Immune responses of Helicoverpa armigera to different kinds of pathogens

Background Insects react against pathogens through innate immunity. The cotton bollworm Helicoverpa armigera (H. armigera) is an important defoliator and an extremely destructive pest insect of many crops. The elucidation of the mechanism of the immune response of H. armigera to various pathogens can provide a theoretical basis for new approaches to biologically control this pest. Results Four kinds of pathogens Bacillus thuringiensis, Klebsiella pneumoniae, Candida albicans, and Autographa californica multiple nucleocapsid nucleopolyhedrovirus harbored green fluorescence protein and polyhedron (AcMNPV-GFP) were used to challenge the insect. The cellular and humoral immune responses to the pathogens were analyzed in the challenged H. armigera. The results show that in the five kinds of haemocytes, only granulocytes phagocytized the Gram-negative and Gram-positive bacteria and fungi. All haemocytes can be infected by AcMNPV. Fourteen immune-related genes including pattern recognition receptors (PRRs) such as peptidoglycan recognition proteins (HaPGRP and HaPGRP C) and Gram-Negative Bacteria-Binding Protein (HaGNBP), and antimicrobial peptides (AMPs) such as cecropin-1, 2 and 3 (HaCec-1, 2 and 3), lysozyme (HaLys), attacin (HaAtt), gallerimycin-like (HaGall), gloverin-like (HaGlo), moricin-like (HaMor), cobatoxin-like (HaCob), galiomicin-like (HaGali), and immune inducible protein (HaIip) appeared in different expression profiles to different pathogen infections. The transcripts of 13 immune related genes (except HaPGRPC) are obviously up-regulated by Gram-positive bacteria. HaCec-1 and 3, HaMor, HaAtt, HaLys, HaIip, HaPGRP and HaGNBP are greatly up-regulated after fungal infection. HaGNBP, HaCec-2, HaGall, HaGlo, HaMor, HaCob, HaGali obviously increased in Gram-negative bacterial infection. Only five genes, HaGNBP, HaCec-1, HaGali, HaGlo, and HaLys, are weakly up-regulated after viral infection. The AMP transcripts had higher expression levels than the PRR transcripts after the microbial challenge. Conclusions These data suggest that the granulocytes are the major phagocytes in H. armigera. All haemocytes can be infected by AcMNPV. The transcripts of 14 immune related genes have different expression patterns in H. armigera infected by different pathogens, which means that the immune-related genes may have different functions against various kinds of pathogens.


Background
The immune system is generally divided into innate and adaptive immunity. Innate immunity is characterized by quick reactions that cause immediate immune responses. In contrast, adaptive immunity is a slow reaction with high specificity and memory. Adaptive immunity has remarkable specificity based on somatic gene rearrangement and hypermutation, leading to an extremely large repertoire of T-and B-cell receptors and antibodies. Such adaptive immunity is restricted to jawed vertebrates. Invertebrates only rely on their innate immune defenses. The innate immune system of insects relies on both humoral and cellular responses [1]. Haemocytes are the primary mediators of cell-mediated immunity in insects including phagocytosis, nodulation, encapsulation, and melanization. The humoral response of innate immunity includes three steps: 1) identification of pathogen-associated molecular patterns (PAMPs) on pathogens by pattern recognition receptors (PRRs) [2]; 2) activation of the regulatory pathways; and 3) production of immune effectors including cellular phagocytosis and molecular effectors such as antimicrobial peptides (AMPs) [3].
Effector molecules such as AMPs are found in most of plants and animals, and their functions in innate immunity are well investigated [9,10]. Approximately 1,000 varieties of AMPs have been reported from vertebrates, invertebrates, human beings, and plants since the Boman group reported the first cecropin from Hyalophora cecropia [11]. In most insects, AMPs are synthesized in the fat body and haemocytes and are then released to the haemolymph.
