PC, a Novel Oral Insecticidal Toxin from Bacillus bombysepticus Involved in Host Lethality via APN and BtR-175

Insect pests have developed resistance to chemical insecticides, insecticidal toxins as bioinsecticides or genetic protection built into crops. Consequently, novel, orally active insecticidal toxins would be valuable biological alternatives for pest control. Here, we identified a novel insecticidal toxin, parasporal crystal toxin (PC), from Bacillus bombysepticus (Bb). PC shows oral pathogenic activity and lethality towards silkworms and Cry1Ac-resistant Helicoverpa armigera strains. In vitro assays, PC after activated by trypsin binds to BmAPN4 and BtR-175 by interacting with CR7 and CR12 fragments. Additionally, trypsin-activated PC demonstrates cytotoxicity against Sf9 cells expressing BmAPN4, revealing that BmAPN4 serves as a functional receptor that participates in Bb and PC pathogenicity. In vivo assay, knocking out BtR-175 increased the resistance of silkworms to PC. These data suggest that PC is the first protein with insecticidal activity identified in Bb that is capable of causing silkworm death via receptor interactions, representing an important advance in our understanding of the toxicity of Bb and the contributions of interactions between microbial pathogens and insects to disease pathology. Furthermore, the potency of PC as an insecticidal protein makes it a good candidate for inclusion in integrated agricultural pest management systems.


Results
PC is pathogenic to silkworms. Silkworm larvae were orally infected with Bb to induce larval death, leading to black chest septicaemia: a peutz appears on the thoracoabdominal region or the 1 st to 3 rd segment of the abdomen (Fig. 1Ab) and then expands to encompass the entire body (Fig. 1Ac). To determine whether or not the pathogenic mechanism of Bb was associated with PC, PC was purified from bacterial cultures using methods similar to those used to isolate Bt parasporal crystals (Fig. 1Ba). Because the protoxin can be hydrolyzed by midgut proteases to form an activated molecule that causes lethal toxicity to insect larvae 22 , the purified PC was digested with trypsin at 37 °C for 6 h and then separated by SDS-PAGE (Fig. 1Bb). All of the bands were excised from the gel and subjected to matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry analysis. The filled triangle band represented PC after trypsin digestion, and MALDI-TOF analysis identified only one protein component in this band, namely, a recombinant PC protein that was expressed and purified (Fig. 1C). To confirm the pathogenicity of PC protein, mortality was determined by oral infection of the 5 th -instar silkworms with 50 μ g PC protein per larva. Beginning at 15 h after treatment, the PC-treated larvae began to die (Fig. S1B), but the wild-type (WT) larvae survived (Fig. S1A). We found that PC induced approximately 29% mortality compared to 0% in WT within 4 days (Fig. 1D), with an LD 50 of 140.512 μ g per larva for oral administration at the 5 th -instar stage. Furthermore, the mortality rate of 3 rd -instar larvae was 40.3% after 4 days at a dose of 50 μ g PC protein per larva. The approximate LD 50 value calculated for 3 rd -instar silkworms was 77.35 μ g per larva. These results indicated that PC is pathogenic to silkworms.

Identification of PC pathogenicity towards Bt-resistant H. armigera.
Bt crops represent one of the most widespread and controversial applications of pest management 23 . Many studies have addressed pest resistance to Bt crops and resistant pests that can overcome the lethality of Bt protein 24 . To investigate the pathogenic capacity of PC towards Bt-resistant insects for future pest control applications, we performed bioassays on larvae from one Bt-susceptible (LF) and three Cry1Ac-resistant H. armigera strains (LF5, LF10, and L240). Cry1Ac-resistant H. armigera strains were established by Prof. Kongming Wu et al. by screening field-collected populations with different doses of Cry1Ac toxins (LF5: 5 μ g Cry1Ac/ Scientific RepoRts | 5:11101 | DOi: 10.1038/srep11101 mL artificial diet, LF10: 10 μ g Cry1Ac/mL artificial diet, and LF240: 240 μ g Cry1Ac/mL artificial diet) under laboratory conditions 25 . PC was fed to 1 st -instar larvae at a dose of 50 μ g toxin/mL artificial diet. PC produced a mortality rate above 60% in the LF species within 5 days (Fig. 1E). Interestingly, three Cry1Ac-resistant strains showed pathogenicities similar to the LF species (Fig. 1F). The mortality rate analysis after oral infection with resistant larvae showed that ~70% of the LF5, LF10, and LF240 infected with PC succumbed within 5 days. The insecticide activity of PC in Cry1Ac-resistant strains indicated its potential use as a bioinsecticide or transgenic crop gene to target Bt-resistant insects.
