Mechanism of membrane damage by Clostridium perfringens alpha-toxin.

The effect of Clostridium perfringens alpha-toxin on liposomes prepared from phosphatidylcholine (PC) containing the fatty acyl residues of 18 carbon atoms was investigated. The toxin-induced carboxyfluorescein (CF) leakage and phosphorylcholine release from multilamellar liposomes increased as the phase transition temperature of the phosphatidylcholines containing unsaturated fatty acyl residues decreased. However, there was no difference between the sensitivity of the different phosphatidylcholines solubilized by deoxycholate to the phospholipase C (PLC) activity of the toxin. However, the toxin did not hydrolyze solubilized distearoyl-L-alpha-phosphatidylcholine (DSPC) or phosphatidylcholine containing saturated fatty acyl residue, and caused no effect on liposomes composed of DSPC. These results suggest that the activity of the toxin is closely related to the membrane fluidity and double bond in PC. The N-terminal domain of alpha-toxin (AT1-246) and variant H148G did not induce CF leakage from liposomes composed of dioleoyl-L-alpha-phosphatidylcholine (DOPC). H148G bound to the liposomes, but AT1-246 did not. However, the C-terminal domain (AT251-370) conferred binding to liposomes and the membrane-damaging activity on AT1-246. These observations suggest that the membrane-damaging action of alpha-toxin is due to the binding of the C-terminal domain of the toxin to the double bond in the PC in the bilayer and hydrolysis of the PC by the N-terminal domain.

Site-specific mutagenesis of Clostridium perfringens alpha-toxin: replacement of Asp-56, Asp-130, or Glu-152 causes loss of enzymatic and hemolytic activities.

The current study has investigated the role of D-56, D-130, and E-152 in zinc ion binding properties, as well as the hemolytic, phospholipase C (PLC), and sphingomyelinase (SMase) activities of Clostridium perfringens alpha-toxin, based upon crystallography studies of the Bacillus cereus PLC, which had suggested these residues might be important for these functional activities. The replacement of D-56 in alpha-toxin resulted in complete loss of hemolytic, PLC, and SMase activities. The variant toxins at D-130 showed an approximately 100-fold reduction of biological activities compared to that of the wild-type toxin. The substitution of glutamine or glycine for E-152 caused complete loss of these activities, but substitution of aspartic acid for E-152 reduced but did not completely inhibit these activities. The variant toxins at D-56 and D-130, as well as the wild-type toxin, possessed approximately 2 mol of zinc atoms per mol of the protein, but E152G and E152Q contained approximately 1 mol of zinc metal per mol of the protein. On the other hand, the zinc content in E152D was calculated as about 1.4 mol in the toxin molecule. The replacement of D-56, D-130, or E-152 had no effect on binding to sheep erythrocytes and uptake of free zinc ion from the solution. The variant toxins at D-130 showed partial antigenic identity with the wild-type toxin on a double gel diffusion test. These observations suggest that D-56 in alpha-toxin is required for catalytic activity of alpha-toxin, D-130 is essential for maintenance of structure, and the carboxyl group of E-152 tightly ligands one zinc ion, which is essential for catalytic activity of the toxin.

Production and purification of Clostridium perfringens alpha-toxin using a protein-hyperproducing strain, Bacillus brevis 47.

Clostridium perfringens alpha-toxin was produced in a protein-hyperproducing strain, Bacillus brevis 47, by cloning the gene into the constructed expression-secretion vector which has the multiple promoters and the signal peptide coding region of an outer cell wall protein gene. The amount of alpha-toxin produced by the B. brevis 47 transformant carrying the gene was approximately 10 times greater than that produced by a B. subtilis transformant carrying the toxin gene. Biological activities and the N-terminal amino acid sequence of the toxin secreted by the B. brevis 47 transformant were identical to those of wild-type alpha-toxin.

Membrane-damaging action of Clostridium perfringens alpha-toxin on phospholipid liposomes.

The effect of Clostridium perfringens alpha-toxin on multilamellar liposomes prepared from various phospholipids and cholesterol was investigated. The toxin induced carboxyfluorescein leakage from liposomes composed of the choline-containing phospholipids such as egg-yolk phosphatidylcholine and bovine brain sphingomyelin in dose-dependent manner, but did not induce leakage from those liposomes composed of bovine brain phosphatidylethanolamine, egg-yolk phosphatidylserine or phosphatidylglycerol. The toxin-induced carboxyfluorescein leakage from egg-yolk phosphatidylcholine liposomes was increased by addition of divalent cations. The toxin induced carboxyfluorescein release from liposomes composed of phosphatidylcholine containing unsaturated fatty acyl residues or shorter chain length saturated fatty acyl residues (12 or 14 carbon atoms), but did not induce such release from liposomes composed of phosphatidylcholine containing saturated fatty acyl residues of between 16 and 20 carbon atoms. Furthermore, the toxin-induced carboxyfluorescein release decreased with increasing chain length of acyl residues of phosphatidylcholine used. The toxin bound to liposomes composed of phospholipids which are hydrolyzed by the toxin, but did not bind to those composed of phospholipids which are not attacked by the toxin. The toxin-induced carboxyfluorescein release from liposomes composed of dipalmitoleoyl-L-alpha-phosphatidylcholine and cholesterol and the toxin binding to the liposomes decreased with decreasing cholesterol contents. These observations suggest that the specific binding site formed by the choline-containing phospholipids and cholesterol, and membrane fluidity in liposomes are essential for the membrane-damaging activity of alpha-toxin.

Synthesis and damage specificity of a novel probe for the detection of abasic sites in DNA.

The abasic site (apurinic/apyrimidinic site) is the most common lesion in DNA and is suggested to be an important intermediate in mutagenesis and carcinogenesis. We have recently reported a novel assay for the detection and quantitation of abasic sites in DNA [Kubo, K., Ide, H., Wallace, S. S., & Kow, Y. W. (1992) Biochemistry 31, 3703-3708]. In this assay, the aldehyde group in an abasic site is first modified by a probe bearing a biotin residue, called the Aldehyde Reactive Probe (ARP) and then the tagged biotin is quantified by an ELISA-like assay. However, in the previous study, ARP was prepared only in a crude form, and no solid chemical data concerning the structure and specificity of ARP were reported. In this study, an improved method for the preparative synthesis of ARP has been established, and its structure has been unambiguously characterized using spectroscopic means. In order to elucidate the specificity of ARP to DNA damages, ARP was incubated with a variety of damaged bases or nucleosides and the reaction mixtures were analyzed by HPLC. Of the 14 compounds tested for their reactivity to ARP, 2-deoxyribose (a model compound for an abasic site) and 5-formyluracil reacted with ARP. Interestingly, compounds bearing a formamide group such as formamidopyrimidine and deoxyribosylformamide did not react with ARP, indicating that ARP is specific to damages having an alkyl or allyl aldehyde group. Furthermore, the ability of ARP synthesized by the defined chemical route to detect abasic sites has been substantiated using natural DNA containing abasic sites. Potential applications and limitations of the ARP assay are discussed.