Downregulation and Mutation of a Cadherin Gene Associated with Cry1Ac Resistance in the Asian Corn Borer, Ostrinia furnacalis (Guenée)

Development of resistance in target pests is a major threat to long-term use of transgenic crops expressing Bacillus thuringiensis (Bt) Cry toxins. To manage and/or delay the evolution of resistance in target insects through the implementation of effective strategies, it is essential to understand the basis of resistance. One of the most important mechanisms of insect resistance to Bt crops is the alteration of the interactions between Cry toxins and their receptors in the midgut. A Cry1Ac-selected strain of Asian corn borer (ACB), Ostrinia furnacalis, a key pest of maize in China, evolved three mutant alleles of a cadherin-like protein (OfCAD) (MPR-r1, MPR-r2 and MPR-r3), which mapped within the toxin-binding region (TBR). Each of the three mutant alleles possessed two or three amino acid substitutions in this region, especially Thr1457→Ser. In highly resistant larvae (ACB-Ac200), MPR-r2 had a 26-amino acid residue deletion in the TBR, which resulted in reduced binding of Cry1Ac compared to the MPR from the susceptible strain, suggesting that the number of amino acid deletions influences the level of resistance. Furthermore, downregulation of OfCAD gene (ofcad) transcription was observed in the Cry1Ac resistant strain, ACB-Ac24, suggesting that Cry1Ac resistance in ACB is associated with the downregulation of the transcript levels of the cadherin-like protein gene. The OfCAD identified from ACB exhibited a high degree of similarity to other members of the cadherin super-family in lepidopteran species.

Ostrinia nubilalis (Hbn.) (Lepidoptera: Crambidae) [29]. Alignment of genes encoding CADs in O. nubilalis, H. armigera and B. mori revealed the following basic structures in common: an extracellular domain that contains a signal peptide and 9-12 cadherin repeats (CRs); a membrane proximal extracellular region (MPR); a transmembrane region (TM); and a small cytoplasmic domain (CD). However, this structure does not appear to be present in the three previously established cadherin categories. The classical CADs are usually located within adherent junctions involved in cell-cell adhesion, whereas lepidopteran CADs are usually located at the base of the microvilli along the length of the midgut and at the apical tip of microvilli in the middle and posterior regional membrane of the midgut [30,31], where the Cry1A toxins are also targeted [30,32]. Although the physiological functions of the Bt-related CADs described in lepidopteran insects have not been fully elucidated, their distribution increases the contact of membrane CADs with Cry1A-type toxins in the midgut fluid, thus facilitating the binding and toxicity of these toxins.
Cry1A toxin binding regions in the CADs have been reported in many lepidopteran insects, such as B. mori [33], M. sexta [34,35], H. armigera and P. gossypiella [36]. Mutations in genes encoding CAD proteins, which affect the toxin binding region to yield shortened proteins, are known to be tightly linked with resistance to Cry1A toxins [23,25,37]. The toxin binding regions play a key role in the interaction between CADs with Cry toxins, where they may either act as synergists [38][39][40][41] or antagonists [42] by enhancing or reducing, respectively, Cry toxicity against lepidopteran, dipteran or coleopteran insect larvae. Not only the mutations, but also downregulation of a midgut CAD (DsCAD1) in resistant (Cry1Ab-RR) strains of Diatraea saccharalis (F.) (Lepidoptera: Crambidae) was shown to be functionally correlated with a decrease in Cry1Ab susceptibility, suggesting a strong association between CADs and the resistance to Cry toxins [43].
Asian corn borer (ACB), Ostrinia furnacalis (Guené e) (Lepidoptera: Crambidae), is a key insect pest of maize in China. Yield losses to this insect are estimated to be 10%-20%, but may be more than 30% or may even result in no harvest at all in an outbreak year [44,45]. Field trials demonstrate that Cry1Ac-expressing maize can offer effective control of ACB [46]. Using co-immunoprecipitation (Co-IP), we have previously shown that a CAD acts as a putative Cry1A type binding protein in ACB [47]. We have also previously selected for a Cry1Ac-resistant strain of ACB under laboratory conditions, which exhibits various levels of cross-resistance to Cry1Ab, Cry1Ah and Cry1F [48]. The objective of the present study is to understand the potential role of CAD in the development of resistance to the Cry1Ac toxin in ACB.

