Modification of the cell surface with neuraminidase increases the sensitivities of cells to diphtheria toxin and Pseudomonas aeruginosa exotoxin

A biologically active lectin of enteroaggregative Escherichia coli

The pathogenesis of enteroaggregative Escherichia coli, a major contributor to paediatric diarrhoea, is still not clearly understood. A complex carbohydrate specific lectin was identified from the culture supernatant of an enteroaggregative E. coli strain. The lectin was purified to 660-fold by a combination of sequential saturated ammonium sulphate precipitation and gel filtration chromatography in the FPLC system. The homogeneity of the purified lectin was established by analytical isoelectrofocusing [pI 6.75]. Hemagglutination of rabbit erythrocytes by the purified lectin was best inhibited by fetuin. The N-terminal sequence of the 41.7 kDa subunit showed homology to the outermembrane porins and the 23.4 kDa subunit showed homology to a hypothetical protein of Yersinia pestis and secreted Hcp protein. This protein could induce extensive morphological changes in HEp-2 cells and significant amount of fluid accumulation in rabbit ileal loop. GM1 showed maximum binding to the lectin among all other gangliosides. This purified protein showed cross-reactivity to the binding subunit of cholera toxin in western immunoblot. The presence of this toxin in some of the clinical isolates of enteroaggregative E. coli was also observed. The structural and functional characteristics of the toxin revealed that it is a novel virulence determinant of aggregative E. coli.

Diphtheria toxin translocation across endosome membranes. A novel cell permeabilization assay reveals new diphtheria toxin fragments in endocytic vesicles

By using cells overexpressing diphtheria toxin (DT) receptor and a novel method of permeabilizing the plasma membrane with a bacterial pore-forming toxin, specific translocation of fragment A to the cytosol was observed, whereas full-size DT and other minor species of DT-derived fragments were left in intracellular vesicles. The translocation competence of DT proteins with mutations in the transmembrane domain is consistent with their cytotoxicities. The charge-reversal mutants E349K and D352K do not translocate their fragment A to the cytosol, whereas D352N is partially competent. ADP-ribosyltransferase activity of fragment A is not required for translocation. Novel fragments of DT with apparent molecular masses of 28 and 35 kDa were detected in endocytic vesicles. The 28-kDa fragment consists of fragment A and an N-terminal piece of fragment B, whereas the 35-kDa fragment contains part of fragment B and may be complementary to the 28-kDa fragment. Time course studies show that the 28-kDa fragment appears in endocytic vesicles prior to translocation of fragment A to the cytosol, raising the possibility that the 28-kDa fragment is an intermediate in translocation. We present a model for translocation of fragment A that incorporates the observations made using the novel permeabilization method.

The different behavior of diphtheria toxin, modeccin and ricin in HeLa cells infected with Trypanosoma cruzi

We have studied the action of diphtheria toxin, modeccin and ricin on HeLa cells infected by Trypanosoma cruzi. Parasitized HeLa cells were resistant to diphtheria toxin and modeccin, whereas non-parasitized cells from the same cultures and control cultures showed cytopathological alterations. Protein synthesis, assayed by the incorporation of labelled methionine, diminished in toxin-treated control cultures but remained unaltered in the infected ones, compared to synthesis by untreated infected cells. Ricin, on the other hand, is a toxin that enters the cytoplasm by endocytosis. It has greater cytopathological effects in parasitized cells than in non-parasitized ones from the same cultures or uninfected control cells. Protein synthesis was inhibited in infected cultures treated with ricin.

Identification of diphtheria toxin receptor and a nonproteinous diphtheria toxin-binding molecule in Vero cell membrane

Two substances possessing the ability to bind to diphtheria toxin (DT) were found to be present in a membrane fraction from DT-sensitive Vero cells. One of these substances was found on the basis of its ability to bind DT and inhibit its cytotoxic effect. This inhibitory substance competitively inhibited the binding of DT to Vero cells. However this inhibitor could not bind to CRM197, the product of a missense mutation in the DT gene, and did not inhibit the binding of CRM197 to Vero cells. Moreover, similar levels of the inhibitory activity were observed in membrane fractions from DT-insensitive mouse cells, suggesting the inhibitor is not the DT receptor which is specifically present in DT-sensitive cells. The second DT-binding substance was found in the same Vero cell membrane preparation by assaying the binding of 125I-labeled CRM197. Such DT-binding activity could not be observed in membrane preparation from mouse L cells. From competition studies using labeled DT and CRM proteins, we conclude that this binding activity is due to the surface receptor for DT. Treatment of these substances with several enzymes revealed that the inhibitor was sensitive to certain RNases but resistant to proteases, whereas the DT receptor was resistant to RNase but sensitive to proteases. The receptor was solubilized and partially purified by chromatography on CM-Sepharose column. Immunoprecipitation and Western blotting analysis of the partially purified receptor revealed that a 14.5-kD protein is the DT receptor, or at least a component of it.

