Expression of a Toll Signaling Regulator Serpin in a Mycoinsecticide for Increased Virulence

Abstract

Serpins are ubiquitously distributed serine protease inhibitors that covalently bind to target proteases to exert their activities. Serpins regulate a wide range of activities, particularly those in which protease-mediated cascades are active. The Drosophila melanogaster serpin Spn43Ac negatively controls the Toll pathway that is activated in response to fungal infection. The entomopathogenic fungus Beauveria bassiana offers an environmentally friendly alternative to chemical pesticides for insect control. However, the use of mycoinsecticides remains limited in part due to issues of efficacy (low virulence) and the recalcitrance of the targets (due to strong immune responses). Since Spn43Ac acts to inhibit Toll-mediated activation of defense responses, we explored the feasibility of a new strategy to engineer entomopathogenic fungi with increased virulence by expression of Spn43Ac in the fungus. Compared to the 50% lethal dose (LD₅₀) for the wild-type parent, the LD₅₀ of B. bassiana expressing Spn43Ac (strain Bb::S43Ac-1) was reduced ∼3-fold, and the median lethal time against the greater wax moth (Galleria mellonella) was decreased by ∼24%, with the more rapid proliferation of hyphal bodies being seen in the host hemolymph. In vitro and in vivo assays showed inhibition of phenoloxidase (PO) activation in the presence of Spn43Ac, with Spn43Ac-mediated suppression of activation by chymotrypsin, trypsin, laminarin, and lipopolysaccharide occurring in the following order: chymotrypsin and trypsin > laminarin > lipopolysaccharide. Expression of Spn43Ac had no effect on the activity of the endogenous B. bassiana-derived cuticle-degrading protease (CDEP-1). These results expand our understanding of Spn43Ac function and confirm that suppression of insect immune system defenses represents a feasible approach to engineering entomopathogenic fungi for greater efficacy.

INTRODUCTION

Entomopathogenic fungi are ubiquitously distributed in almost all ecosystems, where they act as important natural regulators of insect populations (1, 2). Due to their insecticidal activities, there has been much interest in broad-host-range entomogenous fungi, such as Metarhizium anisopliae and Beauveria bassiana, for use as environmentally friendly alternatives to chemical pesticides in insect control, and both are currently approved for commercial use by the Environmental Protection Agency (3, 4). In B. bassiana, the pathogenic process is facultative and not required for completion of the fungal life cycle. Mycosis involves several steps that include (i) adhesion of spores (conidia) to the host cuticle, (ii) germination on the insect surface and hyphal penetration of the exoskeleton into the hemocoel, (iii) proliferation and immune evasion in the hemocoel, and (iv) outward hyphal growth and sporulation on the host cadaver (5). Infection is affected by a wide range of factors, including temperature, humidity, UV exposure, nutrient availability, and even the physical state of the host, and it can take between 6 and 14 days for the fungus to kill target insects, which imposes a significant limitation regarding the use of these fungi as a biological control agent (6, 7). In addition, a number of insect targets display various degrees of resistance to infection by B. bassiana via expression of antifungal compounds or other aspects of innate immune activation. These include hemocyte activation and phagocytosis, encapsulation, reactive oxygen species (ROS) generation, and melanization, and these can be activated in response to complex environmental factors, including insect population densities (8, 9).

Within this context, the Spaetzle/Toll/Cactus signaling cascade pathway in Drosophila melanogaster has been shown to mediate responses to fungal infection (10). The transmembrane receptor Toll does not recognize fungal determinants directly but is activated by a proteolytic fragment of Spaetzle, a cytokine-like molecule, which in Drosophila has been shown to directly bind Toll to establish signaling (11). Spaetzle can be activated via recognition of fungal cell wall components through the pattern recognition receptor (GNBP-3) or through detection of fungal (protease) virulence factors which activate the serine protease Persephone (Psh) (12). A family of proteinase inhibitors known as serpins, in turn, negatively regulates the serine proteinases (13–15). The biochemical functions and structural features of a number of serpins have been at least partially defined; in Manduca sexta, serpin-5 has been shown to regulate prophenoloxidase (proPO) activity (16), and Anopheles gambiae serpin-2 functions as a negative regulator of melanization (17). Far more extensive studies have been performed in Drosophila, where the loss of phenoloxidase (PO) activity has been shown to decrease survival (18), and the serpin Spn43Ac (also known as the nec gene product) has been shown to inhibit Psh and thus act as a negative regulator of Toll-mediated immune defense response signaling (15). Indeed, constitutive activation of Toll-mediated antifungal defense is seen in the absence of the appropriate serpin (19).

