Antimicrobial peptides (AMPs) and lysozymes are the main effectors of the insect immune system, and they are involved in both local and systemic responses. Among local responses, midgut immune reaction plays an important role in fighting pathogens that reach the insect body through the oral route, as do many microorganisms used in pest control. Under this point of view, understanding how insects defend themselves locally during the first phases of infections caused by food-borne pathogens is important to further improve microbial control strategies. In the present study, we analyzed the transcriptional response of AMPs and lysozymes in the midgut of Spodoptera exigua (Lepidoptera: Noctuidae), a polyphagous pest that is commonly controlled by products based on Bacillus thuringiensis (Bt) or baculovirus. First, we comprehensively characterized the transcripts encoding AMPs and lysozymes expressed in S. exigua larval midgut, identifying 35 transcripts that represent the S. exigua arsenal against microbial infection. Secondly, we analyzed their expression in the midgut after ingestion of sub-lethal doses of two different pore-forming B. thuringiensis toxins, Cry1Ca and Vip3Aa, and the S. exigua nucleopolyhedrovirus (SeMNPV). We observed that both Bt toxins triggered a similar, wide and in some cases high transcriptional activation of genes encoding AMPs and lysozymes, which was not reflected in the activation of the classical systemic immune-marker phenoloxidase in hemolymph. Baculovirus ingestion resulted in the opposed reaction: Almost all transcripts coding for AMPs and lysozymes were down-regulated or not induced 96 hours post infection. Our results shed light on midgut response to different virulence factors or pathogens used nowadays as microbial control agents and point out the importance of the midgut immune response contribution to the larval immunity.
Citation: Crava CM, Jakubowska AK, Escriche B, Herrero S, Bel Y (2015) Dissimilar Regulation of Antimicrobial Proteins in the Midgut of Spodoptera exigua Larvae Challenged with Bacillus thuringiensis Toxins or Baculovirus. PLoS ONE 10(5): e0125991. https://doi.org/10.1371/journal.pone.0125991
Academic Editor: Juan Luis Jurat-Fuentes, University of Tennessee, UNITED STATES
Received: January 7, 2015; Accepted: March 27, 2015; Published: May 18, 2015
Copyright: © 2015 Crava et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This research was funded by the Spanish Ministerio de Ciencia e Inovación (projects AGL2009-13340-C02-01, AGL2011-30352-C02-02 and AGL2012-39946-C02-01), by the Generalitat Valenciana (GVPrometeo2011-044 and ISIC/2013/004) and by the European FEDER funds.
Competing interests: The authors have declared that no competing interests exist.
Antimicrobial peptides (AMPs) and lysozymes are small proteins produced by a wide range of organisms (from bacteria to humans) which have direct antimicrobial activity against pathogens like bacteria, fungi and viruses [1,2]. In eukaryotes, they are part of the innate immune system and they act as chemical effectors to arrest microbial infections. In insects, they are induced after pathogen recognition and they are involved in two different innate immune defenses: the local and humoral responses . Insects rely on both immune responses to fight infections, and one of the main differences between them is that their effects are either local or systemic.
Local defenses are constituted by physical barriers that prevent microbial penetrance (such as epithelia), and by local production of AMPs, lysozymes and reactive oxygen species, at the site of infection. Hence, AMPs and lysozymes represent one of the first chemical barriers against infections. Systemic defenses are activated later and include a series of orchestrated mechanisms which range from the production of AMPs and lysozymes by the fat body, which are finally secreted into hemolymph to clear invading microbes inside body cavity (humoral response), to the cellular response lead by specialized cells like hemocytes (phagocytosis and encapsulation of foreign intruders), to the melanization and coagulation of the hemolymph [3,4]. Since the discovery of the first insect AMP , the role of antimicrobial proteins in humoral response has been broadly investigated , however their contribution to the local response was discovered later  and to date it has only been studied in model organisms [3,7,8]. Besides their functions in the direct clearance of invading microbes in immune response, antimicrobial proteins like AMPs and lysozymes also exert other important functions as immuno-modulatory activities  and maintenance of the endosymbiotic bacteria homeostasis .
Since the first AMP was discovered in the 1980s in a giant silk moth , a large number of AMPs have been identified to date. Insect AMPs generally share common features: they are cationic molecules of less than 100 amino acid residues, and their expression is generally induced by injury or microorganism infection . The importance of AMPs and lysozymes in insects defense is thought to be reflected in the expansion of genes coding for such proteins in insect genomes . Accordingly to this hypothesis, the increasing availability of transcriptomic and genomic data from several insect species has led to the discovery of a variety of immune-related genes which highlights the evolutionary diversity and amplitude of the AMP arsenal in insects. It has been proposed that expansion or loss of AMPs in different insect species is a consequence of the different pathospheres in which insects live, and that the remarkable plasticity of genes coding for AMPs represents an adaptive response to biotic and abiotic stresses . To date, the best characterized AMPs are those from the model organism Drosophila melanogaster, whose response patterns are well known. About 20 immune-inducible AMPs, which can be grouped in seven classes, are expressed in D. melanogaster after immune-challenges. Diptericins, drosocins and attacins are very effective against Gram-negative bacteria. Defensins are important for killing Gram-positive bacteria whereas drosomycins and metchnikowins are induced after fungal infection. Finally, cecropins have both antibacterial and antifungal properties [3,4]. In another model organism, the Lepidoptera Bombyx mori, around 30 AMPs and four lysozymes have been described so far .
Spodoptera exigua (Hübner) is a polyphagous pest that attacks many important economic crops, such as cotton, alfalfa, tomato, sugar beet and many ornamental plants worldwide. It has been controlled by chemical pesticides during decades leading to the problem of development of resistance . This, together with the public concerns about the ecological risks related with the use of chemical pesticides, has promoted the adoption of more eco-friendly methods. Nowadays, two effective microbial alternatives are commercially available for the control of S. exigua: products containing Bacillus thuringiensis (Bt) Berliner spores and crystal toxins or baculovirus-based insecticides. B. thuringiensis is a Gram-positive bacterial pathogen that produces several insecticidal toxins during its life cycle, which target the midgut epithelium of susceptible insects (mainly larval stages from orders Coleoptera, Lepidoptera and Diptera). The most studied, as well as the most used Bt toxins, are called Cry proteins, which are produced as crystal inclusions during sporulation . Cry toxins are widely used in countries that allow transgenic crop cultivation, since they are expressed in transgenic plants to provide resistance against several Lepidoptera and Coleoptera pests. A second class of Bt toxins (also expressed in transgenic plants) includes Vip proteins, which are soluble proteins produced during the vegetative phase of B. thuringiensis life cycle . Toxins from both classes are toxic after ingestion and although their mode of action has not been completely clarified, is thought that they both act as pore-forming toxins, inserting in the midgut epithelial membrane of susceptible insects and causing an osmotic shock that leads to cellular death. Whereas the mode of action of Cry proteins has been extensively studied and several Cry-binding proteins have been identified from the midgut epithelia , little is known about the details of the mode of action of Vip proteins [17,18]. Baculoviruses are large DNA viruses that infect arthropods, mainly from the orders Lepidoptera, Diptera and Hymenoptera. During their replication cycle, they produce two distinct morphological and functional forms, the occlusion-derived viruses (ODVs) and the budded viruses . Several ODVs are normally included in a single polyhedron occlusion body (OB) which is orally consumed by larvae in natural conditions. The alkaline environment of larval midgut favors ODV release, which subsequently cross the peritrophic membrane and fuse with the cell membrane of midgut epithelial cells. The virus replicates in these cells forming budded viruses that allow for systemic spread of the infection within the host .
Larval midgut is targeted by both B. thuringiensis and baculovirus, which use it as portal of entry to later establish a systemic infection within insect body; hence midgut immune response may play an important role in defense against both pathogens, and contribute to the success or failure of the infection at early phases. A better knowledge of the diversity of antimicrobial proteins expressed in the larval midgut as well as the deciphering of their specific activity in midgut immune response during food-borne pathogen challenge will broaden our understanding of how caterpillars have adapted to pathogens and environment. Studies which include insects other than model organisms, especially crop pest species which are routinely controlled by microbial pesticides, will shed light on the insect response to the microbes currently used in insect control, and potentially lead to the improvement of the microbial insecticides.
In the present study we focused on the characterization of the main arsenal of antimicrobial proteins (AMPs and lysozymes) expressed in the midgut of S. exigua. We first characterized transcripts coding for AMPs and lysozymes, describing for the first time several new AMPs expressed in this insect. Quantitative real-time RT-PCR (qRT-PCR) was then used to determine the local regulation of a large subset of AMP and lysozyme transcripts in larvae after sub-lethal oral intoxication with diverse Bt virulence factors like Cry1Ca and Vip3Aa toxins, and with a specialized pathogen, the S. exigua baculovirus (SeMNPV). Our results showed that midgut transcriptional response of antimicrobial proteins was wide and robust when Bt pore-forming toxins were ingested in contrast to what happened when the ingested pathogen was a baculovirus, which exerts toxicity through a completely different process.
