Mining of new insecticidal protein genes plus determination of the insecticidal spectrum and mode of action of the bacillus thuringiensis vip3ca protein

Resumen

Bacillus thuringiensis is an entomopathogenic bacterium that belongs to the Bacillus cereus group and produces a wide variety of insecticidal proteins along with other virulence factors contributing to its pathogenicity. Bacillus thuringiensis has been considered as the most successful bioinsecticidal agent during the last century. Currently, it is widely used as a microbial agent of the major insect pests. In the present doctoral thesis, we performed the screening for new B. thuringiensis insecticidal protein genes from selected B. thuringiensis isolates based on their gene content. Also, we studied in detail different aspects (insecticidal spectrum, cross-resistance, mode of action and cell death response of Spodoptera exigua intoxicated with the Vip3Ca protein) of the new Vip3Ca protein family. Regarding the mining of new insecticidal proteins, we found one new couple of binary Vip-like proteins (Vip2Ac-like_1-Vip4Aa-like_1), two new Vip-like proteins (Vip2Ac-like_1 and Vip4Aa-like_2); one Sip1A-like protein (Sip1A-like_1) and eight Crystal-like proteins (Cry23A-like, Cry45Aa-like_1, Cry45Aa-like_2, Cry45Aa-like_3, Cry32Ea-like, Cry32Da-like, Cry32Eb-like and Cry73Aa-like). The Vip-like proteins have been detected in the supernatant, marginally expressed, while the Crystal-like proteins have been found in the parasporal crystal (E-SE10.2: 2.5 % - 30 % ; O-V84.2: 7.0 % - 9.8 % Cry45-like proteins, 30.4 % - 30.5 % Cry32-like proteins and 2.8 % - 4.25 % Cry73Aa-like), except the Cry23Aa-like protein, that has also been found in the supernatant at 24 h and 48 h. In addition, the toxicity of the supernatant (E-SE10.2: 24 h and 48 h; O-V84.2: 24 h and 48 h) and parasporal crystal (E-SE10.2 and O-V84.2) has been tested in S. exigua and Spodoptera littoralis, where the Cry23Aa-like showed some toxicity to both insect species. In the case of the Vip3Ca protein, we first extended the study of its insecticidal activity to ten lepidopteran pests (Cydia pomonella, Grapholita molesta, Sesamia nonagrioides, Galleria mellonella, Plutella xylostella, Pectinophora gossypiella, Ephestia kuehniella, Plodia interpunctella and Ostrinia furnacalis), one aphid (Nezara viridula) and one model organism (Drosophila melanogaster). Several methodologies of biossays were used. For the species tested by surface contamination, C. pomonella was the most susceptible one, followed by G. mellonella, with percentages of mortality higher than 50%. Regarding the four species tested by diet incorporation, O. furnacalis was the most susceptible one, with an LC50 value of 0.31 µg/g. For the other types of bioassays, Tuta absoluta showed some susceptibility to high concentrations of Vip3Ca (30 % mortality). Also, several Cry1A, Cry2A, Dipel, Vip3 and Vip3/Cry2Ab resistant insect colonies were tested against the Vip3Ca protein to determine if they presented cross-resistance. The results showed that the insect colonies resistant to Cry1Ac (Helicoverpa armigera, Trichoplusia ni, O. furnacalis and P. interpunctella) or Cry2Ab (H. armigera and T. ni) were not cross-resistant to Vip3 proteins (Vip3Aa and Vip3Ca). In contrast, the H. armigera colonies resistant to Vip3Aa or Vip3Aa/Cry2Ab showed cross-resistance to the Vip3Ca protein. We should mention that, as a secondary work arising from the search for susceptible species, arose the monographic study on the susceptibility of G. molesta to those formulated by B. thuringiensis, its toxins (Cry and Vip3) and synergies between the Cry and Vip3 proteins. As a result, we found that G. molesta is highly susceptible to B. thuringiensis, Dipel® and Xentari® formulates, and to the B. thuringiensis toxins Cry1Aa, Cry1Ac, Cry1C, Vip3Aa and Vip3Af. Once determined that G. molesta is susceptible to Cry1Aa, Cry1Ac, Cry1C and Vip3Aa toxins, the possible synergies between these proteins were evaluated in dose-response assays. The combinations of Cry1Aa-Vip3Aa and Cry1C-Vip3Aa proteins were antagonistic while the combination of Cry1Ac-Vip3Aa proteins shows an additive effect. It is interesting to mention that the antagonism detected in the Cry1Aa-Vip3Aa and