Genetic engineering has become a powerful tool to improve fungal performance (20–23), and the (over)expression of a variety of cuticle-degrading enzymes in M. anisopliae and/or B. bassiana, including chitinases, proteases, and chitinase-protease hybrids, has led to improvements in the virulence of these fungi (24–27). More recently, small peptides, including trypsin-modulating oostatic factor (TMOF) from the mosquito Aedes aegypti and pheromone biosynthesis-activating neuropeptide (PBAN) from the fire ant Solenopsis invicta, were expressed in B. bassiana and resulted in increased virulence against their respective targets (28, 29). In the present paper, the consequences of the expression of a serpin, Spn43Ac, from Drosophila melanogaster in B. bassiana under the control of the constitutive gpdA promoter were assessed in terms of increased fungal virulence and development within insect hosts, as were its effect on host PO activity. Spn43Ac was also expressed and purified from an Escherichia coli heterologous host, and its activity was examined.

MATERIALS AND METHODS

Microbial strains and media.

B. bassiana strain CGMCC7.34 (China General Microbiological Culture Collection Center) was isolated from infected cadavers of Pieris rapae butterflies in China in 1997 and has been conserved as a mixture of dry conidia at −80°C. Fungal strains were routinely grown on potato dextrose broth (PDB) or potato dextrose agar (PDA), Sabouraud dextrose broth (SDB) or Sabouraud dextrose agar (SDA), and/or Czapek-Dox broth (CZB) or Czapek-Dox agar (CZA). Escherichia coli DH5α was used for plasmid propagation, and E. coli BL21(DE3) was used for protein expression. E. coli strains were cultured in Luria-Bertani (LB) medium supplemented with ampicillin (50 μg/ml) or kanamycin (50 μg/ml), on the basis of the plasmid selection markers used.

Molecular manipulations and transformation.

The complete open reading frame (ORF) of the D. melanogaster Spn43Ac gene (GenBank accession no. NM_080112) was cloned via fusion PCR using genomic DNA derived from D. melanogaster as the template. Primer pairs P1/P2 and P3/P4 (Table 1) were used to amplify the fragments upstream and downstream of a single intron present in the Spn43Ac gene, respectively, with primers P2 and P3 being designed to contain 20-bp overlapping sequences. The full length of Spn43Ac was produced by PCR in a reaction mixture containing 5 μl 5× Phusion Taq polymerase buffer, 2 μl 2.5 mM deoxynucleoside triphosphates, 30 ng upstream fragment, 30 ng downstream fragment, and 0.4 U Phusion Taq DNA polymerase in a total volume of 25 μl. PCR cycling conditions were as follows: 98°C for 2 min, followed by 25 cycles of 98°C for 20 s, 56°C for 30 s, and 72 for 1 min with a final extension at 72°C for 5 min. Primer pair P1 and P4 was then used to obtain the full-length Spn43Ac ORF using the assembled products as the template. The PCR product was cloned into the pEASY-blunt vector (Transgen, China) to yield pEASY-Spn43Ac, and the integrity of the insert was verified by sequencing.

TABLE 1.

Primers used in this study

Primer	Sequencea	Restriction enzyme site
P1	GAATTCATGGCGAGCAAAGTCTCGATCCTTCTCCTGC	EcoRI
P2	GGGGTACGAACTTGGCATAGGAAGCTGCCGAGGCC
P3	GGCCTCGGCAGCTTCCTATGCCAAGTTCGTACCCC
P4	CCCGGGTTAGACGCTCATGGGCGTGGGAT	SmaI
P5	TCTAGAGAGCTCATCGCTTGGCAACG	XbaI
P6	AAGGAAAAAAGCGGCCGCATGGCTCCTTTTCTTCAAACC	NotI
P7	GGACTAGTGCCGGCTCGCGGCGCCAAGGG	SpeI
Pserpinrt1	TCTACTTCCAGGGTCGTTGG
Pserpinrt2	GGTCATGTCCTGCTCGAACT
Pser-Exp1	GAATTCGAGCTCATCGCTTGGCAACG	EcoRI
Pser-Exp2	GCGGCCGCTTAGACGCTCATGGGCGTGG	NotI

^aUnderlined sequences are the restriction enzyme sites.