Materials and Methods
Sequence mining and phylogenetic analyses
S. exigua transcriptome  was mined for the presence of unigenes coding for AMPs and lysozymes. Similar nucleotide sequences were aligned with ClustalX 2.1 with default settings  and visualized in GeneDoc . If necessary, alignments were manually edited (in case of misleading assemblies which resulted in unigenes containing two coding sequences with two different Blast hits or by removing zones with Phred values lower than 20). In order to obtain a non-redundant set of transcripts, unigenes with nucleotide similarity higher than 90% (comprising both ORF and UTRs) were considered as alleles of the same gene. The existence of the assembled unigenes coding for AMPs and lysozymes was confirmed by looking for their presence in our Sanger library  (S1 Table). In case that no correspondent sequences were present in the Sanger library, we designed specific primers to amplify S. exigua cDNA and confirm the presence of such AMP or lysozyme by sequencing (S2 Table). Hence, all the sequences described in this study were verified by Sanger sequencing therefore discarding the possibility that artifacts derived from misleading contig assembly could be present. However, it should be noted that we cannot discard that some transcripts might be allelic variants of the same gene, since heterogeneity in immune system genes is highly correlated with pathogen diversity via host-parasite coevolution . Different pathogens can maintain host genetic variation by favoring the distribution of alternative genotypes  and these could be reflected in our transcriptome containing RNA from three different S. exigua colonies .
Amino acid sequences, as well as molecular weight of the putative proteins, were predicted by the EditSeq software (DNAStar, Madison, WI, US). Predicted amino acid sequences were aligned with MUSCLE software  and visualized in GeneDoc. Phylogenetic analyses were carried out using the neighbor-joining algorithm with p-distance method and 100 bootstrap replicates on the MEGA5 program . Phylogenetic trees were visualized with MEGA5. Sequences from other species employed in phylogenetic analyses were retrieved after Blast-search against different databases: NCBI’s non-redundant database (nr), SPODOBASE (http://bioweb.ensam.inra.fr/spodobase/) and Silkworm Genome Database (http://silkworm.genomics.org.cn/). To obtain Spodoptera frugiperda amino acid sequences, we used S. exigua nucleotide sequences as queries in the Blast search implemented in SPODOBASE. For each search, the ORF from the sequence with the highest e-value was translated to protein using the software EditSeq and used for the analysis. All the S. exigua sequences identified in this study have been deposited in GenBank (accession numbers KP056508-KP056542).
S. exigua larvae challenged with Bt toxins were obtained from the FRA colony, which was kindly provided by Dr. M. López-Ferber (INRA, St.-Christol les Alés, France)  and maintained at the University of Valencia (Spain). Baculovirus-free S. exigua larvae were kindly provided by Dr. P. Caballero from the Public University of Navarra (Pamplona, Spain). The Baculovirus-free colony was regularly checked for the presence of baculovirus infection by qRT-PCR and proved to be negative for SeMNPV. Both insect colonies were reared at 25°C, relative humidity of 70% and a photoperiod of 16 h:8 h (light:dark), on artificial diet, as described in Bel et al. .
Detection of AMP and lysozyme transcripts in larval midgut
To ensure that all the transcripts identified after transcriptome sequence mining were expressed in S. exigua larval midgut, specific primers were designed (S2 Table). cDNA was prepared from total RNA extracted from larval midguts using RNAzol RT reagent (Sigma-Aldrich, St. Louis, MO, US), following the manufacturer’s instructions. Purified RNA was then treated with DNase I (Invitrogen, Carlsbad, CA, US) and subsequently reverse-transcribed to cDNA using oligo-(dT) primers and the SuperScript II Reverse Transcriptase (Invitrogen), according to the manufacturer's protocol. PCRs were carried out in a final volume of 25 μl containing 400 ng of cDNA, 0.75 U of Taq polymerase (Biotools, Madrid, Spain), 1X reaction buffer, 160 μM dNTPs and 0.32 μM of each primer. Amplification conditions were the following: 94°C for 5 min, followed by 35 cycles of 94°C for 30 s, 60°C for 1 min and 72°C for 1 min and a final elongation at 72°C for 5 min. Correct amplicon sizes were checked after running PCR products in a 2% agarose gel stained with GelRed (Biotium, Hayward, CA, US).
Bt toxin production and purification
Vip3Aa protein was produced by an Escherichia coli WK6 strain which harbored the recombinant plasmid pGA85, kindly supplied by Bayer CropScience N.V. (Ghent, Belgium). Protein production, lysate preparation and Vip3Aa quantification were carried out as described by Bel et al. . Filtered lysates were stored at -20°C until use. Cry1Ca protein was also produced by an E. coli WK6 strain, which contained the recombinant plasmid pDB150, kindly supplied by Dr. R. de Maagd (Wageningen University, the Netherlands). Inclusion body purification, Cry1Ca protein solubilization and trypsin activation were performed according to the protocol described by Herrero et al. . Cry1Ca protein was subsequently purified by anion-exchange chromatography using the ÅKTA Explorer System (GE Healthcare Life Sciences, Uppsala, Sweden). The protein was first dialyzed overnight at 4°C against 20 mM Tris/HCl buffer (pH 8.6), centrifuged and filtered before to be loaded on a 5 ml HiTrap Q column (GE Healthcare Life Sciences). Cry1Ca was then eluted from the column by a NaCl gradient. Fractions were recovered and checked for the presence of Cry1Ca by 12% SDS-PAGE. Fractions containing Cry1Ca were pooled, quantified by Bradford assay and stored at -20°C until use.
Bt toxin treatments
Treatments with Bt toxin were carried out according the method previously used to analyze the response of S. exigua to Vip3Aa toxin by microarray analysis . We used newly molted L4 larvae (±16 h) and toxin concentrations able to produce 99% growth inhibition. In the case of Vip3Aa we used a concentration of 111 ng/cm2  whereas in the case of Cry1Ca we used 900 ng/cm2 (S1 Fig). To obtain results comparable to those described by Bel et al. (2013) , we used crude E. coli lysate containing Vip3Aa. However, Cry1Ca protein has different biochemical properties than Vip3Aa and it is produced as an insoluble protein (inclusion body) in E. coli. Hence, to achieve results comparable between the two toxins, Cry1Ca protein previously purified by anion-exchange chromatography was mixed with a known concentration of crude lysate total proteins from an empty E. coli WK6 strains (S2 Fig). The crude lysate from the empty E. coli WK6 was then used as control for both Vip3Aa and Cry1Ca, as well as the phosphate buffer (pH 7.4) containing 0.5 M NaCl used for cell lysis.
Three independent replicates were performed for each treatment. In each replicate, 16 weighed larvae were exposed to surface-contaminated diet placed in 2 cm2 wells of a 128-well bioassay tray (Bio-Cv-16, C-D International, Pitman, NJ, US). After 24 h larvae were weighted again to calculate growth inhibition using the formula previously described  and six larvae of each treatment were selected for hemolymph extraction and gut dissection. After dissection, guts were pooled and total RNA was purified using RNAzol RT reagent, following the manufacturer′s protocol. RNA was first treated with Recombinant DNase I from Takara (Saint-German-en-Laye, France) and then 2 μg of DNase-treated RNA were reverse-transcribed to cDNA using the PrimeScript II 1st strand cDNA Synthesis Kit (Takara), following the manufacturer′s instructions. To improve cDNA synthesis from total RNA, each reaction included both oligo-(dT) primers and random 6-mers.
S. exigua newly molted L4 larvae were individually challenged with SeMNPV at a dose of 104 OBs/larvae by adding virus suspension to a diet plug, as previously described . Control larvae were fed with water. Five larvae were used per each treatment and condition, and the experiment was performed in triplicate. Guts from infected and control larvae were collected at 96 hours post infection (hpi), pooled, and used for total RNA extraction using RNAzol RT reagent. 1 μg of RNA was used for cDNA synthesis. RNA was first treated with DNase I (Invitrogen) and subsequently reverse-transcribed to cDNA using oligo-(dT) primer and SuperScript II Reverse Transcriptase (Invitrogen) according to the manufacturer's protocol.
Quantitative real-time RT-PCR (qRT-PCR)
Primers for gene expression analysis were designed using Primer Express Software from Applied Biosystems (Carlsbad, CA, US) (S2 Table) and synthesized by Sigma-Aldrich. For families whose members showed a high degree of similarity (e.g. cecropins or diapausins), regions with high variability were used (most of them corresponding to the 3’end of the ORFs and to the first part of the 3’UTRs). However, in the case of c-LYZ1 the amplified fragment comprised the three isoforms 1A, 1B and 1C. Similarly, the amplified fragment for cecropin F covered both isoforms F1 and F2. Subunit C of ATP synthase was used as reference gene for normalization [28,31,32]. qRT-PCR reactions were carried out in optical 96-well plates (Applied Biosystems) on an ABI PRISM 7000 machine (Applied Biosystems) using the EvaGreen dye (Biotis, Vilnius, Lithuania), following the manufacturer's instructions. Each reaction was performed in a total volume of 20 μl, which contained 4 μl of cDNA diluted 1:10 (corresponding to ~ 300 ng). Forward and reverse primers were added to a final concentration of 300 pM. The specific amplification of transcripts was verified by dissociation curve analysis. Each PCR reaction was run twice if the SD between the two replicates was lower than 0.35 and if the Cts were not higher than 30, otherwise further replicates were added. Prior to quantifying differential expression among different treatments, the efficiency of each pair of primers was evaluated by performing 5-fold dilution series experiments (S3 Table). Primer efficiencies were taken into account during calculation of the expression ratios (-fold change) by using the REST MCS software (version 2) . Only in the case of c-LYZ3, we used a default efficiency value of 2 since it was not possible to test primer efficiency because the transcript was always expressed at very low levels in all tested tissues. Results were visualized with GraphPad Prism v. 5.1 (GraphPad Inc., La Jolla, CA, US).