Insect innate immunity has been studied in D. melanogaster, Anopheles gambia, and Maduca sexta [12,13], but fewer studies have been conducted on the cotton bollworm, Helicoverpa armigera. H. armigera is a worldwide pest that has developed strong resistance to chemical and biological pesticides. New approaches to control this pest can be designed if the mechanism of its immune responses to pathogens is understood. We have reported the function of a pattern recognition receptor (C-type lectin) in the insect [14,15]. To further investigate the mechanism of H. armigera's response to pathogen infections, its cellular responses to four kinds of pathogens, namely, Gram-positive bacterial, Gram-negative bacterial, fungal, and viral pathogens which are Bacillus thuringiensis (G + -bacterium), Klebsiella pneumoniae (G --bacterium), Candida albicans (fungus), and Autographa californica multiple nucleocapsid nucleopolyhedrovirus harbored green fluorescence protein and polyhedron (AcMNPV-GFP) (virus), respectively, were investigated. Simultaneously, quantitative real-time PCR (qRT-PCR) analysis was performed at different times after the challenge by these pathogens. Fourteen immunity-related genes including PRRs and antimicrobial peptide genes were chosen for the analysis. The results demonstrated that the innate immune responses of H. armigera larvae to the pathogens included cellular phagocytosis and upregulated expression of immune effector genes.

Cellular responses to pathogens in vivo
Five kinds of cells were distinguished in H. armigera, namely, prohaemocytes (Pr), plasmatocytes (Pl), granulocytes (Gr), oenocytoids (Oe), and spherulocytes (Sp) (Fig. 1A). After the injection of acridine orange labeled bacteria and fungi, it was observed that only the granulocytes could phagocytize the pathogens including G + organism B. thuringiensis, the Gorganism K. pneumoniae, and the fungus C. albicans. Other haemocytes did not phagocytize (Fig. 1B-d). This result indicated that the granulocytes were the major phagocytes in H. armigera.
When the budded virus, AcMNPV-GFP was injected into the larval haemocoel, haemocytes released green fluorescence at 36 h after injection. The morphology of the haemocytes was changed and could not be identified as a recognizable cell type. The green fluorescence increased in intensity proportional to infection time till 6 d post infection. The haemocytes appeared aggregated and destroyed. However, the larvae did not die, but were arrested in the metamorphically committed stage till 6th instar 6 d. These results suggested that the haemocytes could be infected by the AcMNPV-GFP virus and the virus could replicate inside the insect body (  The expression patterns of the three pattern recognition protein genes had some differences responding to the pathogens. The expression of HaPGRP was increased 3 h post-injection of the G + bacteria and fungi; there was no significant increase in expression after Gbacteria injection and virus injection compared with the PBS-injected insects (Fig. 4A). In contrast, HaPGRP C showed a trend of decreased expression after the injection of G + bacteria, Gbacteria, fungi, or virus (Fig. 4B). The HaGNBP was up-regulated to the highest expression sequentially by four pathogens 3 h post-Gbacterial injection, 12 h post-G + bacterial, fungal and -viral injection. The expression level of HaGNBP is higher than HaPGRP (Fig. 4C). These results indicated that the expressions of PRRs were regulated after the infections of the pathogens.
Three cecropins (HaCec-1, 2 and 3) had different expression patterns responding to the pathogens. HaCec-1 and 3 were up-regulated by four or three pathogens, respectively. However, higher expression levels of HaCec-1 and 3 were found in the fungal injected larvae (Fig. 5A, C). In contrast, HaCec-2 was only up-regulated by two kinds of bacteria with lower expression levels, 3 h post-injection by G + bacteria and 6 h post-injection by Gbacteria. Almost no variation was observed for fungal and viral challenge compared with the mock (PBS) challenge (Fig. 5B). The HaCec-1 and 3 had high similarities in amino acid sequence comparison ( Fig. 3A), therefore they had similar expression pattern after pathogen infection. These results also suggested that HaCec-1 and 3 responded to fungus well and that HaCec-2 responded to bacteria well.
