPLA), and Vero cell assays for the presence of enterotoxin. Although the most time consuming method, ELISA was the most specific and reproducible. RPLA was slightly more sensitive than ELISA, but it showed some non-specific reactions. The Vero cell assay was the least sensitive and least reproducible method, being affected by some non-specific cytotoxic and cytotonic reactions. Normal rabbit serum should be included in the Vero cell assay as a control for the neutralisation of cytotoxic effects.     

Site-directed mutagenesis of Glu-141 and His-223 in Pseudomonas aeruginosa elastase: catalytic activity, processing, and protective activity of the elastase against Pseudomonas infection.

Both Pseudomonas aeruginosa elastase and Bacillus thermoproteolyticus thermolysin are zinc metalloproteases. On the basis of the high homology of the P. aeruginosa elastase with the Bacillus thermolysin, we hypothesized that Glu-141 and His-223 are the key residues for catalytic activity of the Pseudomonas elastase. To test this possibility, we replaced Glu-141 with Asp, Gln, and Gly and His-223 with Gly, Glu, and Leu by site-directed mutagenesis. These substitutions dramatically diminished the proteolytic activities of the mutant elastases when they were expressed in Escherichia coli cells. Although these mutant elastase precursors (proelastases) were produced, no appreciable processing was observed with these mutants. The possibility that autocatalysis is involved in both the processing and activation of elastase is discussed. Furthermore, by immunizing mice with vaccines made from these mutant elastase, we were able to obtain good protection against an intraperitoneal P. aeruginosa challenge.     

The effects of Clostridium perfringens enterotoxin on rat and rabbit ileum: an electron microscopic study.

Intestinal epithelial damage caused by Clostridium perfringens enterotoxin in rats and rabbits was identified by light microscopy and compared at the surface (scanning electron microscopy), and the ultrastructural (transmission electron microscopy) levels. Under the light microscope damage to the epithelial layer of villus tips was clearly evident in cross-sections. Whole tissue viewed under the scanning electron microscope showed comparable tip localization of morpholigic damage in the form of collapsed tips and a dense covering of rounded blebs on the tips. Ulstructuctural observations included partial and sometimes complete disappearance of microvilli structures, budding of the terminal web region into the lumen, and even complete destruction of epithelial cells. These data suggest that C. perfringens enterotoxin attacks the epithelial cells with a preference for cells at the villus tips and causes damage at least in part by altering the cells' apical membranes. This then leads to cellular sloughing, death, and lysing.