Hemolymph phenoloxidase (PO) activity was determined using 2 μl of hemolymph obtained from each individual larva as previously described . Hemolymph was collected, immediately diluted 1:10 in ice-cold PBS (137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH7.4) and centrifuged at 16,000 x g for 3 min at 4°C to obtain cell-free plasma. Samples were then snap frozen in liquid N2 and conserved at -80°C until use. PO activity was assayed by adding 8 μl of cell-free plasma to 200 μL of 1 mM L-Dopa (Sigma-Aldrich) in PBS. Measurements were taken by an ELISA Multiskan Ascent reader (Thermolab Systems) at 490 nm every minute throughout 45 minutes. Since the observed absorbance curve was linear from 20–45 min after adding the substrate, the slope of the curve from 25–45 min of the reaction was used to calculate the absorbance increase per min. Measurements were repeated twice for each sample.
Results and Discussion
Identification of S. exigua transcripts encoding AMPs and lysozymes in larval midgut
Extensive data mining of the S. exigua transcriptome  was performed to identify transcripts encoding AMPs and lysozymes. The effort resulted in the identification of a wide and diverse spectrum of antimicrobial proteins that constitute a considerable part of the defensive arsenal of S. exigua against pathogen infection (Table 1).
All the transcripts identified after transcriptome sequence mining were amplified using cDNA synthesized from midgut RNA, revealing that these antimicrobial proteins are actively transcribed in S. exigua midgut in normal conditions. This may be due to the need for broad spectrum local response molecules to control different food-borne pathogens which may continuously reach the insect body cavity through midgut, since this is the only part of the gastrointestinal tract that is not covered by a cuticular layer. Moreover, feeding behavior of lepidopteran larvae exposes them to a wider range of potential pathogens compared to piercing-sucking insects, since they ingest a high quantity of plant material including plant surface, where a large number of bacteria is most probably located. Also, a pool of diverse antimicrobial proteins may be needed in midguts to maintain homeostasis of the commensal bacteria community since lepidopteran larvae seem to be able to shape the phylotypes of the gut microbiota, as recently reported . In any case, our results show that S. exigua midgut expresses a large set of transcripts encoding antimicrobial proteins, highlighting the important role of these peptides in larval midgut homeostasis and immunity.
It has been recently proposed that, amongst insects, generalist herbivores may have a more efficient immune defense strategy than specialist herbivores , and in the case of S. exigua this may be reflected in the expansion of AMP and lysozyme gene families. Further analyses of the EST database SPODOBASE revealed that such an arsenal is widely shared with another generalist related species, S. frugiperda, suggesting a common evolution for AMPs and lysozymes between these two species. Below, the different families of S. exigua AMPs and lysozymes with the description of their main features are reported, including the new transcripts retrieved in this work.
In Lepidoptera, two classes of glycine-rich AMPs have been identified: attacins and gloverins. It has been reported that both mainly target Gram-negative bacteria and fungi . Attacins have been found in both Lepidopteran and Dipteran species whereas gloverins are typical of Lepidoptera. Attacins were first purified from the hemolymph of Hyalophora cecropia and separated in two groups based on their amino acid properties: the basic and the acidic isoforms . Our transcriptome data mining led to the identification of three transcripts encoding attacins in S. exigua (Table 1), in contrast to the two genes present in B. mori genome . SeAttacin1 (previously described by Bang et al., ) corresponded to the acidic form (calculated pI 6) whereas SeAttacin2 and SeAttacin3 represented the basic forms (calculated pI 9). Homologs from S. frugiperda were retrieved by Blast searches in SPODOBASE (acc. nr. SfAttacin1, Sf2H08805-3-1; SfAttacin 2, Sf1P24355-5-1; SfAttacin 3, Sf2H07836-5-1) and amino acid identities between the two species ranged from 85 to 90%. Phylogenetic analysis revealed that lepidopteran attacins formed a highly divergent gene family, with no close orthologs or apparent subfamily structure conserved across lepidopteran families (Fig 1).
Upper panel) Neighbor-joining tree of Lepidoptera attacins. D. melanogaster attacins were used as outgroup. Bootstrap values (above 75) are indicated on each node of the tree (100 replicates). Sequences from nr database, SPODOBASE and Silkworm Genome Database were used. Accession number of each sequence is indicated. Abbreviations: Am: A. mylitta, Ap: Antheraea pernyi, Ay: Antheraea yamamai, Bm: B. mori, Dm: D. melanogaster; Dp: D. plexippus, Ha: H. armigera, Hv: H. virescens, Hc: H. cecropia, Hcu: Hyphantria cunea, Lo: Lonomia obliqua, Ms: Manduca sexta, Of: O. furnacalis, Pp: Papilio polytes, Px: Papilio xuthus, Sc: Samia cynthia, Se: S. exigua, Sf: S. frugiperda, Tn: T. ni. Lower panel) ClustalX alignment of SeAttacins. Black shadows highlight residues conserved between the three isoforms whereas grey shadows highlight residues conserved among two isoforms. Arrows indicate conserved glycine residues.
Gloverins are Lepidoptera glycine-rich AMPs that have not been identified in other insect orders . Up to four different gloverin genes have been found in B. mori genome so far  whereas in the S. exigua transcriptome only one transcript coding for gloverin was identified (Table 1), which corresponded with the gene previously described by Hwang & Kim . The in silico translated protein shared 92% identity with its S. frugiperda homolog (acc. nr. Sf2H01396-5-1), retrieved from SPODOBASE after Blast search.
Cysteine-rich AMPs are cationic AMPs widely distributed among different organisms, from plants to humans. The presence of at least six cysteine residues forming at least three disulfide bridges is their main structural feature . Within insect class, the best-studied cysteine-rich AMPs have six conserved cysteine residues which form a cysteine-stabilized alpha beta motif (CS-αβ). CS-αβ lepidopteran AMPs have been well studied in S. frugiperda and four different transcripts belonging to four different classes have been described [41,42]. In S. exigua transcriptome, we found up to five transcripts coding for CS-αβ AMPs that could be grouped into four classes. For each of the first two classes, SeDefensin (alternatively named SeSpodoptericin according to nomenclature in Spodoptera genus) and SeGallerimycin, we identified only one transcript, as described in S. frugiperda. The amino acid identities between the predicted proteins and their S. frugiperda homologs (acc. nr. AAQ18895 and AAQ18896, respectively) were 93 and 85%, respectively. Interestingly, in B. mori genome only one gene coding for defensin has been retrieved, whereas no gallerimycin orthologs have been found so far .
Another class of CS-αβ lepidopteran AMPs are the x-tox proteins, which are large pre-pro-proteins composed by several CS-αβ motifs (the structural scaffold of invertebrate defensins) . Only one transcript coding for SeX-tox was retrieved in S. exigua transcriptome, and it was slightly different from its ortholog from S. frugiperda (acc. nr. AFC87713). S. frugiperda x-tox is composed of 11 CS-αβ whereas the in silico translated sequence from S. exigua only had 8 repeats (S3 Fig). It is thought that genes encoding x-tox in Lepidoptera undergo accelerated evolution, by acquisition of new repeat motifs or by duplication of preexisting repeat motifs. To support this hypothesis, besides finding different numbers of CS-αβ motifs between species (six in G. mellonella, five to six in B. mori and three in H. virescens), it has been shown that the number of CS-αβ motifs in S. frugiperda can also vary between individuals from the same species . Hence, although we identified an 8 repeat SeX-tox, we cannot rule out that this protein may be present with a different number of repeats in other S. exigua individuals or populations.
Differently from the other classes of CS-αβ AMPs, two different copies of cobatoxin were identified in S. exigua transcriptome (Table 1), and the corresponding mature peptides shared 41% amino acid identity. In S. frugiperda, the homolog to SeCobatoxin A had been already described (acc. nr. AAP69839), whereas the homolog to SeCobatoxin B was retrieved from SPODOBASE after Blast search (acc. nr. Sf1P05363-5-1). The amino acid identity between the two species ranged from 72 to 83% for cobatoxin A and B, respectively. Phylogenetic analysis including lepidopteran cobatoxin protein sequences available in GenBank revealed that at least two cobatoxins were commonly present in Noctuidae species (Fig 2) whereas no orthologs have been reported in B. mori genome so far .
