The development of transgenic crop plants for managing pest populations represents one of the most significant developments in pest management in the last 40 years, and is likely to have profound effects on agricultural production of food and fiber. The primary impacts of this technology include the significant enhancement of long term productivity, higher quality and greater stability of US agricultural production, and insurance against the sporadic affects of severe pest damage otherwise not manageable with conventional pest management techniques. Additional advantages of genetically engineered plants include the reduced costs from not having to apply synthetic pesticides as well as minimizing the environmental impact and human health risks that arise from the use of non-selective neurotoxic insecticides. Because of current reliance on environmentally disruptive techniques for controlling crop pests, the development of transgenic plant varieties may help to improve the profitability and sustainability of U.S. agriculture.
Most known insecticidal proteins that have been tested for utility in transgenic plants are those derived from the soil bacterium, Bacillus thuringiensis (Bt). In their microbial form, Bt strains comprise the active agents in a variety of microbial insecticide products that have been used for over 50 years. These biological insecticides are most widely used for the control of many economically important lepidopteran, coleopteran, and dipteran pests (Höfte and Whitely 1989). Bt is a species of gram-positive bacteria that produces highly toxic crystal proteins that are specific to a narrow range of species (Knowles 1994). During sporulation the bacterium produces a proteinaceous crystalline inclusion, which consists of one or more proteins called endotoxins or insecticidal crystalline proteins (ICPs). Once ingested, the ICP is dissolved in the midgut of target pests, and the liberated protoxin is proteolytically activated to a toxic fragment. This toxin crosses the peritrophic membrane and binds to high-affinity receptors on the midgut epithelium (Van Rie et al. 1990, Gill et al. 1992, Knowles 1994). Cessation of feeding occurs minutes after ingestion of the toxin (Dulmage et al. 1978). Once bound, the protein inserts into the membrane, which causes an opening or pore to form, and cell death results from osmotic lysis. Cell lysis of the gut epithelium allows the contents of the gut to enter the hemoceol causing septicemia and subsequent larval death (Knowles 1994).
Although Bt insecticides have been used in spray application for more than 50 years organic crop production (Cannon 1993), they have not been widely adopted by growers in large-scale crop production because of their high cost, narrow spectrum of activity, rapid environmental degradation, and inconsistent control of pests (ILSI 1998). The usefulness of Bt for pest management has been enhanced by development of new strains and improved formulations. Additionally, recent advances in genetic engineering have allowed the insertion and expression of Bt toxins in major crop plants (Gasser and Fraley 1989, Koziel et al. 1993, Fischhoff 1996). Crop plants that express high levels of Bt toxins throughout the growing season circumvent many of the inadequacies of Bt-based microbial insecticides.
Several genes encoding Bt insecticidal proteins have been isolated and characterized from many different Bt strains. These Bt genes, called cry genes, have been categorized into families based on sequence homology and insecticidal spectrum (Höfte and Whitely 1989). The Bt insecticidal proteins available in commercial Bt formulations are those encoded by the cry1, cry2, cry3, and cry4 genes. In a general sense, cry1-encoded proteins are toxic to lepidopterans, cry2 proteins are toxic to lepdipterans and dipterans, cry3 proteins are toxic to coleopterans, and cry4 proteins are toxic to dipterans (Gill et al. 1992). The cry1-endcoded proteins are the most common in Bt strains, with at least 12 different Bt proteins identified to date that differ in their primary structure and in their potency to lepidopteran insects (Fischhoff 1996).
The ability to genetically alter plants has resulted from the convergence of two areas of molecular biology: 1) the ability to engineer a variety of plant species to contain and efficiently express heterologous genes and 2) the isolation and characterization of single genes that encode insecticidal proteins (Fischhoff 1996). To date, transgenic crop plants expressing insecticidal Bt toxins that have been registered for use in the U.S. include maize, cotton, and potato. The cry1Ac gene expressed in transgenic cotton is effective against the major lepidopteran pests of this crop (Perlak et al. 1990, Fischhoff 1996). The cry3 genes in transgenic potatoes have proven effective for control of the Colorado potato beetle, Leptinotarsa decemlineata (Say) (Fischhoff 1996). Similarly, the cry1Ab, cry1Ac, and cry9c genes in transgencic field corn provide effective and selective control of most major lepidopteran corn pests (Koziel et al. 1993, Fischhoff 1996, Jansens et al. 1997). In addition to those Bt crop plants already registered for commercial use in the U.S., there are a number of registrations currently pending or have been granted experimental use permits and others such as Bt rice, soybean, and sunflower are being developed for foreign markets.