cDNA of the Cadherin-Like Protein in Asian Corn Borer (ACB)
A cDNA sequence coding for ofcad (GenBank Accession No. EU022587.1), a CAD gene from the larval midguts of ACB-BtS, was obtained. The open reading frame (ORF) sequence consisted of 5154 nucleotides and encodes for OfCAD (1717 amino acid residues) with a predicted molecular weight of 191.92 kDa and an isoelectric point of 4.21. The OfCAD was shown to have the common features of lepidopteran CADs, including a signal peptide, which contained 21 amino acids residues, 11 cadherin repeats (CRs), a membrane proximal extracellular region (MPR), a transmembrane region (TMR) and a small cytoplasmic region (CPR) (Figure 1).

Figure 1.
Deduced amino acid sequence of OfCAD (GenBank Accession No. EU022587.1). Protein sequence analysis was carried out using the Swiss Institute for Experimental Cancer Research (ISREC) ProfileScan Server [49]. SIG is the putative signal peptide sequence, CR1-CR11 are cadherin repeats, MPR is the membrane-proximal extracellular region, TMR is the transmembrane spanning region (indicated by a thin underline) and CPR is the small cytoplasmic region. Potential N-glycosylation sites are indicted by arrows.

Recombinant Expression of OfCAD Internal Peptides and Binding Studies
Based on the cDNA sequence of cadherin from ACB-BtS, nine cDNA fragments ( Figure 2B) were generated by PCR, which were inserted into the pET30a (+) expression vector. Nine recombinant peptides were subsequently expressed and purified ( Figure 2C). These peptides matched the predicted molecular weights (Table 1) and could be detected by antibodies against the His-tag, which formed part of the expression construct. Ligand blots revealed the presence of three peptides ( Figure 2D), indicating that Cry1Ac had bound to these peptides, designated 6, 7 and 8. Each of these peptides contained the following domains, CR11-MPR, MPR and CR8-MPR, respectively, demonstrating that the minimum binding region mapped to MPR.

MPR Mutants in ACB-Ac200 and Their Binding Ability to Cry1Ac
The cDNAs coding for MPR in ACB-Ac200 were obtained by amplification with primers F5/R5 ( Table 2). The deduced amino acid sequences of these cDNAs were aligned with the MPR amino acid sequence in ACB-BtS ( Figure 3) using DNAMAN, which revealed that they were either amino acid substitutions or omission mutants, i.e., there were two, three and three amino acid substitution mutations in MPR-r1, MPR-r2 and MPR-r3, respectively. The substitution of Thr 1457 →Ser occurred uniformly. In addition, 26 amino acid residues were absent in MPR-r2, thus resulting in a lower molecular weight ( Figure 4). Differences in the respective binding abilities between MPR-r2 and MPR to Cry1Ac were revealed by Co-IP and shown to be weaker in MPR-r2 compared to MPR (Figure 4).  Figure 3. The ACB-Ac200 survivors harboured three mutant alleles in the MPR of OfCAD, namely MPR-r1, MPR-r2 and MPR-r3. The amino acid sequences were aligned with MPR from ACB-BtS using DNAMAN 6.0.

Figure 4.
Co-immunoprecipitation assays demonstrating the binding ability of MPR (Lane 1) and MPR-r2 (Lane 2) to Cry1Ac. The fainter band shows that less MPR-r2 is able to bind to the same amount of Cry1Ac compared to MPR.