A secretion signal peptide (Bbsp) derived from B. bassiana Chit1 (GenBank accession no. AY145440) was added to the cloned Spn43Ac ORF. Bbsp was amplified using primers P6 and P7 and B. bassiana genomic DNA as the template and cloned into the pGEM-T Easy vector (Promega), yielding pGEM-Bbsp. Primers P5 and P4 were then used to amplify the Spn43Ac ORF using pEASY-Spn43Ac as the template. pGEM-Bbsp was digested with NotI and SpeI to produce the Bbsp fragment, and the Spn43Ac PCR product was digested with XbaI and SmaI. The Bbsp and Spn43Ac fragments were cloned into the pGEM-Pgpda-TtrpC vector containing the glyceraldehyde phosphate dehydrogenase (gpd) promoter from Aspergillus nidulans and the TrpC terminator sequences to produce Pgpda-Bbsp:Spn43Ac-TtrpC. The Pgpda-Bbsp:Spn43Ac-TtrpC fragment was subsequently cloned into the pBAR-GFP vector (24), yielding pBAR-GFP-Pgpda-Bbsp:Spn43Ac-TtrpC for use in fungal transformation. B. bassiana fungal transformation and screening of transformants were conducted as previously described (24). Integration of the vector into the transformants was confirmed by PCR using primers P4 and P5.

For construction of the E. coli heterologous expression vector, the ORF of Spn43Ac was amplified by PCR using primers Pser-Exp1/Pser-Exp2 and pGEM-Spn43Ac as the template. The PCR product was cloned into the pGEM-T Easy vector, and the integrity of the insert was verified by sequencing. The fragment containing the Spn43Ac gene was subcloned from the pGEM-T Easy vector into the corresponding sites of pGEX-6p-1 (GE Healthcare) by EcoRI and NotI restriction enzyme digestion to produce pGEX-GST-Spn43Ac. For protein expression and purification, the recombinant plasmid (pGEX-GST-Spn43Ac) was transformed into E. coli BL21(DE3) cells. The empty vector pGEX-6p-1 was also transformed into BL21(DE3) cells and used as a negative control.

Gene expression analysis of Spn43Ac in transformants.

Total RNA was extracted from wild-type and fungal transformants (grown in 0.5× SDB for 3 days) using an Aurum total RNA minikit (Bio-Rad), and the extraction included a step of on-column removal of genomic DNA per the manufacturer's instructions. cDNA was synthesized using a RevertAid First-Strand cDNA synthesis kit (MBI Fermentas, Canada). Real-time PCR was performed using an iCycler iQ multicolor real-time PCR detection system with SYBR green (Bio-Rad). Reaction mixtures contained 5 μl of iQ SYBR green Supermix (Bio-Rad), 0.5 μl each of primers Pserpinrt1 and Pserpinrt2 (10 μM), and 4 μl of 1:10-diluted cDNA template and were incubated as follows: a 5-min denaturation step at 95°C, followed by 40 cycles of 95°C for 15 s, 56°C for 30 s, and 72 for 30 s. The relative expression levels of Spn43Ac were normalized using actin (GenBank accession no. HQ232398), gpd (GenBank accession no. AY679162), and cypA (GenBank accession no. HQ610831) as reference genes in iQ5 optical system software (version 2; Bio-Rad). The primers used in the real-time PCR are listed in Table 1.

Colony growth and conidial germination assays.

Fungal growth and conidial germination of the wild-type and Bb::S43Ac strains were analyzed as described previously (30). Briefly, conidial suspensions (1 μl of 1 × 10⁷ conidia/ml) were spot inoculated onto PDA, SDA, and CZA plates and examined over a 5-day interval. Conidial germination was observed microscopically by inoculating 10- to 100-μl aliquots of conidial suspensions (1 × 10⁶ conidia/ml) onto PDA plates, followed by culture at 26°C for 12 to 24 h and direct visualization of germ tubes via light microscopy. Conidial germination and vegetative growth were also tested on cicada (Cryptotympana atrata Fabricius) hind wings. The wings were surface sterilized in oxymethylene overnight and rinsed 4 times (for 5 min each time) in sterile distilled water. The wings were placed on wet filter paper and inoculated with 1.0 ml of conidia (5 × 10⁷ conidia/ml, prepared with sterile distilled water). Samples were incubated at 26°C for 22 h, and the percent spore germination and hyphal growth/appressorium formation was examined microscopically.

Insect bioassays.

Fungal cultures were grown on PDA plates for 14 to 20 days. Aerial conidia were harvested into sterile 0.05% Tween 80. The 3rd instar larvae of Galleria mellonella (wax moth) and adult Myzus persicae (green peach aphid) were treated via topical application of conidial solutions. Briefly, the larvae were immersed in fungal conidial solutions at various concentrations (5 × 10⁶, 3 ×10⁷, and 1 × 10⁸ conidia/ml) for 3 to 6 s and then taken out and placed on a dry paper towel to remove the excess liquid on the insect bodies. Controls were treated with sterile 0.05% Tween 80. Treated larvae were placed in large (150-mm) petri dishes and incubated at 26°C. All experiments were repeated three times, and each replicate contained a minimum of 25 insects. The number of dead insects was recorded daily, and the median (50%) lethal time (LT₅₀) and lethal dose (LD₅₀) were calculated by probit analysis.

