[21I Bacillus thuringiensis in Biological Control B. A. FEDERICI Department of Entomology and Interdepartmental Graduate Programs in Genetics and Microbiology University of California Riverside, California
Whereas it is widely known in the pest control field that B.t. kills insects through the action of protein toxins, specific knowledge of its complexity, diversity, and mode of action as well as its potential for further development is often lacking. Due to the present and growing importance of B.t., my purpose in this chapter therefore is to provide an overview of the basic biology of B.t. and how it is being used in biological control, both in the form of bacterial insecticides and transgenic plants. The chapter is designed primarily to provide information for non-B.t, specialists. After a general discussion of the basic biology and systematics of B.t., I focus on the mode of action of the key types of insecticidal proteins and the composition of the most common isolates used in commercial formulations of bacterial insecticides. I then cover the use of selected B.t. proteins in recombinant bacteria and transgenic plants, and close with short sections on resistance and the prospects for further use of B.t. in the future.
The insecticidal bacterium Bacillus thuringiensis (B.t.) is the most successful commercial biological control agent of insect pests. Moreover, aside from the parasites and predators that have been effective in classical biological programs, it is the most widely used biological control agent in the world. B.t. is currently used in many countries to suppress numerous lepidopteran and coleopteran pests of forests, vegetables, and field crops; and to control the larvae of many species of vector and nuisance mosquitoes and blackflies. The principal reasons for the success of B.t. include the high efficacy of its insecticidal proteins, the existence of a diversity of proteins that are effective against a range of important pests, its relative safety to nontarget insect predators and parasites, its ease of mass production at a relatively low cost, and its adaptability to conventional formulation and application technology. In addition, knowledge of B.t. and plant molecular biology, recombinant DNA technology, and plant transformation techniques have been used to develop insect-resistant B.t.-transgenic varieties of major crops such as cotton, corn, potatoes, and rice, creating a new multibillion dollar industry. New types of insecticidal B.t. proteins are being discovered routinely, and other advances in knowledge about B.t. proteins and plant molecular biology indicate the use of these proteins will be extended in the future to control a greater range of invertebrate pests. Thus, among the various types of insect pathogens, parasites, and predators, B.t. has excellent prospects for more widespread use in future insect control programs. An advantage of these programs is that the successful use of B.t. markedly reduces the use of synthetic chemical insecticides.
GENERAL BIOLOGY OF BACILLUS THURINGIENSIS B.t. is a spore-forming bacterium that can be readily isolated on simple media such as nutrient agar from a variety of habitats including soil, water, plants, grain dust, dead insects, and insect feces. Compared with other insect pathogens, its life cycle is simple. When nutrients are sufficient for growth, the spore germinates producing a vegetative cell that grows and reproduces by binary fission. The bacterium continues to multiply until one or more nutrients, such as sugars, certain amino acids, and oxygen, become insufficient for continued vegetative growth. Under these
Copyright 9 1999 by Academic Press. All rights of reproduction in any form reserved.
conditions, the bacterium sporulates producing a spore and parasporal body, the latter composed primarily of insecticidal protein toxins (Fig. 1). Though B.t. can be isolated from many environmental sources and is typically referred to as a "soil bacterium," it has several features indicating that its principal ecological niche is insects (Federici, 1993; Meadows, 1993). The original isolations of B.t., for example, were made from diseased caterpillars (Ishiwata, 1901; Berliner, 1915), and these remain a good source of isolates. More importantly, B.t. produces a range of toxins and toxin synergists that are very effective at killing certain species of insects, especially larvae of lepidopterous insects, providing a rich substrate for B.t. reproduction. The principal toxins are the Cry and Cyt protein 6-endotoxins, but many isolates of B.t. also produce the /3-exotoxin, the synergist zwittermycin, and enzymes such as phospholipases that enhance the activity
FIGURE 1 Sporulatingcells and endotoxin-containingparasporal bodies of Bacillus thuringiensis. (A) Phase-contrast micrograph of a wet-mount preparation of sporulating B.t. cells illustrating ovoidal spores and adjacent parasporal bodies (arrowheads). (B) Scanning electron micrograph of the bipyramidal and cuboidal endotoxin inclusions characteristic of the HD1 isolate of B. thuringiensis subsp, kurstaki (H3a3b). Commercially,this is the most successful strain of B.t. with more than 100 formulations marketed in different countries around the world to control caterpillar pests. (C) Transmission electron micrograph of the parasporal body of the ONR60A isolate of B. thuringiensis subsp, israelensis (H14) used widely to control the larvae of mosquitoes and blackflies. Each crystalline inclusion contains a different endotoxin. The arrowheadspoint to the parasporal envelope that holds the inclusions together.
of the 6-endotoxins. In addition, the spore itself can synergize the activity of Cry proteins in some insects. In many insect species, especially grain-feeding lepidopterans, the bacterium reproduces to very high levels after insect death, with millions of spores being produced per cadaver (Fig. 2). Thus, B.t. ecology and reproductive biology suggest that its toxins and toxin synergists evolved to debilitate or kill directly a range of insect species, thereby providing a substrate for reproduction of this bacterium (Federici, 1993). An interesting aspect of B.t. general biology is that unlike most other insect pathogens, with the possible exception of certain grain-feeding lepidopterans, B.t. does not cause natural epizootics in insect populations.