Although genetically altered plants that produce their own protective pesticides provide an important new alternative to existing pest control technologies, there is considerable concern that that a large-scale introduction and wide-spread adoption of this technology will result in intense selection pressures that will rapidly lead to the development of resistance in pest populations. Insects possess a remarkable capacity to adapt to selective pressures. Pest adaptations are common and often are not noticed because they have no sudden or dramatic effect on crop production. However, in some instances pest adaptations reduce the effectiveness of specific management tactics and may result in crop loss, greater management expenses, and negative environmental side effects.
Exposure to synthetic insecticides has resulted in more than 4,000 instances of resistance in about 500 species of insects (Georghiou and Lagunes-Tejeda 1991). Resistance is defined as the evolved capacity of an organism to survive in response to a selective pressure from exposure to a pesticide. Insect populations are able to acquire resistance after repeated exposure to insecticides. It has been proposed that transgenic crops pose a serious threat to resistance development because they produce toxin throughout much, if not all, of the plant’s life (Gould et al. 1998). Constant long-term exposure of pest populations to Bt encourages survival of individual pests that are genetically resistant to the toxin. Compared to conventional insecticides, delivery of insecticidal compounds via genetically engineered crops greatly increases the duration and intensity of exposure to the selective agent and dramatically increases the risk of resistance development.
Genes that allow insects to survive exposure to certain toxins (resistance genes) are normally considered to exist within pest populations at very low frequencies before selection. However, the frequency of resistance genes prior to a control failure is generally not known and is difficult to estimate. In the absence of other data, initial resistance allele frequencies have been assumed to range from 10-2 (Georghiou and Taylor 1977) to 10-13 (Whitten and McKenzie 1982) based on theoretical assumptions about the equilibrium between mutation and selection (Roush and Daly 1990). These assumptions relate to a balance between the generation of new alleles (i.e., mutation rate), the selective disadvantage of heterozygous genotypes and the dominance of the resistance allele. Although some data exist for the fitness of resistant heterozygotes in the absence of pesticides and the dominance of resistance alleles, there is a relative absence of data concerning the rate of mutations. What is certain, however, is that the mutation of a susceptible allele (+) to the R allele occurs at low frequencies, and that in the early stages of resistance development, the majority of resistance alleles are present in heterozygotes (McKenzie 1996).
Bt resistance has been documented for several pest insects that have been repeatedly selected for resistance in the laboratory. In the specific case of Lepidoptera, laboratory selection has resulted in significant increases in LC50's (up to 1000-fold) for cabbage looper, Tricoplusia ni Hübner (Estada and Ferré 1992), diamondback moth, Plutella xylostella (L.) (Tabashnik et al. 1991), Indian meal moth, Plodia interpunctella (Hübner) (McGaughey and Whalon 1992), beet armyworm, Spodoptera exigua (Hübner) (Moar et al. 1995), tobacco budworm, Heliothis virescens (F.) (Gould et al. 1995), cotton leaf worm, Spodoptera littoralis Boisduval (Müller-Cohn et al. 1996) and European corn borer, Ostrinia nubilalis (Huang et al. 1997). The diamondback moth is notable as the only insect to evolve high levels of resistance in the field as a result of repeated use of formulated Bt (Tabashnik 1994). However, the Indian meal moth probably evolved low levels of resistance in stored grain treated with Bt (McGaughey and Whalon 1992).