Fungal proliferation within the insect hemocoel.

Hyphal bodies represent a biochemically distinct yeast-like fungal cell type produced in the insect hemolymph after hyphal penetration through the integument (31, 32). For collection of hemolymph, infected and control larvae were anesthetized on ice and the rear leg was cut off with a scissor at 12-h intervals starting 60 h after topical inoculation (10⁷ conidia/ml). The hemolymph that exuded from the wound was collected. Fungal hyphal bodies present in the insect hemolymph were observed via bright-field microscopy (Olympus IX81 microscope), and the number of hyphal bodies was determined by direct counting using a hemocytometer. Three replicates were performed for each treatment, and each replicate contained four randomly picked larvae.

Phenoloxidase assay.

Hemolymph from larvae was collected and placed into 1.5-ml microcentrifuge tubes on ice by cutting the rear leg and collecting the drops. For the PO assay, samples were centrifuged at 10,000 × g for 20 min at 4°C, and 50 μl of the supernatant was added to 150 μl of phenoloxidase substrate solution (5 g/liter dopamine in 50 mM sodium phosphate, pH 6.9). Samples were incubated at 25°C for 30 min, and the absorbance at 490 nm (A₄₉₀) was measured. One unit of PO activity was defined as a change in the A₄₉₀ of 0.01 after 30 min, as described previously (27). PO activity was determined in insects subjected to the following treatments: 2 μl (1 × 10⁷conidia/ml) of wild-type and Bb::S43Ac conidial suspensions prepared in 0.05% Tween 80 was injected into the hemocoel of G. mellonella larvae, with injections of 0.05% Tween 80 used as controls. In order to test the effect of inactive conidia on PO activity, fungal conidia were treated at 65°C for 20 min and then placed on ice for 5 min, after which 2 μl (1 × 10⁷conidia/ml) of the conidial suspensions was injected into the hemocoel of the larvae. Experiments using each strain were repeated three times, and each repeat contained 30 larvae. Hemolymph was collected and assayed for PO activity over time (0, 2, 8, 12, 24 h) after injection.

Purification of Spn43Ac.

E. coli cell growth, induction, and harvesting were conducted according to the manufacturer's instructions (Invitrogen). Briefly, E. coli BL21(DE3) cells harboring plasmid pGEX-GST-Spn43Ac or the empty vector control were grown in LB medium supplemented with 50 μg/ml ampicillin to an optical density at 600 nm of 0.5 (4 × 10⁸ cells/ml), after which isopropyl-β-d-thiogalactoside (IPTG) was added and the culture was allowed to grow as indicated until the cells were harvested via centrifugation. Expression parameters, including the IPTG concentration (0.05 to 1.0 mM), induction temperature (16 to 37°C), and induction time (0.5 to 8 h), were analyzed. Harvested cells were washed once in buffer (phosphate-buffered saline [PBS]), and cells were lysed by use of the MagneGST cell lysis reagent (Promega). Cell lysates were centrifuged (12,000 × g, 10 min), and the supernatant was designated the crude extract. Soluble GST-Spn43Ac protein in the crude extract was purified using a MagneGST protein purification system (Promega). Glutathione S-transferase (GST) from the pGEX-6p-1 empty vector was also purified using the same method. The purified proteins were dialyzed against PBS (pH 6.9) four times at 4°C and analyzed by SDS-PAGE. Western blot assays were performed using standard protocols, and the blots were probed using a GST antibody (Abcam, England) and visualized using Clarity Western ECL substrate (Bio-Rad). The effect of Spn43Ac on the activity of the B. bassiana cuticle-degrading protease CDEP-1 was analyzed with the synthetic peptide succinyl–(alanyl)₂–prolyl–phenylalanine–p-nitroanilide [Suc-(Ala)₂-Pro-Phe-NA; Sigma] as the substrate (33). The CDEP-1 enzyme was produced and purified from Pichia pastoris as indicated in our previous study (34).

The effect of various purified proteins on PO activity was analyzed in reaction mixtures containing 150 μl 25 mM PBS (pH 6.9), 0.25 mg l-3,4-dihydroxyphenylalanine (l-DOPA), and the following: (i) purified GST, GST-Spn43Ac, or inactivated GST-Spn43Ac (boiled in water for 10 min) at concentrations ranging from 0 to 6 μM and (ii) the PO activator lipopolysaccharide (LPS), trypsin, laminarin, or chymotrypsin (final concentration, 0.25 mg/ml), which was added to activate the PO cascade. In some experiments, chymostatin (final concentration 2.5 ng/ml), an inhibitor of chymotrypsin, was added as a control.