Molecular characterization of PC. The first 29 amino acids of PC corresponded to a putative signal peptide as predicted by the SignalP program. An analysis of putative post-translational modifications using the NetNglyc and YinOYang servers suggested the absence of N-linked glycosylation and a putativeO-(beta)-GlcNAc glycosylation event at G 74 . To further explore the relationship between PC and Bt Cry toxins (Cry1A-Cry28A), we used ClustalX and Mega 4 to visualize phylogenetic relationships. Sequence alignment showed that some sites of PC are conserved with Bt Cry toxins (Fig. S2), showing . PC was extracted from Bb cultures according to the purification method used for Bt toxin and was digested by trypsin at 37 °C for 6 h. The filled triangle band represents PC after trypsin digestion. (C) Purification of PC from a prokaryotic expression system. (D) Mortality of newly exuviated 5 th -instar silkworm larvae following oral infection with PC. WT, unchallenged. (E) Mortality observed feeding PC into Bt-susceptible Helicoverpa armigera (LF) strains. (F) The mortality rate of three Cry1Ac-resistant Helicoverpa armigera strains characterized to show different resistance levels after oral infection with PC. Error bars depict ± SEM. Statistically significant differences from control samples are indicated; ** P < 0.01.
Scientific RepoRts | 5:11101 | DOi: 10.1038/srep11101 an identity of 8.6% with Bt Cry6Aa. However, as to Bt Cry6Aa, Bt Cry15Aa and Bt Cry23A, phylogenetic tree showed that PC resides on its own deep branch (Fig. S3), indicating that it should be considered a distinct toxin. The deduced amino acid sequence of PC displayed limited sequence similarity (≤11.4%) with a range of bacterial pore-forming toxins as determined by a BlastP search.
Binding assays to detect PC interaction with BmAPN4. Bt Cry toxins kill many insects by interacting with receptors such as APN or cadherin, leading to pore formation and ultimately cell lysis 14,22,26 . Bb toxins induce larvae poisoning responses in the midgut that are similar to those induced by Bt 20 . Surprisingly, PC accumulation in host intestinal epithelial cells and pores could be observed in the optical membrane of the larvae midgut after Bb bacterium infection 20 . A gene encoding an APN protein (BmAPNN4) was obtained by biopanning against Bb bacterium from a silkworm midgut T7 phage display cDNA library 27 . Additionally, some midgut-expressed APNs can be modulated by Bb infection 20,28 . Therefore, we hypothesized that silkworm APNs interact with PC to cause pathogenicity.
To test our hypothesis, far-western blot analysis was performed. The gene encoding BmAPN4 was cloned by RT-PCR, and recombinant BmAPN4 protein with a glutathione S-transferase (GST) tag at the N-terminus was expressed and purified. The trypsin-activated PC proteins were separated and transferred to PVDF membranes. The membranes were separately incubated with BmAPN4 protein before the addition of anti-GST antibody (Fig. 2Aa). Membranes incubated with GST or bovine serum albumin (BSA) protein were used as negative controls (Fig. 2Ab,c), and another negative control, a membrane without PC, was incubated with BSA to exclude the possibility of a direct interaction between the anti-GST antibody and the PVDF membrane (Fig. 2Ad). Positive bands were observed only when trypsin-activated PC/BmAPN4 complexes were present (Fig. 2Aa). Far-western blot analysis with trypsin-activated PC and BmAPN4 showed that trypsin-activated PC was capable of interacting with silkworm BmAPN4. We used co-immunoprecipitation (Co-IP) assays to determine whether BmAPN4 interacted with trypsin-activated PC. As shown in Fig. 2B, the Co-IP assay indicated that PC bound to BmAPN4. Furthermore, ELISA binding saturation assays showed that PC bound to BmAPN4 (Fig. 2C), which confirmed the interaction between PC and BmAPN4.

