- The cytolytic properties of the mosquitocidal 25 Kd protein deltaendotoxin of Bacillus thuringiensis subsp. israelensis (Bti) was studied using the human erythrocyte as a target cell. Both a fixedtime and a kinetic hemolytic bioassay were developed in order to screen a variety of treatments for their influence on the toxinerythrocyte interaction. Upon addition of 4 to 8 nM toxin to 2.5 X 10⁷ cells/ml, at 37°C, after an initial 10 minute lag, hemolysis began and reached a maximum value in 60 minutes. The extent of hemolysis was independent of cell concentration over an 18.4-fold range. Toxin was equally active from pH 6 to 9.5, but hemolysis was reduced at pH 5 and 10. Lowering the pH from 10 to 9 restored about 40% of the activity. Hemolysis was temperature dependent; lowering the temperature below 37°C caused an increase in the reaction lag time and a reduction in total hemolysis in 60 minutes. Raising the temperature to 47°C did not alter the extent or kinetics of hemolysis as compared to 37°C. Potassium release preceded hemoglobin release, suggesting that if a pore was formed, it was not stable for a long period of time. Hemoglobin release followed a single phase first-order efflux, but potassium efflux appeared to be multiphasic. 40 mM calcium inhibited hemolysis, regardless of the time that it was added during the assay. Magnesium also inhibited hemolysis but it was 6 times less potent than calcium. However, very low concentrations of magnesium stimulated lysis. When toxin was removed in the early stages of the assay, a reduction of hemolysis was seen, which suggested that the toxin was binding weakly to the membrane. From the lack of the effect of DTT, EDTA, and EGTA on hemolysis, it appeared that the toxin-erythrocyte interaction did not require intact disulfide bonds or divalent metals. Lack of hemolysis inhibition by BSA and sucrose suggested the lysis was not due to a colloid osmotic effect. Substitution of KC1 for NaCl did not alter lysis kinetics. Addition of ATP to the assay did not alter hemolysis, suggesting that positive charges on the toxin or erythrocyte were not required at pH 7.4. Addition of a purified monoclonal antibody to the Bti toxin did not inhibit lysis. Bti toxin was not interacting with a sialic acid-containing receptor because neuraminidase-treated erythrocytes failed to alter the extent of lysis. The target for the Bti toxin was neither an ion channel nor the Na⁺/K⁺ ATPase because specific ion channel and pump antagonists failed to inhibit lysis. This finding was also supported through the use of ATP-depleted cells, which did not affect lysis, suggesting that cell lysis was not dependent on an ATP-requiring membrane pump. Biotinylated toxin was biologically inactive on erythrocytes, suggesting that lysines were essential for activity. Toxin labelled with FITC demonstrated some binding to lysed cells, but not to intact cells. Toxin was also able to lyse nucleated erythrocytes, (chicken cells), but they were less sensitive than human cells. The Bti toxin interaction with Aedes albopictus insect cells was sensitive to heat at 50°C for 50 minutes. NaCl protected this inactivation at 50°, but not at 60°C. The heat-denatured toxin may have been aggregated, as it failed to migrate into a native polyacrylamide gel and it bound 2.5 times more Coomassie blue dye. Protein structure changed with heat treatment, as the fluorescence spectra showed a decrease in peak height but no change in the wavelength. The toxin-erythrocyte interaction was also heatsensitive, as 50°C for 60 minutes inactivated the toxin by 40%; no renaturation was seen for up to 4 hours after heat removal. The Bti toxin interaction with insect cells was also sensitive to 6 M urea treatment as the physical properties of the protein changed, causing a loss of fluorescence peak height, a red wavelength shift, and loss of biologic activity. Upon urea removal, there was a partial restoration of the fluorescence parameters, but not of the biological activity. Excess dye-binding and lack of gel migration suggested, like heat denaturation, that the toxin was forming an aggregate or undergoing some other type of structural change. Bti toxin was examined for phospholipase C activity which proved to be negative. A tryptophan content analysis of the Bti toxin showed three tryptophans per toxin molecule. From all of these studies and reports of other investigators, a tentative model for the mode of action of the Bti toxin-erythrocyte lysis process is suggested, which provides a bests for additional studies.