Statistical analysis.

All statistical analyses were conducted using SPSS software (version 17.0). A P value of <0.05 was considered statistically significant.

RESULTS

Construction of B. bassiana Spn43Ac expression strains.

Due to the presence of a single intron in the ORF of the D. melanogaster Spn43Ac gene, a fusion PCR approach was used to construct the vector for expression in B. bassiana, as detailed in Materials and Methods. The A. nidulans gpd promoter was used to drive the expression of Spn43Ac, and the construct contained a TrpC terminator sequence. The A. nidulans gpd promoter has been demonstrated to be functionally active and essentially constitutively produce recombinant protein in B. bassiana (24, 34). In addition, a secretion signal peptide derived from the B. bassiana Chit1 (chitinase) gene was fused to the N terminus of the Spn43Ac protein to drive extracellular secretion of the protein. The expression vector was transformed into B. bassiana strain CGMCC7.34, and putative transformants were further screened for genomic insertion of the Spn43Ac construct by PCR (see Fig. S1 in the supplemental material). To determine whether expression of Spn43Ac resulted in any effects on fungal morphology or development, transformants were examined for growth on PDA, SDA, and CZA. No significant differences in germination, radial and mycelial growth, or conidiation were seen between the transformants and the wild-type strain (data not shown). The wild-type and transgenic strains showed similar germination rates on PDA medium (at 14 h postinoculation, 95% versus 96%; P > 0.05). In addition, no significant differences in growth or appressorium formation of the fungal strains grown on cicada wings were noted (Fig. 1). In order to select for strains with the highest levels of heterologous protein expression, transformants were screened using real-time PCR for quantification of Spn43Ac expression levels. All tested transformants showed production of Spn43Ac transcripts (which was absent in the wild-type strain), although a significant variation in expression levels was noted between different isolates (Fig. 2). One strain, designated Bb::S43Ac-1, exhibited the highest expression level and was chosen for further study.

FIG 1.

Micrographs of B. bassiana (wild-type [WT] and Bb::S43Ac-1) adhesion to cicada (C. atrata Fabricius) wings for 22 h. Arrows, the appressorium. Bars = 50 μm.

FIG 2.

Spn43Ac gene expression analysis. Quantitative reverse transcriptase PCR analysis of Spn43Ac expression normalized to gpd, actin, and cypA expression in B. bassiana. WT, B. bassiana wild-type strain; Bb::S43Ac-1 to Bb::S43Ac-3, three transformants expressing the Spn43Ac gene. All strains were cultured in 0.5× SDB for 3 days, and then total RNA was extracted and quantitative reverse transcriptase PCR was performed as described in the Materials and Methods section.

Expression of Spn43Ac in B. bassiana increases virulence.

Topical bioassays, which represent the natural route of infection, revealed enhanced virulence of the Bb::S43Ac-1 strain compared to that of the wild-type parent, with LT₅₀s (3 × 10⁷ conidia/ml) of 84 h and 110 h for the Bb::S43Ac-1 and wild-type strains, respectively, against G. mellonella (Fig. 3A; Table 2). Topical bioassays using Myzus persicae adults resulted in an LT₅₀ of 98 h for the Bb::S43Ac-1 and an LT₅₀ of 117 h for the wild type (at 1 × 10⁷ conidia/ml) (Fig. 3B). These data indicate a 16 to 24% decrease in the LT₅₀ for the Spn43Ac-expressing strain compared to that for the wild-type parent. Expression of Spn43Ac also decreased the LD₅₀. Using G. mellonella, the LD₅₀ for Bb::S43Ac-1 was 2.0 × 10⁷ conidia/ml, and that for the wild type was 6.0 × 10⁷ conidia/ml, indicating a 3-fold reduction in the number of spores required for the same level of mortality (Table 2).

FIG 3.

Insect bioassays. Survival curves of G. mellonella (A) (3 × 10⁷ conidia/ml) and M. persicae (B) (1 × 10⁷ conidia/ml) infected with the wild-type and Bb::S43Ac-1 strains are shown.

TABLE 2.

Calculated LD₅₀s and LT₅₀s of WT and Spn43Ac-expressing B. bassiana strains against G. mellonella and M. persicae

Strain	Host	LD50a (no. of conidia/ml)	LT50 (h)
Wild-type B. bassiana	G. mellonella	6.0 ± 0.7 × 107a	110.5 ± 3.8b
Bb::S43Ac	G. mellonella	2.0 ± 0.2 × 107a	84 ± 3.1b
Wild-type B. bassiana	M. persicae	NDd	117 ± 4.5c
Bb::S43Ac	M. persicae	ND	98 ± 4.2c

^aThe LD₅₀ was calculated from the 96-h time point.