Upper panel) Neighbor-joining tree of Lepidoptera cobatoxins. Bootstrap values (above 75) are indicated on each node of the tree (100 replicates). Sequences from nr database and SPODOBASE were used. Accession number of each sequence is indicated. Ha: H. armigera, Se: S. exigua, Sf: S. frugiperda, Tn: T. ni. Lower panel) ClustalX alignment of SeCobatoxins. Black shadows highlight residues conserved between the two isoforms. Dotted box indicates the sequences of the mature peptides . Arrows indicate conserved cysteine residues from CS-αβ motif.
Other types of cysteine—rich AMPs are diapausins, which differ from CS-αβ AMPs in their primary structure, although tridimensional folding is similar . The first diapausin described was identified in the leaf beetle Gastrophysa atrocyanea , and was named as diapausin because it was found highly expressed in adults during diapause. This type of AMP has been poorly described in insects; in Lepidoptera there is only one report describing a diapausin gene (from S. exigua) . In the S. exigua transcriptome, we identified eight different types of transcripts coding for SeDiapausins (Table 1). Based on amino acid similarity, these sequences were divided into two clusters that we named SeDiapausin A and SeDiapausin B (Fig 3). Seven different transcripts were classified as diapausins type A whereas only one was assigned to the type B group. Differences among transcripts encoding SeDiapausins A were mainly due to variability in the 3’ UTR sequences whereas the ORFs were conserved. SeDiapausins A and B shared an amino acid identity ranging from 77 to 79%. The major differences between these two types of proteins were mainly due to changes in the signal peptide and in C-terminal sequences (Fig 3). Blast searches on SPODOBASE allowed us to identify two S. frugiperda transcripts coding for proteins homologous to SeDiapausin A (acc. nr. Sf2H03719-5-1) and B (acc. nr. Sf1P04653-5-1) (amino acid identity from 97 to 87%, respectively) (Fig 3). Interestingly, no diapausin homologs were retrieved from the B. mori genome database after Blast search.
Upper panel) Neighbor-joining tree of insect diapausins. SeDiapausins A3, A5, A6 and A7 were omitted since they were identical to SeDiapausin A1. Bootstrap values (above 75) are indicated on each node of the tree (100 replicates). Sequences from nr database and SPODOBASE were used. Accession number of each sequence is indicated. Ga: G. atrocyanea, Ok: Oiketicus kirbyi, Se: S. exigua, Sf: S. frugiperda, Sli: S. littoralis, Sl: S. litura. Lower panel) ClustalX alignment of SeDiapausins. Black shadows highlight residues with 100% identity whereas grey shadows highlight residues with more than 80% identity. Arrows indicate conserved cysteine residues (CX3CX9CCX10CX6C motif ).
Proline-rich AMPs have a different mode of action with respect to other AMPs, since it seems that they penetrate susceptible pathogen cells and act intracellularly . In Lepidoptera, the best-characterized AMPs belonging to this family are lebocins and it is commonly thought that they are synthesized as large polypeptide precursors that are subsequently split into four or five peptides by proteolytic processing . Data mining of S. exigua transcriptome rendered two transcripts encoding two lebocins which shared 79% amino acid identity (Table 1). Both S. exigua lebocin amino acid sequences displayed three conserved RXXR motifs, which is important for the proteolytic activation  (Fig 4). Only one lebocin sequence, the SeLebocin2 homolog, was found in S. frugiperda EST database (acc. nr. Sf2H01005-5-1, 80% amino acid identity). Likewise only one gene coding for a lebocin has been found in the B. mori genome . Phylogenetic analysis showed that SeLebocin1 is closer to SeLebocin2 than to other lebocins from other lepidopteran genus (Fig 4) suggesting that these two copies may be paralogs originated from gene duplication within the S. exigua species.
Upper panel) Neighbor-joining tree of Lepidoptera diapausins. Bootstrap values (above 75) are indicated on each node of the tree (100 replicates). Sequences from nr database and SPODOBASE were used. Accession number of each sequence is indicated. Am: A. mylitta, Ap: A. pernyi, Bm: B. mori, Ms: M. sexta, Pi: P. includens, Pr: Pieris rapae, Sc: S. cynthia, Se: S. exigua, Sf: S. frugiperda, Tn: T. ni. Lower panel) ClustalX alignment of SeLebocins. Black shadows highlight residues conserved between the two sequences, arrows indicate proline residues. Dotted boxes indicate RXXR motifs which are involved in proteolytic activation .
Amphipathic peptides are a class of α-helical AMPs present across different insect orders . The first animal immune-inducible AMP described belongs to this family: it was discovered in H. cecropia and therefore was named cecropin . It is thought that multiple cecropin copies are present in Lepidoptera species, and in fact, in the B. mori genome, thirteen genes have been identified so far . A previous phylogenetic analysis of cecropins in S. exigua with 31 S. exigua cecropin unigenes  grouped them into six different subfamilies. We performed a more exhaustive analysis that included-38 S. exigua unigenes with homology to cecropins, this resulted in the identification of eight different subfamilies of transcripts, that encoded seven different pre-proteins (ORFs of SeCecropin A1 and A2 were identical) (Table 1). Alignment of the deduced mature amino acid sequences encoded by the eight transcripts showed that they could be grouped into six highly similar protein classes (amino acid identities ranged from 36 to 96%) that corresponded to the six subfamilies previously described . The number of protein classes increased to seven when pre-protein sequences were compared since the amino acid sequences of pre-SeCecropins F1 and F2 differed by two amino acid substitutions located in the signal peptide. In the closely related species S. frugiperda, up to thirteen different types of transcripts encoding cecropins had been identified and grouped into four sub-families named A, B, C and D , which respectively corresponded to the families B, A, E and C described in the present study (amino acid identity ranged from 86 to 100%). However, homologs to SeCecropin D (acc. nr. Sf2H07985-5-1), F1 (acc. nr. Sf1H09041-3-1) and F2 (acc. nr. Sf2H04303-5-1) are also present in S. frugiperda; their sequences were retrieved by Blast searches on SPODOBASE (amino acid identity ranged from 94 to 97%), suggesting that cecropin gene number in Spodoptera genus is conserved.
Another class of lepidopteran amphipathic AMPs is moricin. Moricin was first described in B. mori  and, differently from cecropins, it has so far only been found in Lepidoptera. BmMoricin has a broad spectrum of activity since it is active against both Gram positive and negative bacteria . The B. mori genome contains only one gene coding for moricin, but eight genes encoding moricin-like proteins . Data mining of S. exigua transcriptome led to the identification of a unique type of transcript encoding moricin (which shared 97% amino acid identity with its S. frugiperda homolog retrieved from SPODOBASE, acc. nr. Sf2H03805-5-1) (Table 1), whereas no homologs to B. mori moricin-like proteins were found.
Lysozymes are antimicrobial enzymes that have been extensively studied in both plant and animal kingdoms. They have the ability to cleave the β-1,4 glycosidic linkages between N-acetylmuramic acid and N-acetylglucosamine in peptidoglycans, which are components of bacterial cell walls . In insects, lysozymes are divided into two categories according to the presence of catalytic amino acid residues essential for muramidase activity: the classical c-type lysozymes (c-LYZs) and the lysozyme-like proteins (LLPs) . In this study 32 unigenes encoding lysozymes were clustered into seven types of transcripts that coded for 5 c-LYZs and 2 LLPs (Table 1), expanding the number of lysozyme classes previously identified in S. exigua . Within the SeC-LYZ clusters, three types of transcripts differed in the 3’ UTR but encoded the same protein, SeC-LYZ1 (SeC-LYZ1A, 1B and 1C diverged by one amino acid substitution). The other two transcripts coded for two different proteins, SeC-LYZ2 and SeC-LYZ3. Transcripts encoding five different proteins (three c-LYZs and two LLPs) were also found in S. frugiperda , indicating that the lysozymes, as well as AMPs, were conserved between these two closely related species. Alignment of the predicted proteins showed that amino acid identities between S. exigua and S. frugiperda lysozymes ranged from 84 to 97% for c-LYZs and from 77 to 87% for LLPs. In the B. mori genome, one gene coding for a c-LYZ isoform and three genes coding for LLPs have been described .
S. exigua midgut-specific AMPs and lysozyme transcriptional responses to B. thuringiensis toxins
Gene expression changes in a large subset of AMPs and lysozymes in the S. exigua midgut were analyzed more in detail by qRT-PCR after intoxication of larvae with Bt toxins. First of all, we observed that the ingestion of E. coli WK6 proteins (present in Vip3Aa samples, and added to the Åkta-purified Cry1Ca samples) (S2 Fig) did not cause any moderate or strong changes in AMP or lysozyme gene expression (Fig 5). Although REST analysis showed that the level of four transcripts (SeDiapausinA1, SeDiapausinA2, SeC-LYZ1 and SeLLP1) significantly decreased after WK6 protein ingestion (less than 2.5-fold), and a fifth one (SeCecropinB) was slightly up-regulated (less than 2-fold), we considered that the overall gene regulation was not influenced by the presence of E. coli proteins in cell lysate. Moreover, WK6 proteins did not trigger any growth inhibition effects in bioassays (S4 Fig).
Gene expression of each transcript after the different treatments (WK6 lysate, Cry1Ca, and Vip3Aa) was compared to its control (used as calibrator). Fold-changes were determined by using the REST MCS software. Each bar represents the mean of three independent experiments (± SEM). Dotted lines indicate 1 and -1 fold-changes. A) Cysteine-rich peptides; B) glycine- and proline-rich peptides; C) amphipathic peptides; D) lysozymes.
