Biochemical basis of synergism between pathogenic fungus Metarhizium anisopliae and insecticide chlorantraniliprole in Locusta migratoria (Meyen)

We challenged Locusta migratoria (Meyen) grasshoppers with simultaneous doses of both the insecticide chlorantraniliprole and the fungal pathogen, Metarhizium anisopliae. Our results showed synergistic and antagonistic effects on host mortality and enzyme activities. To elucidate the biochemical mechanisms that underlie detoxification and pathogen-immune responses in insects, we monitored the activities of 10 enzymes. After administration of insecticide and fungus, activities of glutathione-S-transferase (GST), general esterases (ESTs) and phenol oxidase (PO) decreased in the insect during the initial time period, whereas those of aryl acylamidase (AA) and chitinase (CHI) increased during the initial period and that of acetylcholinesterase (AChE) increased during a later time period. Activities of superoxide dismutase (SOD), catalase (CAT) and peroxidase (POD) decreased at a later time period post treatment. Interestingly, treatment with chlorantraniliprole and M. anisopliae relieved the convulsions that normally accompany M. anisopliae infection. We speculate that locust mortality increased as a result of synergism via a mechanism related to Ca2+ disruption in the host. Our study illuminates the biochemical mechanisms involved in insect immunity to xenobiotics and pathogens as well as the mechanisms by which these factors disrupt host homeostasis and induce death. We expect this knowledge to lead to more effective pest control.


Results
Virulence of chlorantraniliprole in combination with M. anisopliae against L. migratoria. The efficacy of chlorantraniliprole mixed with M. anisopliae against L. migratoria is presented in Table 1. The LC 50 value of M. anisopliae alone against this insect was 0.15 mg/L (at ~7.5 × 10 6 spores/mL). Mixing chlorantraniliprole with M. anisopliae in Treatments 1 and 2 resulted in higher mortality rates with LC 50 values of 0.01 and 0.02 mg/L and co-toxicity coefficients of 1646 and 1619, respectively. In contrast, when formulated in Treatment 3, an antagonistic interaction was observed between chlorantraniliprole and M. anisopliae, resulting in a co-toxicity coefficient of 34. In a separate experiment, L. migratoria was treated with M. anisopliae alone for 3 d before introducing wheat sprouts treated with chlorantraniliprole in Treatment 4, the resulting co-toxicity coefficient was 127 ( Table 2).
Effect of chlorantraniliprole and M. anisopliae on enzyme activities in L. migratoria. The activities of ESTs in L. migratoria are shown in Fig. 1. Chlorantraniliprole increased the activities of ESTs in L. migratoria during the initial post-treatment period; likewise, the activities of ESTs significantly increased during the initial period following M. anisopliae infection. However, the activities of ESTs markedly decreased after treatment with a mixture of M. anisopliae and chlorantraniliprole during the initial days of the experiment.
The activities of GSTs with different substrates were assessed, as shown in Figs 2 and 3. When treated with chlorantraniliprole, L. migratoria locust nymphs exhibited high GST activity (with CDNB or DCNB) during the early period, but this GST activity decreased during the later period. The activities of GSTs also increased during the initial period following M. anisopliae infection. However, GST activities decreased significantly during the initial period of the experiment when L. migratoria was treated with a combination of M. anisopliae and chlorantraniliprole simultaneously.
The activities of PO in L. migratoria after single and dual treatments were assessed, as shown in Fig. 4. When locust nymphs were treated with chlorantraniliprole alone, PO activities increased during the initial days of the experiment. When locust nymphs were treated with M. anisopliae, PO activities were high during the early period but decreased during the later period. In contrast, PO activities decreased during the initial days but increased during the later period when locust nymphs were treated with a combination of M. anisopliae and chlorantraniliprole.
The activities of MFO under single and dual treatment conditions were assessed, as shown in Fig. 5. MFO activity increased during the later period for locusts treated with M. anisopliae alone. MFO activity also increased during the later period when M. anisopliae and chlorantraniliprole were administered simultaneously; however, this increase in MFO activity was much lower than that obtained for treatment with M. anisopliae alone.
