Abstract: Natural variability in susceptibility of field collected rice yellow stem borer (Scirpophaga incertulas) and rice leaffolder (Cnaphalocrocis medinalis) to Bacillus thuringiensis (Bt) δ-endotoxins was investigated to establish a geographic baseline for comparison of future population responses to the increased use of Bt based insect control products like Bt transgenic rice and commercial formulations. Populations of S. incertulas and C. medinalis were evaluated for their susceptibility to purified Bt Cry1Aa, Cry1Ab, Cry1Ac, Cry1C and Cry2A toxins. The range of LC50's among selected populations in response to variety of different Bt toxins were between 19-88 fold for S. incertulas and 2-20 fold for C. medinalis. The data provide a strong basis for monitoring changes in susceptibility of S. incertulas and C. medinalis populations to future use of Bt toxins in Bt rice program.
Introduction
The development of insecticides based on Bacillus thuringiensis δ-endotoxin proteins has increased in response to the need of efficacious, environmentally safe and selective pesticides with unique modes of action. Advances in Bt formulation technology and genetic engineering and the discovery of Bt strains with a broader spectrum of activity, have resulted in new microbial products with increased potency and greater stability (Khan et al., 1995a, b). In addition, genetically modified crops such as cotton, corn and potato have already been commercialized (James and Krattiger, 1996). The advantages of genetically engineered plants include more efficient delivery of active ingredient to the target pest, less input in terms of the costs of pesticide application and reduced environmental impact from pesticide use. Although, genetically altered plants producing their own protective insecticides provide an exciting new approach to insect control, a large scale introduction of these crops could rapidly lead to the development of resistance to Bt toxins within pest populations (Tabashnik, 1994). The combined impact of transgenic rice and increased use of Bt formulation for rice lepidopteran pests management increase the likelihood for resistance development which would negate the economic and environmental benefits of this important management option. Once commercially available, these new microbial and plant products will play an important role in insect management (Karim and Riazuddin, 1997).
Selection for rice yellow stem borer (Scirpophaga incertulas) (Lepidoptera: Pyralidae) and rice leaffolder (Cnaphalocrocis medinalis) (Lepidoptera: Pyralidae) to Bt toxins is expected to be intense and is likely to result in the evolution of resistance to Bt endotoxins (Karim et al., 1999a). An effective resistance management program will also be needed to foresee the long-term utility of Bt technology (Karim and Riazuddin, 1999b). The US Environmental Protection Agency (EPA) has approved conditional registrations for several Bt crops and is requiring the development of scientifically sound resistance management strategy by the year 2001 (Mellon and Rissler, 1998). The currently favored resistance management strategy for Bt crops is the "high dose/refuge strategy" (Gould, 1994, 1998).
S. incertulas and C. medinalis are important economic targets for many Bt based products but variation in susceptibility among populations has not been investigated. One of the most important research needs currently in Bt resistance management programs is the development of baseline susceptibility studies of key insect pests in all ecosystems where Bt will be used (Whalon, 1992). The present study reports the results from experiments to determine if differences exist in susceptibility to Cry toxins among S. incertulas and C. medinalis field populations collected from Basmati rice growing area (Kalar Tract) in Punjab province of Pakistan.