The HaGali, HaGlo, and HaLys were up-regulated by four microbial challenges. However, they exhibited different expression patterns. The expression of HaGali was up-regulated by four microbial challenge 3 h postinjection, with the highest expression after fungal challenge at 12 h (Fig. 5D). The mRNA of HaGlo was also up-regulated by four microbes 3 h post-injection but with the highest expression after viral challenge at 3 h (Fig. 5E). The expression of HaLys was up-regulated by four pathogens with different patterns. It was up-regulated 3 h post-injection of the G + bacteria and gradually recovered to normal level from 6 h to 24 h, and it was up-regulated 3 h post-injection of the fungus and  and by Gbacteria 24 h post-injection (Fig. 6A). The expression of HaCob was up-regulated to the highest levels 3 h post-G + bacterial injection, 6 h post-fungal injection, and 24 h post-Gbacterial injection (Fig. 6B). HaAtt was increased 3 h to 6 h post-injection of the G + bacteria and gradually decreased to normal level from 12 h to 24 h post-injection. In contrast, HaAtt appeared to have an increase trend from 3 h to 12 h post-injection of fungus. There was a low-level increase of the HaAtt at 24 h post-Gbacterial injection. The expression levels of HaAtt to fungus and G + bacteria were much higher than to Gbacteria (Fig. 6C). These results suggested that the genes had different initiation response times to different pathogens and had different expression patterns after infection of various pathogens.
The HaGall and HaIip only responded to two pathogens. HaGall was greatly up-regulated by Gbacteria 6 h after injection and was weakly up-regulated by fungus at 12 h post-injection (Fig. 6D). HaIip was greatly upregulated by fungus 6 h after injection and was weakly up-regulated by G + bacteria at 3 h to 12 h post-injection (Fig. 6E). These two genes could play roles against Gbacteria infection and fungus infection, respectively.

Discussion
This work first investigated the immune responses at the cellular and humoral levels of H. armigera to four pathogens, B. thuringiensis, K. pneumoniae, C. albicans, and AcMNPV-GFP, which represent G + bacterial, Gbacterial, fungal, and viral pathogens, respectively. The  results revealed that the innate immunity of H. armigera larvae responded both on cellular and humoral levels against these pathogens.
Regarding the functions of haemocytes relating to innate immunity, it is known that plasmatocytes form the bulk of capsules around foreign bodies. Granular haemocytes phagocytize foreign pathogens, and oenocytoids produce phenoloxidases (PO), which are responsible for haemolymph melanization [16]. Plasmatocytes are involved in the phagocytosis of non-self microsphere beads, whereas granulocytes are involved in the phagocytosis of self-dead cells. Other studies showed that phagocytes of Manduca sexta had similar functions as vertebrate phagocytes [17]. This study demonstrated that granulocytes were the major cells that phagocytized G + and Gbacteria, and fungi. No phagocytosis function was observed for other kinds of haemocytes in H. armigera. This suggested that haemocytes played different roles in different insect species.
In addition to cellular immunity, humoral immunity also plays significant role in insect immunity, which involves the induction of a robust AMPs that kill invading pathogens [9]. Eleven AMP genes were found from H. armigera in this study: HaCec-1, 2 and 3, HaGall, HaGlo, HaLys, HaMor, HaCob, HaAtt, HaGali, and HaIip. They belong to 9 kinds of AMPs.
Some functions of these AMPs are known in other species. For example, the cecropins are linear, cationic AMPs that cause lysis of both Gram-positive and Gramnegative bacteria [18]. HaCec-1 appeared to have very high expression levels in fungus C. albicans-injected larvae and AcMNPV-injected larvae in addition to G + and Gbacteria, suggesting that HaCec-1 not only responded to G + and Gbacteria but also to fungus and virus, and might play very important roles in H. armigera immunity.