The activities of AChE in L. migratoria under different treatment conditions are shown in Fig. 6. AChE activity increased in locusts during the early period following treatment with chlorantraniliprole alone. When treated with M. anisopliae alone, locust AChE activities increased during the early period but decreased during the later period. When locusts were treated with a combination of M. anisopliae and chlorantraniliprole, AChE activities increased throughout the entire period, except the third day.
The activities of CHI (Fig. 7) in L. migratoria increased during the initial days of the experiment when locusts were treated with chlorantraniliprole alone. When locusts were treated with M. anisopliae alone, CHI activities increased throughout the entire analysis period. When locusts were treated with a combination of M. anisopliae and chlorantraniliprole, CHI activities increased during the early period but decreased during the later period.
The activities of AA (Fig. 8) increased during the initial days after independent treatment with chlorantraniliprole or M. anisopliae. AA activities also increased during the initial days after locusts were treated with a combination of M. anisopliae and chlorantraniliprole.
SOD, POD and CAT activities (Figs 9-11) increased during the initial days after independent chlorantraniliprole or M. anisopliae treatment. SOD and POD activities increased to high levels when locusts were treated with a combination of M. anisopliae and chlorantraniliprole; however, under these conditions, CAT activities decreased to low levels during the initial period, and, near the end of the experiment, CAT, SOD and POD activities decreased to low levels.

Discussion
Previous studies have indicated that combinations of entomopathogenic fungi and insecticides can have synergistic, antagonistic or additive physiological and mortality effects on insects 10,12,47 . However, these investigations of fungal-insecticide interactions, mortality, and/or LC 50 values were obtained using only one dose of pathogen or insecticide, such that the most efficacious synergistic formulations could not be determined [10][11][12] .  Table 1. The virulence of Metarhizium anisopliae fungi combined with chlorantraniliprole insecticide against Locusta migratoria. Ch = chlorantraniliprole, Ma = Metarhizium anisopliae. "-" signifies "no co-toxicity. " T1, T2 and T3 = Treatment1, Treatment 2 and Treatment 3, respectively. Mortalities (% ± SE) contained within a column followed by the same letters are not significantly different (LSD test, P < 0.01).
The present study utilised co-toxicity coefficients to calculate insecticide incorporation rates and estimate the efficacies of different chlorantraniliprole+ M. anisopliae formulations. We found both synergistic and antagonistic interactions when chlorantraniliprole and M. anisopliae were administered together. Treatments 1, 2, and 4 revealed synergistic interactions with co-toxicity coefficients of 1646, 1619, and 127, respectively. These high co-toxicity coefficients, which were accompanied by insect mortalities > 97% for some treatments, illustrate the effectiveness of this dual-attack method of insect pest control. In contrast, an antagonistic interaction (co-toxicity coefficient = 33) was observed for Treatment 3, possibly because of the high proportion of insecticide in this treatment. Our results demonstrate that chlorantraniliprole influences the virulence of M. anisopliae against L. migratoria relatively early in the initial infection period, producing a strong synergistic interaction with very high co-toxicity coefficients that vary depending on the proportions of the two agents and the time of application. As such, co-toxicity coefficients can be used to evaluate the effect of interactions between entomopathogenic fungi and insecticides, such that the best dose-combinations and time of application can be determined. Our study also illuminated some of the biochemical consequences of a paired insecticide-pathogen challenge in insects. The accompanying enzymatic responses are numerous and complex, in part because they include attempts to detoxify the insecticide as well as the xenobiotic (the insect's immune response to fungal attack) and  Table 2. Mortality in Locusta migratoria following 3-d exposure to Metarhizium anisopliae followed by addition of chlorantraniliprole. Ch = chlorantraniliprole, Ma = Metarhizium anisopliae. "-"signifies "no co-toxicity. " T4 = Treatment 4. Mortalities (% ± SE) within a column followed by the same letters are not significantly different (LSD test, P < 0.01).