Bacillus thuringiensis as a Bacterial Species The species B.t. was first isolated from diseased larvae of the silkworm, B o m b y x mori, in Japan by Ishiwata (1901). It was not officially described, however, until it was reisolated by Berliner (1915) from diseased larvae of the Mediterranean flour moth, Anagasta kuehniella, in Thuringia, Germany, hence the derivation of species name thurin-
FIGURE 2 Occurrence of Bacillus thuringiensis in larvae of the navel orangeworm, Amyelois transitella. The top panel shows larvae killed by B. thuringiensis subsp, aizawai (H7) during an epizootic in a laboratory colony fed on grain (wheat). The bottom panel shows a pure culture of the B. thuringiensis subsp, aizawai strain isolated from a small piece of fat body from a dead larva. These larvae contain as many as 108 spores per cadaver.
Bacillus thuringiensis in Biological Control
giensis. Though commonly referred to in the singular as
B.t., B. t h u r i n g i e n s i s as currently recognized is actually a complex of subspecies, all of which are characterized by the production of a parasporal body during sporulation. The parasporal body is the principal characteristic used to differentiate this species from the closely related species, B. cereus, and other bacilli. This parasporal body contains one or more proteins, typically as crystalline inclusions, and most of these are highly toxic to one or more species of insects (Fig. 3). The toxins are known as endotoxins, and occur in the parasporal body as protoxins, which after ingestion dissolve and are converted to active toxins through cleavage by proteolytic enzymes in the insect gut. The activated toxins bind to the midgut microvillar membrane in sensitive insects, lyse the cells, and destroy much of the midgut epithelium, causing insect death. Thus, B.t. endotoxins are stomach poisons selective for insects and certain other invertebrates. At present, there are more than 60 subspecies of B.t. (Table 1), distinguished from one another on the basis of immunological differences in flagellar (H antigen) serotype (de Barjac & Franchon, 1990). Each subspecific name corresponds with a specific H antigen number. For example, B. t h u r i n g i e n s i s subspecies k u r s t a k i is H3a3b,
whereas B. t h u r i n g i e n s i s subspecies m o r r i s o n i is H8a8b. In the literature, the term "variety" is also used in place of subspecies, as is occasionally the term "strain." Because the H antigen serotype-subspecific name often does not correlate with insecticidal properties, acronyms and numbers are often used to designate specific isolates, especially those with important insecticidal properties. For example, HD1 (isolate number 1 from Howard Dulmage) is used to designate a specific isolate of B. t h u r i n g i e n s i s subspecies kurstaki (H3a3b) that produces four major endotoxin proteins and has a broad spectrum of activity against lepidopteran pests. Another isolate of B. t h u r i n g i e n s i s subsp, k u r s t a k i (H3a3b) is HD73. This produces only a single endotoxin protein and as a result has a much narrower spectrum of activity against insects than does HD1. Historically, HD1 was the first B.t. isolate developed commercially for the control of lepidopterous pests, and remains the most widely used in commercial products today. This isolate has also been the source of much of our knowledge of B.t. genetics and molecular biology as well as endotoxin genes used in transgenic bacteria and plants. Despite the large number of B.t. subspecies that have been described, the taxonomic validity of these as well as maintaining B.t. as a species separate from B. c e r e u s has been in question for many years. At the species level, the
H antigen b 1 2
3a 3a3b 4a4b 4a4c 5a5b 5a5c 6 7
8a8b 8a8c 8b8d FIGURE 3 Endotoxin complexity in a parasporal body of Bacillus thuringiensis. (a) and (b) show, respectively, scanning and transmission electron micrographs of the parasporal body of B. thuringiensis subsp, morrisoni (H8a8b), isolate PG-14, a mosquitocidal isolate of this subspecies. In (c 1), the protein composition of B. thuringiensis subspecies israelensis parasporal body is compared with that of the PG-14 parsporal body (c2). Note that the latter isolate contains the four major proteins characteristic of most mosquitocidal isolates of B.t. (Cry4A, 135 kDa; Cry4B, 128 kDa; CryllA 65 kDa; and CytlA, 27 kDa), and in addition a protein of 144 kDa. The four mosquitocidalproteins are encoded by a large transmissible plasmid of 125 kb, whereas the 144-kDa protein is encoded on a different plasmid. Isolates of B.t. with a broad spectrum of activity against insect species typcially contain four or more endotoxin proteins in the parasporal body.