After ingestion of both Cry1Ca and Vip3Aa toxins at a dose inhibiting 99% of growth, almost all the transcripts coding for AMPs and lysozymes were up-regulated, showing moderate (2–20 fold) to high transcriptional activation (up than 20 fold) (Fig 5). These results are in agreement with the studies comparing transcriptional response (by transcriptome profiling or microarrays) of lepidopteran or coleopteran insects to different Cry toxins. Most of these analyses showed that, among others, genes involved in immune-response were usually up-regulated after Bt challenge, particularly genes encoding AMPs [28,32,54–61].
Interestingly, AMP and lysozyme expression responded in a similar way after either Cry1Ca or Vip3Aa intoxication (Fig 6), with only few quantitative differences observed. It is worth noting that the transcriptional induction was not strictly gene family-dependent, e.g., some amphipathic peptides (SeMoricin) were largely up-regulated after intoxication with both toxins while others were not regulated or even down-regulated (SeCecropin E) (Fig 5). Only one transcript (SeCecropin E) showed a down-regulation, though slight (3-fold), after Cry1Ca ingestion (Fig 5C).
Correlation analysis between the average fold change (3n) after Cry1Ca and Vip3Aa treatments (Pearson r and p-value are shown).
When the midgut transcriptional response to Bt toxin challenge was analyzed in detail, we observed that the most responsive AMPs were the cysteine-rich peptides. All the tested members of this family were up-regulated after ingestion of both toxins, with no major differences between Cry1Ca and Vip3Aa (Fig 5A). A more exhaustive analysis of cysteine-rich AMP response revealed that classical CS-αβ AMPs (spodoptericin and cobatoxins) showed a moderated level of induction (from 4 to 15 fold) whereas the majority of the members of diapausin family showed high levels of induction (up than 40 fold) (Fig 5A). It has been shown that classical CS-αβ AMPs are involved in humoral response against several bacteria and fungi in a number of Lepidoptera species such as B. mori, S. cynthia, S. littoralis, S. frugiperda [41,62–65]. Among these studies, only one (performed with H. armigera larvae) assessed their role in response to B. thuringiensis infection, showing that transcripts coding for defensin and a cobatoxin (homolog to SeCobatoxin A) were up-regulated after intrahemocoelic injection of B. thuringiensis . Interestingly, the most responsive cysteine-rich peptide family was diapausin, also one of the most expanded AMP families in S. exigua (up to 8 transcripts were retrieved in the present study). Diapausins increased their transcription 121-fold (i.e. SeDiapausin A6 after Vip3Aa intoxication) (Fig 5A). Similar results have been reported using microarrays to investigate transcriptional changes in S. exigua after Vip3Aa intoxication . Two unigenes annotated as diapausin precursors (namely SE_U33476 and SE_U10027 that correspond to SeDiapausin A3 and A4, respectively) were found among the most up-regulated genes, and other two diapausins (SE_U14458 and SE_U22863, which correspond to SeDiapausin A2 and A1, respectively) were moderately up-regulated. Noticeably, we showed that the transcriptional response of diapausins to Cry1Ca intoxication was similar to the Vip3Aa response (Fig 6). To date, insect diapausins have received only limited study. The recombinantly-expressed SeDiapausin A1 showed antimicrobial activity against the Gram positive bacterium, Bacillus megaterium . However, in the leaf beetle G. atrocyanea, transcripts encoding diapausins did not accumulate after pathogen-challenge . Given the high number of transcripts coding for SeDiapausins that were expressed in S. exigua larval midgut and up-regulated after Bt toxin challenge and that some diapausins showed the strongest transcriptional activation, it is suggestive that these AMPs play a major role in S. exigua local response to prevent the establishment of infection and to maintain midgut immunity.
Glycine-rich and proline-rich AMPs showed inter-sample variability after Cry1Ca or Vip3Aa ingestion (Fig 5B). REST statistical analysis showed that SeGloverin and SeLebocin1 expression levels were significantly different from control after either Cry1Ca or Vip3Aa ingestion whereas SeAttacin1 and 3 did not exhibit such clear induction pattern. Interestingly, it has been shown that dsRNA-mediated silencing of SeGloverin enhanced toxicity of B. thuringiensis, suggesting a role of this AMP in controlling Bt infection in S. exigua . Likewise, expression of gloverin, as well as an attacin (homolog to SeAttacin1) and a lebocin (homolog to SeLebocin1) were more frequently detected in T. ni larvae that ingested Bt preparations although, surprisingly, such transcripts were not detected in 4th instar midguts . In H. armigera, attacin (homolog to SeAttacin2/3) and gloverin increased their expression in larval body after injections of B. thuringiensis . To date, glycine-rich and proline-rich peptides have been the best studied AMPs in response to Bt challenge and results indicate their involvement in response to Bt ingestion. However, their contribution to the midgut immune response (both to B. thuringiensis and to its virulence factors) needs to be clarified.
Amphipathic peptides (moricin and cecropins) showed the most divergent response to Cry1Ca and Vip3Aa (Fig 5C). Previously, up-regulation of moricin after Bt intrahemocoelic injections was observed in H. armigera larvae , although in S. litura, moricin was not detected in the midgut after bacterial challenge while it was induced in fat-body and hemocytes . In the present study moricin showed a transcriptional response to both Cry1Ca and Vip3Aa ingestion (from 16- to 27-fold up-regulation respectively) whereas the response pattern of the other amphipathic peptides, cecropins, was not uniform. SeCecropin E showed a slight down regulation only after Cry1Ca ingestion. SeCecropin C and D increased their expression after both Bt toxin treatments, but the responses showed a high inter-sample variability and were not statistically significant (Fig 5C). These results led us to speculate that cecropins could play a limited role in the local immune response after Bt toxin ingestion, although they are broad-range AMPs and thought to be active against both Gram positive and negative bacteria [4,49].
Finally, four out of five lysozymes were up-regulated following both Cry1Ca and Vip3Aa treatments (Fig 5D). SeC-LYZ1 and SeLLP2 were the highest induced transcripts after both treatments (around 4.5-fold) whereas SeC-LYZ2 and SeLLP1 were only slightly induced (from 2 to 2.5-fold). SeC-LYZ3 was the lysozyme detected at lowest levels in all the samples tested and it never responded to Bt toxin challenge (Fig 5D). In the S. exigua close relative S. frugiperda, expression of c-LYZ and LLP homologs was tested in different tissues after intrahemocoelic injection of non-pathogenic (E. coli and Micrococcus luteus) and pathogenic (Pseudomonas luminescens) bacteria. Results showed that expression after bacterial challenge differed among the tissues tested and the only transcripts up-regulated in midguts after microbial treatments were c-LYZ1 and c-LYZ3 . Likewise, in T. ni, c-LYZ expression (homolog to SeC-LYZ1) increased after ingestion of B. thuringiensis formulations  and in H. armigera, c-LYZ1 transcript level increased after injection of B. thuringiensis vegetative cells .
The transcriptional profile of S. exigua larval midgut after challenging with two different Bt toxins showed a wide and, in some cases, high up-regulation of transcripts coding for antimicrobial proteins, which was not substantially different between the two toxins tested (Fig 6). Both Cry1Ca and Vip3Aa are pore-forming toxins which are produced at different growth stages of B. thuringiensis as virulence factors to breach the larval midgut barrier to allow later spread into insect hemocoel . More specifically, Cry toxins are important for spore toxicity whereas Vips are important for vegetative cells toxicity. Even though both toxins exert their toxicity action through the same sequence of events (activation, binding to specific receptors in the apical membrane of midgut epithelial cells and pore formation), the mode of action of these two proteins differs since they bind to different proteins [16–18], and pores seem to be structurally and functionally distinct ; however we observed that both trigger a common midgut immune response, which seems independent from the respective modes of action.
The cascade of events that activates the local immune response starts with the recognition of invading bacteria (like B. thuringiensis) by pathogen recognition proteins (like PGRP and GNBP), which recognize microbial molecules such bacterial peptoglycans (PGNs) or lipopolysaccharides (LPSs) [3,4]. However, in our study, we found that local immune response was not triggered after ingestion of WK6 E. coli lysate indicating any PNGs and LPS possibly present in our lysates were not influencing the immune status of S. exigua midgut. When Vip3Aa or Cry1Ca were present, a high and wide range of transcriptional activation of immune-related genes coding for AMPs or lysozymes was observed. This may be explained by the significant tissue damage that both Bt toxins provoke in susceptible larval midguts as a consequence of pore-formation and cell lysis, and by the probable production of damage-associated endogenous pattern molecules (DAMPs), which are another class of immune response elicitors . Under this point of view, the local immune response observed would be triggered by the consequences of the action of the Bt proteins more than by their presence in the midgut. A recent study has shown that a large number of AMPs were activated in Tribolium castaneum larvae after cuticle pricking . Although cuticle is a different tissue than midgut, it may be possible that the phenomenon observed is somehow correlated to what we detected in S. exigua midgut after Bt toxin injuries. Also, we should consider the presence of enteric bacteria (normally benign) that may exert pathogenic effects in case of midgut damage, and that may contribute to the activation of local immune response after pore-formation induced by the Bt toxins. An earlier study showed that loss of M. sexta gut integrity due to the action of Bt toxins enabled a gut commensal bacteria, Enterococcus faecalis, to translocate from the gastrointestinal tract into the hemolymph where it induced a cellular immune response and larval death . Therefore, midgut production of antimicrobial proteins after wounding may be a defense mechanism used by larvae to prevent septicemia caused by a “commensal-to-pathogen” switch.