Figure 1. Linear-regression analysis of the interrelation between fold changes of esterases activities and different treatment days involving chlorantraniliprole, Metarhizium anisopliae, and M. anisopliae combined with chlorantraniliprole.
because they involve biochemical/physiological disruptions due to the insecticide-and pathogen-induced pathology. Overall, most of the enzymes studied showed moderate to strong responses. MFO, GST and ESTs are the major enzymes involved in detoxifying penetrating xenobiotics in insects 18,42 . Chlorantraniliprole has been found to increase the activities of ESTs and GST in Cry1Ac-susceptible and -resistant strains of Helicoverpa armigera caterpillars and to inhibit GST and MFO activities in Plutella xylostella larvae 43,48 . Sublethal doses of chlorantraniliprole have been observed to increase MFO and EST activities but to decrease GST activity in the Lepidoptera H. armigera, Spodoptera exigua and Choristoneura rosaceana 45,49,50 . In the present study, increased activities of GST and ESTs following chlorantraniliprole treatment suggest that these enzymes may act to detoxify this insecticide. Reported exceptions to these phenomena may be due to differences in the insect species as well as the concentrations of chlorantraniliprole used 45,48,50 .
Increased detoxifying enzyme activities against mycoses and other infections represent the insect's response to bodily intoxication by metabolites or the host-tissue-degrading products of pathogens 51 . In the present study, we found that the activities of ESTs and GST were significantly increased during the initial period following M. anisopliae infection. MFO activity significantly increased during the later period following M. anisopliae infection. In Lepidopteran Galleria mellonella caterpillars, the activities of GST and ESTs in the haemolymph significantly increase during the first three days after M. anisopliae infection 18 . In the Hemipteran Eurygaster integriceps,  the activities of GST and ESTs increase four to five days after infection with B. bassiana spores and secondary metabolites 42 . Alterations in the activities of GST and ESTs in G. mellonella and L. migratoria infected with entomopathogenic fungi are considered to constitute nonspecific body responses to integument damage, induction of additional isoenzymes or fungal toxins 21,51,52 . Based on these observations, GST and ESTs appear to participate in the defensive reaction of L. migratoria to M. anisopliae during the initial period of infection.
When L. migratoria were treated with a combination of M. anisopliae and chlorantraniliprole simultaneously, the activities of ESTs and GST decreased significantly during the initial period. However, MFO activity increased during the later period. Our results are consistent with those from a previously published investigation of the system components involved in insect detoxification. For example, when the Colorado Potato Beetle Leptinotarsa decemlineata is treated with a mixture of M. anisopliae and an organophosphate insecticide, EST and GST activities significantly decrease 2 d after inoculation 46 . Hall has reported that the pathogen sickens the pest, thereby lowering its chemical resistance, and the chemical, in turn, sufficiently weakens the pest and increases its susceptibility to pathogen infection 53 . Furlong and Groden reported that the interactions between fungi and insecticides are probably mediated during certain periods of the infection process, such as conidial attachment, conidial germination, cuticular penetration, or the initial proliferation of hyphal bodies in the haemocoel 54 . We believe that chlorantraniliprole can be used as a stressor in L. migratoria to increase its susceptibility to M. anisopliae and  that toxic metabolites produced during the initial period of M. anisopliae infection may block the activation of detoxification components in the host.
PO is a key enzyme in insect immunity against pathogens. PO converts phenols to quinones, which subsequently polymerise to form melanin, which functions against both parasites and pathogens 32,34 . For example, in Spodoptera exempta, melanic larvae have higher PO activities and are more resistant to the entomopathogenic fungus, Beauveria bassiana, than non-melanic larvae 55,56 . Interestingly, high PO levels in insects might also increase resistance to insecticides. For example, insecticide-resistant diamondback moths display higher PO activities than susceptible moths 36 . In the present study, PO activities increased following treatment with either chlorantraniliprole or M. anisopliae, alone. In stark contrast, PO activities decreased significantly during the initial period after combined treatment with chlorantraniliprole and M. anisopliae, suggesting an antagonistic interaction of the insecticide and the pathogen in the host. Hiromori and Nishigaki have reported that mixed application of M. anisopliae and the insecticides teflubenzuron or fenitrothion inhibit PO activity in larvae of the beetle Anomala cuprea, and that granular cells are supressed 3 . These authors speculated that the observed interaction might be due to inhibition of the larval humoral defence and cellular immune systems. Moreover, joint treatment with M. anisopliae and an insecticide caused significant decrease in encapsulation intensity in  the beetle Leptinotarsa decemlineata 46 . The interaction observed in our L. migratoria experiments might also be caused by PO inhibition and weakened humoral defence or cellular immune systems.