^bThe bioassay was performed using a spore concentration of 3 × 10⁷ conidia/ml.

^cThe bioassay was performed using a spore concentration of 1 × 10⁷ conidia/ml.

^dND, not determined.

Hyphal body development in infected G. mellonella larvae.

G. mellonella hemolymph was analyzed over time after inoculation with Bb::S43Ac-1 or wild-type conidia as described in the Materials and Methods section. Microscopic observation of hemolymph samples showed greater hyphal body proliferation in larvae inoculated with strain Bb::S43Ac-1 than in larvae inoculated with the wild type at 84 h postinoculation (Fig. 4A). Quantification of hyphal body production over time postinoculation (60 to 108 h) indicated a 2- to 4-fold increase in hyphal bodies during the later stages of infection for the Bb::S43Ac-1 strain compared to that for the wild type (Fig. 4B).

FIG 4.

Fungal hyphal body production in G. mellonella hemocoel. The 3rd instar larvae of G. mellonella were topically inoculated with fungal spores (10⁷ conidia/ml) in 0.05% Tween 80. (A) Hyphal bodies observed at 84 h postinoculation using light microscopy. Arrows, hyphal bodies. (B) Time course of hyphal body proliferation postinoculation of wild-type and Bb::S43Ac-1 spores. Hyphal body concentrations were determined by direct counting using a hemocytometer over the indicated time course. The experiment was repeated three times, and each treatment contained four replicates. Error bars indicate SDs.

In vivo inhibition of phenoloxidase activation.

Phenoloxidase activity in G. mellonella hemolymph was analyzed over time after challenge with fungal conidia. At 2 h postinjection, PO activity was similar between all samples (Fig. 5). However, significant differences (P < 0.01) in PO activity between Bb::S43Ac-1 and the other strains were seen at 8 h postinfection. PO activity peaked at 8 h in control injections and injections with heat-inactivated (65°C, 20 min) wild-type or Bb::S43Ac-1 conidia, after which the activity tapered off, reaching basal levels at 24 h. Compared to the results for the control, injection of wild-type conidia resulted in an ∼50% suppression of the peak activity at 8 h, whereas injection of Bb::S43Ac-1 conidia resulted in the same level of PO activity seen at the baseline.

FIG 5.

PO activity in G. mellonella hemolymph. Larvae were injected with wild-type, Bb::S43Ac-1, heat-inactivated wild-type (IAW), and heat-inactivated Bb::S43Ac-1 (IAS) spores and 0.05% Tween 80 (control). To inactive fungal spores, the spore suspension was treated at 65°C for 20 min and then placed on ice for 5 min. Hemolymph samples were collected over the indicated time course, and the PO activity was measured as described in Materials and Methods. The experiment was repeated three times, and each treatment contained three replicates. Error bars indicate SDs.

Heterologous expression and purification of Spn43Ac from E. coli and its inhibition of PO activation.

The Spn43Ac ORF was subcloned into the pGEX-6p-1 vector as an N-terminal glutathione S-transferase-tagged fusion protein. Optimal production of Spn43Ac in the soluble crude extract fraction was found to occur when induction of protein expression with 0.2 mM IPTG occurred at 26°C for 4 h. The fusion protein (as well as GST for use in control experiments) was purified to homogeneity using GST affinity chromatography. The molecular mass of GST-Spn43Ac matched the calculated value based on its amino acid sequence (∼74 kDa), as detected by SDS-PAGE and Western blotting (Fig. 6). A concentration-dependent inhibition of chymotrypsin activation of PO activity was observed (Fig. 7A), with an ∼75% decrease in PO activity seen in reaction mixtures containing 6 μM GST-Spn43Ac compared to that for the controls that included GST and inactivated GST-Spn43Ac (95°C, 10 min). The effect of GST-Spn43Ac on various other PO activators was also examined. Addition of lipopolysaccharide resulted in a 2-fold increase in PO activity over control (PBS treatment) levels, and this activity was only moderately (∼14%) inhibited by addition of GST-Spn43Ac (Fig. 7B). In contrast, activation of PO activity by trypsin or chymotrypsin was inhibited by 28 to 44% when GST-Spn43Ac was added, whereas laminarin activated PO levels ∼2.5-fold above that for the control (the highest seen), and addition of GST-Spn43Ac suppressed this activation by ∼25% (Fig. 7B). We also determined the effect of GST-Spn43Ac on the activity of the B. bassiana cuticle-degrading protease CDEP-1, which has been shown to be a virulence factor (27). As shown in Fig. 8, the activity of CDEP-1 was not affected by addition of the GST-Spn43Ac protein.