Materials and Methods
S. incertulas and C. medinalis were collected from 4 administrative districts representing traditional Basmati rice growing areas (Kalar Tract) of Punjab, Pakistan (Fig. 1, Table 1). Field collections were made from booting stage of local Basmati rice cultivars. Bt has never been used for insect control in these areas (Karim and Riazuddin, 1999a). Therefore, these populations had no previous Bt exposure. Adult female moths of S. incertulas settled on top of rice plants were captured in the early morning hours. Moths were immediately transferred to TN1 rice plants in plastic cages. Moths were oviposit for one night and egg masses were cut from rice leaves. Egg masses were kept in scintillation vials and incubated in the Insectary at 28 ± 2°C (Karim et al., 1994) for hatching of larvae. Adult moths of C. medinalis were collected by netting and infested on 40 days old TN1 rice plants in their vegetative stage covered with plastic cages (Karim et al., 1999a). Collections of C. medinalis moths were carried out in bright sunlight. Larvae infested on rice plants were kept in the Insectary for egg laying. Eggs were allowed to hatch on the rice plants. Second instar larvae of C. medinalis maintained on TN 1 plants and neonatal stage of S. incertulas were used for bioassay studies. The Cry1Ac, Cry1Ab, Cry1Ac, Cry1C and Cry2A proteins were obtained as recombinant proteins expressed in Escherichia coli (Karim et al., 1999b). Bt δ-endotoxins were purified from E. coli as documented (Karim et al., 1999b). Inclusion bodies from Cry1Aa, Cry1Ab, Cry1Ac, Cry1C and Cry2A were solubilized in Alkalic buffer (50 mM Sodium carbonate, 10 mM Dithiothreitol, pH 9.5) for 2 hrs at 37°C. The solubilized protoxin of Cry2A was dialyzed against 50 mM sodium carbonate buffer (pH 10.5). The purity of protoxins and toxins was examined by Sodium dodecyl sulfate 10% Polyacrylamide gel electrophoresis (Laemmli, 1970). Protein concentrations of the protoxins and toxins were determined by the Bradford method (Bradford, 1976).
SDS-PAGE was carried out by the method reported by Laemmli (1970). Toxins were run on 10% Polyacrylamide gel and stained with Coomassie brilliant blue stain (0.25% Coomassie brilliant blue, 45.5% methanol and 9% Glacial acetic acid) at 65°C and destained at the same temperature (Fig. 2).
'TN1' rice seedlings, 18-25 days old were used for the toxicity assays of Cnaphalocrocis medinalis larvae. Roots of five tillers were wrapped in wet cotton to prevent drying. The leaves of rice seedlings were initially dipped in 0.02% Triton x-100 and then rinsed in water for several times. After air drying, seedlings were dipped in different dilutions of trypsin activated Bt Cry1Aa, Cry1Ab, Cry1Ac, Cry1C and Cry2A toxins. Seedlings were dried under ceiling fan and were placed in plastic petriplates lined with moist filter paper to keep the plant materials fresh. These petriplates were then infested with 2nd instar Cnaphalocrocis medinalis larvae. In case of Scirpophaga incertulas fresh 'TN1' rice stems of equal length were inoculated with different dilutions of Bt toxins and were placed in petri plates. The neonatal larvae were allowed to bore in the inoculated stems and incubated at 28±2°C and 65% humidity. Untreated stems and seedlings were used as control. Mortality was recorded after 3 days.
Data generated through biotoxicity assays of S. incertulas and C. medinalis were statistically analysed through Probit analysis using computer programme Quantal software (LeOra Software, 1987). IRRISTAT software was used to estimate ANOVA and correlation for individual toxin and insect strain effect (IRRI, 1997).
Results
Insect population sensitivity was investigated to Bacillus thuringiensis Cry1Aa, Cry1Ab, Cry1Ac, Cry1C and Cry2A toxins. Figure 2 shows a 10% SDS-PAGE containing the trypsin activated Cry toxins used in these studies. The significant difference was found in sensitivity of populations of S. incertulas and C. medinalis to Cry toxins. However, no obvious geographic trends in susceptibility for S. incertulas and C. medinalis were observed (Table 2-6), nor was the level of population susceptibility to one Bt toxin consistently predictive of the toxins relative susceptibility to the other.