Cytotoxic activity of PC in Sf9 cells expressing BmAPN4. Having determined that PC binds
to silkworm BmAPN4, we investigated the biological significance of silkworm BmAPN4 in silkworms infected with Bb. To ascertain whether silkworm APNs are involved in PC pathogenicity, the cytotoxic activity of PC in Sf9 cells transfected with BmAPN4 gene was tested. BmAPN4 was expressed in the membrane fractions of Sf9 cells transfected with psl1180[hr3-BmAct4-BmAPN4-SV40] vectors (Fig. 3A, lane 2). Additionally, immunofluorescence was used to evaluate the location of BmAPN4 in cells (Fig. 3B). Separately, the incubation of live, viable cells expressing BmAPN4 with trypsin-activated PC at 50 μ g/mL induced distinct morphological changes. A significant proportion of the cells exhibited dramatic cytological changes, including swelling and lysis (Fig. 3Cd). Such effects were not observed in the control Sf9 cells (Fig. 3Ca), Sf9 cells transfected with silkworm BmAPN4 that were not exposed to toxin (Fig. 3Cb), or control Sf9 cells incubated with trypsin-activated PC (Fig. 3Cc), suggesting that the expressed BmAPN4 directly interacted with trypsin-activated PC to induce morphological aberrations and cell lysis. The number of cells with such alterations after PC administration were counted in 50 different optical fields, indicating that the proportion of viable cells was reduced to ~15.8% (Fig. 3D). Susceptibility to trypsin-activated PC was further confirmed using a cytotoxicity assay based on the extracellular detection of LDH activity. Trypsin-activated PC caused a 35.3% increase in the release of total intracellular LDH in BmAPN4-transfected Sf9 cells into the culture medium (Fig. 3E). Untransfected Sf9 cells were not affected by PC. Together with the results of the binding assay (Fig. 2), this finding indicates that BmAPN4 primarily serves as a functional receptor involved in PC pathogenicity.
Interaction of PC with Bt-R175. The interaction with APN concentrates Bt Cry toxin in the midgut brush border membrane vesicles (BBMV), where the toxin is able to bind with high affinity to cadherin 29 . Binding to cadherin is complex and involves three cadherin fragments that correspond to the extracellular regions CR7, CR11, and CR12 30 . Therefore, to determine whether cadherin is involved in Bb infection in silkworms, qRT-PCR was performed to analyse the expression pattern of cadherin (BtR-175) in the midgut after Bb infection. Similar to Bt infection, the induced peaks of the gene encoding BtR-175 appeared at 3 and 24 h, corresponding to the induced expression profiles of the gene encoding BmAPN4 after Bb infection (Fig. 4A) and suggesting that BtR-175 expression is associated with responses in Bb-infected silkworms. However, whether BtR-175 is involved in PC pathogenicity is unknown.
To examine the role of BtR-175 in the process of toxin pathogenicity, we cloned three cadherin fragments that correspond to CR7 (residues M 811 -A 927 ), CR11 (residues L 1257 -T 1398 ), and CR12 (residues N 1368 -G 1484 ) of silkworm BtR-175. The recombinant CR7, CR11, and CR12 proteins with His tags at N-termini were purified for use in binding assays. Far-western blot analysis of trypsin-activated PC and His-CR7, His-CR11, or His-CR12 resulted in positive bands that reacted with anti-His antibody only when trypsin-activated PC/His-CR7 or trypsin-activated PC/His-CR12 complexes were present, indicating that PC bound to CR7 and CR12 cadherin fragments, but not to CR11 (Fig. 4B). The purified His-CR7, His-CR11, and His-CR12 proteins were each incubated with trypsin-activated PC. After incubation, His-CR7-bound protein(s), His-CR11-bound protein(s), and His-CR12-bound protein(s) were purified  by affinity chromatography. As a negative control, Ni-NTA beads incubated with trypsin-activated PC were used to exclude the possibility of PC directly binding to the NTA resin. As shown in Fig. 4C, PC was not detected by His-CR11 pull down but could be observed in the CR7 and CR12 pull-down assays. Furthermore, ELISA binding assays confirmed that trypsin-activated PC could bind to CR7 or CR12, but not to CR11 (Fig. 4D). These results showed that PC interacted with silkworm BtR-175 by binding to the cadherin fragments CR7 and CR12.