Transcription Levels of ofcad
Expression of the ofcad transcripts from the ACB-BtS and ACB-AcR strains during larval development (first to fifth larval instars) was investigated by qPCR ( Figure 5). In the ACB-BtS strain, the levels of ofcad transcripts were significantly (F = 4.23; df = 1,4; P = 0.017) greater in larvae fed the control diet compared to when fed the diet containing a sub-lethal dose of Cry1Ac toxin (0.015 μg/g, Cry1Ac/diet; LC 50 = 0.03 μg/g) ( Figure 5B). With the exception of the fourth instar larvae, where transcription levels increased by approximately two-fold, there was an overall decline in ofcad transcripts in larvae fed the control diet. This contrasts with larvae fed the sub-lethal dose, where the trend was for a slight increase in transcript level throughout development. Interestingly, there was no significant difference in the number of ofcad transcripts in the fifth instar larvae of the ACB-BtS strain, irrespective of diet ( Figure 5A,B). These results are in contrast to those for the ACB-AcR strain, where there were no significant differences in ofcad transcript levels between those fed control diet and those fed Cry1Ac toxin from the second to fourth instars ( Figure 5C). However, by the fifth instar, the ofcad transcript levels in the control fed larvae increased approximately five-fold compared to those fed the diet containing Cry1Ac ( Figure 5A). Furthermore, the results also showed that transcription levels for ofcad were much lower during the first to fourth larval instars when fed the diet containing Cry1Ac, irrespective of the strain of ACB (i.e., ACB-AcR larvae fed Cry1Ac selecting diet and ACB-BtS larvae fed a sub-lethal dose of Cry1Ac toxin) compared to ACB-BtS larvae fed the control diet ( Figure 5A,B).