9 10 11a 11b 11a 11c
Subspecies/Serovarieties of Bacillus thuringiensis a Subspecies/ serovarietyC, ar thuringiensis finitimus alesti kurstaki sotto kenyae galleriae e
canadensis entomocidus aizawai morrisoni ostriniae nig e riensis tolworthi darmstadiensis toumanoffi kyushuensis thompsoni
H antigen 13 14
15 16 17 18 19
20a20b 20a20c 21
22 23 24 25 26 27f
Subspecies/ serovariety pakistani israelensis dakota indiana tohokuensis kumamotoensis tochigiensis yunnanensis pondicheriensis colermi shandongiensis japonensis ne oleonensis coreanensis silo mexicanensis
aFrom de Barjac, H. & Franchon, E. (1990). Entomophaga, 35, 233240. bFlagellar antigen used to prepare antibodies. CAlso referred to in the literature as "variety" and occasionally "strain." dH antigens and subspecies in bold indicate they are used in commercial products. eUsed in the former Soviet Union and People's Republic of China. fSee Thiery & Frachon (1997) for a more complete list.
primary problem is that the only phenotypic character that clearly differentiates B. cereus from B.t. is the parasporal body synthesized by the latter species during sporulation (Baumann et al., 1987). The information for parasporal body production is encoded on large transmissible plasmids in most B.t. subspecies [see Hofte and Whiteley (1989) for review]. When these plasmids are lost naturally or are cured from B.t. subspecies by growing cells at 42~ no parasporal body is produced. Cured B.t. strains that hosted these plasmids cannot be reliably distinguished from B. cereus. At the subspecies level, phenotypic biochemical differences among many are minor, indicating that many subspecies may not be valid using accepted standards of differentiating bacterial subspecies. In other cases, the differences in biochemical properties and insecticidal activity are so significant that some subspecies could be viewed as distinct species. Molecular S y s t e m a t i c s of B. thuringiensis Attempts to clarify the relationships between B.t. and B. cereus, and among subspecies of B.t., have used the meth-
ods of molecular systematics. These studies have provided strong evidence that B.t. and B. cereus are so closely related that they could be considered the same species. For example, in a study of small subunit ribosomal RNA sequences, Ash et al. (1991) found that B. cereus and B.t. fell into the same clade, and were much more closely related to each other than to 48 of the 51 Bacillus species they studied. Even more definitive evidence for the close relationship of B.t. and B. cereus was found by Carlson and Kolsto (1993), who used D N A - D N A hybridization techniques and chromosome mapping to study the relationship of these species. Under stringent hybridization conditions they found that 19 random B. cereus probes hybridized strongly to B. t. subsp. thuringiensis (HD2) fragments, but did not hybridize with or gave only weak signals to fragments of B. subtilis under conditions of low stringency. Moreover, they found that when the NotI chromosomal map of B. t. subsp, thuringiensis was compared with that of four different strains of B. cereus, the B.t. map was more similar to two of the B. cereus strains than these were to the two other B. cereus strains. With respect to the relationships among different subspecies, Carlson and Kolsto (1993) also compared the NotI chromosomal fragment patterns of 10 different B.t. subspecies and found that they differed markedly. In another study using DNA fragment patterns, Priest et al. (1994) found considerable variation within and among the restriction fragment length polymorphisms (RFLPs) of ribosomal gene clusters of several different B.t. subspecies. Yet they also found that in subspecies B. thuringiensis subsp, israelensis and B. thuringiensis subsp, aizawai, the RFLPs for each were so unique and characteristic among different isolates that they could be used to assign isolates to these subspe-
cies. Care must be taken, however, in interpreting the differences in chomosomal and ribosomal gene clusters to mean that these subspecies actually differ much from one another because even minor differences in large DNA sequences can lead to major differences in RFLPs that do not correspond with significant biological differences. From the standpoint of B.t. being used as an insecticide, taxonomic studies, especially flagellar serotyping, have aided isolate classification but have proved unreliable as accurate predictors of insecticidal activity. Perhaps the best example of this is found in B. thuringiensis subsp, morrisoni (H8a8b), which includes isolates active against lepidopteran (isolate HD12), coleopteran (isolate DSM2803 and others), or dipteran (isolate PG14) insects. Another example of this is found in the occurrence of the 125-kb plasmid that encodes mosquitocidal Cry4 and Cytl proteins among numerous subspecies of B.t. including israelensis, morrisoni, entomocidus, kenyae, and thompsoni (Lopez-Meza et al., 1996; Thiery et al., 1996). The absence of an absolute correlation between subspecies/serotype and insecticidal activity continues to make it essential that the toxicity spectrum of new isolates be determined through bioassays, and ultimately be characterized in terms of the nucleotide sequences of genes encoded and expressed by individual B.t. isolates. Other methods have been examined for classifying existing and new isolates, including typing isolates with batteries of monoclonal antibodies to known endotoxins or using the polymerase chain reaction to identify genes carried by isolates. However, these have not proved better than H antigen serotyping for identification and classification of isolates. Moreover, though useful for identifying new toxins, these techniques carry the risk of missing unknown proteins with novel host spectra. To summarize the current status of B.t. systematics, most molecular evidence is in agreement with more classical biochemical and physiological studies that indicate B.t. and B. cereus are the same species, and that the latter becomes the former when it acquires one or more plasmids that express genes for insecticidal proteins. Nevertheless, maintaining B.t. as a separate species has practical value because of its insecticidal properties, as does dividing B.t. isolates into subspecies based on flagellar antigens. The latter are useful in cataloging the more than 20,000 isolates of B.t. collected to date. However, to understand the insecticidal properties of an isolate, regardless of its subspecies/serotype or other designation, knowledge of the insecticidal protein genes encoded and expressed is required.