Effect of BtR-175 on PC pathogenicity in B. mori.
To further characterize BtR-175 biological function during the toxin pathogenicity, we used CRISPR/Cas9 genome editing to knock out the BtR-175 gene for bioassay. A guiding sequence of BtR-175 gRNA was synthesized and inserted into pUC57-gRNA to form pUC57-BtR-175. The plasmids pUC57-hA4-Cas9 and pUC57-BtR-175 were introduced into silkworm embryos by co-microinjection to induce BtR-175 mutagenesis. Finally, homozygous mutants were obtained from G3 moths using T7EI genotyping and TA clone sequencing to reveal the exact mutant sequences (Fig. 5A). A homozygous mutant (Δ BtR-175) was inserted into 2 bp with disrupted BtR-175 leading to translation termination and loss-of-function (Fig. 5B). In total, 50 μ g PC per larva was used to calculate the mortality of 3 rd -instar Δ BtR-175 strain for oral administration toxicity assays. Remarkably, Δ BtR-175 silkworms were much more potent against PC. The mortality of WT silkworms orally infected with toxins at 4 days (40.2 ± 2.4%) was significantly elevated ( ** P < 0.01) compared to that of the Δ BtR-175 strain (5.9 ± 3.25%) (Fig. 5C). Moreover, no significant mortality was observed in WT silkworms (2.8 ± 2.4%) and in the Δ BtR-175 strain. This result indicated that knocking out BtR-175 gene increased the resistance of silkworms to PC. Together with the results of binding assay (Fig. 4), these data show that BtR-175 is involved in PC pathogenicity.

Discussion
Many Gram-positive and Gram-negative bacterial pathogens produce toxins that contribute to their virulence 31 . Bb, a typical natural pathogen of silkworms, is a model of insect host-pathogen interactions based on triggering a strong host response 20 . Here, we described the cloning and functional characterization of a novel toxin that contributes to the virulence of Bb, which we named PC. Bioinformatic analysis showed that PC does not exhibit any homology to other bacterial toxins and appears to be distinct from previously identified insecticidal PFTs such as Cry toxins (Fig. S3). Nevertheless, amino acid sequence analysis showed the presence of conserved residues in PC and Bt Cry toxins. Bioassays based on oral activity confirmed that this toxin isolated from Bb is an insecticidal toxin and that this toxin was lethal to B. mori (Figs 1 and 5). How might PC induce silkworm death? The insect midgut represents the first line of resistance and triggers the immune response, which deploys multiple innate defence mechanisms to fight microbial intruders. However, many bacterial pathogens are equipped with clever infectious stratagems that circumvent these defence systems. The interplay between pathogens and their hosts involves the specific recognition of surface molecules that modulate cell recognition, membrane insertion, or internalization, which determine the fate of bacterial infections and disease outcomes. Toxins are important virulence factors for bacterial pathogenicity. Interactions with single receptors are a general strategy, although several toxins such as diphtheria, Cry, or aerolysin toxins target more than one surface molecule in their binding activities and action modes 8 . Here, we demonstrated that binding to BmAPN4 provokes the pathogenicity of PC (Figs 2 and 3), although the kinetic parameters of the interaction between PC with APN have not been fully analysed. APNs are known to play key roles in the mechanisms responsible for the toxicity of various classes of Bt Cry toxins 32 . Both the monomer and oligomer Bt Cry toxins interact with the APN receptor. The binding of oligomer to APN induces its insertion into cell membrane, leading to pore formation and cell lysis, which is a key determinant of toxicity 26,33 . APNs are highly abundant proteins that are anchored to membranes by a glycosyl phosphatidylinositol anchor 34 . In Trichoplusia ni, different APNs are associated with resistance to Bt Cry1Ac toxin 35 . In Spodoptera exigua, the absence of APN causes resistance to Bt Cry1Ca toxin 36 . In B. mori, BmAPN4 is specifically expressed in the midgu 28,37 , which explains why the injection of PC in the body cavity at high doses had no impact on silkworm survival (data not shown), while oral infection induced silkworm death. In this study, we report the discovery that BmAPN4, which is specifically expressed in the midgut, functions as a receptor and is involved in Bb recognition and toxin-related pathogenicity. Additionally, the binding of toxin to cell surface is the key step in causing toxicity. Thus, the loss or alteration of this receptor could play a pivotal role in the natural development of toxin resistance in silkworms.