FIG 6.

SDS-PAGE and Western blot analysis of Spn43Ac expressed in E. coli BL21(DE3). (A) SDS-PAGE analysis of purification of GST-Spn43Ac from an E. coli host. Lanes M, molecular mass markers; lane 1, crude extract from untransformed E. coli control; lane 2, crude extract from E. coli BL21(DE3) transformed with plasmid pGEX-GST-Spn43Ac; lane 3, purified GST-Spn43Ac protein. The samples in both lanes 1 and 2 were induced with 0.2 mM IPTG at 25°C for 4 h before harvesting and analysis. (B) Western blot analysis of GST-Spn43Ac probed with anti-GST antibody as described in the Materials and Methods section (arrow).

FIG 7.

Inhibition of PO activation by purified GST-Spn43Ac. (A) Concentration-dependent inhibition of chymotrypsin-mediated PO activation by GST-Spn43Ac (0 to 6 μM). G. mellonella hemolymph was collected and chymotrypsin (0.25 mg/ml) was added to activate the PO cascade. Spn43Ac-IA, GST-Spn43Ac protein inactivated by heat denaturation (boiling, 10 min, 6 μM). (B) Effect of purified GST-Spn43Ac on PO activation by different stimulants. PO activity in G. mellonella-derived hemolymph was induced by PBS (0.1 M, pH 6.9, control), LPS, trypsin, laminarin, or chymotrypsin in the absence or presence of GST-Spn43Ac. Negative controls included a reaction mixture with GST, and a reaction mixture with chymostatin, a known inhibitor of chymotrypsin, was used as a positive control. The experiment was repeated three times, and each treatment contained three replicates. Error bars indicate SDs.

FIG 8.

Effects of Spn43Ac on the activity of a cuticle-degrading protease, CDEP-1, of B. bassiana. The activity of CDEP-1 (1.5 μM) was analyzed using Suc-(Ala)₂-Pro-Phe-NA (1 mg/ml; Sigma) as the substrate with different concentrations of GST-Spn43Ac (0 to 4 μM) as the inhibitor. Phenylmethylsulfonyl fluoride (PMSF; 15 μM), a serine protease inhibitor, was used as a negative control. No inhibition activity of GST-Spn43Ac on CDEP-1 was observed. Each treatment contained three replicates.

DISCUSSION

The insect cuticle and subsequent interactions of the fungus with the immune system represent the two predominant obstacles to successful mycosis. Cuticular defenses can include both endogenous factors (e.g., antimicrobial peptides, fatty acids, wax esters, and quinones) and exogenous factors (e.g., beneficial microbes), as well as communal behaviors, such as grooming and antiseptic washing mechanisms, that pose a formidable barrier for most microbes (5). In some instances, when integrated with cuticular defenses, the insect immune system can recognize and attempt to eliminate invading microbes via a series of immune reactions that can include hemocyte activation and phagocytosis, encapsulation, melanization, and activation/expression of antimicrobial proteins and molecules (8, 9). B. bassiana, however, has evolved mechanisms for overcoming these barriers that include secretion of cuticle-degrading enzymes, toxic metabolite production, mechanical penetration, and host immune evasion and suppression (35–38). However, the length of time required to kill target insects and the high spore concentrations required have constrained the use of these fungi for insect control. Increased expression of cuticle-degrading hydrolases has been shown to increase virulence, resulting in 6 to 40% decreases in LT₅₀ values, depending upon the enzyme expressed (24, 26, 34, 39). In this study, expression of an immune-regulating serpin gene, Spn43Ac, from Drosophila melanogaster in B. bassiana significantly increased fungal virulence, resulting in an ∼16 to 24% decrease in the LT₅₀ and a 3-fold decrease in the LD₅₀, indicating that 2 times fewer conidia are needed to achieve the same level of control. These data indicate the feasibility of genetic engineering of entomopathogenic fungi to target the insect immune system for increased efficacy.

