INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Bacillus thuringiensis strains pathogenic to insects produce two distinct types of toxin proteins, Cry and Cyt proteins (9, 28). Generally, the genes encoding these proteins are located on large plasmids, and the proteins are synthesized and form crystalline inclusions during sporulation. More than 100 different cry genes have been identified and sequenced, and significant homologies among the amino acid sequences of this group, in combination with experimental studies, suggest they have a common mode of action, colloid-osmotic lysis (10, 14). Cyt toxins, however, have only been found in B. thuringiensis strains that are mosquitocidal, have amino acid sequences that are unrelated to those of the Cry toxins, and can lyse a variety of cell types in vitro (14, 34). The mode of action of Cyt toxins has not been fully elucidated. However, research suggests that Cyt toxins are also involved in colloid-osmotic lysis (14), but may differ in the mechanism by which lesions are formed in the cell membrane (6, 7, 9).

One of the more interesting features of the Cry and Cyt toxin combination that is found in B. thuringiensis subsp. israelensis, the subspecies upon which many commercial mosquito larvicides are based, is the effect of this combination on toxicity. This subspecies produces a crystalline parasporal inclusion containing four major toxic proteins, Cry4A (134 kDa), Cry4B (128 kDa), Cry11A (66 kDa), and Cyt1A (27 kDa). These four proteins are assembled into separate inclusions that are enveloped together to form the parasporal body. The proportion of each protein in the parasporal body, based on sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electron microscopy (15), is approximately 40% Cyt1A and 20% for each of the three Cry proteins. Studies of these proteins revealed that the toxicity of individual proteins in B. thuringiensis subsp. israelensis was much less than that of the intact parasporal crystal. Subsequently, it was shown that the Cry toxins in B. thuringiensis subsp. israelensis interact synergistically with the Cyt1A toxin, as well with each other, to produce this high level of activity (9, 15, 24, 41). This synergism has also been shown to be important in the relatively low rate of resistance development toward B. thuringiensis subsp. israelensis in Culex mosquitoes (12) and can suppress high levels of resistance to Cry4 and Cry11 toxins (37, 38). The mechanism of this synergism is not understood, but we have postulated that Cyt1A aids these toxins in binding to or inserting into the mosquito microvillar membrane (37). The synergistic capacity of Cyt1A was extended by the recent observation that sublethal concentrations of Cyt1A, combined with B. sphaericus, an unrelated mosquitocidal bacterium which does not have Cry-type toxins, were synergistic and toxic toward highly resistant Culex quinquefasciatus (40). Because this combination was synergistic against resistant mosquitoes that have lost the capacity to bind B. sphaericus toxins (21), we postulated that this same mixture might be synergistic toward a mosquito species that is naturally insensitive to B. sphaericus.

To evaluate this possibility, we tested Cyt1A in combination with B. sphaericus against larvae of the mosquito Aedes aegypti. Although highly toxic to Anopheles and Culex species, B. sphaericus is not active against A. aegypti (18, 19). Thus, in A. aegypti, enhanced toxicity of B. sphaericus in the presence of Cyt1A would be due to the effect of the Cyt1A toxin. Here, we report that a mixture of Cyt1A with B. sphaericus strain 2362 was >3,600-fold more toxic than B. sphaericus toward a laboratory strain of A. aegypti.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Mosquitoes. A laboratory colony of A. aegypti was obtained from M. Mulla, Department of Entomology, University of California, Riverside. This colony was established in culture in our laboratory and maintained by standard procedures described previously for this species (39). Early fourth-instar larvae were used in the bioassays.

Bacterial strains and toxins. Toxin preparations for this study were lyophilized powders of spore-crystal mixtures. Technical powder of B. sphaericus 2362 was obtained from Abbott Laboratories (North Chicago, Ill.). The Cyt1A powder was derived from a recombinant strain of B. thuringiensis subsp. israelensis that produces only the Cyt1Aa toxin (42). Two powders of Cyt1A were used in these experiments because the quantity of the first powder was insufficient to complete the bioassays. For the first powder, cells were grown on liquid media as previously described (22, 37). For the second powder, the cells were grown on wheat-germ media (Berdsk Co., Berdsk, Russia). After fermentation, the sporulated cells were washed in distilled water and sedimented, and the resultant pellets were lyophilized. Lyophilized powders of purified Cyt1A crystals were also used (42).

