Contributions of gut bacteria to Bacillus thuringiensis-induced mortality vary across a range of Lepidoptera

Background: Gut microbiota contribute to the health of their hosts, and alterations in the composition of this microbiota can lead to disease. Previously, we demonstrated that indigenous gut bacteria were required for the insecticidal toxin of Bacillus thuringiensis to kill the gypsy moth, Lymantria dispar. B. thuringiensis and its associated insecticidal toxins are commonly used for the control of lepidopteran pests. A variety of factors associated with the insect host, B. thuringiensis strain, and environment affect the wide range of susceptibilities among Lepidoptera, but the interaction of gut bacteria with these factors is not understood. To assess the contribution of gut bacteria to B. thuringiensis susceptibility across a range of Lepidoptera we examined larval mortality of six species in the presence and absence of their indigenous gut bacteria. We then assessed the effect of feeding an enteric bacterium isolated from L. dispar on larval mortality following ingestion of B. thuringiensis toxin. Results: Oral administration of antibiotics reduced larval mortality due to B. thuringiensis in five of six species tested. These included Vanessa cardui (L.), Manduca sexta (L.), Pieris rapae (L.) and Heliothis virescens (F.) treated with a formulation composed of B. thuringiensis cells and toxins (DiPel), and Lymantria dispar (L.) treated with a cell-free formulation of B. thuringiensis toxin (MVPII). Antibiotics eliminated populations of gut bacteria below detectable levels in each of the insects, with the exception of H. virescens, which did not have detectable gut bacteria prior to treatment. Oral administration of the Gram-negative Enterobacter sp. NAB3, an indigenous gut resident of L. dispar, restored larval mortality in all four of the species in which antibiotics both reduced susceptibility to B. thuringiensis and eliminated gut bacteria, but not in H. virescens. In contrast, ingestion of B. thuringiensis toxin (MVPII) following antibiotic treatment significantly increased mortality of Pectinophora gossypiella (Saunders), which was also the only species with detectable gut bacteria that lacked a Gram-negative component. Further, mortality of P. gossypiella larvae reared on diet amended with B. thuringiensis toxin and Enterobacter sp. NAB3 was generally faster than with B. thuringiensis toxin alone. Conclusion: This study demonstrates that in some larval species, indigenous gut bacteria contribute to B. thuringiensis susceptibility. Moreover, the contribution of enteric bacteria to host mortality suggests that perturbations caused by toxin feeding induce otherwise benign gut bacteria to exert pathogenic effects. The interaction between B. thuringiensis and the gut microbiota of Lepidoptera may provide a useful model with which to identify the factors involved in such transitions.