AChE is a key enzyme that terminates nerve impulses by catalysing the hydrolysis of the neurotransmitter acetylcholine in the nervous system of various organisms 57 . As such, AChE is the primary target site in the central nervous system for organophosphate and carbamate insecticides 41 . In our study, AChE activities decreased during the later period after M. anisopliae infection. A previous study has demonstrated the inhibitory effect of secondary metabolites on AChE in synapses of B. bassiana 42 . Inhibition of AChE causes acetylcholine to accumulate at synapses, such that post-synaptic membranes remain in a state of permanent stimulation; the result is paralysis, ataxia, a general lack of coordination in the neuromuscular system and eventual death 58 . In our experiment, however, AChE activities significantly increased throughout the entire experimental period, except for the third day, when locusts were treated with a combination of M. anisopliae and chlorantraniliprole. We speculated that the synergy between chlorantraniliprole and M. anisopliae affected AChE activity, thus mitigating symptoms of paralysis and ataxia.
CHIs are present in both fungi and insects. In fungi, they are involved in cell growth and division. In insects, they degrade chitin in the peritrophic membrane and exoskeletal cuticle and play an important role in moulting. Accumulating evidence suggests that CHI may also have roles in organismal defence against pathogenic fungi and parasites [59][60][61] . AA catalyses the hydrolysis of various anilide derivatives and esters and transfers an acetyl group to aniline, which acts as an acetyl acceptor 62 . In larvae of the diamondback moth, the specific activity of AA is significantly inhibited by chlorantraniliprole 43 . In our study, CHI and AA activities increased during the initial period after combined chlorantraniliprole and M. anisopliae treatment, thus helping the fungi to colonise the host. However, CHI activities decreased later in the experiment. We speculated that the synergy between chlorantraniliprole and M. anisopliae affected CHI and AA activities and disrupted the insect's moulting process.
The major components of an insect's antioxidant defence system include several antioxidant enzymes, such as superoxide dismutase (SOD), catalase (CAT) and peroxidase (POD) 27 . These enzymes remove damaging reactive oxygen species (ROS), such as superoxide anions (O 2 •− ), hydrogen peroxide (H 2 O 2 ), singlet oxygen molecules (O 2 • ) and hydroxyl radicals (OH •− ) 30 . ROS are regularly synthesised in the guts of insects following natural bacterial infections. The dynamic cycle of ROS generation and elimination appears to be a continuous and essential process in insects 63,64 . Studies in the grasshopper Oxya chinensis have shown increased SOD and CAT activities after treatment with phoxim, malathion and chlorpyrifos insecticides 65,66 . Likewise, SOD, CAT and POD activities have been shown to increase after selection by chlorpyrifos for eight generations in Nilaparvata lugens planthoppers 67 . SOD activity increases in the midgut of Galleria mellonella L. moth larvae on days 1-3 after infection by Bacillus thuringiensis, whereas CAT activity decreases during this time 68 . In our study, when L. migratoria were treated with M. anisopliae alone, CAT, SOD and POD activities increased during the initial period of infection. It should be noted that SOD and CAT, together, take part in stepwise oxygen reduction. Enhanced SOD activity in the infected nymph should elevate H 2 O 2 concentrations and increase CAT activity. When chlorantraniliprole was applied in combination with M. anisopliae, SOD and POD activities increased but CAT activities decreased during the initial period post infection. We assumed that CAT was inhibited by the accumulation of superoxide radicals generated during the destruction processes. Research has shown that SOD1 protects calcineurin (CaN) from inactivation by ROS 69 . The enhanced activities of SOD and CAT then lead to the elimination of ROS 68 . In our study, the decreased activities of SOD, POD and CAT suggested a sharp decline in the ability to eliminate ROS during the later phase of infection. Large quantities of generated ROS can rapidly denature a wide