The pathogenicity of Bt Cry toxin is the result of sequential interactions with at least two receptor molecules, cadherin and APN 26 . Interactions with APN concentrate Bt Cry toxin at the midgut brush border membrane, where the toxin can bind to cadherin receptor 33 . The binding of Bt Cry toxin to cadherin triggers toxin oligomerization to form a pre-pore structure, which interacts with APN and results in pore formation that leads to insect death 14,26,38 . Pores form in the silkworm intestinal epithelium after Bb oral infection, after which the infiltration becomes prominent, and material exchange throughout silkworm body is disrupted similar to Bt Cry toxin poisoning 20 . Nevertheless, whether the formation of pores related to APN occurs following the binding of PC to BmAPN4 is unknown. However, in this study we found that PC interacts with two regions of silkworm BtR-175 receptor that are located in the CR7 and CR12 cadherin repeats (Fig. 4), and does not bind to CR11. Cadherin repeats have been recognized as important regions of cadherin that mediate Bt Cry toxicity. In Manduca sexta, Bt Cry1A toxins bind to cadherin Bt-R 1 receptor via interactions with CR7, CR11, and CR12 cadherin repeats, which are essential for toxicity 30 . Resistance to Bt Cry toxin in some insects is linked with mutations that disrupt a toxin-binding cadherin protein 39 . Using a CRISPR/Cas9 system to knock out the binding region of BtR-175 gene for bioassay, we identified that BtR-175 mutant exhibited reduced PC pathogenicity in vivo (Fig. 5). The enhanced activity of PC in the presence of silkworm BtR-175 correlates with the mechanism of action of the toxin. These demonstrated that PC bound to silkworm BtR-175, which is located on the luminal membrane of midgut epithelial cells 40 , to induce host death.
Bacterial toxins secreted by pathogens diffuse towards the target cell by binding to specific receptors that are ultimately inserted in cell membranes and that form a pore 8 . Aerolysin, a pore-forming toxin, initially interacts with a high-abundance and low-affinity molecule and subsequently interacts with a low-abundance and high-affinity binding site 41 . A "ping pong" binding mechanism has been reported in which Bt Cry toxin is involved in the first binding event with the high-abundance and low-affinity APN receptor and then interacts with the high-affinity cadherin receptor 33 . APN is abundant in the M. sexta midgut, in contrast to cadherin, which is present at a much lower concentration 42 . Therefore, we speculate that PC binds to the high-abundance BmAPN4 receptor before interacting with Bt-R175. However, little is known regarding the structure of PC or its loop regions and receptor binding regions. In Bt Cry toxin, differences in sequence, length, and loop conformation are thought to be key determinants of its insect midgut BBMV binding ability and toxin selectivity. Thus, mapping the specificity of binding regions in PC will help to improve the applications of PC to insect pest control with Bt toxin.
The results shown here indicate that PC, a novel toxin, is pathogenic to silkworms. BmAPN4 primarily acts as a functional receptor that participates in PC pathogenicity and BtR-175 can interact with the toxin to affect pathogenicity. Based on our data, a schematic overview of Bb infection in the silkworm 20 and previous studies of the mechanism of action of Cry toxins in lepidopterans at the molecular level 22,26 , we propose a schematic model for the process by which Bb damages the silkworm midgut and the haemolymph (Fig. 6). Larvae ingest Bb, which produce PC. PC can be digested by gut proteases. The digested PC passes through the PM to bind to the high-abundance APN receptor, after which the toxin is in close proximity to the membrane of the midgut. This interaction is followed by high-affinity binding to the BtR-175 receptor. Interactions with BtR-175 trigger oligomerization of the toxin, which then binds to the receptors and leads to pore formation (Pathway 1). This sequence of events is similar to the sequential model for Cry toxins activity proposed by Bravo et al. 22,26,43 . Additionally, unknown toxins might play a part in larval death (Pathway 2). In these models, all of the receptors would work together to disrupt the balance of infiltration and material exchange throughout the insect, resulting in larval death caused by Bb infection.