The mechanisms of Bt resistance that have been reported in most cases involve some form of binding site modification. For insect species where adaptation results from altered toxin binding, reduced affinity of the Bt toxin for the brush border membrane of the midgut epithelium has been identified as a primary mechanism of resistance. Studies with radioactively labeled Cry1Ab showed that a 50-fold reduction in binding was correlated with a >100-fold reduction in toxicity of Cry1Ab in resistant versus susceptible strain of P. interpunctella (Van Rie et al. 1990). Similarly, a strain of P. xylostella from the Philippines showed >200-fold resistance to Cry1Ab accompanied by little or no binding of Cry1Ab toxin to the midgut epithelial membrane compared with a susceptible strain (Bravo et al. 1992a, Bravo et al. 1992b, Ferré et al. 1995). In contrast to the results for P. interpunctella and P. xylostella, two independent studies of H. virescens found no clear association between toxin binding and resistance to Cry1Ab or Cry1Ac (Gould et al. 1992, MacIntosh et al. 1991).
Perhaps the biggest concern regarding resistance development to Bt transgenic crops is that it could render some Bt insecticide formulations ineffective in areas where resistance has developed (Mellon 1998). Because microbial insecticides occupy a small percentage of the insecticide market, the loss of microbial Bt insecticides would have little effect on most agricultural production. However, if resistance to a broad spectrum of Bt endotoxins does develop, members of the organic farming industry could lose an important insect control option (ILSI 1998). It should be recognized that organic farmers have few “organically certifiable” insecticides, and the loss of Bt products would represent a serious loss in some cases. The possibility of replacing Bt with a new certifiable insecticide is very limited and loss of Bt insecticides may have implications for the organic farmer well into the future. However, generalizing the potential impact of resistance development is problematic because a specific impact depends greatly on the insect affected, Bt gene used, organic crops being grown, and proximity of the organic crop (ILSI 1998).
The development of insecticide resistance among pest populations can best be described as an evolutionary response to a selective pressure. Unlike most evolutionary phenomena, however, insecticide resistance has practical and economic significance (Denholm and Rowland 1992). The evolution of resistance in pest species poses a formidable challenge in view of current difficulties in discovering and, in particular, developing pest control technologies, including transgenic crops. There is an increasing recognition that resistance is an unavoidable consequence of any pest management strategy, and that unless appropriate measures are taken to delay the onset of resistance, a given technology is likely to have limited market life.
Because evolution of resistance poses a threat to effective pest management, the proactive development and implementation of resistance management (RM) principles for transgenic crops combined with education of growers, consultants, and extension educators presents a new challenge to the agricultural community (ILSI 1998). Although RM has been conceptually discussed for at least two decades, there have been few attempts to implement a comprehensive, science-based approach to RM techniques. An effective resistance management program must consider numerous factors, including characteristics of the Bt product, target-pest biology, the pest complex, and crop systems. Among the many tactics proposed for resistance management, spatial refuges (where pests are not exposed to insecticides) combined with high dose expression of toxin has received the most attention (ILSI 1998). This approach provides a refuge from exposure to the Bt crop allowing survival of susceptible insects, which decreases the intensity of selection for resistance. Ideally, relatively large numbers of susceptible insects from refuges survive and mate with very few resistant insects that survive on transgenic crops. The other component of this strategy, that is “high-dose,” requires that the transgenic insecticidal crop kill essentially all susceptible homozygotes (SS) and a very high percentage of heterozygotes (RS), leaving only the very rare resistant homozygotes to survive (Tabashnik 1994). The high dose of toxin expressed in transgenic plants ensures that the resistance is inherited as a functionally recessive trait (i.e., RS individuals susceptible to the level of expression in the transgenic plant). The refuge provides large numbers of SS individuals to mate with the rate RR insect survivors from the transgenic crop (Alstat and Andow 1995, Roush 1996, Andow and Hutchison 1998, Gould 1998).
The development and deployment of effective resistance management programs must acknowledge the complexity of resistance evolution and limitations of current knowledge and experience. Efforts to avoid rapid pest adaptation to Bt crops will require that decisions be made with less than comprehensive data and theory. As experimental and survey data accumulates, it will be possible to test current assumptions about RM strategies and to develop more robust resistance management plans. However, long-term predictions regarding resistance evolution will always be subject to uncertainty, and therefore, target insects should be monitored for unexpected ecological and genetic changes so that RM plans can be modified to deal with such changes (Gould et al. 1998).