Discussion
A phylogenetic tree generated by ClustalW alignment of CAD amino acid sequences from 11 different lepidopteran insect species showed that the OfCAD identified from ACB exhibited a high degree of similarity to other members of the cadherin super-family in lepidopteran species (Figure 6). Pairwise distance was also analysed to estimate the evolutionary divergence between these sequences [50].  Previous studies have reported the presence of two Cry1A binding sites in M. sexta Bt-R1, designated TBR1 ( 865 NITIHITDTNN 875 ) and TBR2 ( 1323 IPLPASILTVTV 1342 ), which interact with loop α2 and loop α8 of domain II of the 3D toxin [35]. Similarly, in O. nubilalis, these Cry1A binding sites have been shown to be located on CR7 ( 861 DIEIEIIDTNN 871 ) and CR11 ( 1328 IPLQTSILVVTV 1339 ), but with several amino acid substitutions [51]. Results from the present study showed that the homolog in O. furnacalis was also within the CR7 and CR11 domain, with the same amino acid residues as for O. nubilalis. In order to investigate the ability of these two regions to bind the Cry1Ac toxin, both the CR7 and CR11 domains of ACB OfCAD were expressed as recombinant proteins in E. coli. However, the results failed to show any binding in ligand blot assays. Similar findings have been reported for both pink bollworm [36] and M. sexta [34,52]. In the pink bollworm, the homologs of TBR1 and TBR2 (located in domain CR6 and CR10, respectively) were also expressed in E. coli, but likewise, both peptides failed to demonstrate any binding of Cry1Ac under denaturing conditions, although the recombinant CR10 peptide did show some binding in dot binding assays; however, it is not clear whether this is the result of non-specific binding or possible binding to breakdown products.
In H. virescens, residues 1412 GVLSLNMQ 1418 were previously reported as crucial for binding to Cry1Ac toxin, based on analyses using hydropathic complementarity to the loop 3 region of Cry1Ac [53]. Based on alignment studies, similar homologous peptides were identified in M. sexta and H. armigera as 1416 GVLTLNIQ 1423 and 1423 GVLSLNFQ 1430 , respectively. Furthermore, in H. armigera, the Cry1Ac toxin binding region has been mapped to residues 1,217 to 1,461 [28]. These results are in broad agreement with those of the present study, where the minimum binding region was mapped to MPR and consisted of the following eight amino acid residues, 1412 GVLSLNMQ 1419 . The importance of these eight amino acid residues in Cry1Ac binding was confirmed by the loss of binding ability following deletion of this epitope in the ninth peptide. This finding corroborates previous studies [53].
Evolution of resistance in target pests to Cry toxins threatens the durability of Bt crops. To date, field resistance has been reported for Bt corn [54]. One of the most important mechanisms of insect resistance to Bt Cry toxins is due to mutation of the receptor in the insect midgut, thus affecting the binding of the Bt toxin [11]. Thus, identification of receptors is fundamental, both for understanding the mode of action of Bt Cry toxins and the molecular mechanisms of insect resistance to these particular toxins. Cry1A-binding proteins have been widely identified as CADs. Although the relative role of these putative receptor molecules in insects has yet to be conclusively confirmed, Cry1A resistance in a number of lepidopteran insects has been attributed to changes in these cadherin receptors. Such changes may be due to: (1) alternations in expression of CADs, as reported for Cry1Ab resistance in D. saccharalis [43] and Cry1Ac resistance in H. armigera [28]; (2) mutation of CAD genes, resulting from amino acid substitutions in the TBR of CADs, as reported in H. virescens [53]; (3) mutation of the CAD genes resulting from amino acid residue deletions, such as reported for the laboratory selected Cry1Ac resistant strain, GYBT, and the field strain of H. armigera [55], as well as in the laboratory selected Cry1Ab resistant strain of O. nubilalis [56]; (4) mutation of CAD genes resulting from premature stop codons, such as in H. virescens [23]. Further examples also include the laboratory selected Cry1Ac resistant strain of H. armigera [37,57], the field evolved Cry1Ac resistant population of P. gossypiella [25] and the laboratory selected Cry1Ab resistant strain of O. nubilalis [56].
In the present study, screening of cadherin-receptor gene mutants in ACB-AcR focused on the Cry1Ac TBR, which was based on the hypothesis that resistance might result from mutation in this region. Alignment of OfCAD from ACB-AcR and ACB-BtS showed that there were only three amino acid substitutions (D 1379 →N, I 1384 →T and T 1457 →S) in the MPR. However, a mutant with 26-amino acid residue deletions in the MPR (MPR-r2) was detected in ACB-Ac200 larvae. Furthermore, the ability of binding with Cry1Ac was significantly reduced in the MPR-r2 compared with MPR. These findings suggest that Cry1Ac resistance in ACB is associated with mutations in the TBR of the OfCAD and that the number of amino acid residue deletions influences the level of resistance, as seen for ACB-Ac200 larvae, which had 26 such deletions in the MPR.
Despite the fact that most studies to date have demonstrated that Cry1A resistance in Lepidoptera is linked to mutations in the CAD genes [23,25,37,53,[55][56][57][58], a few studies have suggested that resistance in D. saccharalis and H. armigera to Cry1Ab [43] and Cry1Ac [28], respectively, is due to downregulation of genes encoding CADs. Similarly, results from the present study have shown a significant reduction in transcription levels of ofcad, i.e., ofcad in the midgut of ACB during the first to fourth larval instars from both the Cry1Ac resistant strain (ACB-AcR) larvae and ACB-BtS larvae exposed to the Cry1Ac toxin. Theoretically, this reduction in expression of OfCAD in ACB-BtS larvae exposed to sub-lethal doses of Cry1Ac toxin may be a stress-induced response as a consequence of an unsuitable food source. Moreover, the reduced expression of OfCAD in ACB-AcR larvae fed on a normal diet was also responsible for a fitness cost in terms of longer development period, reduced pupal masses, etc. These results suggest that Cry1Ac resistance in ACB is primarily associated with the downregulation of CADs.

Insect Strains and Experimental Treatments
Two strains of Asian corn borer (ACB), O. furnacalis (Guené e), were used in this study, i.e., a Bt-susceptible strain (ACB-BtS) and a Cry1Ac-resistant strain (ACB-AcR). ACB-BtS larvae were reared on a semi-artificial diet as described by Zhou et al. [59]. ACB-AcR larvae were selected and maintained using activated Cry1Ac toxin as described by Han et al. [48]. An ACB-Ac200 colony was selected from ACB-AcR by exposing neonates to a diet containing Cry1Ac toxin at a concentration of 200 μg/g for 7 days. All survivors (12 out of 96 larvae exposed to Cry1Ac at a concentration of 200 μg/g) were then transferred to a new agar-free semi-artificial diet without Cry1Ac and were then reared on this diet until the 4th instar.