Remarkably, PC exhibits oral toxicity against H. armigera (Fig. 1) similar to B. mori. H. armigera, which is one of the most severe insect pests in Asia, attacks wheat, corn, vegetables, and cotton. The oral potency of PC and its insecticidal activity make it a good candidate for deployment against H. armigera and possibly other pest insect species. Because PC is a genetically encoded protein, engineering transgenes that encode this protein in plants should be possible. The most significant advancement in the use of bio-resources as bio-control agents in recent years is the use of transgenes encoding Bt Cry proteins, which have been incorporated into many crops to manage key insect pests successfully and to protect crops such as cotton and corn 14 . More than 420 million hectares are planted with Bt crops worldwide. Furthermore, in China, several Bt-producing crop varieties are waiting to be approved by the agricultural ministry for commercial use. However, to date, Bt Cry toxin resistance has been reported in some key pest species 15,44,45 , and the increasing use of Bt crops will accelerate the emergence of resistant insects and threaten long-term benefits. PC shows a pathogenic capacity towards three Cry1Ac-resistant H. armigera strains through oral infection (Fig. 1E), indicating that this toxin may provide farmers with a new weapon against the development of Bt Cry toxin resistance in insect pests. Thus, PC might be a good candidate for transgene engineering to replace or use in combination with Bt Cry toxins for integrated pest management systems with a strong biological control component.
The discovery of PC as a candidate for insecticides provides opportunities to develop PC into an effective insect pest control agent. Nevertheless, the most serious threat to the continued efficacy of PC is low insecticidal activity. PC has an LC 50 value of 77 μ g/per larvae. In contrast, Cry1A toxin has an LC 50 value of 1 ng per square cm or a few hundred nanograms for B. mori instar larvae (less than 200 ng/cm 2 ) indicating that PC is at least 400-fold less toxic 8,46 . The shortcoming of low toxicity of PC may be a limitation for its utility in pest control. Thus, one important future direction of research on PC is the provision of approaches to enhance its insecticidal efficacy. These possible strategies include the following. (i) The localization of the toxic portion of PC may be useful for enhancing PC activity. The goal of this strategy is to circumvent PC activation, resulting in improved toxicity against pests; thus, basic knowledge regarding the toxic portion of PC is helpful for the design of a toxin to control pests. (ii) Because the processing PC into its active form by insect proteases is crucial to its toxicity, enhanced processing by insect proteases may increase its efficacy against pests, such as the modification of PC to facilitate proteolytic activation. (iii) Binding to insect receptors is an important step in inducing toxicity; thus, increased binding to target insect receptors could improve insecticidal efficacy. Several approaches have been used to modify the binding affinity of Bt Cry toxins with the ultimate goal of producing designer toxins that target pests. According to these studies 14,47 , the alteration of PC binding affinity can be broken into the following categories: incorporation of binding fragments or domain swapping between PC and Bt Cry toxins, and site-directed mutagenesis of PC. This approach has provided important insight into PC residues that interact with insect midgut receptors. (iv) Phage display screens for mutated PC offer a potentially powerful method for engineering PC and for the subsequent selection of mutant PC with enhanced activity. Overall, these strategies for toxin modification will help to increase PC insecticidal efficacy against pests. Incorporation of these approaches into current management strategies will maximize the use of PC as a tool for crop protection and pest management.
Insect pests that lead to crop losses are a significant limiting factor in food production. In modern agriculture, insect pest control has undergone a revolution. In the United States, China, India, and Australia, the engineering of crop plants for enhanced resistance to insect pests has been a success for Bt crops that has resulted in reduced reliance on chemical insecticides. Second-generation Bt crops, termed pyramids, produce two or more Bt Cry toxins that are selected for resistance to one toxin that does not cause cross-resistance to the other toxins, which can delay or prevent the emergence of resistant insects 48 . In the past decade, farmers in Australia, India, and the United States have switched from planting first-generation transgenic crops to using pyramids against a particular pest 15 . Therefore, the goal of using PC transgenes in crops with Bt Cry toxins could be achieved by designing the toxin to enhance the efficacy of insect-and Bt-resistant pests, thus providing levels of crop protection that are more durable.