Bioassay procedures. Groups of 20 early fourth-instar larvae were exposed to a range of concentrations of the spore-crystal powders in 100 ml of deionized water in 237-ml plastic cups. This range, usually 5 to 10 different concentrations, resulted in mortality between 2 and 98% after 24 h for Cyt1A and 48 h for B. sphaericus. Bioassays were replicated on five different days. The different ratio combinations of B. sphaericus and Cyt1A were based on the dry weight of the respective powders. Cyt1A powder 1 was used for the tests of the ratios 3:1 and 10:1, whereas Cyt1A powder 2 was used for the tests of the ratios 20:1 and 50:1. Mortality in these bioassays was evaluated at both 24 and 48 h. Because of the limited quantity of Cyt1A crystal, bioassays utilizing the crystal in combination with B. sphaericus technical powder took place in 30-ml plastic cups, with 10 ml of deionized water and 10 early fourth-instar larvae. These bioassays were replicated on three different days. Mortality was evaluated after 48 h.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

B. sphaericus exhibited only a low level of toxicity against fourth-instar larvae of A. aegypti, yielding a 50% lethal concentration (LC₅₀) of 860 µg/ml (Table 1). In contrast, the powders of sporulated Cyt1A cells were much more toxic, with an LC₅₀ of 5.90 µg/ml for Cyt1A powder 1 and an LC₅₀ of 14.1 µg/ml for Cyt1A powder 2. These two values were not significantly different. Bioassays in the smaller cups yielded lower lethal concentrations for both B. sphaericus and Cyt1A, with LC₅₀ values of 12.8 and 1.86 µg/ml, respectively (Table 2).

When lyophilized spore-crystal powders of the Cyt1A strain were combined with B. sphaericus technical powder and tested against A. aegypti, the resulting toxicity was much greater than that observed for B. sphaericus alone. As stated above, the LC₅₀ of B. sphaericus was 860 µg/ml. In the presence of a 3:1 ratio of B. sphaericus to Cyt1A, this value was 5.23 µg/ml (Table 1; Fig. 1). When the proportion of Cyt1A in the combination was reduced, the mixtures continued to be highly toxic, with LC₅₀ values of 2.46, 1.75, and 3.17 µg/ml for the ratios 10:1, 20:1, and 50:1, respectively. Similar high activity was observed at the LC₉₅ (Table 1; Fig. 1). The concentration of Cyt1A in the latter three ratios was sublethal, whereas for the 3:1 ratio, Cyt1A could have contributed between 12 and 60% of the observed mortality. The 10:1 ratio of B. sphaericus and Cyt1A was the most active combination tested. Although toxicity was slightly lower at the 20:1 and 50:1 ratios, these combinations were still highly toxic.

View larger version (16K): [in this window] [in a new window]   FIG. 1.   Dose-response regression lines of B. sphaericus (strain 2362) technical powder and combinations of different ratios of B. sphaericus and Cyt1A spore-crystal powder from B. thuringiensis subsp. israelensis toward larvae of the mosquito A. aegypti.

The B. sphaericus and Cyt1A mixtures had an additional effect on toxicity. In contrast to B. sphaericus, which normally requires 48 h for maximum toxicity, high toxicity was observed 24 h after exposure to the combination of B. sphaericus and Cyt1A. In general, lethal concentrations did not increase significantly between 24 and 48 h of exposure to these combinations. However, when Cyt1A was present at a lower proportion in the mixture, 20:1 and 50:1, a significant increase in mortality after 48 h was noted at the LC₅₀. Slopes were higher for the ratios 3:1 and 10:1 than for the 20:1 and 50:1 ratios, which may reflect differences between the two powders of Cyt1A (Table 1; Fig. 1).

Calculation of the synergism factors from the 24-h data for the various combinations indicated that all of the mixtures were synergistic. Synergism factors ranged from 2.9 to 17.0 at the LC₅₀ and from 20.4 to 172 at the LC₉₅. The synergism factors, in agreement with the toxicity levels, were greatest with the 10:1 ratio of B. sphaericus to Cyt1A.

When B. sphaericus technical powder was combined with purified Cyt1A crystals at a 10:1 ratio and tested in the smaller bioassay system, we observed increased toxicity toward A. aegypti, with LC₅₀ and LC₉₅ values of 3.8 and 31.5 µg/ml, respectively, and synergism factors of 2.1 and 8.6, respectively (Table 2).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Low concentrations of the Cyt1A protein from B. thuringiensis subsp. israelensis dramatically altered the toxicity of B. sphaericus against the normally insensitive mosquito species, A. aegypti. A ratio of 10:1 of B. sphaericus technical powder and Cyt1A spore-crystal powder was 3,600-fold more toxic at the LC₉₅ against A. aegypti than was B. sphaericus, and this high level of activity resulted from synergism between the Cyt1A toxin and B. sphaericus.