Role of glycosylation in expression of functional diphtheria toxin receptors

We have previously demonstrated, by using a detergent-solubilized system, the existence of specific diphtheria toxin-binding glycoproteins on the surface of toxin-sensitive cells. We have now tested the effect of tunicamycin treatment on the sensitivity of cells in culture to diphtheria toxin and have investigated the toxin sensitivity of mutant cells with known defects in glycosylation of asparagine-linked glycoproteins. Treatment of CHO-K1 cells with tunicamycin, which blocks the synthesis of both high-mannose-type and complex-type oligosaccharide chains of asparagine-linked glycoproteins, resulted in a 50- to 100-fold decrease in sensitivity to diphtheria toxin. In contrast, CHO-K1 mutants, defective in the synthesis of either high-mannose-type or complex-type oligosaccharides, showed no difference in toxin sensitivity compared with that of their parental cell lines. When we used an acid shock system, which is believed to result in receptor-dependent direct toxin penetration at the cell surface, the toxin sensitivity of tunicamycin-treated cells was not restored to that of untreated cells, suggesting that tunicamycin treatment results in a decrease in functional toxin receptors. Direct binding studies with 125I-labeled toxin demonstrated that this decrease in functional receptors is due to a decrease in the affinity of the receptors rather than to a change in the number of receptors. Taken together, these data are consistent with the interpretation that the diphtheria toxin receptor is a glycoprotein and suggest that the toxin binds neither to carbohydrate residues unique to the high-mannose-type oligosaccharides nor to those unique to the complex-type oligosaccharides. Furthermore, these data are consistent with the hypothesis that diphtheria toxin binds to the peptide backbone of the glycoprotein receptor.

Chromosomal assignment of the gene for human elongation factor 2

Elongation factor 2 (EF-2), polypeptidyl -tRNA translocase, is an essential factor for protein synthesis in eukaryotic cells and Archebacteria . We isolated diphtheria toxin-resistant human primary embryo cells that contain EF-2 that cannot be ADP-ribosylated by diphtheria toxin and Pseudomonas exotoxin A (PA toxin). Somatic cell hybrids were constructed from mouse L cells and toxin-resistant human embryo cell mutants. Forty-one hybrid clones were isolated, of which 15 clones were resistant to PA toxin. Karyotypic analysis and isozyme studies revealed that there was an absolute correlation between human chromosome 19 and resistance to PA toxin in the hybrids. On subcloning of PA toxin-resistant hybrid cells, we obtained one PA toxin-resistant hybrid subclone containing human chromosome 19 as the only human chromosome. Furthermore, the resistance to PA toxin of hybrid cell strains was lost after infection with poliovirus, for which sensitivity is conferred by human chromosome 19. It was confirmed by using two-dimensional gel electrophoresis that PA toxin resistance in hybrid cells was caused by the presence of EF-2 resistant to ADP-ribosylation by fragment A of diphtheria toxin. These facts suggest that the gene encoding EF-2 is located on human chromosome 19.

Entry of diphtheria toxin into cells: possible existence of cellular factor(s) for entry of diphtheria toxin into cells was studied in somatic cell hybrids and hybrid toxins

Ehrlich ascites tumor cells were found to be very insensitive to diphtheria toxin. We formed 37 hybrids from Ehrlich tumor cells and diphtheria toxin-sensitive human fibroblasts. The effects of diphtheria toxin on protein synthesis in those hybrids were examined. The hybrids were divided into three groups on the basis of toxin sensitivity. Group A hybrids were as sensitive to diphtheria toxin as human fibroblasts, Group C were as resistant as Ehrlich tumor cells, and Group B had intermediate sensitivity. Group A hybrids had diphtheria toxin-binding sites but Group B and C had no detectable binding sites. Elongation factor-2 of all the hybrids was susceptible to ADP-ribosylation by fragment A of diphtheria toxin. Cells of Group A and B became more sensitive to CRM 45 (cross-reacting material 45 of diphtheria toxin) after they were exposed to low pH (pH = 4.5). The resistance of Group C to CRM 45 was not affected by the same treatment. Group A and B hybrids and human fibroblasts had similar sensitivities to a hybrid toxin composed of wheat germ agglutinin and fragment A of diphtheria toxin, but Group C and Ehrlich tumor cells were resistant to this hybrid toxin. All the hybrids and Ehrlich tumor cells were more sensitive to a hybrid toxin composed of wheat germ agglutinin and subunit A of ricin than were human fibroblasts. On subcloning of Group B hybrids, one Group C hybrid was obtained, but no Group A hybrid. These facts suggest that Ehrlich ascites tumor cells differ from human fibroblasts in the expression of a factor(s) that is involved in entry of fragment A of diphtheria toxin into the cytoplasm after the toxin binds to its surface receptors.

Quantitation of binding and subcellular distribution of Clostridium perfringens enterotoxin in rat liver cells

Binding of enterotoxin from Clostridium perfringens type A was studied in suspensions of parenchymal and nonparenchymal cells from rat liver. In hepatocytes, 1.5 X 10(6) specific binding sites per cell with an association constant of 3.2 X 10(6) M-1 were found. About 1% of the added toxin was nonspecifically bound to the hepatocytes. At concentrations of toxin below 0.1 micrograms/ml, 80% of the toxin density of 7 X 10(6) cells per ml. Binding did not increase after the cells became permeable to the toxin. Subcellular fractionation in a sucrose gradient produced no evidence for binding to parts of the cell other than the plasma membrane. The degree of binding to nonparenchymal cells was less than 10% of the binding to hepatocytes.