Phenoloxidase response to B. thuringiensis Cry1Ca and Vip3Aa oral intoxication
In insect hemolymph, melanization is a systemic immune response controlled by a cascade of serine proteases that leads to the final activation of the pro-phenoloxidase (proPO) to its active form phenoloxidase (PO), which is responsible for the formation of melanin around intruding microorganisms . The PO level is believed to be correlated with insects' immunocompetence, especially against invading fungi or insect parasitoids . In the present study, the humoral response of S. exigua larvae after ingestion of a sub-lethal dose of Cry1Ca or Vip3Aa (99% growth inhibition) were investigated by measuring hemolymph PO activity. Results showed that neither the two toxins nor the WK6 lysate (used as control) had any effect on hemolymph PO activity (Fig 7), even though both Bt toxins strongly induced local (midgut) up-regulation of transcripts encoding immune response effectors.
Larvae were fed on artificial diet supplemented with: phosphate buffer, WK6 lysate, Vip3Aa and Cry1Ca. PO activity (absorbance increase per min per μl of hemolymph) was measured in hemolymph samples. Results represent individual values (mean of two technical replicates) and the mean of each treatment (n = 18).
The absence of PO response (measured 24 h after Bt toxin ingestion) may be due to the lack of bacterial infection in the hemolymph at that time, since no Bt vegetative cells or spores able to colonize insect body were present in our preparation. Freitak et al.  showed that in T. ni, after ingestion of a non-pathogenic bacterial cocktail, expression of seventeen-immune related genes (among which ten AMP and lysozyme transcripts) were up-regulated in the midgut whereas PO activity in hemolymph was even lower than in control larvae. This may suggest a local response activated in Lepidoptera larval midgut when a higher bacterial load or wounding are detected, but does not imply an automatic activation of the systemic response. Such a response may be later triggered only when bacteria reaches the body cavity as a consequence of a successful infection establishment.
In Lepidoptera, few studies have recorded PO activation after Bt oral ingestion and they are somewhat contradictory. For example, in G. mellonella both LC15 and LC50 doses of Bt spores and crystals resulted in significantly elevated hemolymph PO levels compared to control (even if PO activity was lower after exposure to LC50 than to LC15 dose) . Other authors reported that ingestion of a mix of spores and crystals significantly increased hemolymph PO activity in Plodia interpunctella whereas in H. virescens and T. ni the increase was around 2-fold. Notably, no increase was observed for S. exigua but it was the only species that showed a 20% mortality at the used Bt dose, compared with 50% mortality of the other insects . In another study with T. ni, hemolymph PO activity in susceptible larvae was even reduced after ingestion of a sub-lethal doses of a mix of spores and crystals . Some authors speculated that the reduction in PO activity after ingestion of sub-lethal doses of Bt spores and crystals could be due to the contamination of the hemolymph by bacterial cells which can overwhelm the cellular immune-reactions by depleting the hemocytes (which are the main responsible for proPO production) .
S. exigua midgut-specific AMP and lysozyme transcriptional responses to baculovirus challenge
The transcriptional response of AMPs and lysozymes was also analyzed in S. exigua midgut after baculovirus infection. Results showed that only one out of 23 transcripts coding for effector molecules was up-regulated after viral infection, namely SeCecropin D, whose transcript was 8-fold more abundant than in control larvae (Fig 8). Expression of the remaining effectors was down-regulated or did not show a response to the infection. This response was in agreement with that described by Jakubowska et al.  working with guts of S. exigua baculovirus infected larvae, where a down- or a non-regulation of immune-related gene expression (including four genes encoding AMPs) was observed. In the present study, the highest levels of repression were observed for SeC-LYZ1 (37-fold reduction), followed by SeAttacin3 (18-fold), SeCobatoxin A (15-fold), and SeAttacin3 and SeCecropinE (around 10-fold). SeDiapausinA2, SeCecropinC and SeC-LYZ2 were the least down regulated transcripts (around 4-fold reduction). It is noteworthy that decreases in transcription after viral infection were not family-dependent, since in each antimicrobial protein family, only few members were specifically down-regulated. The overall picture that emerged from this comprehensive analysis of immune effector expression regulation after baculovirus infection was that only few peptides or enzymes with antimicrobial activity were transcribed in midguts of infected larvae (with the exception of SeCecropin D) in contrast to what happened after Bt toxin ingestion. Manipulation of the host local immune response has been proposed as a method that would improve baculovirus fitness by promoting the increase of enteric bacteria load in the midgut . Our results are in agreement with this hypothesis, since after baculovirus infection, a microbial-friendly environment is maintained in S. exigua midgut.
Gene expression of each transcript after baculovirus treatment was compared to its control (used as calibrator). Fold-changes were determined by using the REST MCS software. Each bar represents the mean of three independent experiments (± SEM). Dotted lines indicate 1 and -1 fold-change. * p-value < 0.1; ** p-value < 0.05.
This work has addressed the response of local (midgut) immune system effectors in S. exigua after ingestion of different virulence factors (Bt toxins such Cry1Ca and Vip3Aa) and a viral pathogen (SeMNPV) used in microbial control agricultural practices. Here, we have shown that larvae responded locally to the toxic effects associated to the Bt protein by widely inducing transcription of their AMP and lysozyme arsenal in the midgut most probably to prevent septicemia produced by commensal bacteria, without the contemporaneous activation of the systemic defense (represented by PO activity). On the other hand, baculovirus infection led to a down-regulation of some AMPs and lysozymes in the midgut and to no or little changes in the other ones, confirming the importance for the virus to establish its infection in a “microbial-friendly” gut environment to increase its fitness . Taken together, these results show the importance of the midgut immune response in shaping the first step of infections triggered by ingested pathogens, thus highlighting the need to take it into account in host-pathogen interaction studies.
S1 Fig. Growth inhibition dose-response curve of S. exigua newly molted L4 larvae challenged with Cry1Ca.
Growth inhibition values were calculated following Herrero et al. . Three biological replicates of the experiment (using 8 larvae per dose) were performed.
S2 Fig. SDS-PAGE profiles of Bt proteins and controls used for S. exigua L4 treatments.
Lane 1) crude WK6 lysate containing 1 μg Vip3Aa; 2) crude empty WK6 lysate; 3) 1 μg of Åkta purified Cry1Ca; 4) 1 μg of Åkta purified Cry1Ca mixed with crude empty WK6 lysate; 5) Precision Plus Protein Unstained Standards (Bio-Rad, hercules, CA, US).
S3 Fig. ClustalX alignment of Lepidoptera x-tox proteins.
Black box indicates the signal peptide sequences; dotted black boxes indicate CS-αβ motifs. In each motif, conserved six cysteines are highlighted in green. Accession number of each sequence is indicated. Bm: B. mori, Ha: H. armigera, Hv: H. virescens, Se: S. exigua, Sf: S. frugiperda.
S4 Fig. Comparison of growth rate values from larvae that ingested phosphate buffer or WK6 lysate proteins.
Each dot represents an individual observation. Lines represent mean ± SD. Mann-Whitney U and p-value are shown.
S1 Table. Blastn results using S. exigua antimicrobial proteins as query and S. exigua Sanger library as database. S. exigua Sanger sequences acc. nr. from JK786560 to JK789862.
S2 Table. Primers for sequence confirmation, qRT-PCR analyses and/or transcript detection.
We thank Rosa Maria González-Martínez for her excellent help with insect rearing and laboratory management.
Conceived and designed the experiments: CMC YB SH BE. Performed the experiments: CMC AKJ YB. Analyzed the data: CMC. Wrote the paper: CMC YB.
- 1. Kingsolver MB, Huang Z, Hardy RW (2013) Insect antiviral innate immunity: pathways, effectors, and connections. J Mol Biol 425: 4921–4936. pmid:24120681
- 2. Yi HY, Chowdhury M, Huang YD, Yu XQ (2014) Insect antimicrobial peptides and their applications. Appl Microbiol Biotechnol 98: 5807–5822. pmid:24811407
- 3. Lemaitre B, Hoffmann J (2007) The host defense of Drosophila melanogaster. Annu Rev Immunol 25: 697–743. pmid:17201680
- 4. Davis MM, Engstrom Y (2012) Immune response in the barrier epithelia: lessons from the fruit fly Drosophila melanogaster. J Innate Immun 4: 273–283. pmid:22237424
- 5. Steiner H, Hultmark D, Engstrom A, Bennich H, Boman HG (1981) Sequence and specificity of two antibacterial proteins involved in insect immunity. Nature 292: 246–248. pmid:7019715
- 6. Ferrandon D, Jung AC, Criqui M, Lemaitre B, Uttenweiler-Joseph S, Michaut L, et al. (1998) A drosomycin-GFP reporter transgene reveals a local immune response in Drosophila that is not dependent on the Toll pathway. EMBO J 17: 1217–1227. pmid:9482719
- 7. Liehl P, Blight M, Vodovar N, Boccard F, Lemaitre B (2006) Prevalence of local immune response against oral infection in a Drosophila/Pseudomonas infection model. PLoS Pathog 2: e56. pmid:16789834
- 8. Wu S, Zhang X, He Y, Shuai J, Chen X, Ling E (2010) Expression of antimicrobial peptide genes in Bombyx mori gut modulated by oral bacterial infection and development. Dev Comp Immunol 34: 1191–1198. pmid:20600274
- 9. Easton DM, Nijnik A, Mayer ML, Hancock RE (2009) Potential of immunomodulatory host defense peptides as novel anti-infectives. Trends Biotechnol 27: 9.