Methods
Insect and bacterial strains. Silkworm larvae (DaZao P50 strain) were reared on fresh mulberry leaves at a stable temperature of 25 °C. The H. armigera strains (LF, LF5, LF10, and LF240) were kindly provided by Prof. Kongming Wu (State Key Laboratory for Biology of Plant Diseases and Insect Pests, Chinese Academy of Agricultural Science, China) and reared using methods reported by Liang et al. 49 . . PC can be digested by gut proteases. The digested PC pass through the peritrophic membrane (PM) to bind the high-abundance APN receptor, allowing the toxin to be located in close proximity to the membrane. This interaction is followed by high-affinity binding to the BtR-175 receptor by the cadherin fragments CR7 and CR12. Interactions with BtR-175 trigger the oligomerization of toxin that binds to the receptors and leads to pore formation (Pathway 1). However, unknown toxin(s) might also induce larvae death (Pathway 2). All of these toxins and receptors could work together, leading to aberrant gut infiltration and material exchange throughout the insect body, resulting in the death of larvae that is caused by B. bombysepticus infection.
Scientific RepoRts | 5:11101 | DOi: 10.1038/srep11101 The LF strain was started with approximately 100 3 rd to 6 th instars collected from Bt cotton in Langfang, Hebei Province, China in 1998. The LF5 50 , LF10 and LF240 strains were generated with resistance by selecting insects from LF strains grown with different concentrations of MVPII (Dow AgroSciences), a commercial formulation of Cry1Ac protoxin incorporated in the diet, over more than a decade. Bb was kindly provided by Prof. Yangwen Wang (Silkworm Disease Laboratory of Shandong Agriculture University, China).

Purification of PC from Bb. The method for purifying PC was based on the procedure used for Bt
Cry toxin purification [51][52][53] . Bb was grown on 100 μ g/mL ampicillin at 37 °C. Cells were centrifuged at 8000 rpm for 20 min, and the pellet was washed three times with ice-cold 1 M NaCl and then washed five times with distilled water. The PC was solubilized in 50 mM sodium carbonate buffer (pH 9.5) containing 0.2% 2-mercaptoethanol and 25 mM EDTA on ice for12 h 54 . After the mixture was incubated, it was centrifuged at 8000 rpm for 30 min at 4 °C. The supernatant was subjected to a second round of centrifugation to remove insoluble components. The remaining supernatant was adjusted to pH 4.5 using 4 M CH 3 COONa-CH3COOH buffer and acetic acid, and was maintained on ice for 4 h. After the sample was centrifuged at 12,000 rpm for 30 min at 4 °C to remove the supernatant, the pellet was washed two times with distilled water. The PC protein was solubilized in 50 mM sodium carbonate buffer (pH 10) at 4 °C. The solubilized PC protein was digested with trypsin at trypsin: toxin ratio of 1:25 (w/w) at 37 °C for 6 h. The proteins were purified by adding them to1/10 (v/v) 4 M CH 3 COONa-CH 3 COOH buffer and incubating them at 4 °C for 15 min before centrifugation at 12,000 rpm for 10 min. The pellet was then solubilized in distilled water. These steps were repeated five times. Finally, the pellet was solubilized in 50 mM sodium carbonate buffer. SDS-PAGE gels were run and then visualized using Coomassie Brilliant Blue (CBB) staining.

Expression vector construction and recombinant protein expression. The truncated ORF of
PC and the cadherin repeats CR7, CR11, and CR12 were cloned into pET-28a expression vectors with a 6× His tag on the N-terminal ends of the target sequences. BmAPN4 was cloned into a pGEX-4T-1 expression vector with a GST tag on the N-terminus. These recombinant plasmids were transformed into the Escherichia coli BL21 (DE3) strain to express PC, GST-tagged BmAPN4 (GST-BmAPN4), His-tagged CR7 (His-CR7), His-tagged CR11 (His-CR11), and His-tagged CR12 (His-CR12) proteins. The PC, His-CR7, His-CR11, and His-CR12 proteins were purified by affinity chromatography using Ni-NTA (GE Healthcare). GST-BmAPN4 was purified by GSTrapFF (GE Healthcare). All of these purified proteins were used for subsequent experiments.