Furthermore, Galiomicin, a defensin first found in Galleria mellonella, has antibacterial and antifungal activities [19]. The results of this study were consistent to the report except for the detection of some up-regulation of HaGali after virus infection. Gloverins, a family of glycine-rich polypeptides, are effective against Gbacteria and are inactive against G + bacteria, yeasts, mammalian cell lines, and pestivirus [20]. However, upregulation of this gene was detected after four kinds of pathogens were injected into H. armigera including G + and Gbacteria, fungus, and virus. The highest expression of HaGlo after AcMNPV virus injection suggests that this gene may play a role in antivirus defense. Lysozymes are a type of hydrolase which can digest the peptidoglycan layer of the bacteria cell wall by hydrolyzing the β-1,4-glycosidic linkage between N-acetylglucosamine and N-acetylmuramic acid. Lysozymes also hydrolyze the β-1, 4-linkages of chitooligosaccharides in the fungal cell wall. Therefore, lysozymes have antibacterial and antifungal activities [21]. In addition to the antibacterial and antifungal activities, the results of this study indicate that HaLys also highly respond to virus, which implies its role in antivirus defense.
Moricin is a highly basic AMP first reported in the silkworm Bombyx mori. It shows antibacterial activity against several Gand G + bacteria [22]. The results in H. armigera showed that HaMor also had a higher expression after C. albicans (fungus) induction, suggesting that HaMor played a role in antifungus response. Cobatoxin is a toxin present in the venom of the Centruroides noxius scorpion that blocks two K + -channel subtypes: voltage-gated and Ca 2+ -activated channels [23]. The results of this study show HaCob responded to Gand G + bacteria and fungus infections in H. amigera. Attacin is an AMP, originally isolated from the immune haemolymph of Hyalophora cecropia. The antibacterial effect of attacin was found to be limited to some species of Gram-negative bacteria [24]. In H. armigera, HaAtt responded to G + bacteria and fungus better than Gbacteria, suggesting that HaAtt participated in the anti-G + bacteria and fungus response in addition to Gbacteria.
Gallerimycin is an antifungal peptide from the greater wax moth G. mellonella. Its expression is up-regulated after stimulation with bacterial lipopolyscaccharides (Gbacteria). Gallerimycin is active against the entomopathogenic fungus Metarhizium anisopliae but not against yeast and Gram-negative or Gram-positive bacteria in G. mellonella [25]. Helicoverpa gallerimycin-like AMP had a very high expression after the Gbacteria (K. pneumoniae) challenge, suggesting that it played a role in Gbacteria defense. HaIip is an antimicrobial peptide with Knot1 domain representing plant lectins http://smart.embl-heidelberg.de. The up-regulation of HaIip in H. armigera after the induction of fungus (C. albicans) suggests that it played an important role in antifungal defending.
Innate immunity was formerly thought to be a nonspecific immune response characterized by phagocytosis. However, innate immunity has considerable specificity and is capable of discriminating between self and nonself, as proposed in the concept of PRRs of the host. These PRRs can recognize conserved PAMPs shared by large groups of microorganisms, therefore successfully defending invertebrates and vertebrates against infection [2]. PGRP was first purified from silkworm (Bombyx mori) hemolymph based on the high affinity of PGRP to bacterial peptidoglycans [26]. GNBPs have been originally described in the silkworm B. mori, and one of them is known to bind strongly to the surface of Gram-negative bacteria [27]. The expression of HaPGRP was up-regulated by G + bacteria and fungi; the HaGNBP was upregulated by four kinds of pathogens. These results indicated that the PRRs in H. armigera might have different functions in the pattern recognition.