INSECTICIDAL PROTEINS OF
BACILLUS THURINGIENSIS There are two principal active components in commercial preparations of B.t., the spore and parasporal body. For
Bacillus thuringiensis in Biological Control
most insect pests, the parasporal body, which consists of one or more insecticidal toxin proteins, accounts for most of the formulation's activity, including initial paralysis followed by death. The lethal effects of these proteins have been known since their discovery in the 1950s (Hannay, 1953; Angus, 1954). They are generally referred to as 6endotoxins, "6" designating a particular class of toxins, and endotoxin referring to their localization within the bacterial cell after production, as opposed to being secreted. In the early 1980s, shortly after the development of recombinant DNA techniques, it was discovered that B.t. 6-endotoxins were encoded by genes carried on plasmids. This discovery led quickly to a major research effort in many laboratories around the world aimed at understanding the genetic and molecular biology of these toxins. This effort resulted in cloning and sequencing of numerous B.t. genes, over 100 to date, and characterization of the toxicity of the proteins each gene encodes (Crickwore et al., 1998). Pertinent information derived from these studies through 1988 was summarized in the excellent review by Hofte and Whiteley (1989). At the time Hofte and Whiteley wrote their review, a wide variety of confusing names and acronyms were being used to refer to B.t. endotoxin genes and proteins. Computer analyses showed that the nucleotide sequences of most of these genes were quite similar. To standardize the terminology, Hofte and Whiteley (1989) proposed a simplified nomenclature for naming all insecticidal B.t. genes and proteins. In this nomenclature, the proteins are referred to as Cry and Cyt proteins. Though modified by Crickmore et al. (1998), this system is still in use today, and is described later along with its modifications and additions. Cry a n d Cyt Protein N o m e n c l a t u r e The Cry and Cyt nomenclature developed by Hofte and Whiteley (1989) was originally based on the spectrum of activity of the proteins as well as their size and apparent relatedness as deduced from nucleotide and amino acid sequence data, and protein gel analyses. At that time, with the exception of a 27-kDa cytolytic protein gene from B. thuringiensis subsp, israelensis, all genes appeared to be related, and probably derived from the same ancestral gene. Thus Hofte and Whiteley termed these "cry" (for crystal) genes and the proteins they encode "Cry" proteins. This designation was followed by a Roman numeral that indicates pathotype (I and II for toxicity to lepidopterans, III for toxicity to coleopterans, and IV for toxicity to dipterans), followed by an uppercase letter indicating the chronological order in which genes with significant differences in nucleotide sequences were described. The I and II for lepidopteran-toxic proteins also indicate size differences, with the I referring to proteins with a mass in the range of 130 kDa, and the II designating those with a mass from 65 to 70 kDa. Some epithets also include a lowercase letter in parentheses, which indicates minor differences in the nucle-
otide sequence within gene or protein type. Thus, CrylA referred to a 130-kDa protein toxic to lepidopterous insects for which the first gene (crylA) was sequenced, whereas CrylVD referred to a 72-kDa protein with mosquitocidal activity for which the encoding gene was the fourth from this pathotype sequenced. Though not perfect, this system was preferable to the chaos that existed before it was proposed. The 27-kDa CytA protein first isolated from B. thuringiensis subsp, israelensis differs from other B.t. proteins not only in its smaller size, but also in that it is highly cytolytic to a wide range of cell types in vitro, including those of vertebrates [see Federici et al. (1990) for reviews]. In addition, this shares no apparent relatedness with Cry proteins. Because of these differences and its broad cytolytic activity, Hofte and Whiteley (1989) referred to this as the CytA protein encoded by the cytA gene. In their revision of B.t. gene nomenclature, Hofte and Whiteley (1989) listed 38 published gene sequences that encoded 13 different Cry proteins and the single CytA protein. Since their publication, the number of B.t. Cry endotoxins has more than tripled, and several new cyt genes have also been described (Crickmore et al., 1998). As more and more cry genes were sequenced and analyzed, it was decided to name genes based on their relatedness as determined primarily from the degree of their deduced amino acid identity. As