Abstract
Cry4 Aa produced by Bacillus thuringiensis subsp. israelensis (Bti) is a mosquitocidal toxin that exhibits a specific high toxicity to Anopheles, Aedes, and Culex larvae. This work describes the hyper-expression and the functional roles of three major loops in domain II of Cry4Aa.
In order to produce a large amount of the Cry4Aa protein, the expression level of Cry4Aa was checked in Escherichia coli. The expression of wild-type Cry4Aa in E. coli is relatively low, which is a major disadvantage in its development as a bioinsecticide. In this study, to establish an effective production system, a 1914-bp modified synthetic gene (cry4Aa-SI) encoding the Cry4Aa protein was designed in accordance with the G+C content and codon preference of E. coli genes without altering the encoded amino acid sequence. The cry4Aa-S1 gene allowed a significant improvement in the expression level, over 5-fold, compared to that of the original cry4Aa gene. The cry4Aa-S1 gene product showed the same level of insecticidal activity against Culex pipiens larvae as that from cry4Aa. This suggested that unfavorable codon usage was one of the reasons for poor expression of cry4Aa in E. coli, and therefore, changing the cry4Aa codons to accord with the codon usage in E. coli led to efficient production of Cry4Aa. The efficient production in E. coli can be a powerful measure to prepare a sufficient amount of Cry4Aa protein for both analytical and applied researches.
The molecular mode of action of Cry4Aa is poorly understood in comparison with the lepidopteran specific Cryl-type protein. In general, domain II of the Cry toxin is believed to be important for insecticidal specificity. In order to elucidate the functional mechanisms for Cry4Aa, especially to understand the toxin-receptor interactions and to identify the epitopes in the Cry4Aa molecule responsible for those interactions, three loops in the domain II of Cry4Aa were targeted and the mosquitocidal activities were analyzed both by loop-exchange and site directed mutagenesis methods.To analyze the biological functions of loops 1, 2, and 3 of Cry4Aa, mutants were
constructed in which one of the loops was replaced with either of the other two loops. A bioassay using Culex pipiens larvae revealed that the mosquitocidal activity was virtually lost upon replacement of the loop 2. The mutants in which the loops 1 and/or 3 were replaced showed decreased but some significant insecticidal activities. This suggested that the loop 2, but not the loops 1 and 3, was essential for the mosquitocidal activity of Cry4Aa. Proteolytic digestion revealed the involvement of loops in the stability of the Cry4Aa structure. No significant differences were observed between the wild type and the mutants of Cry4Aa in binding to the BBMVs prepared from the C. pipiens larvae.
Identification of the loop 2 in domain II as an essential element for mosquitocidal toxicity of Cry4Aa suggests that this loop may be a functional motif for the interaction with its potential receptor(s). Therefore, to characterize a potential receptor-binding site more precisely, I have constructed a series of Cry4Aa mutants in which amino acid residues in the loops 1, 2 and 3 were replaced with alanine. A bioassay using Culex pipiens larvae revealed that the replacement of some residues in loop 2 sequence depressed the mosquitocidal activity of Cry4Aa, but the effect was limited. This was inconsistent with the previous results suggesting that the replacement of the Cry4Aa loop 2 causes a significant loss of the mosquitocidal activity. Therefore, I constructed additional mutants in which multiple (5 to 6) residues in the loop 2 were replaced with alanine. Although the replacement of multiple residues also caused some decrease in the mosquitocidal activity, the mutants still showed relatively high activity. On the other hand, the alanine mutants of loop 1 and 3 did not show so much decrease in the toxicity as compared with the wild-type. Since the insecticidal spectrum of Cry4Aa is specific, Cry4Aa must have a specific receptor on the surface of the target tissue and loss of binding to the receptor should cause a complete loss of the mosquitocidal activity. The present results suggested that, unlike the well-characterized Cryl, the receptor-binding site of Cry4Aa is different from the loops 1, 2, and 3 or consists of multiple binding sites that work cooperatively for receptor binding.