The sensitivity of S. incertulas populations to Bt Cry1Aa, Cry1Ab, Cry1Ac, Cry1C and Cry2A toxin proteins is mentioned in Table 2-6. LC50s ranged from 31.13 ng/ml (Sheikupura) to 1633 ng/ml (Narowal); this difference represented an 52-fold in susceptibility to Cry1Aa toxin (Table 2). The range (19-23 fold) in sensitivity to Cry1Ab, Cry1Ac and Cry2A toxins was narrower than Cry1Aa and Cry1C toxins. Sialkot population was found most susceptible to Cry1Ac and Cry2A toxins with LC50 values 28.09 and 33.83 ng/ml respectively. Narowal population of S. incertulas was least susceptible to all toxins. Cry1C toxin was found the most potent and also high variation in its toxicity. ANOVA on the basis of LC50 revealed significant variation among S. incertulas populations for their susceptibility (F 1.980 > Prob 0.162) and among Bt toxins for their efficacy ( 33.079 > 4.403) (Table 7-8). Results were evaluated at 0.05, 1-tail with critical value +/- 0.8221. Variation in susceptibility to purified toxins was comparatively less than that of S. incertulas populations from Kalar tract in Punjab. LC50 value of Cry1Aa for C. medinalis ranged 3 fold from 31.37 ng/ml (Narowal) to 112.58 ng/ml (Sheikupura) (Table 2). Cry1C toxin found most toxic to C. medinalis with LC50 value of 8.74 ng/ml and as high range of 21 fold with 175.13 ng/ml for Sialkot population. Cry1C toxin also found the most potent and also highly variable among C. medinalis populations. ANOVA on the basis of LC50 also revealed significant variation among C. medinalis populations For their susceptibility (F 1.708 > Prob 0.218) and among Bt toxins for their efficacy (F 0.821 > Prob 0.536)(Table 9-10). Correlation among different populations for their behavior towards Bt d-endotoxins was also evaluated at 0.05, 1-tail with critical value +/- 0.8221.
Discussion
Significant differences in susceptibility to Bacillus thuringiensis δ-endotoxins exist among the S. incertulas and C. medinalis populations tested in the present study. Briese (1981), Georghiou (1988) and Rossiter et al. (1990) suggested that, if the natural variation in susceptibility within a population is heritable, resistance can develop. Because of the extensive resistance of insects to chemical insecticides monitoring for the potential development of resistance to Bt is crucial. The results of present investigation indicate that similar natural variation exists among S. incertulas and C. medinalis and may be responsible for differences in susceptibility unrelated to prior exposure of Bt toxin.
The difference in susceptibility among S. incertulas and C. medinalis populations was almost 3-fold minimum and 88-fold maximum (Table 2-6). Correlation coefficient provided a qualitative evaluation for the behavior of insect population towards Bt toxins (Table 7-10) in relation to each other. These populations were collected from an area not known to have received prior exposure to Bt and probably do not represent a resistant population. Surveys of susceptibility to Bt in other pest species have shown that considerable inter-population variation exists reviewed by Tabashnik (1994). Although, we have not done an extensive sampling of target pests, variation has been reported in other Lepidoptera including the diamond back moth, Plutella xylostella (Tabashnik et al., 1990), Indian meal moth, Plodia interpunctella (Hubner) (Pyralidae) (Kinsinger and McGaughey, 1979), cotton bollworm, Heliothis virescens (F.) (Noctuidae), corn earworm, Helicoverpa zea (Boddie) (Noctuidae) (Stone and Sims, 1993) and gypsy moth, Lymantria dispar L. (Lymantridae) (Rossiter et al., 1990), European Corn borer, Ostrinia nubilalis (Pyralidae) (Siegfried et al., 1995) and Spruce budworm, Christoneura fumiferana (Clemens) (Tortricidae) (Van Frankenhuyzen et al., 1995). Such differences in tolerance are probably the result of natural variation among the populations (Robertson et al., 1995) and unrelated to prior exposure to the Bt based insecticide.
The results obtained from a limited number of S. incertulas and C. medinalis populations sampled suggest that variability in susceptibility to Bt may exist among geographically distinct populations (Table 2-6). However, large unexplained variations in susceptibility of field populations have been observed (Sawicki, 1987; Ffrench-Constant and Roush, 1990). Further investigation is required to establish that differences in susceptibility are genetically controlled and not the result of other unidentified factors. We recommend that stably transformed Bt rice be used as standards for comparison with purified proteins.