a result, the nomenclature developed by Hofte and Whiteley (1989) has been modified in the following manner. The cry and cyt descriptors have been maintained, but the Roman numerals have been replaced with Arabic numbers to indicate major relationships (90% identity), with higher degrees of identity being indicated by following uppercase letters (95% identity), and minor variations of these alleles being designated by lowercase letters, with the parentheses around the latter eliminated. Thus, for example, what was CrylA(c) is now CrylAc (and the corresponding gene, crylAc), a relatively minor change. However, for some genes and proteins the changes are greater. For example, CrylVD does not cluster with the earlier CrylV (now Cry4) proteins, and thus is now the taxon CryllAa. The previous and new nomenclature for representative examples of the most commonly studied and used endotoxin proteins is presented in Table 2. The complete list of currently recognized B.t. genes and proteins and references to the literature describing these can be obtained from the following website: http://www.biols.susx.ac.uk /Home/Neil__Crickmore/B t/index.html. Though the new designations supposedly carry no specific information concerning insecticidal spectrum, because the numbers have been maintained for many of the genes listed by Hofte and Whiteley (1989), and because a high degree of correlation between relatedness and insecticidal spectrum remains, primary insecticidal activity can often be inferred.
To illustrate this, Cry l still refers to lepidopteran toxicity; Cry2, to lepidopteran toxicity and in some, dipteran activity; Cry3, to coleopteran toxicity; and Cry4, to dipteran toxicity.
M o d e of Action a n d Structure of Cry Proteins The spore can play an important role in the pathogenicity of B.t. to certain insect species, but the parasporal body causes the rapid paralysis and ultimate death of most target species [for references see Huber and Luthy (1981), Aronson et al. (1986), Hofte and Whiteley (1989), Moar et al. (1989)]. The parasporal body consists of one or more Cry 6-endotoxins, and though the three-dimensional structure of three of these is now known from X-ray crystallographic studies, the mode of action of these toxins has not been resolved at the molecular level. Therefore, I will first review our knowledge of the mode of action of Cry endotoxins derived from early and later studies, and then consider the three-dimensional structure and insight provided into the mode of action.
Mode of Action
In the typical B.t., B. thuringiensis subsp, kurstaki, for example, the parasporal body dissolves after ingestion encountering the alkaline (pH 8 to 10) juices of the midgut. Dissolution requires the reduction of disulfide bridges that stabilize the Cry molecules in the parasporal crystal (Aronson, 1993). Most Cry toxins are actually protoxins of about 130 to 140 kDa (e.g., Cry l and Cry4 proteins) from which an active toxin "core" in the range of 60 to 70 kDa is released in the midgut by proteolytic cleavage from the Cterminal half of the molecule (Fig. 4). These activated toxin molecules pass through the peritrophic membrane and bind to specific receptors on the apical microvillar brush border membrane of midgut epithelial cells, which lies just outside the peritrophic membrane. Binding is an essential step in the intoxication process, and in susceptible insects the toxicity of a particular B.t. protein is correlated with the number of specific binding sites (i.e., receptors) on microvilli and with the affinity of the B.t. molecules for these sites (Hofmann et al., 1988; Van Rie et al. 1989, 1990). However, binding by itself, even high-affinity binding, does not
N o m e n c l a t u r e for Representative Insecticidal Proteins and Their Encoding G e n e s from
Bacillus thuringiensis Old nomenclature a
New nomenclature b
crylA(a) crylA(b) crylA(c) crylB crylC crylD crylIA crylIB crylIIA crylIIB crylVA crylVB crylVC crylVD jeg80 cytA cytB
CrylA(a) CrylA(b ) CrylA(c) CrylB CrylC CrylD CrylIA CrylIB CrylIIA CrylIIB CrylVA CrylVB CrylVC CrylVD Jeg80 CytA CytB
crylAa crylAb crylAc crylBa crylCa crylDa cry2Aa cry2Ab cry3Aa cry3Ba cry4Aa cry4Ba crylOAa cry11Aa cryl 1Ba cytlAa cyt2Aa
Cry 1Aa Cry 1Ab Cry 1Ac Cry 1Ba Cry 1Ca Cry 1Da Cry2Aa Cry2Ab Cry3Aa Cry3Ba Cry4Aa Cry4B a Cry 10Aa Cry 11Aa Cry 11B a CytlAa Cyt2Aa
Lepidoptera Lepidoptera Lepidoptera Lepidoptera Lepidoptera Lepidoptera Lepidoptera/Diptera Lepidoptera Coleoptera Coleoptera Diptera c Diptera Diptera Diptera Diptera Diptera/others d Diptera
Number of amino acids
1176 1155 1178 1207 1189 1165 633 633 644 649 1180 1136 675 643 248
133.2 131.0 133.3 138.0 134.8 132.5 70.9 70.8 73.1 74.2 134.4 127.8 77.8 72.4 80 27.4