Insect innate immunity involves identification of pathogen-associated molecular patterns (PAMPs) that include bacterial LPSs, peptidoglycans (PGs), and fungal cell wall carbohydrates, e.g., β-1,3-glucans (40, 41). Pattern recognition involves membrane receptors and soluble proteins and with respect to fungi appears to be mediated via Toll-like receptor signaling pathways that result in hemocyte activation (phagocytosis, nodule formation, or encapsulation), synthesis of antimicrobial peptides, and activation of prophenoloxidase (proPO) and melanization pathways. In the mosquito malaria vector Anopheles gambiae, the melanization response has been implicated in defense against B. bassiana, with silencing of two mosquito-positive regulators of melanization, Tep1 and CLIPA8, resulting in increased susceptibility to infection by the fungus (42). Toll-mediated activation is controlled by (serine) protease cascades that, in the absence of a signal (i.e., a PAMP), are kept in check by the activities of serpins, i.e., serine protease inhibitors. Several insect viruses produce PO cascade inhibitors that are thought to aid them in their ability to overcome host immune reactions (43). Thus, we reasoned that expression of a serpin by a fungal pathogen during the invasion process might inhibit innate immune activation to increase the virulence of the fungus.

Transformation of B. bassiana with a vector containing the Drosophila serpin Spn43Ac, implicated in regulating Toll-mediated signaling, under the control of a constitutive A. nidulans gpd promoter and with a B. bassiana Chit1 signal peptide was used to drive expression of the heterologous gene. A greater than 10-fold difference in Spn43Ac transcript levels was seen between various clones, indicating the potential for significant genome localization effects regarding integration of the vector into the host fungal genome. The localization effects were also observed in our previous study for protease CDEP-1 and hybrid protease CDEP:BmChBD (BmChBD is a chitin binding domain from Bombyx mori) (34). The B. bassiana transformant displaying the highest Spn43Ac expression levels displayed both reduced LT₅₀ and LD₅₀ values against a lepidopteran host (G. mellonella) and a hemipteran aphid (M. persicae). During the normal course of infection, after the fungus has breached the insect integument, proliferation of single-celled propagules known as hyphal bodies is seen in the hemolymph in vivo (31, 32). Infection by the Spn43Ac-expressing B. bassiana strain resulted in a 2- to 4-fold increase in hyphal body production, suggesting that innate immune systems do attempt to limit the microbial infection. The mechanism underlying these observations was further probed via demonstration of PO suppression in the hemolymph of Bb::S43Ac-1-infected hosts. Heterologous expression and purification of the Drosophila Spn43Ac in E. coli confirmed the concentration-dependent inhibition of PO activation by various proteases (trypsin and chymotrypsin). Intriguingly, only a low level of Spn43Ac inhibition (14%) of PO activation by LPS and a moderate level of Spn43Ac inhibition (∼25%) of PO activation by laminarin were noted, indicating regulation of these activation pathways by other serpins. Hyphal body proliferation was enhanced in the Spn43Ac-expressing strain, indicating an increased ability to evade the host immune system. We also observed some fluctuation in PO activity in G. mellonella hemolymph after challenge with the various fungal strains. PO activity increased to a maximum at 8 h after infection (by the form of injection) and then decreased to basal levels at 24 h. These data indicate robust host immune responses to fungal challenge leading to an increase of PO activity. Challenge with the Spn43Ac-expressing strain suppressed PO activation, consistent with the more rapid hyphal body proliferation in the hemocoel seen with this strain. Not too surprisingly, even with challenge by the wild type, suppression of PO activity was seen over time (i.e., by 24 h postinoculation), confirming that the fungus also contains endogenous mechanisms for immune suppression and/or evasion.

Although Spn43Ac is a serine protease inhibitor, our results indicate that expression in B. bassiana did not affect spore germination, growth, or hyphal penetration of the insect exoskeleton. The last process involves the expression of fungus-derived serine proteases required for degrading parts of the target cuticle. Our data showed that the protease activity of CDEP-1, a serine protease from B. bassiana involved in cuticle degradation, was not affected by addition of purified Spn43Ac in vitro (Fig. 8). In addition, Spn43Ac-expressing and wild-type strains showed similar phenotypes on artificial medium and during growth on cicada wings (Fig. 6). Finally, the hyphal bodies of the wild-type and Spn43Ac strains appeared in the hemolymph at the same time (72 h postinoculation), although they were more numerous in the engineered strains, indicating that they have similar abilities to penetrate the insect cuticle. These data suggest that Spn43Ac has some specificity regarding target serine proteases, as has been observed for other serpin proteins. For example, the serpin from Tenebrio molitor (spn48) could inhibit the amidase activity of Spaetzle-processing enzyme but not that of thrombin or trypsin (44).

Overall, our data indicated an improvement in infection rates, allowing control to be achieved with fewer spores, while decreasing the opportunity for the insect to feed, continue development, and/or reproduce. Due to the significant difference between vertebrate and invertebrate immunity pathways, targeting of insect immune systems may provide a safe and effective means of pest control. Use of fungal pathogens as a vehicle for delivery of immune-modulating molecules can also shed light on the functions of those molecules themselves. Given the diversity of serpins, future work to evaluate different serpins within the host-pathogen context is warranted.