Insecticidal activity of two proteases against Spodoptera frugiperda larvae infected with recombinant baculoviruses

Trichoplusia ni insect cells (BTI-Tn5B1-4) [47] and/or S. frugiperda IPLB-Sf21-AE (Sf-21) [48] were kept at 27°C in TC-100 medium supplemented with 10% fetal bovine serum (GIBCO-BRL). These cell lines were used for the in vitro propagation of AcMNPV and the recombinant vSynVI-gal, which contains the β-galactosidase (lac-Z) gene in place of the polh gene [49], and were also used for the construction of the recombinant viruses containing the ScathL and Keratinase genes, respectively.

The cathepsin-L (ScathL) gene from S. peregrina was amplified by PCR using specific oligonucleotides (Protease F 5'-CCACCAGCAACCATCACCTTAAGCTT TAACAC-3') (Protease R 5'-GAATTC AATTGAAAAAGGCAG-3') and DNA from the pKYH5 plasmid (courtesy of Dr. Robert Harrison, Iowa State University, USA). The Protease F oligonucleotide anneals at positions -10 to -35 and relative to the start codon (ATG) and the Protease R oligonucleotide anneals to positions +76 to +91 relative to the last nucleotide of the stop codon (TAA) of the ScathL gene. The position of the Hin dIII and Eco RI restriction sites are shwon in italics, respectively. The amplified fragment was then cloned into vector pGEM^®-T following the manufacturer's instructions (Promega). The plasmid pGEMScathL containing the gene for ScathL was digested with Nco I (Invitrogen) and Not I (Promega), the resulting fragment was separated by electrophoresis in an agarose gel (0.8%) and the fragment of 1,100 bp, corresponding to the ScathL gene was purified using the DNA extraction Perfect Gel Cleanup kit, according to manufacturer's instructions (Eppendorf). Next, we carried out a T4 DNA polymerase reaction (Invitrogen) using the purified fragment in order to create blunt ends, following the manufacturer's instructions (Invitrogen) and ligated the fragment to the transfer vector pSynXIVVI+X3 [49], which enables insertion of the heterologous gene under the control of two promoters in tandem (pSyn and pXIV) [49], and previously digested with Sma I and dephosphorylated, according to the manufacturer's protocol (Promega). Escherichia. coli DH5α cells were transformed with the ligation by electroporation [50] and the recombinant plasmid (pSynScathL) was obtained. The plasmid pGEMKerat containing the Keratinase gene from A. fumigatus [46] was amplified in DH5α cells of E. coli and purified using the DNA extraction Concert kit, according to manufacturer's instructions (Invitrogen). The plasmid was digested with Eco RI (GE), the DNA fragment corresponding to 1,200 bp was purified from an agarose gel (0.8%), using the GFX DNA extraction kit, according to the manufacturer's instructions (GE). The purified fragment was ligated with the Eco RI-digested and dephosphorilated transfer vector pSynXIVVI+X3 [49], using the Rapid DNA Ligation^® kit following the manufacturer's instructions (Promega). The ligation product was then used to transform DH5α cells in order to obtain the transfer vector pSynKerat.

The plasmid DNAs from pSynScathL and pSynKerat (1 μg each) were separately co-transfected with the DNA (0.5 μg) of the Bsu 36I-linearized vSynVI-gal recombinant virus in BTI-TN-5B1-4 cells (10⁶), using liposomes following the manufacturer's instructions (CellFectin^®, Invitrogen ).

Seven days after co-transfection, the supernatants of the co-transfected cells were collected and used for the isolation of the recombinant viruses vSynKerat and vSynScathL by serial dilution in 96-well plates [51].

Thirty 3^rd instar S. frugiperda larvae (for each virus) were injected with 10 μl of each viral stock (approximately 1 × 10⁶ pfu) into the hemolymph, as a negative control, thirty S. frugiperda larvae were also injected with culture medium and the experiment was repeated three times. The inoculated larvae were placed individually in plastic cups with artificial diet and observed twice daily until death. Statistical analysis was performed using the Polo Plus program (LeOra Software).

Bioassays with occluded viruses were conducted using the droplet feeding method [52] with five different concentrations of occlusion bodies per nanoliter (10², 10¹, 1.0, 0.1, 0.01 occlusion bodies/nL). Thirty neonate larvae of S. frugiperda were used for oral inoculation with the different viral doses from each of the recombinant viruses, the wild type AcMNPV and with only dye (2% phenol red) as negative control. Mortality was scored until 10 d.p.i. and the data analyzed by probit analysis using the Polo Plus program (LeOra Software). The insects were monitored every eight hours for ten days. The inoculated larvae were placed individually in plastic cups with artificial diet and the experiment was repeated three times. The mean time to death (TD) was calculated according to Morales et al. [53].