- 10. Eleftherianos I, Atri J, Accetta J, Castillo JC (2013) Endosymbiotic bacteria in insects: guardians of the immune system? Front Physiol 4: 46. pmid:23508299
- 11. Tanaka H, Ishibashi J, Fujita K, Nakajima Y, Sagisaka A, Tomimoto K, et al. (2008) A genome-wide analysis of genes and gene families involved in innate immunity of Bombyx mori. Insect Biochem Mol Biol 38: 1087–1110. pmid:18835443
- 12. Vilcinskas A (2013) Evolutionary plasticity of insect immunity. J Insect Physiol 59: 123–129. pmid:22985862
- 13. IRAC (2014). http://www.irac-online.org/pests/spodoptera-exigua/
- 14. Schnepf E, Crickmore N, Van Rie J, Lereclus D, Baum J, Feitelson J, et al. (1998) Bacillus thuringiensis and its pesticidal crystal proteins. Microbiol Mol Biol Rev 62: 775–806. pmid:9729609
- 15. Estruch JJ, Warren GW, Mullins MA, Nye GJ, Craig JA, Koziel MG (1996) Vip3A, a novel Bacillus thuringiensis vegetative insecticidal protein with a wide spectrum of activities against lepidopteran insects. Proc Natl Acad Sci U S A 93: 5389–5394. pmid:8643585
- 16. Vachon V, Laprade R, Schwartz JL (2012) Current models of the mode of action of Bacillus thuringiensis insecticidal crystal proteins: a critical review. J Invertebr Pathol 111: 1–12. pmid:22617276
- 17. Lee MK, Walters FS, Hart H, Palekar N, Chen JS (2003) The mode of action of the Bacillus thuringiensis vegetative insecticidal protein Vip3A differs from that of Cry1Ab delta-endotoxin. Appl Environ Microbiol. 69: 4648–4657. pmid:12902253
- 18. Singh G, Sachdev B, Sharma N, Seth R, Bhatnagar RK (2010) Interaction of Bacillus thuringiensis vegetative insecticidal protein with ribosomal S2 protein triggers larvicidal activity in Spodoptera frugiperda. Appl Environ Microbiol 76: 7202–7209. pmid:20833785
- 19. van Oers MM, Vlak JM (2007) Baculovirus genomics. Curr Drug Targets 8: 1051–1068. pmid:17979665
- 20. Passarelli AL (2011) Barriers to success: How baculoviruses establish efficient systemic infections. Virology 411: 10.
- 21. Pascual L, Jakubowska AK, Blanca JM, Cañizares J, Ferre J, Gloeckner G, et al. (2012) The transcriptome of Spodoptera exigua larvae exposed to different types of microbes. Insect Biochem Mol Biol 42: 557–570. pmid:22564783
- 22. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H, et al. (2007) Clustal W and Clustal X version 2.0. Bioinformatics 23: 2947–2948. pmid:17846036
- 23. Nicholas KB, Nicholas HB Jr, Deerfield DW II (1997) GeneDoc: analysis and visualization of genetic variation. EMBNEW. NEWS 4: 14.
- 24. Lazzaro BP, Little TJ (2009) Immunity in a variable world. Philos Trans R Soc Lond B Biol Sci 364: 15–26. pmid:18926975
- 25. Edgar RC (2004) MUSCLE: a multiple sequence alignment method with reduced time and space complexity. BMC Bioinformatics 5: 113. pmid:15318951
- 26. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S (2011) Molecular Evolutionary Genetics Analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28: 9.
- 27. Hernández-Martínez P, Ferré J, Escriche B (2008) Susceptibility of Spodoptera exigua to 9 toxins from Bacillus thuringiensis. J Invertebr Pathol 97: 245–250. pmid:18082763
- 28. Bel Y, Jakubowska AK, Costa J, Herrero S, Escriche B (2013) Comprehensive analysis of gene expression profiles of the beet armyworm Spodoptera exigua larvae challenged with Bacillus thuringiensis Vip3Aa toxin. PLoS One 8: e81927. pmid:24312604
- 29. Herrero S, González-Cabrera J, Ferré J, Bakker PL, de Maagd RA (2004) Mutations in the Bacillus thuringiensis Cry1Ca toxin demonstrate the role of domains II and III in specificity towards Spodoptera exigua larvae. Biochem J 384: 507–513. pmid:15320864
- 30. Herrero S, Borja M, Ferré J (2002) Extent of variation of the Bacillus thuringiensis toxin reservoir: the case of the geranium bronze, Cacyreus marshalli Butler (Lepidoptera: Lycaenidae). Appl Environ Microbiol 68: 4090–4094. pmid:12147511
- 31. Jakubowska AK, Vogel H, Herrero S (2013) Increase in gut microbiota after immune suppression in baculovirus-infected larvae. PLoS Pathog 9: e1003379. pmid:23717206
- 32. Hernández-Martínez P, Navarro-Cerrillo G, Caccia S, de Maagd RA, Moar WJ, Ferré J, et al. (2010) Constitutive activation of the midgut response to Bacillus thuringiensis in Bt-resistant Spodoptera exigua. PLoS ONE 5: e12795. pmid:20862260
- 33. Pfaffl MW, Horgan GW, Dempfle L (2002) Relative expression software tool (REST) for group-wise comparison and statistical analysis of relative expression results in real-time PCR. Nucleic Acids Res 30: e36. pmid:11972351
- 34. Hernández-Martínez P, Naseri B, Navarro-Cerrillo G, Escriche B, Ferré J, Herrero S (2010) Increase in midgut microbiota load induces an apparent immune priming and increases tolerance to Bacillus thuringiensis. Environ Microbiol 12: 2730–2737. pmid:20482744
- 35. Mason CJ, Raffa KF (2014) Acquisition and structuring of midgut bacterial communities in gypsy moth (Lepidoptera: Erebidae) larvae. Environ Entomol 43: 595–604. pmid:24780292
- 36. Barthel A, Kopka I, Vogel H, Zipfel P, Heckel DG, Groot AT (2014) Immune defence strategies of generalist and specialist insect herbivores. Proc Biol Sci 281: 20140897. pmid:24943370
- 37. Hultmark D, Engstrom A, Andersson K, Steiner H, Bennich H, Boman HG (1983) Insect immunity. Attacins, a family of antibacterial proteins from Hyalophora cecropia. EMBO J 2: 571–576. pmid:6628360
- 38. Bang K, Park S, Yoo JY, Cho S (2012) Characterization and expression of attacin, an antibacterial protein-encoding gene, from the beet armyworm, Spodoptera exigua (Hübner) (Insecta: Lepidoptera: Noctuidae). Mol Biol Rep 39: 5151–5159. pmid:22160467
- 39. Hwang J, Kim Y (2011) RNA interference of an antimicrobial peptide, gloverin, of the beet armyworm, Spodoptera exigua, enhances susceptibility to Bacillus thuringiensis. J Invertebr Pathol 108: 194–200. pmid:21925182
- 40. Dimarcq JL, Bulet P, Hetru C, Hoffmann J (1998) Cysteine-rich antimicrobial peptides in invertebrates. Biopolymers 47: 465–477. pmid:10333738
- 41. Volkoff AN, Rocher J, d'Alencon E, Bouton M, Landais I, Quesada-Moraga E, et al. (2003) Characterization and transcriptional profiles of three Spodoptera frugiperda genes encoding cysteine-rich peptides. A new class of defensin-like genes from lepidopteran insects? Gene 319: 43–53. pmid:14597170
- 42. Girard PA, Boublik Y, Wheat CW, Volkoff AN, Cousserans F, Brehélin M, et al. (2008) X-tox: an atypical defensin derived family of immune-related proteins specific to Lepidoptera. Dev Comp Immunol 32: 575–584. pmid:17988734
- 43. d'Alencon E, Bierne N, Girard PA, Magdelenat G, Gimenez S, Seninet I, et al. (2013) Evolutionary history of x-tox genes in three lepidopteran species: origin, evolution of primary and secondary structure and alternative splicing, generating a repertoire of immune-related proteins. Insect Biochem Mol Biol 43: 54–64. pmid:23142192
- 44. Kouno T, Mizuguchi M, Tanaka H, Yang P, Mori Y, Shinoda H, et al. (2007) The structure of a novel insect peptide explains its Ca2+ channel blocking and antifungal activities. Biochemistry 46: 13733–13741. pmid:17994764
- 45. Tanaka H, Sato K, Saito Y, Yamashita T, Agoh M, Okunishi J, et al (2003) Insect diapause-specific peptide from the leaf beetle has consensus with a putative iridovirus peptide. Peptides 24: 7.
