Shifting paradigm on Bacillus thuringiensis toxin and a natural model for Enterococcus faecalis septicemia.

The Bt toxin is widely used in agriculture both as a spray and in transgenic plants, yet its precise mode of action against lepidopterans is poorly understood. The study by Mason et al. [mBio 2(3):e00065-11, 2011] revealed that Bt toxin enables an inhabitant of the midgut, Enterococcus faecalis, to enter the hemolymph of larvae and cause septicemia, leading to death.

Commentary

The most parsimonious or logical solution is not always the path that nature follows. Mason et al. (1) have demonstrated one such example in their recent investigation of the widely used insecticide Bt toxin, in which they discovered the toxin employs a more complex mode of action than had been previously appreciated.

The Bt toxin, produced by the spore-forming bacterium Bacillus thuringiensis, is in widespread use as an agricultural pesticide due to its effectiveness against lepidopteran pests, its ineffectiveness against the natural predators of lepidopterans, and its safety for humans (2). Bt toxin is safe enough for human consumption that it may be applied to food crops as late as 1 day before harvest. Sprays containing spores of B. thuringiensis are applied to control such economically important pests as the tobacco budworm and the cotton bollworm, a scourge that causes over 300 million dollars of damage to the cotton harvest each year.

Bt toxin is not only applied as a spray. Molecular characterization has made it possible to introduce the gene that encodes Bt toxins, cry, into the genomes of crop plants (2) like maize and cotton, enabling them to synthesize the toxin themselves, thus providing protection. “Bt cotton” reduces pesticide spraying by roughly 70% and increases cotton yield by 80%. Between its application in spray form and its use in transgenic plants, Bt toxin is now widely used in agriculture, so understanding how the toxin works is important.

Bt toxin is a δ-endotoxin that forms crystalline inclusions on the surface of B. thuringiensis spores (3). Upon ingestion by insects, the spores germinate and alkaline conditions in the digestive tract dissolve the inclusions, releasing and solubilizing the protoxin. Proteases in the midgut of the insect activate the protoxin, which then binds to a receptor on gut epithelial cells of lepidopterans and forms pores in the gut. Insects that succumb to Bt toxin fall into three categories: type I insects exhibit general paralysis and loss of integrity of the midgut caused by the toxin, which causes rapid death; type II insects exhibit gut paralysis without leakage of the gut content, which results in slow death; and type III insects die from septicemia, not the toxin, after B. thuringiensis spores germinate in the midgut (4).

Among insects susceptible to Bt toxin, type II insects, which die a slow death from gut paralysis without feeding, are the most common. The widely accepted mechanism for the slow death of the insect has been that the toxin forms pores in the gut membrane, which is thought to disrupt food absorption and starve the caterpillar to death.

The current report in mBio builds on a report by Broderick et al. (5) that revealed that B. thuringiensis requires the presence of enteric bacteria to have a toxic effect. Like all animals, caterpillars are not sterile but harbor microorganisms in their gut. This community harbors fewer species than the digestive tracts of most mammals and harbors many beneficial symbionts that are always found in the midgut of the insect, as well as organisms that pass through the gut. Using gypsy moths (Lymantria dispar), Broderick et al. demonstrated that the slow killing of type II insects by Bt toxin requires the presence of the bacteria that live in the midgut of the insects. Antibiotic-treated animals were not susceptible to B. thuringiensis, but administering the bacterium together with an Enterococcus sp. restored the ability of B. thuringiensis to kill the insects. So what are the enterococci doing and why are they required for the Bt toxin to work?

Mason et al. set out to investigate these questions. After establishing that Enterococcus faecalis is a normal member of the gut community of Manduca sexta, the tobacco hornworm, and does not kill the caterpillars when fed to M. sexta larvae, these researchers standardized the toxin inoculation. Larvae were reared on antibiotic-containing food to clear the microbiome of the midgut and then were force-fed the Bt toxin to standardize the amount consumed. Interestingly, Mason et al. used purified toxin for inoculation instead of a spore preparation, which can deliver more-variable doses of toxin. This approach greatly increased the reproducibility of the study—a challenge faced in previous studies of toxin action.

In the key experiment, larvae were fed Bt toxin alone, E. faecalis alone, both Bt toxin and E. faecalis, or phosphate-buffered saline (PBS). Larvae fed E. faecalis survived, while larvae fed Bt toxin alone began to die 6 days after feeding—at the same time the starved animals died. However, the animals that were fed the combination of Bt toxin and E. faecalis began to die after 2 days. Interestingly, E. faecalis could be recovered from the hemolymph of only those animals fed both Bt toxin and E. faecalis, which is indicative of a septicemic infection. When E. faecalis was injected directly into the hemolymph, the M. sexta larvae died. The aggregation of hemocytes (macrophage-like immune cells from insects) in the hemolymph further indicates that the larvae launch an immune response to E. faecalis. Taken together, these observations indicate that Bt toxin enables E. faecalis to escape the midgut and enter the hemolymph, where it overpowers the innate immune system and kills the insects by septicemia (Fig. 1). Hence, the otherwise harmless symbiont speeds the killing process: the larvae that were administered E. faecalis and toxin died much more quickly than when the toxin was administered alone.

FIG 1

Schematic of B. thuringiensis toxin mode of action. A cross section of a lepidopteran larva, highlighting the midgut, is shown. (1) Cocci representing E. faecalis are present throughout the midgut. (2) Small diamonds representing the toxin bind to the midgut surface. (3) E. faecalis enters the hemolymph. (4) Dead caterpillar is shown (dark color due to melanization).

Clearly, the location of a microorganism within an insect is an important determinant of the type of interaction that occurs. Simply by moving a few micrometers from the lumen of the gut, where the bacterium is a beneficial symbiont that aids the host in the digestion of nutrients, across the epithelial cell barrier to the hemolymph enables that same bacterium to become a serious health concern. Here the authors describe a natural model for studying this process. When B. thuringiensis and E. faecalis cooccur in the midgut of M. sexta, the Bt toxin enables E. faecalis to penetrate the epithelial barrier and cause septicemia. The innate immune system, including hemocytes, gears up to defend against the invading pathogen. Studying such processes in invertebrates has several important advantages: (i) innate immune responses can be studied in the absence of the adaptive immune response, (ii) comparative studies allow evolutionarily conserved processes to be revealed, and (iii) because these interactions are likely to occur in nature, the molecular dialog between the microorganism and the host has probably coevolved. Important mechanistic insights can be gained by investigating invertebrate model system that can direct future research in vertebrate or mammalian models.