range of biomolecules, including lipids, proteins and nucleic acids, thereby threatening virtually all cellular processes Based on previous and current results, we suggest that L. migratoria normally activates enzymatic defences against either pathogen or insecticide assault. Although these biochemical defences are sometimes successful, the combination of these two agents is able to overcome this defence. What is the chemical basis of this synergistic effect? One possibility is the disruption of Ca 2+ concentrations in locust cells, given that both chlorantraniliprole and M. anisopliae are known to strongly disrupt Ca 2+ balances in insects. For example, PO functions in an insect's immune response to pathogens, but PO activity is regulated by Ca 2+ concentration 70,71 . Hence, disrupting the Ca 2+ balance could disrupt PO activity, thereby weakening the host's immunity to M. anisopliae. Our results also reveal some of the complex biochemical processes that underlie the synergistic action of the two agents tested herein and illuminate the enzymatic mechanisms involved in insect immunity to pathogens and detoxification of insecticides. Understanding the response and activity of each enzyme alone and in combination will allow the design of novel and more effective means for pest control. Our results should provide guidance for future studies of the biochemical mechanisms underlying M. anisopliae's disruption of host immunity.
Locusts were originally collected from Cangzhou, Hebei Province, China, then reared for 13 generations in the laboratory without exposure to insecticides. The eggs were hatched in a growth chamber maintained at 30 ± 2 °C and 60 ± 5% relative humidity (RH) for 2 wks. After hatching, the nymphs were fed fresh wheat sprouts and were maintained in a cage (60 × 50 × 70 cm) in the laboratory at 30 ± 2 °C and 60 ± 5% RH under a 14:10-h light:dark photoperiod.

Fungus and synthetic insecticide. The entomopathogenic fungus Metarhizium anisopliae (Metschnikoff)
Sorokin IMI330189 was cultured on potato dextrose agar yeast extract (PDAY) at 27 ± 1 °C and 75 ± 10% RH for 7-10 d under constant light. Conidia were harvested from culture plates by scraping the surface of the PDAY with a sterile mounted needle and were then placed into plastic centrifuge tubes containing 0.1% Tween in sterile water. An oscillator was used to break up any aggregates. The spore concentration was determined using an improved Neubauer haemocytometer and then adjusted to a concentration of 2.5 × 10 8 spores/mL. For each experiment, this procedure was repeated to obtain a fresh suspension of spores. Chlorantraniliprole powder (96%) was provided by the Pesticide Science Group of the Institute of Plant Protection, Chinese Academy of Agricultural Sciences, Beijing, China. Preliminary trials showed that, at the doses tested (below) chlorantraniliprole did not significantly positively or negatively influence fungal growth on agar plates, nor did it influence fungal enzymes. Experimental procedures. Experiment I. Efficacy of chlorantraniliprole against L. migratoria under laboratory conditions. The leaf-dip bioassay method described by Shelton et al. and Liang et al. 72,73 was adopted for the toxicity bioassay of chlorantraniliprole. Wheat sprout bundles (3-cm diameter) were cut near the roots and dipped in various concentrations of chlorantraniliprole (shown in Table 3) prepared with distilled water. A bundle dipped in distilled water was used as a control. Each bundle was dipped for 10 s and allowed to air dry at room temperature, and three replicates were performed for each concentration analysed. Each bundle was placed inside a separate plastic container (30 × 22 × 9 cm) in which 30 third-instar nymphs were confined. Nymph mortality was recorded at 24-h intervals for 6 d. Nymphs were recorded as dead if they did not move when probed with a camel-hair brush. L. migratoria were fed only wheat sprouts treated with chlorantraniliprole for 1 d; on the other 5 d they were fed fresh, non-contaminated wheat sprouts. We used a similar protocol to obtain insecticide-challenged locusts (chlorantraniliprole, 5 mg/L, 0.5 mg/L and 0.05 mg/L) for our enzyme studies. Any surviving nymphs were used for the enzyme activity analyses (see below).