The results of expression patterns revealed that most examined genes were significantly up-regulated after infection by pathogens. However, the responses of the genes to different kinds of pathogens showed specific characters. Although some genes could respond to all four tested pathogens, the intensity of their expression to different pathogens varied. For instance, HaCec-1, HaCec-3, and HaIip were greatly up-regulated after fungi infection, which suggests that these genes mostly responded to fungal infection in H. armigera, although they also did respond to other pathogens. HaCec-2 was up-regulated by G + and Gbacteria, which suggests that it responds to bacteria. HaGall obviously increased in Gbacterium (K. pneumoniae) infection, and HaGlo also obviously increased in AcMNPV infection, which might imply that these genes were involved in response to the Gram-negative bacterium or virus, respectively. In addition, only five genes were up-regulated after virus infection: HaGNBP, HaCe1, HaGali, HaGlo, and HaLys, but the expression levels of the genes against the virus were much lower except HaGlo and HaLys. These phenomena suggest that the same AMP gene might respond to different pathogens by different expression patterns.
Lemaitre et al. [28] found that the challenge of fruit flies with a fungus results in the biosynthesis of antifungal AMPs, whereas the challenge by a Gbacterium results in the up-regulation of AMPs appropriate for the destruction of such bacteria. It seems that there is no strict correspondence in the immune response of H. armigera against different pathogens in mRNA level, although there are specific characters as described above. The expression patterns and functions of the immune related genes at protein levels need to be investigated in the future work.
There is growing interest in antiviral mechanism in Drosophila [29][30][31]. Dostert et al. identified 140 genes that were upregulated in adult Drosophila 24 and 48 h post injection with Drosophila C virus. Two thirds of these genes were not upregulated in response to bacterial or fungal infections. However, most genes encoding antimicrobial peptides regulated by the Toll and Imd pathways were not upregulated or were only weakly upregulated by DCV infection [31]. Same situation was found in H. armigera, only HaGlo and HaLys were weakly upregulated by viral challenge. This might suggest that the insect used a distinct mechanism to combat viral infection. H. armigera is a good model to analyze the viral infection. It is worth further investigation for the antiviral mechanism.

Conclusions
Five kinds of hemocytes were distinguished in H. armigera, the granulocytes in the insect are the major phagocytes. All haemocytes can be infected by AcMNPV. The expression patterns of 3 upstream PRR genes and 11 downstream effector genes in the insect were analyzed by qRT-PCR. The transcripts of 14 immune related genes have different expression patterns in H. armigera infected by different pathogens. Most of the genes were upregulated by bacteria and fungi, and only five genes were upregulated by the virus AcMNPV. These data suggested that the immune-related genes might have different functions against various kinds of pathogens.

Insect
The eggs of H. armigera were obtained from the Wuhan Institute of Virology, Chinese Academy of Science, Wuhan, China. After incubation, the larvae were maintained in the laboratory at 26 ± 1°C under light and dark conditions for 14:10 h and were reared on the artificial diet described by Zhao et al. [32].

Microbial pathogens
Bacillus thuringiensis, Klebsiella pneumoniae, and Candida albicans were from Shandong Agricultural University. AcMNPV-GFP was constructed in the researchers' laboratory [33].
Phagocytosis of the pathogens in vivo in H. armigera larvae B. thuringiensis, K. pneumoniae, and C. albicans were collected from the LB medium by centrifugation at 6,000 rpm for 10 min, fixed in 95% ethanol for 10 min, suspended in PBS (140 mM NaCl, 2.7 mM KCl, 10 mM Na 2 HPO 4 , 1.8 mM KH 2 PO 4 , pH7.4), added to acridine orange at 1 μg/ml for 10 min, centrifuged, and washed in PBS to clear. The microbes stained with acridine orange were injected into the haemocoels of three 5th instar 24 h larvae. Each larva was injected with 1 × 10 5 colony forming units (cfu) in 5 μl of PBS. The haemocytes were collected by cutting the proleg of the caterpillar 3 h post-injection, mixed a 200 μl PBS with 1 mM reduced glutathione and covered by glass cover slip, and observed under microscope.