- 46. Wan H, Yuan MA, You H, Li JH, Jin BR (2012) Molecular cloning and characterization of a diapause-specific peptide in the beet armyworm, Spodoptera exigua. J Asia-Pac Entomol 15: 369–373.
- 47. Scocchi M, Tossi A, Gennaro R (2011) Proline-rich antimicrobial peptides: converging to a non-lytic mechanism of action. Cell Mol Life Sci 68: 2317–2330. pmid:21594684
- 48. Rayaprolu S, Wang Y, Kanost MR, Hartson S, Jiang H (2010) Functional analysis of four processing products from multiple precursors encoded by a lebocin-related gene from Manduca sexta. Dev Comp Immunol 34: 638–647. pmid:20096726
- 49. Duvic B, Jouan V, Essa N, Girard PA, Pages S, Abi Khattar Z, et al. (2012) Cecropins as a marker of Spodoptera frugiperda immunosuppression during entomopathogenic bacterial challenge. J Insect Physiol 58: 881–888. pmid:22487443
- 50. Furukawa S, Tanaka H, Nakazawa H, Ishibashi J, Shono T, Yamakawa M (1999) Inducible gene expression of moricin, a unique antibacterial peptide from the silkworm (Bombyx mori). Biochem J 340: 265–271. pmid:10229682
- 51. Jolles P, Jolles J (1984) What's new in lysozyme research? Always a model system, today as yesterday. Mol Cell Biochem 63: 165–189. pmid:6387440
- 52. Gandhe AS, Janardhan G, Nagaraju J (2007) Immune upregulation of novel antibacterial proteins from silkmoths (Lepidoptera) that resemble lysozymes but lack muramidase activity. Insect Biochem Mol Biol 37: 655–666. pmid:17550822
- 53. Chapelle M, Girard PA, Cousserans F, Volkoff NA, Duvic B (2009) Lysozymes and lysozyme-like proteins from the fall armyworm, Spodoptera frugiperda. Mol Immunol 47: 261–269. pmid:19828200
- 54. Behrens S, Peuss R, Milutinovic B, Eggert H, Esser D, Rosenstiel P, et al. (2014) Infection routes matter in population-specific responses of the red flour beetle to the entomopathogen Bacillus thuringiensis. BMC Genomics 15: 445. pmid:24908078
- 55. Meunier L, Prefontaine G, Van Munster M, Brousseau R, Masson L (2006) Transcriptional response of Choristoneura fumiferana to sublethal exposure of Cry1Ab protoxin from Bacillus thuringiensis. Insect Mol Biol 15: 475–483. pmid:16907834
- 56. van Munster M, Prefontaine G, Meunier L, Elias M, Mazza A, Brosseau R, et al. (2007) Altered gene expression in Choristoneura fumiferana and Manduca sexta in response to sublethal intoxication by Bacillus thuringiensis Cry1Ab toxin. Insect Mol Biol 16: 25–35. pmid:17257206
- 57. Lei YY, Zhu X, Xie W, Wu QJ, Wang SL, Guo Z, et al. (2014) Midgut transcriptome response to a Cry toxin in the diamondback moth, Plutella xylostella (Lepidoptera: Plutellidae). Gene 533: 180–187. pmid:24120626
- 58. Yao JX, Buschman LL, Lu NY, Khajuria C, Zhu KY (2014) Changes in Gene Expression in the Larval Gut of Ostrinia nubilalis in Response to Bacillus thuringiensis Cry1Ab Protoxin Ingestion. Toxins 6: 1274–1294. pmid:24704690
- 59. Oppert B, Elpidina EN, Toutges M, Mazumdar-Leighton S (2010) Microarray analysis reveals strategies of Tribolium castaneum larvae to compensate for cysteine and serine protease inhibitors. Comp Biochem Physiol part D Genomics Proteomics 5: 280–287. pmid:20855237
- 60. Sparks ME, Blackburn MB, Kuhar D, Gundersen-Rindal DE (2013) Transcriptome of the Lymantria dispar (Gypsy Moth) larval midgut in response to infection by Bacillus thuringiensis. PLoS ONE 8: e61190. pmid:23658687
- 61. Rodríguez-Cabrera L, Trujillo-Bacallao D, Borras-Hidalgo O, Wright DJ, Ayra-Pardo C (2008) Molecular characterization of Spodoptera frugiperda Bacillus thuringiensis Cry1Ca toxin interaction. Toxicon 51: 681–692. pmid:18222513
- 62. Kaneko Y, Tanaka H, Ishibashi J, Iwasaki T, Yamakawa M (2008) Gene expression of a novel defensin antimicrobial peptide in the silkworm, Bombyx mori. Biosci Biotechnol Biochem 72: 2353–2361. pmid:18776686
- 63. Seufi AM, Hafez EE, Galal FH (2011) Identification, phylogenetic analysis and expression profile of an anionic insect defensin gene, with antibacterial activity, from bacterial-challenged cotton leafworm, Spodoptera littoralis. BMC Mol Biol 12: 47. pmid:22067477
- 64. Wen HX, Lan XQ, Cheng TC, He NJ, Shiomi K, Kajiura Z, et al. (2009) Sequence structure and expression pattern of a novel anionic defensin-like gene from silkworm (Bombyx mori). Mol Biol Rep 36: 711–716. pmid:18360778
- 65. Hashimoto K, Yamano Y, Morishima I (2008) Cloning and expression of a gene encoding gallerimycin, a cysteine-rich antifungal peptide, from eri-silkworm, Samia cynthia ricini. Comp Biochem Physiol B Biochem Mol Biol 150: 229–232. pmid:18436462
- 66. Wang Q, Liu Y, He HJ, Zhao XF, Wang JX (2010) Immune responses of Helicoverpa armigera to different kinds of pathogens. BMC Immunol 11: 9. pmid:20196874
- 67. Tanaka H, Suzuki K (2005) Expression profiling of a diapause-specific peptide (DSP) of the leaf beetle Gastrophysa atrocyanea and silencing of DSP by double-strand RNA. J Insect Physiol 51: 701–707. pmid:15936770
- 68. Támez-Guerra P, Valádez-Lira JA, Alcocer-González JM, Oppert B, Gómez-Flores R, Támez-Guerra R, et al. (2008) Detection of genes encoding antimicrobial peptides in Mexican strains of Trichoplusia ni (Hubner) exposed to Bacillus thuringiensis. J Invertebr Pathol 98: 218–227. pmid:18359041
- 69. Oizumi Y, Hemmi H, Minami M, Asaoka A, Yamakawa M (2005) Isolation, gene expression and solution structure of a novel moricin analogue, antibacterial peptide from a lepidopteran insect, Spodoptera litura. Biochim Biophys Acta 1752: 83–92. pmid:16115804
- 70. Nielsen-LeRoux C, Gaudriault S, Ramarao N, Lereclus D, Givaudan A (2012) How the insect pathogen bacteria Bacillus thuringiensis and Xenorhabdus/Photorhabdus occupy their hosts. Curr Opin Microbiol 15: 220–231. pmid:22633889
- 71. Moreno-García M, Conde R, Bello-Bedoy R, Lanz-Mendoza H (2014) The damage threshold hypothesis and the immune strategies of insects. Infect Genet Evol 24: 25–33. pmid:24614506
- 72. Mason KL, Stepien TA, Blum JE, Holt JF, Labbe NH, Rush JS, et al. (2011) From commensal to pathogen: translocation of Enterococcus faecalis from the midgut to the hemocoel of Manduca sexta. Mbio 2: e00065–11 pmid:21586646
- 73. Eleftherianos I, Revenis C (2011) Role and importance of phenoloxidase in insect hemostasis. J Innate Immun 3: 28–33. pmid:21051882
- 74. Freitak D, Wheat CW, Heckel DG, Vogel H (2007) Immune system responses and fitness costs associated with consumption of bacteria in larvae of Trichoplusia ni. BMC Biol 5: 56. pmid:18154650
- 75. Grizanova EV, Dubovskiy IM, Whitten MM, Glupov VV (2014) Contributions of cellular and humoral immunity of Galleria mellonella larvae in defence against oral infection by Bacillus thuringiensis. J Invertebr Pathol 119: 40–46. pmid:24735783
- 76. Valádez-Lira JA, Alcocer-González JM, Damas G, Núñez-Mejía G, Oppert B, Rodríguez-Padilla C, et al. (2012) Comparative evaluation of phenoloxidase activity in different larval stages of four lepidopteran pests after exposure to Bacillus thuringiensis. J Insect Sci 12: 80. pmid:23414117
- 77. Ericsson JD, Janmaat AF, Lowenberger C, Myers JH (2009) Is decreased generalized immunity a cost of Bt resistance in cabbage loopers Trichoplusia ni? J Invertebr Pathol 100: 61–67. pmid:19026655