aFrom Hofte, H., & Whiteley, H. R. (1989). Microbiological Review, 53, 242-255. bFor a complete list of crystalline endotoxin proteins see the following website: http://www.biols.susx.ac.uk/Home/NeiL.Crickmore/Bt/index.htm1 These proteins are protoxins that are activated in vivo by proteolytic cleavage in the insect midgut after ingestion. The Cry proteins are cleaved to form activated toxins in the range of 60 to 65 kDa, with most of the protein (ca. 600 amino acids) in the 130-kDa size range, proteins being cleaved from the C terminus. A small amount of cleavage also occurs at the N terminus of all Cry proteins. c cry4 and Cry l 1 proteins are the most toxic known proteins to dipteran insects, and have only been reported to have significant activity against members of the suborder Nematocera (e.g., insects such as mosquitoes, blackflies, chironomid midges, psychodid flies, and crane flies). din vitro, the CytA protein has been shown to be cytolytic for a wide range of cell types including those from invertebrates as well as vertebrates. Proteolysis yields a cytolytic protein of 25 kDa.
Bacillus thuringiensis in Biological Control
always lead to toxicity, indicating that insertion and probably postinsertional processing in the midgut membrane are required to obtain toxicity. For example, in three different studies it has been demonstrated that the Cry lAc toxin can bind with high affinity to microvillar membrane vesicles from Lymantria dispar (Wolfersberger, 1990), Spodoptera frugiperda (Garczynski et al., 1991) and Heliothis virescens (Gould et al., 1992), but with little or no subsequent toxicity. This indicates that insertion and likely postinsertional processing of Cry proteins are essential to intoxication. In highly sensitive insect species, the microvilli lose their characteristic structure within minutes of toxin insertion, and the cells become vacuolated and begin to swell (Huber & Luthy, 1981; Luthy & Ebersold, 1981). This swelling continues until the cells lyse and slough from the basement membrane of the midgut epithelium. As more and more cells slough, the alkaline gut juices begin to leak into the hemocoel where, as a result, the hemolymph pH rises by a half unit or more. This causes the paralysis and eventual death of the insect (Heimpel & Angus, 1963; Heimpel, 1967). Though this general picture of the mode of action has been known for many years, the actual process of intoxication at the molecular level remains unresolved, especially the series of events that occurs after the toxin binds to the receptor and inserts into the microvillar membrane. There
60 - 70
125 - 145
I I I i i-e
Proteolysis ActivatedToxin i 1
13451 Selectivity Domain
FIGURE 4 Schematic illustration of the structure and activation of typical Cry toxins of Bacillus thuringiensis. Cry 1 and Cry2 toxins are typically active against lepidopteran insects, and the latter occur naturally at about half the size of the former. When ingested by a larva, midgut proteases activate the protoxin, releasing the active toxic peptide by progressive cleavage of peptide fragments from both the N and C termini. Similar processing occurs to activate Cry3 (toxic to coleopteran insects) and Cry4 (toxic to nematoceran dipterans) toxins in the midgut of sensitive insects. AA indicates amino acid residues, whereas the arrows indicate protease cleavage sites. The numbers and hashed boxes indicate the approximate location of the five converged blocks of amino acids that are present in many Cry toxins. [After Hofte H., & Whitelely, H. R. (1989). Microbiological Reviews, 53, 242-255].
is good evidence for an influx of cations, especially potassium, and water into the columnar cells, which continues until cell lysis. To explain this influx, it has been proposed that activated Cry molecules insert into the apical microvillar membrane forming transmembrane pores that are cationselective at the alkaline pH that exists in the lepidopteran midgut (see Wolfersberger, 1989; Knowles & Ellar, 1987; Knowles & Dow, 1994; and Schwartz et al., 1993). For pores to form, it is postulated that during membrane insertion, the molecules undergo conformational changes that permit insertion. After insertion, the toxin molecules aggregate to form a pore composed of six toxin molecules. As more and more Cry molecules enter the membrane, additional pores form, accelerating the influx of cations and water, followed by cell hypertrophy and death. A more detailed discussion of the mode of action of Cry toxins and the data supporting various pore-formation models can be found in the reviews by Gill et al. (1992), Knowles and Ellar (1987), and Schnepf et al. (1998).