Construction of knock-out mutations. The plasmids of pUC57-hA4-Cas9 and pUC57-gRNA were kindly provided by Dr. Sanyuan Ma (State Key Laboratory of Silkworm Genome Biology, Southwest University, China). Guiding sequence of BtR-175 gRNA were synthesized and inserted to BbsI treated pUC57-gRNA, forming pUC-BtR-175. The microinjection of B. mori embryos, crossing strategy, T7 endonuclease I assay and sequencing of the mutations were performed according to the method of Ma et al. 55 .
Determination of PC insecticidal activity on silkworm and H. armigera larvae. The mortality of silkworms after oral inoculation with PC was investigated. A fresh mulberry leaf was cut into square pieces 2 cm long and coated with a dose of 50 μ g purified PC recombinant proteins per silkworm larva. Newly exuviated 3 rd -instar and 5 th -instar larvae that ate leaves with toxin were selected for continuous rearing. After 24 h, the larvae were transferred to a fresh diet (without toxin). Mortality was recorded after 4 days. We tested 30 larvae per test for each treatment, and this bioassay was replicated three times. The LC 50 was calculated using GraphPad Prism software version 5.0 for Windows (GraphPad Software, San Diego, CA). Statistical analysis was performed with GraphPad using Student's t-tests. No significant differences between the sessions were observed. P > 0.05 and statistically significant differences are indicated; * P < 0.05, ** P < 0.01.
For H. armigera bioassay, the method was based on the procedure by Cao et al. 25 . In brief, the 1 th -instar susceptibility (LF) and three Cry1Ac-resistant larvae (LF5, LF10, and LF240) were placed individually on PC-contaminated artificial diet (doses of PC: 50 mg toxin/mL artificial diet) in plastic wells in 24-well plates. And 144 larvae were test per PC concentration. After five days, we recorded the mortality rate. The statistical analysis were performed according to above described.
Far-western blot, co-immunoprecipitation, pull-down assay and ELISA binding assay. Far-western blots were performed according to the method of Wu et al. 58 . A summary of the Co-IP protocol is described below. The proteins (100 μ g) were incubated with IP antibody for 1 h at room temperature. The complexes were bound to Protein A magnetic beads (Thermo Fisher Scientific) for 1 h at room temperature and then BS3 (Thermo Fisher Scientific) was used for cross-linking. After washing, the beads were incubated with PC (100 μ g) overnight at 4 °C. The complexes were eluted after washing. The methods used for His pull-down experiments were the same as those described for the Pierce Protein Interaction Pull-Down Kit (Thermo Fisher Scientific). A summary of the ELISA protocol used was described previously by Pacheco 33 .
Cytotoxicity assay of PC in Sf9 cells. Sf9 cells were harvested and transferred to a 48-well tissue culture dish (Costar) at 0.1 × 10 6 cells per well. After the cells settled, they were transfected with pSL118 0[hr3-BmAct4-BmAPN4-SV40]. After 48 h, the cell layer was overlaid with PC at 50 μ g/mL in PBS after three gentle washes with PBS. The cells were incubated with the proteins for 6 h, after which they were observed under an Olympus TH4-200 microscope. The total number of cells per field was counted. Final values represent an average of 50 fields that were randomly selected for each treatment; each treatment was replicated three times. This assay measures the amount of stable cytosolic lactate dehydrogenase (LDH) released from the lysed cells using a coupled enzymatic assay that results in the conversion of a tetrazolium salt into a red formazan product according to the methods of Wang et al. 59 and Tsuda et al. 60 .

RNA extraction and quantitative real-time PCR after Bb challenge. RNA extraction and
qRT-PCR analysis of the midgut expression pattern of the gene encoding BtR-175 after Bb infection were performed as described by Lin et al. 28 .