The mechanism of synergism between Cyt1A and B. sphaericus is not known. Previous work has demonstrated that Cyt1A has little or no effect on the toxicity of B. sphaericus toward susceptible C. quinquefasciatus (40). However, in a laboratory-selected, resistant population of that species, low concentrations of Cyt1A combined with B. sphaericus suppressed >30,000-fold resistance to B. sphaericus. Resistance in this strain of mosquitoes is due to failure of the binary B. sphaericus toxin to bind to midgut microvilli (21). B. sphaericus is poorly toxic toward A. aegypti because this species either lacks a specific receptor for the binary toxin of B. sphaericus or has an extremely low concentration of such receptors (20). Other factors that could affect the activity of B. sphaericus, such as the rate of ingestion of the toxins or differences in proteolytic activation, have not been found to be significantly different between C. quinquefasciatus and A. aegypti (1, 5). Our results, therefore, suggest that Cyt1A may enhance toxicity by facilitating the binding or insertion of the binary toxins to the microvillar membrane.

Our tests revealed that purified Cyt1A crystals combined with B. sphaericus technical powder were more toxic to A. aegypti than B. sphaericus alone, and the combination was synergistic. However, the levels of synergism were lower than those observed with Cyt1A spore-crystal powder in the large-cup bioassay system. The reduction in the synergism factor primarily resulted from the increased toxicity of B. sphaericus in the smaller cups relative to its toxicity in the larger cups (67-fold more toxic at the LC₅₀). Because larval density can have a significant effect on the toxicity of B. sphaericus (21) and B. thuringiensis subsp. israelensis (2), the higher density in the small cups was probably a contributing factor. Moreover, Cyt1A crystals in the small cups were only three- to fourfold more toxic than Cyt1A spore-crystal powder in the large cups. This is not unexpected because of the proportionally higher concentration of Cyt1A in the crystal powder (on a weight basis) relative to spore-crystal powder. Because the highest proportion of Cyt1A-to-spore-crystal was associated with the lowest synergism level, the higher proportion of Cyt1A in tests using crystal may also have contributed to the decreased synergism. However, the role of spores in the observed synergism cannot be ruled out because B. sphaericus 2362 spores have been shown to be toxic to Aedes (4).

Previous attempts have been made to improve the activity and host range of B. sphaericus by incorporating toxin genes from B. thuringiensis into the former species and expressing the combined products. Cyt1Ab1 from B. thuringiensis subsp. jegathesan was cloned and expressed in B. sphaericus 2297; however, no improvement in activity toward A. aegypti was detected (33). Cry toxins, or combinations of Cry toxins and Cyt1A, also have been incorporated into B. sphaericus. The activity of these engineered strains was extended to A. aegypti and toxicity was enhanced between 10- and 100-fold (3, 23, 25, 29, 35). However, these improvements in toxicity were considerably lower than the levels we have reported here, and the contribution of B. sphaericus to the observed activity was not established. Synergism, although implied, was not demonstrated directly.

From a practical perspective, Cyt1A may prove useful in expanding the host range of B. sphaericus against mosquito species that have heretofore been refractory to control with this bacterium. For example, the LC₉₅ of a 10:1 ratio of B. sphaericus to Cyt1A spore-crystal powder was 5.36 µg/ml. This toxicity level is 11.5-fold higher than that of B. sphaericus toward C. quinquefasciatus (40), a primary target of B. sphaericus-based insecticides. Our data indicate that even extremely low concentrations of Cyt1A combined with B. sphaericus were beneficial in enhancing toxicity toward A. aegypti. As many Aedes species are sensitive to B. thuringiensis subsp. israelensis and therefore are likely to be sensitive to Cyt1A, this combination may provide a practical mechanism for extending the host range of B. sphaericus. The B. sphaericus and Cyt1A combination may also have the potential to reduce the risk for developing B. sphaericus resistance. Studies of the development of resistance to B. thuringiensis subsp. israelensis have provided strong evidence that Cyt1A can delay resistance to the Cry toxins in this material (12). This is particularly important in view of the cases of B. sphaericus field resistance that have been reported in several parts of the world (26, 30; G. Sinègre, M. Babinot, J. M. Quernal, and B. Gaven, VII Eur. Meeting Soc. Vector Ecol., abstract, 1994).

The Cyt toxin group now contains several different toxins, including Cyt1Aa and Cyt2Ba from B. thuringiensis subsp. israelensis (13, 36), Cyt2Aa from B. thuringiensis subsp. kyushuensis (17), Cyt1Ab1 from B. thuringiensis subsp. medellin (33), and Cyt2Bb from B. thuringiensis subsp. jegathesan (8). These Cyt proteins vary in their toxicity to mosquitoes and their ability to synergize toxins. The differences among the Cyt proteins may provide insight into their mode of action, as well as a practical means to manage resistance and extend the host range of mosquitocidal bacterial insecticides.