Cry1C-Vip3Aa protein combinations increased as the amount of Vip3A protein in the mixture was higher; this fact suggests that, though these proteins do not compete for the same binding sites, they may interact in events prior to the binding, making more or less accessible the binding to the receptor. To study in more detail the mode of action of the Vip3Ca protein, we chose M. brassicae because of its relatively high susceptibility to this toxin. (I) Proteolysis of the Vip3Ca protein and oligomerization The proteolytic pattern of the Vip3Ca protein with trypsin and midgut juice showed that theVip3Ca protein was processed into two main fragments of ~70 kDa and ~20 kDa. The kinetics of Vip3Ca with the midgut juice from insect species with different susceptibility correlated with the susceptibility of the insect species to this toxin. To discard that the differences in the toxicity of the Vip3Ca protein was due to inappropriate activation of the protein, the Vip3Ca protein was processed with midgut juice of O. nubilalis or M. brassicae and this did not significantly change its toxicity to either insect species. In an attempt to determine if the Vip3Ca protein makes oligomers, the size of the Vip3Ca digested with trypsin and midgut juice of M. brassicae was determined by size exclusion chromatography. As a result, we found that the Vip3Ca protein, either activated with trypsin or midgut juice, forms an oligomer composed of 4 units, that is stable up to two days in solution. Also, the toxin fragments of ~70 kDa and ~20 kDa coeluted together, suggesting that both protein bands are needed to form the oligomer. (II) Histopathological effects and in vivo / in vitro binding of Vip3Ca The histopathological effects of Vip3Ca were determined in midgut sections of M. brassicae intoxicated for different intervals with the Vip3Ca protein. The results indicated that the Vip3Ca protein was able to disrupt the midgut epithelium of M. brassicae. To know if the Vip3Ca protein binds to the midgut apical membrane, the larvae were fed with Vip3Ca and, after 3 h of exposure, the binding of the Vip3Ca protein to the midgut epithelium was detected. With the aim to demonstrate that the Vip3Ca protein bound specifically to the midgut epithelium of M. brassicae and compete for the same binding sites than Vip3Aa, binding assays were performed with biotin-labelled Vip3 proteins. The homologous competition showed that the Vip3Aa and Vip3Ca proteins bound specifically to the brush border membrane vesicles (BBMV) because the respective unlabelled Vip3 proteins were able to displace the homolog proteins. The heterologous competition experiments indicated that Vip3Ca and Vip3Aa share binding sites. (III) Cell death response of S. exigua intoxicated with Vip3Ca2 The election of S. exigua was due to the fact that a previous publication described a set of 47 S. exigua genes that responded to the intoxication by the Vip3A protein. The gene expression analysis indicated that the number of the S. exigua genes is dose-dependent. Regarding the up-regulated genes, these were involved in the immune system and hormone modulation, whereas the down-regulated genes were those involved in the digestion process and peritrophic membrane permeability. In addition, the gene encoding a component of the Jak-Stat pathway was found down-regulated after 24 h of exposure at the higher dose tested. The effect of the Vip3Aa and Vip3Ca proteins on the midgut epithelium indicated that the release of aminopeptidase N (APN) is higher when the larvae are exposed to Vip3Aa rather than Vip3Ca, at concentrations that produce growth inhibition > 99 %. Also, the fitness cost associated to the intoxication with the Vip3Ca and Vip3Aa protein is affected by an increase in the time of pupation and a reduction of the percentage of pupation. To determine if the down-regulation of the Jak-Stat pathway is involved in the apoptosis, the expression level of the caspases and TUNEL staining of midgut sections was performed. The results showed the presence of TUNEL-positive cells in the S. exigua midgut sections exposed to sublethal concentrations of Vip3Ca protein and the Se-caspase-4 was up-regulated at all the time intervals, while the Se-caspase-1 and Se-caspase-2 were up-regulated at 12 h.