Structure Analysis of B.t. cry gene sequences conducted toward the end of the 1980s showed that the active portion of Cry toxin molecules (i.e., essentially amino acids 30 to 630 for most Cry molecules) contained five blocks of conserved amino acids distributed along the molecule, and a highly variable region within the C-terminal half (Hofte & Whiteley, 1989). The variable region was thought to be responsible for the insect spectrum of activity, and experimental evidence from recombinant DNA studies was obtained in support of this hypothesis. For example, by swapping the highly variable regions of CrylAa and CrylAc, the insect spectrum of these molecules could be reversed (Ge et al., 1989). In addition, binding studies showed that in most cases the degree of sensitivity of an insect to a particular Cry molecule was directly correlated with the number of high-affinity binding sites on the midgut microvillar membrane (Hofmann et al., 1988; Van Rie et al., 1989). Studies over the past decade have built upon this foundation and considerably improved our knowledge of B.t. protein molecular biology as a result of determination of the crystal structure of three Cry proteins (Cry3A, CrylAa, and Cry2A), identification of regions on the Cry molecule involved in midgut binding and specificity, and identification of several glycoprotein receptors for B.t. molecules on the insect microvillar membrane. The structure of the Cry3A molecule was the first solved, and similarities in conserved amino acid blocks among Cry toxins and conservation in hydrophobicity indicate this structure is likely a good general model for Cry toxins. Thus, the information discussed later combines interpretations from the crystal structure of the Cry3A molecule with recombinant DNA experiments on Cry 1 molecules.
The solution of the crystal structure of the Cry3A molecule (Li et al., 1991) showed that this protein is basically wedge shaped and consists of three domains (Fig. 5). Domain I is composed of amino acids 1 to 290 and contains a hydrophobic seven-helix amphipathic bundle, with six helices surrounding a central helix. This domain contains contains all the first conserved block and a major portion of the second conserved block of amino acids noted earlier. Theoretical computer models of the helix bundle of this domain show that after insertion and rearrangment, aggregations of six of these domains could form a pore through the microvillar membrane (Li et al., 1991). Domain II extends from amino acids 291 to 500 and contains three antiparallel/3 sheets around a hydrophobic core. This domain contains most of the hypervafiable region and most of conserved blocks three and four. The crystal structure of the molecule together with recombinant DNA experiments and binding studies indicates that the three extended loop structures in the/3 sheets are responsible for initial recognition and binding of the toxin to binding sites on the microvillar membrane (Lee et al., 1992). Domain III is composed of amino acids 501 to 644 and consists of two antiparallel/3 sheets, within which is found the remainder of conserved block number three along with blocks four and five. The three-dimensional structure resolved by Li et al. (1991) indicates this domain is involved in maintaining the structural integrity of the molecule, and site-directed mutagenesis studies of conserved amino acid block 5 in the Cry lAa molecule suggest this domain may also play a role in pore formation (Chen et al., 1993). Though the structure of the molecule, experiments, and computer modeling indi-
cate binding is attributed primarily to the hypervariable region of this domain, studies by Wu and Aronson (1992) also show that positively charged amino acids in domain I are important to binding and toxicity. Moreover, domain III has also been involved in receptor binding (DeMaagd et al., 1996). An important aspect of understanding B.t. molecular biology and determining its mode of action at the molecular level is identification of the proteins to which the toxin molecules bind on the microvillar membrane. Sangadala et al. (1994) have contributed to this area with the identification of the first two proteins in the midgut of Manduca sexta to which the Cry lAc toxin binds. They identified an aminopeptidase of 120 kDa as a major binding protein, and a second protein of 65 kDa, an alkaline phosphatase, as a minor binding protein. More recently, a cadherin-like Z10-kDa protein in M. sexta had been identified as a receptor for the CrylAb toxin (Francis & Bulla, 1997). These findings are important because they indicate that the initial recognition signal on the midgut of a sensitive insect is not an ion channel, but instead one or more "housekeeping" enzymes that extend from the microvillar membrane into the gut lumen. Thus, if these observations hold for other Cry proteins, the midgut receptors are then essentially docking proteins for the toxins, indicating the toxins do not directly affect ion permeability by binding to ion channels. These studies of the structure of B.t. toxins and their binding properties have identified key regions of the molecule responsible for binding and putatively the first binding proteins on the insect midgut epithelium. While further studies are certainly needed to more clearly define the specificity of binding and binding sites, the most enigmatic area of B.t. molecular biology with respect to its mode of action is how the toxin causes toxicity after binding to the membrane. Understanding this should provide information that will enable the development of insecticidal Cry proteins for use against insects against which we currently have none, such as cockroaches and grasshoppers. M o d e of Action a n d Structure of Cyt Proteins
FIGURE 5 Schematicillustration of the three-dimentional structure of the Cry3Aamolecule (from Knowles & Dow, 1994).