ACB Larval Midgut cDNA Synthesis
Total RNAs were isolated using the RNeasy Mini Kit (Qiagen, Hilden, Germany) from the midguts of larvae from the 1st to the 5th instars. At least 12 larvae were used for each experiment. The first strand cDNA used for PCR was synthesized using the TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix Kit (Transgen, Beijing, China). For the construction of the expression plasmid for OfCAD fragments, total RNA extracted from ACB-BtS served as the template to synthesise the cDNA. For screening the MPR mutants, total RNA isolated from the ACB-Ac200 colony served as the template. For qPCR, the total RNAs extracted from larvae of the 1st to 5th instar from both ACB-BtS (fed on a diet either containing Cry1Ac toxin at 0.015 μg/g (sub-lethal dose) or no toxin) and ACB-AcR (fed on a diet either containing Cry1Ac toxin at 24 μg/g (ACB-AcR selecting dose) or no toxin) served as templates, with the exception of 1st instar ACB-AcR larvae (fed on the diet without Cry1Ac), where RNA was not available.

Cloning of a Full-Length cDNA Coding for Cadherin-Like Proteins in ACB
The full-length cDNA coding for OfCAD was cloned by RACE using the 5' RACE System for Rapid Amplification of cDNA Ends (Invitrogen, USA) and 3'-Full RACE Core Set (Takara, Dalian, China). Degenerate primers (ofcad-F and ofcad-R) were used to obtain a middle fragment, and the GSPs (GSP1, GSP2 and ofcad-3F) combined with the primers supplied by the kits (AUPU and 3 sites adaptor) were used to amplify both ends of the cDNA. Primers (Table 2) used in this section were designed by Oligo 6 (Primer Analysis Software, version 6.71, Wojciech & Piotr Rychlik) and synthesized by Sangon Biotech Co. Ltd. (Shanghai, China).
In both the 5' and 3' RACE reactions, PCR products were subcloned into the pEASY-T1 simple cloning vector (Transgen, Beijing, China), transformed into Trans1-T1 Phage Resistant Chemically Competent Cells (Transgen, Beijing, China) and cultured on LB solid agar plates according to the manufacturer's instructions. Positive recombinant plasmids were sequenced by Sangon Biotech Co. Ltd. (Shanghai, China). The sequences were analysed by DNAMAN 6, and full-length cDNA sequences were assembled. Protein sequence analysis was carried out using the ISREC Profile Server [49], and the presence of the signal peptide was tested by the SignalP Server [60]. Various physical and chemical parameters were computed using ProtParam [61]. N-Glycosylation sites were predicted using the NetNGlyc Server [62].

Expression and Purification of OfCAD Fragments
Primers used to construct the expression vectors were designed based on the cDNA of ofcad (Accession No. EU022587.1) cloned in this study using Oligo 6 (see Table 2). PrimeSTAR ® HS (Premix) (Takara, Dalian, China) was used to clone the ofcad fragments. PCR products were purified by the EasyPure PCR Purification Kit (Transgen, Beijing, China), digested with EcoR1 and Xho1 restriction endonuclease (NEB), inserted into the expression vector pET30a(+) (Novagen, Darmstadt, Germany) using the DNA Ligation Kit Ver.2.1 (Takara, Dalian, China) and transformed into Trans1-T1 Phage Resistant Chemically Competent Cells (Transgen, Beijing, China) following the manufacturer's instructions. The recombinant plasmids were sequenced and transformed into Trans BL21 (DE3) Chemically Competent Cells (Transgen, Beijing, China). The target peptides were expressed in E. coli (ArtMedia Protein Expression; Transgen, Beijing, China) containing Kanamycin selection and affinity purified on a Ni-Agarose column (CWBIO, China; CW0894 was used to purify the soluble proteins; CW0893 was used to purify the proteins expressed as inclusion bodies). The purified peptides were concentrated using Centricon-10 centrifugal filters (Millipore, Ireland), separated by 10% SDS-PAGE gel electrophoresis and detected by staining with Coomassie Brilliant Blue. Recombinant protein was desalted using Zeba Desalt Spin Columns (Thermo Scientific, Uppsala, Sweden) and the protein concentration determined using the Easy Protein Quantitative Kit (Transgen, Beijing, China).

