Comparative cellular and molecular responses of insect and human cells exposed to commercial Bacillus thuringiensis bioinsecticides

This thesis describes a detailed in vitro toxicological analysis of commercial Bacillus thuringienisis-based biopesticides (BT) and subcomponents, using bioassays involving insect and human cell lines as targets. The BT products are concentrated cultures, containing B. t. kurstaki (Btk) or B. t. israelensis (Bti) spores (>109/mL), variable amounts of cellular debris and intact parasporal inclusion bodies (PEBs), and polypeptides mainly related to delta-endotoxins. Simulated field doses of 10--7 to 10--3 International Units per target cell were used to study responses in: morphology, monitored by phase contrast and electron microscopy; bioreduction with reporter dyes; protein levels and synthesis, measured by immunoprobes and 35S-methionine incorporation, and DNA integrity. With antibiotic to inhibit spore activity, cell changes were affected by only BT doses >10--3 I.U./cell. Without antibiotic, BT exhibited dose-lag/dose-dependent damage caused by vegetative Bt cells, resulting in target cell shedding, loss in bioreduction, protein markers and DNA integrity, and appearance of nuclear pyknosis and cytolysis. BT dose-induced damage of insect and human cell types was remarkably similar, differing only in response time due to assay temperature (27°C versus 37°C) which affected Btk or Bti proliferation rate. Equivalent exposures with Bacillus subtilis and Escherichia coli produced minor effects. Cytotoxicity of BT supernatants free of spores and PIBs was seen in 2-h bioassays at doses equivalent to ≥ 1 I.U./cell. This activity was ≤ 0.08% of that of vegetative cell culture filtrates (CF) derived from BT products (spores). 35S-methionine labelling of target cells showed that CF-exposure rapidly arrested protein synthesis. Studies with intact PEBs, proteins derived from them (activated delta-endotoxins), Bt phosphatidylinositol phosphohpase C, and Bacillus cereus enterotoxin, all showed that CF was the most toxic agent. This CF toxicity could be abolished by pretreatment with a broad-spectrum protease or heat ≥ 60°C. Column chromatography of CF gave several activity peaks, but 32P-NAD-mediated ribosylation revealed only one product of 45 kDa. These results substantiate that toxification of BT products can be practically compared and quantified by various bioindicators in vitro, and that the most toxic effect arises as an outcome of infection of either insect or human cells by even a single spore ( ∼ 10--9 to 10--10 I.U./cell). The finding that vegetative cells of Btk and Bti produce the most potent cytotoxins of all BT product subcomponents questions simplistic views on mode of action attributed to existing BT products when used in environmental applications.