Ten 3^rd instar S. frugiperda larvae were injected with the recombinant viruses as described above and dissected at different times post infection. The insects were dissected by cutting along their backs with an entomological scissors to expose the gut and other organs and were photographed under a stereomicroscope (Stemi SV 11, Zeiss). An uninfected larvae was used as control. Furthermore, the infected insects were also prepared for scanning electron microscopy, as described in Matos et al. [54]. Briefly, the infected insects were fixed in a solution of 2% glutaraldehyde and 2% paraformaldehyde in sodium cacodylate buffer 0.1 M, pH 6.4 for 2 h at 4°C, washed by 3 cycles of 15 min with cacodylate buffer 0.1 M and post-fixed in osmium tetroxide and 1:1 potassium ferrocyanide for 2 h and then dehydrated with an ascending series of acetone and then dried (Balzer CPD30 critical point drier) and covered with gold in an sputter coater apparatus (Balzer SCD 050). The samples were then analyzed in a scanning electron microscope JEOL JSM 840 at10 kV.

Third-instar S. frugiperda larvae were separately inoculated with BV stocks (10⁸ pfu/mL) with AcMNPV, vSynScathL, vSynKerat and mock infected as described above. At 72 h p.i., haemolymph was collected and placed into 100 μl of anticoagulant buffer (0.098 M NaOH, 0.186 M NaCl, 0.017 M EDTA, 0.041 M Citric acid) and used for detection of phenoloxidase activity. Briefly, hemolymph samples were kept on ice, and hemocytes were pelleted by centrifugation at 3,000 × g for 5 min at 4°C. The cell-free hemolymph, 113 μg, was then transferred to a tube containing 800 μL of 10 mM L-3,4-dihydroxyphenylalanine (L-DOPA) and incubated for 20 min at 25°C and the mixture analyzed in a spectrophotometer at 475 nm.

The ScathL gene from S. peregrina was amplified by PCR from pKYH5 plasmid DNA and cloned into the vector pGEM^®-T Easy (data not shown). The DNA fragment containing the gene was removed from the cloning vector by digestion with restriction enzymes and cloned into the transfer vector pSynXIVVI+X3 forming a new plasmid, called pSynScathL (data not shown). Similarly, the Keratinase gene was removed from a cloning vector by digestion with restriction enzymes and cloned into the transfer vector pSynXIVVI+X3 generating the plasmid pSynKerat (data not shown). The recombinant viruses were constructed by separetely co-tranfecting insect cells with the pSynScathL and pSynKerat DNA and DNA from the recombinant vSynVI-gal in BTI-Tn5B1-4 cells. Within the insect cells, homologous recombination occurred between regions of the plasmid vector and viral genome. The recombinant viruses vSynScathL and vSynKerat were then isolated from the supernatant of co-transfected insect cells by end-point dilution (Figure 1).

Thirty 3^rd instar S. frugiperda larvae were separetely inoculated with aproximately 10⁶ pfu per larvae of each recombinant and wild type virus via hemolymph. The recombinants vSynScathL and vSynKerat were able to induce insect death faster than the wild-type virus (Table 1). The vSynScathL showed a LT₅₀ and a mean time to death (TD) of 47 h and 2.62 days, respectively, while the AcMNPV, a LT₅₀ of 136 h and a TD of 5.37 days, respectively. This represents a significant 65.5% reduction in the time needed to kill the virus infected insects when compared to the wild type virus. The LT₅₀ and TD for the vSynKerat were 91 h and 3.70 days, respectively, with represents a reduction of 32.8% compared to AcMNPV. Moreover, in the final stages of infection, viruses with the ScathL and Kerat genes induced melanization of the cuticle, which was not observed with AcMNPV infected insects (Figure 2).

Droplet feeding bioassays were also carried out with neonate S. frugiperda larvae with different concentrations of occlusion bodies from AcMNPV, vSynScathL and vSynKerat. The recombinant vSynScathL was also shown to induce death in neonate larvae faster compared to wild-type virus (Table 2). The vSynScathL showed a LT₅₀ of 77 h while the AcMNPV, a LT₅₀ of 104 h when inoculated with 10² PIBs/nL. This represents a reduction of 26% in the time needed to kill the infected insects when compared to the wild type virus. The LT₅₀ for the virus vSynKerat was 54 h, with a reduction of 48% compared to the virus AcMNPV. We also analysed the LC₅₀ for the two recombinants but no significant diffference was observed when compared with the wild type virus (Table 3).

S. frugiperda larvae uninfected and infected with AcMNPV, vSynScathL and vSynKerat were examined under a stereomicroscope (Figure 3) and a scanning electron microscope (Figure 4). The larvae infected with AcMNPV showed the presence of fat tissue (Figure 3) and tracheal system firmly attached to the gut of the caterpillar (Figure 4). On the other hand, larvae infected with vSynScathL (Figure 2) and vSynKerat (Figure 2) showed melanization of the cuticle, had little or no fat tissue and tracheal system was loosely connected to the midgut of the insect (Figure 4).

Phenoloxidase activity was determined spectrophotometrically by measuring formation of dopachrome from L-DOPA at 475 nm in haemolymph samples from insects infected with vSynScathL, vSynKerat, AcMNPV and mock infected (figure 5). We observed an expressive increase in phenoloxidase activity in haemolymph from S. frugiperda larvae infected with vSynScathL (0.23) and vSynKerat (0.17) when compared with haemolymph from mock-infected (0.10) and AcMNPV-infected insects (0.05). The experiment was repeated three times.