Ligand Blot Binding Assay
Each (1 μg) of 9 purified peptides were separated by 10% SDA-PAGE and transferred onto nitrocellulose membranes using iBlot™ Gel Transfer Device (Invitrogen, Carlsbad, CA, USA). After blocking in TBST buffer (CWBIO, Beijing, China) containing 3% BSA for 1 h, membranes were incubated with 20 nM Cry1Ac for 2 h and probed with rabbit antiserum against Cry1A (1:2000 dilution) in TBST buffer for 1 h. Goat anti-rabbit IgG, AP conjugate (CWBIO, Beijing, China), was used as the secondary antibody (1:5000 dilution) and developed using Western Blue ® Stabilized Substrate for Alkaline Phosphatase (Promega, Madison, WI, USA). Membranes were washed between each step (3 times), 5 min in TBST buffer. All steps were conducted at room temperature on an orbital shaker.

Variation of Membrane Proximal Extracellular Region (MPR) Screening in ACB-Ac200 Larvae
F5/R5 (Table 2) served as primers to clone the MPR of the ACB-Ac200 colony; the cDNA used in this section was as described in Section 4.3. PrimeSTAR ® HS DNA Polymerase was used here; an extra A was added to the PCR product by incubating with the PCR products at 72 °C for 30 min. The PCR products were then subcloned, transformed and sequenced following the method described in Section 4.4. Bitraditional sequencing was used to verify the cDNAs.

Binding Assay through Co-Immunoprecipitation (Co-IP)
Co-IP was used to detect the difference in the binding ability of Cry1Ac between the recombinant expressed MPR of the ACB-BtS strain and the mutant allele (MPR-r2) screened in Section 4.7. This additional binding assay was carried out, since, in contrast to ligand blots (which are carried out under denaturing conditions, so as to completely expose the epitope), Co-IP is carried out under non-denaturing conditions. Dynabeads ® M-280 Streptavidin (Invitrogen Dynal, Oslo, Norway) was coated with bio-Cry1Ac. The immobilized target protein was then incubated with surplus recombinant MPR or MPR-r2 (previously expressed in E. coli), and the captured peptides were dissociated by boiling, following the manufacturer's instructions. Samples were separated by 10% SDS-PAGE and stained with Coomassie Brilliant Blue.

Quantitative Real-Time PCR (qPCR)
To further compare the transcription levels of ofcad in both ACB-BtS and ACB-AcR strains, qPCR was performed with a set of cDNAs (described in Section 4.3), which served as templates. Amplifications were conducted with three technical replicates for each of the two independent biological samples on an ABI7500 thermal cycler (ABI, Abilene, TX, USA), with β-actin as a reference. The SYBR Premix Ex Taq™ kit (Takara, Dalian, China) was used in qPCR; reaction samples (20 μL) consisted of 10 μL of SYBR Premix Ex Taq (Takara, Dalian, China), following the manufacturer's instructions. Relative transcript abundance was calculated using the 2 −ΔΔC T method [63]. Data from gene expression assays were analysed using analysis of variance (ANOVA), and the means were separated by Fisher's protected LSD test for significance, using the SAS program (SAS Institute Inc., Cary, NC, USA, 1999). The primers used (see Table 2) were validated as described by Livak et al. [63].

Conclusions
The present study identifies a CAD gene ofcad from the larval midguts of ACB-BtS. The open reading frame (ORF) sequence of ofcad consists of 5154 nucleotides and encodes a cadherin-like protein (OfCAD; 1717 amino acid residues), with a predicted molecular weight of 191.92 kDa and an isoelectric point of 4.21. OfCAD has a high degree of similarity to other members of the cadherin super-family in lepidopteran species. The binding region of the Cry1Ac toxin is located in the membrane proximal extracellular region (MPR) of OfCAD. Cry1Ac resistance in laboratory selected ACB is associated with mutations resulting from amino acid substitutions and/or deletions in this toxin-binding region (TBR). The number of amino acid residue deletions influences the level of resistance, as seen for ACB-Ac200 larvae, which has 26 such deletions in the MPR. In addition, Cry1Ac resistance in ACB is primarily associated with the downregulation of CADs.