Infection of larvae by the pathogens and extraction of RNA B. thuringiensis, K. pneumoniae, and C. albicans were collected from the LB medium by centrifugation at 6,000 rpm for 10 min, washed three times with PBS, and then resuspended in PBS. The number of pathogens was calculated by a haemocytometer. The pathogens were injected into the haemocoels of six 5th instar 24 h larvae. Each larva was injected with 1 × 10 5 colony forming units (cfu) in 5 μl of PBS. The budded virus (BV) of AcMNPV-GFP, amplified in Sf21 cells, was injected into the haemocoels at a concentration of 1 × 10 7 pfu/5 μl/larva. AcMNPV virus was quantitated using plaque forming units following the viral plaque assay in the instruction manual of Bac-to-Bac Baculovirus Expression Systems of Invitrogen http://wolfson.huji.ac. il/expression/bac.pdf. Briefly, the monolayers of Sf9 cells were prepared in 6-well plates, Eight-log serial diluted virus (10 -1 -10 -8 ) were produced and the diluted virus (10 -3 -10 -8 ) were added to each well of the plates respectively and incubate 1 h at room temperature. Sequentially the virus inoculum was removed from the wells and replaced with 2 ml diluted agarose and incubated at 27°C in a incubator for 5 days. The number of plaques was counted. The titer of the virus can be calculated by the following formula: pfu/ml of the virus = 1/dilution factor × number of plaques × 1/(ml of inoculum/plate). Controls were injected with PBS or Grace's medium, respectively, for the bacteria and virus. A fluorescence microscope (Olympus BX51) was used to observe the green fluorescence. The total RNAs were extracted from the whole bodies of three randomly chosen larvae per treatment to normalize the individual difference and then they were reverse transcribed to complementary deoxyribonucleic acid (cDNA) for quantitative real-time PCR analysis.

cDNA library construction
The total RNA was extracted from whole body of H. armigera using Unizol reagent following the manufacture's instructions (Biostar, Shanghai, China). Messenger RNA (mRNA) was extracted with the PolyATract mRNA isolation system (Promega, USA). The mRNAs were used to construct a cDNA library. The Creator SMART cDNA Library Construction Kit (Clontech, USA) was used for the cDNA library construction following the manufacturer's instructions. The double strand cDNA was digested and ligated with the pDNR-LIB vector, and then transformed into competent DH5α cells. Individual colonies were randomly selected, and plasmid was extracted for sequencing from the 5'-ends (Beijing Genomics Institute, China).

Quantitative real-time PCR (qRT-PCR)
The primers of the detected genes were designed from the specific sequences of the genes obtained by cDNA library sequencing in the laboratory ( Table 1). The quantitative real-time PCR was performed following the manufacturer's instruction of the SYBR Premix Ex Taq kit (Takara, Japan) with a real-time thermal cycler (Bio-Rad, Hercules, CA). The cDNA templates, obtained in method 2.5, were first examined by semi-quantitative RT-PCR for the qualities of the cDNAs and the gene primers by β-actin and every examined gene. The cDNA templates were then diluted 1:20 in distilled water, and the primers (100 μM) were diluted 1:100 in distilled water. The reaction mixture SYBR®Premix EX mixture 5 μl, primer 2 μl each, template 1 μl, water 2 μl was mixed on ice. The tubes with the reaction mixtures were put into the real-time PCR machine, and the following procedure was followed: one cycle of 95°C for 3 min, 40 cycles of 95°C for 15 s, 62°C for 50 s, read, 72°C for 2 s, read, 82°C for 2 s, read, 72°C for 10 min. The qRT-PCR data from three repeats were analyzed using the Opticon Monitor 2. Furthermore, the expression level of the immune-related gene was analyzed using the comparative C T method. In this method, the discrepancy between the C T for the gene and β-actin (ΔCT) were calculated to normalize the variation in the amount of cDNA in each reaction. The data obtained from three repeats were calculated by 2 -ΔCt and statistically analyzed by student t-test. A significant difference was accepted at P < 0.05.