Experiment II. Efficacy of M. anisopliae against L. migratoria under laboratory conditions. Five conidial suspensions were prepared at various concentrations in sterile water containing 0.1% Tween 80 (shown in Table 3). Sterile water containing 0.1% Tween 80 was used as a control. Third-instar nymphs were used for all experiments. A total of 30 third-instar nymphs were treated per conidial concentration, with 2 mL of each conidial suspension applied as a spray under a Potter Precision Spray Tower (Burkard Manufacturing, Rickmansworth, UK). Three replicates each of five concentrations were tested. After being subjected to the conidial suspension spray treatment, nymphs were confined in each container and fed fresh wheat sprouts. Mortality was recorded at 24-h intervals for 6 d; nymphs were recorded as dead if they did not move when probed with a camel-hair brush. We used a similar protocol to obtain M. anisopliae-challenged locusts (M. anisopliae, 2.5 × 10 8 , 2.5 × 10 7 and 2.5 × 10 6 spores/mL) for our enzyme studies. Any surviving nymphs were used for the enzyme activity analyses (see below).     Table 3). After being sprayed with a conidial suspension, nymphs were confined in each container and fed fresh wheat sprouts for 3 d. On the 4th day, the nymphs were fed wheat sprouts treated with various concentrations of chlorantraniliprole (shown in Table 3, Treatment 4), and for the following 2 d the nymphs were fed fresh wheat. Mortality, which was assessed as described above for earlier experiments, was recorded at 24-h intervals for 6 d. Enzyme activity assays. The activities of ESTs were assayed using a modification of the method described by Han et al. 74 . Enzyme activity was determined by kinetic analysis using a microplate reader (Molecular Devices, LA, CA, USA), with 100 μ L of 1-naphthyl acetate solution (10 mM), 100 μ L Fast Blue RR salt (1 mM) and 90 μ L PBS placed in each microplate well. The reaction was initiated by the addition of 10 μ L of enzyme solution.
Optical densities (ODs) at 450 nm were recorded at 25-s intervals for 10 min. All reactions were carried out at 27 °C. Enzyme activities were calculated as the rate of absorbance change per mg protein (mOD/min/mg). GST activity was measured using a modification of the method described by Oppenoorth and Welling 75 . After pipetting 100 μ L of 1-chloro-2,4-dinitrobenzene (CDNB) (20 mM) or 3,4-dichloronitrobenzene (DCNB) (40 mM) and 100 μ L of GSH (40 mM) into microplate wells, we added 50 μ L of enzyme solution (for DCNB) or 10 μ L of enzyme solution and 90 μ L of PBS (for CDNB). The OD values at 340 nm were recorded at 25-s intervals for 10 min. Enzyme activities were calculated as the rate of absorbance change per mg protein (mOD/min/mg).
MFO activity was assayed using a method modified from Hansen and Hodgson 76 . After pipetting 100 μ L of p-nitroanisole (2 mM) and 50 μ L of nicotinamide adenine dinucleotide 2′ -phosphate reduced tetrasodium salt Scientific RepoRts | 6:28424 | DOI: 10.1038/srep28424 (9.6 mM) into microplate wells, we added 50 μ L of enzyme solution. The OD values at 405 nm were recorded at 25-s intervals for 10 min. Enzyme activities were calculated as the rate of absorbance change per mg protein (mOD/min/mg).
AChE activity was assayed using a modification of the method described by Han et al. 74 . After pipetting 100 μ L of 5,5′ -dithiobis-(2-nitrobenzoic acid) (45 μ M), 100 μ L of acetylthiocholine iodide and 90 μ L of PBS into microplate wells, we added 50 μ L of enzyme solution. The OD values at 405 nm were recorded at 30-s intervals for 40 min. Enzyme activities were calculated as the rate of absorbance change per mg protein (mOD/min/mg).