The data provide an important baseline information on the susceptibility of S. incertulas and C. medinalis to variety of different purified Cry toxins. However, the relationship between population susceptibility to Bt and the potential for field resistance remains to be established.
| Table 1: | Scirpophaga incertulas and Cnaphalocrocis medinalis colonies tested for susceptibility to purified Bt δ-endotoxins from rice Basmati growing area in Pakistan |
| Table 2: | Susceptibility of field collected S. Incertulas and C. medinalis to Bt Cry1Aa δ-endotoxin |
| Table 3: | Susceptibility of field collected S. Incertulas and C. medinalis to Bt Cry1Ab δ-endotoxin |
| Table 4: | Susceptibility of field collected S. Incertulas and C. medinalis to Bt Cry1Ac δ-endotoxin |
| Table 5: | Susceptibility of field collected S. Incertulas and C. medinalis to Bt Cry1C δ-endotoxin |
| Table 6: | Susceptibility of field collected S. incertulas and C. medinalis to Bt Cry2A δ-endotoxin |
| Table 7: | ANOVA for the Variation in the susceptibility among S. incertulas populations to different Bt δ-endotoxins and among the toxicity of individual Bt toxins |
| Table 8: | Correlation among the susceptibility of S. incertulas populations to different Bt δ-endotoxins |
Further monitoring for potential resistance genes in S. incertulas and C. medinalis populations would facilitate the establishment of a discriminating dose based on the frequency of resistance alleles in the populations. This information is critical to the development of an effective resistance monitoring program and implementation of resistance management strategies. However, it is important to document the extent of this variability throughout the known range of S. incertulas and C. medinalis to establish a diagnostic Bt concentration or dose that could be used in more extensive resistance monitoring program. Further, a baseline susceptibility against a variety of Bt toxins, as reported here, provides a basis for determining if resistance is developing as a result of increased exposure to Bt toxins as might be expected with increased use of Bt based products.
| Table 9: | ANOVA for the Variation in susceptibility among C. medinalis populations to different Bt δ-endotoxins and among the toxicity of individual Bt toxins |
| Table 10: | Correlation among the susceptibility of C. medinalis populations to different Bt δ-endotoxins |
| Fig. 1: | Map of collection sites in Basmati growing area of Pakistan |
| Fig. 2: | Commassie Blue-stained 10% SDS-PAGE, Lane 1: high molecular weight protein markers, land 2-5: proteolytically activated Cry1Aa, Cry1Ac and Cry1C, lane 6: solubilized Cry2A toxin |
Such information is essential to develop resistance management strategies designed to maintain the efficacy of these environmentally sound and economically important management options.
In lab, greenhouse and field assays (Anwar et al., 1999; Altosaar et al., 1999) rice genetically modified to contain a Bt δ-endotoxin protein has consistently demonstrated efficacy against both S. incertulas and C. medinalis. These initial trials indicate that genetically modified Bt rice will be effective throughout rice growing countries. The development of strategies for implementing genetically modified Bt rice within integrated pest management offers an exciting challenge to pest management specialists and a considerable potential benefit to rice growers. We analyzed natural variation to variety of different Bt toxins in field collected insects of S. incertulas and C. medinalis. These results suggest that natural tolerance to Bt toxins is present in four different populations of both insects towards Bt toxins. Therefore, it is suggested that high dose/refuge strategy should be evaluated on the basis of natural variations among distinct populations of insect pests. Now, a question arises, if natural variation among pest populations are present, then what would be the range of high dose? Does it vary from locality to locality for same pest? These results could have important implications for resistance management of rice pests in Bt rice. The currently proposed resistance management strategy for Bt rice, the high dose/refuge strategy (Cohen et al., 1999) requires a) plant tissue express enough Bt toxin to kill all heterozygotes, b) the resistance alleles be very rare and c) susceptible insects are within an effective mating distance of resistant insects. The high dose/refuge strategy would not be useful for resistance management if high natural tolerance exit among different populations of insect pests. However, at higher dosages all individuals are expected to be susceptible. The practical importance of this natural variation will depend on whether these insects can survive on Bt rice. In presence of natural variations, the high dose/refuge strategy is subject to a number of stringent prerequisites that may be difficult to meet in practice. More robust resistance management options are imperative.
Acknowledgements
We owe special thanks to Dr. Khudadad Khan, Institute of Statistics, University of the Punjab for his help in statistical analysis. We thank Mr. Tariq Mahmood, Mr. Ali Raza Zaidi, Mr. Muhammad Munir, Mr. Abdul Karim and Mr. Muhammad Kashif for their technical assistance in the lab. The Rockefeller Foundation financially supported research work.