Role of receptors in Bacillus thuringiensis crystal toxin activity

Abstract

Bacillus thuringiensis produces crystalline protein inclusions with insecticidal or nematocidal properties. These crystal (Cry) proteins determine a particular strain's toxicity profile. Transgenic crops expressing one or more recombinant Cry toxins have become agriculturally important. Individual Cry toxins are usually toxic to only a few species within an order, and receptors on midgut epithelial cells have been shown to be critical determinants of Cry specificity. The best characterized of these receptors have been identified for lepidopterans, and two major receptor classes have emerged: the aminopeptidase N (APN) receptors and the cadherin-like receptors. Currently, 38 different APNs have been reported for 12 different lepidopterans. Each APN belongs to one of five groups that have unique structural features and Cry-binding properties. While 17 different APNs have been reported to bind to Cry toxins, only 2 have been shown to mediate toxin susceptibly in vivo. In contrast, several cadherin-like proteins bind to Cry toxins and confer toxin susceptibility in vitro, and disruption of the cadherin gene has been associated with toxin resistance. Nonetheless, only a small subset of the lepidopteran-specific Cry toxins has been shown to interact with cadherin-like proteins. This review analyzes the interactions between Cry toxins and their receptors, focusing on the identification and validation of receptors, the molecular basis for receptor recognition, the role of the receptor in resistant insects, and proposed models to explain the sequence of events at the cell surface by which receptor binding leads to cell death.

FIG. 1.

Crystal structure of Cry1Aa (64) (PDB code, 1CIY), Cry2Aa (127) (PDB code, 1I5P), Cry3Aa (45) (PDB code, 1DLC), and Cry4Ba (11) (PDB code, 1W99). (Adapted from reference with permission from Elsevier.) Domain I, domain II, and domain III are shown in red, green, and blue, respectively. The N-terminal protoxin domain of Cry2Aa is shown in yellow. Images of protein structures in this and subsequent figures were generated using the program PyMol (Warren L. DeLano, DeLano Scientific LLC, San Carlos, CA [http://www.pymol.sourceforge.net]).

FIG. 2.

Crystal structure of colicin N (137) (PDB code, 1A87). The helical bundle with structural similarity to Cry toxin domain I is shown in red.

FIG. 3.

Crystal structure of banana lectin (yellow/green) bound to laminaribiose (red) at two sites. PDB code, 2BMZ (123).

FIG. 4.

Crystal structure overlay of the CBM CmCBM6-2 (blue) in complex with two molecules of cellotriose (yellow), and domain III of Cry1Aa (green). Aromatic residues important for carbohydrate binding are shown in magenta. The PDB codes are 1UYY (140) (CmCBM6-2) and 1CIY (64) (Cry1Aa). Clefts involved in CBM carbohydrate binding are indicated (140).

FIG. 5.

Phylogenetic analysis of lepidopteran midgut APN sequences. (A) Phylogenetic tree of representative lepidopteran midgut APN sequences, created using the programs CLUSTALX and DRAWTREE (PHYLIP package). The species name and GenBank accession number are shown for each protein. APNs boxed in purple indicated those reported to interact with Cry toxins. Classes are as proposed by Herrero et al. (67). Species names abbreviations are as follows: Se, Spodoptera exigua; Ms, Manduca sexta; Ld, Lymantria dispar; Hv, Heliothis virescens; Ha, Helicoverpa armigera; Hp, Helicoverpa punctigera; Bm, Bombyx mori; Sl, Spodoptera litura; Px, Plutella xylostella; Pi, Plodia interpunctella; Ep, Epiphyas postvittana; and Tn, Trichoplusia ni. References for binding studies are as indicated in the relevant section of the text. (B) Average amino acid sequence identity within and among the different APN classes.

FIG. 6.

Schematic representation of a typical lepidopteran APN protein. The proregion and the threonine-rich region, shown with broken lines, have been reported only in some APNs.

FIG. 7.

Comparison of predicted N-linked and O-linked glycosylation sites among representative lepidopteran midgut APNs, sorted by class. Predictions were made using the NetNGlyc 1.0 server and the NetOGlyc 3.1 server (84) (http://www.cbs.dtu.dk/services/). Species name abbreviations and GenBank accession numbers are the same as in Fig. 5.

FIG. 8.

Summary of reported binding between Cry toxins and endogenous (En) or exogenous (Ex) nondenatured (N) or denatured (D) APNs as discussed and referenced in the preceding sections. Binding, no binding, conflicting reports, and absence of data are indicated by green, red, yellow, or white/gray boxes, respectively. APN was expressed exogenously by E. coli (A) in vitro translation (B), S2 cells (C), Sf9 cells (D), Sf21 cells (E), T. ni cells (F), Drosophila or Sf21 cells (G), E. coli or T. ni cells (H), or E. coli or Sf9 cells (I). In cases where the conditions of binding (denaturing or nondenaturing) were not reported, boxes are merged. Species names are abbreviated as in Fig. 5. Species and class are abbreviated “Sp” and “Cl,” respectively.

FIG. 9.

Phylogenetic tree of lepidopteran cadherin-like proteins deposited in GenBank, created using the programs CLUSTALX and DRAWTREE (PHYLIP package). The species name and GenBank accession number is shown for each protein. Cadherins boxed in purple are those reported to bind to Cry toxins, as discussed in text. Species names abbreviations are as follows: Sf, Spodoptera frugiperda; Ms, Manduca sexta; Ld, Lymantria dispar; Hv, Heliothis virescens; Ha, Helicoverpa armigera; Hz, Helicoverpa zea; Bm, Bombyx mori; Px, Plutella xylostella; Pg, Pectinophora gossypiella; On, Ostrinia nubilalis; Cs, Chilo suppressalis; and Ai, Agrotis ipsilon. Only partial sequence information was available for P. xylostella; however, gaps in sequence alignment were excluded from tree construction.

FIG. 10.

Domain structure and putative Cry1A toxin binding sites in lepidopteran cadherin-like proteins. The proteins are labeled as follows: PRO, proprotein region; SP, signal peptide; EC, ectodomain; MPED, membrane-proximal extracellular domain; TM, transmembrane domain; CYTO, cytoplasmic domain. Domains are as defined by Nagamatsu et al. (130) (B. mori), Wang et al. (184) (H. armigera), Gahan et al. (44) (H. virescens), and Dorsch et al. (33) (M. sexta). Putative toxin binding sites are as reported by Nagamatsu et al. (129) (B. mori), Wang et al. (184) (H. armigera), Xie et al. (190) (H. virescens), and Dorsch et al. (33), Gómez et al. (55-57), and Hua et al. (74) (M. sexta). Proteins are illustrated such that homologous regions are horizontally adjacent. Features not expected in the mature form of the protein are outlined with a dashed line.