PO activity was measured using a method modified from that described by Luo and Xue 77 . We pipetted 180 μ L of catechol (10 mM) and 20 μ L of enzyme solution into microplate wells and recorded the OD values at 420 nm at 1-min intervals for 1 h. Enzyme activities were calculated as the rate of absorbance change per mg protein (mOD/ min/mg).
SOD, POD and CAT activities were determined using commercial assay kits (Nanjing Jiancheng, Nanjing, China) according to the manufacturer's instructions. SOD, POD and CAT enzyme activities were measured in units of U/mg, U/mg and U/g, respectively.
CHI activity was measured by using a reducing-sugar assay. A 400-μ L reaction mixture containing 300 μ L of 1% (w/v) colloidal chitin (prepared based on the methods of Hsu and Lockwood 78 ) and 100 μ L of enzyme solution was incubated at 37 °C in an Eppendorf tube. After 4 h of incubation, the mixture was centrifuged at 8,000 × g for 5 min to precipitate the remaining chitin. We mixed 200 μ L of supernatant with 80 μ L of potassium tetraborate (0.8 M, pH 9.1). The reaction was terminated by boiling at 100 °C for 5 min, after which it was cooled to room temperature under running water. The mixture was mixed with 1.2 mL of dimethylamine borane (DMAB) (10 g DMAB diluted to 1,000 mL with distilled water and then diluted 10-fold with glacial acetic acid). This mixture was then incubated at 37 °C for 20 min and subsequently cooled to room temperature under running water. The absorbance was detected at 585 nm, with CHI activity measured as the change of absorbance per mg protein (Δ OD/min/mg).
A method modified from Tang et al. 44 was used to assay AA activity. A reaction mixture comprising 50 μ L enzyme solution, 50 μ L p-nitroacetanilide (1.2 mM in absolute ethanol) and 150 μ L PBS was incubated in water at 35 °C for 30 min. The reaction was terminated by boiling in water for 10 min, and the mixture was then centrifuged at 10,000 × g for 15 min. The p-nitroaniline released into the supernatant was measured at 405 nm, and boiled enzyme was used as a control. AA activity was expressed as the change of absorbance per mg protein (Δ OD/min/mg).
Protein assay. Sample protein concentrations were estimated by using the method described by Bradford 79 .
Bovine serum albumin was used for the calibration curve. Measurements were performed at 595 nm using a microplate reader with SoftMax Pro 6.1 software. Data Analysis. Co-toxicity coefficient values were calculated as described by Sun and Johnson 80 . Co-toxicity coefficients for the mixed formulation were calculated after calculating the LC 50 of each incorporated component. Calculations were performed using the following equations: (1) toxicity index of agent = (LC 50 of standard agent/LC 50 of supplied agent) × 100; (2) theoretical toxicity index of the mixed formulation = (toxicity index of agent 1 × percentage of agent 1 in the mixed formulation) + (toxicity index of agent 2 × percentage of agent 2 in the mixed formulation); (3) co-toxicity coefficient = (actual toxicity index of the mixed formulation/theoretical toxicity index of the mixed formulation) × 100.
A co-toxicity coefficient of 100 indicates that the effect of the mixture is identical to that predicted from the proportions of the two components. A co-toxicity coefficient significantly greater than 100 corresponds to a synergistic effect. In contrast, when a mixture is characterised by a co-toxicity coefficient less than 100, the effect of the mixture is antagonistic.
LC 50 values and concentration-mortality slopes for each bioassay were estimated via probit analysis 81 using POLO-PC software 82 . Significant differences among LC 50 values were determined on the basis of non-overlapping 95% confidence limits. In the figures, each symbol point represents average fold change values of five sub-replicates of enzyme activities compare with control. One-way ANOVAs were also used to analyse the activities of detoxification enzymes (ESTs, GST and MFO), protective enzymes (SOD, CAT and POD) and PO, AChE, CHI and AA in L. migratoria nymphs. Differences among means were compared by using the LSD test at P < 0.05.