Four Cyt proteins are currently recognized, CytlA, Cyt2A, CytlB, and Cyt2B (Crickmore et al., 1998), though most of our knowledge of these proteins is based on studies of CytlA and Cyt2A, and the crystal structure of the latter toxin. These proteins are highly hydrophobic and all have a mass in the range of 24 to 28 kDa. They share no significant amino acid sequence identify with Cry proteins, and are thus unrelated. The first Cyt protein, CytlA, was identified as a component of the parasporal body of B. thuringiensis subsp, israelensis. Initially, it was thought that CytlA played little role in the toxicity of this subspecies to mosquito and blackfly larvae, but there is now general agreement that not only is it toxic to these and related flies
Bacillus thuringiensis in Biological Control
belonging to the dipteran suborder Nematocera, but also it is important as a synergist of Cry4 and Cry l 1 toxins, and as a preventative against developing insect resistance. The Cyt proteins derive their name from being cytolytic to a wide range of invertebrate and vertebrate cells in vitro (Thomas & Eller, 1983). While they are rare in comparison to Cry proteins, since the early work on CytlA in B. thuringiensis subsp, israelensis, Cyt proteins have been reported from mosquitocidal isolates of B. thuringiensis subsp, kyushuensis (Hl lallc), B. thuringiensis subsp, morrisoni (H8a8b), B. thuringiensis subsp, medellin (H30), and B. thuringiensis subsp, jegatheson (H28a28c). Mode of Action
As in the case of Cry proteins, Cyt proteins are synthesized during sporulation as protoxins and are assembled into crystalline inclusions that make up a portion of the parasporal body. They are always associated with Cry proteins, but do not cocrystallize with these, instead forming a separate inclusion. The shape of inclusion formed by Cyt proteins varies from hemispherical to angular because they assemble in a spherical parasporal body simultaneously along with inclusions formed by the other Cry proteins with which they occur. However, when produced alone in a recombinant B.t., the CytlA protein forms a bipyramidal crystal with 12 faces (Wu & Federici, 1993). Based on studies of CytlA, after ingestion the 27.3-kDa protoxin molecules dissolve from the crystal under the alkaline conditions present in the midgut of susceptible insects such as mosquito larvae. Proteolytic enzymes then activate the protoxin by cleaving amino acids at both the C and N terminus, releasing an active toxin of 24 kDa. This molecule passes through the peritrophic membrane and then inserts into the microvillar brush border membrane of midgut epithelial cells. Unlike Cry molecules, it appears that Cyt proteins do not require a glycoprotein receptor for binding. Cyt proteins have a high affinity for the lipid portion of the membrane, specifically for unsaturated phospholipids such as cholesterol, phospholipidcholine, and sphigomyelin (Thomas & Ellar, 1983). After binding to the microvillar membrane, the cells hypertrophy and subsequently lyse. The specific mode of action of Cyt toxins at the molecular level is not known. The leading hypothesis is that Cyt molecules insert into the microvillar membrane and then assemble into clusters of as many as 18 molecules to form pores that act much like those prosposed for Cry toxins (Koni & Ellar, 1994; Gill et al., 1992). Lysis then results from the influx of cations and water. However, evidence also has been provided that Cyt molecules do not form pores, but instead act like detergents, binding to and perturbing the lipid bilayer and disrupting the structural arrangement of membrane proteins (Butko et al., 1996, 1997), which would also cause lysis.
Unlike Cry proteins, Cyt proteins exhibit the same cytolytic activity in vitro against a wide range of invertebrate and vertebrate cell types, though at higher concentrations. However, toxicity is not observed when ingested in vivo by most insects tested, including larvae of lepidopterous species, nonnematoceran flies such as houseflies and fruit flies, and vertebrates such as mice. The mechanism of this in vivo specificity is not known, but may involve specific combinations of unsaturated phospholipids, or an unknown protein receptor. Though Cyt proteins have been thought to be only toxic to nematocerous dipterans, it has been reported that at least one beetle species, Chrysomela scripta, is sensitive to the CytlA protein (Federici & Bauer, 1998). An important property of Cyt proteins, based primarily on studies of CytlA, is the ability to synergize the toxicity of the Cry proteins with which they occur. For example, combination of CytlA with the Cry4A, Cry4B, or Cryl 1A proteins with which it occurs in the parasporal body of B. thuringiensis subsp, israelensis results in toxicities three- to fivefold higher than the specific toxicity of CytlA or any of the Cry proteins alone (Wu & Chang, 1985; Ibarra & Federici, 1986; Wu et al., 1995; Crickmore et al., 1996). The mechanism of this synergism is not known but probably involves cooperativity of the Cyt and Cry proteins in binding to and/or inserting into the microvillar membrane. Relevant to this is the finding that combination of CytlA with Cry4 or C r y l l A proteins enables the latter to overcome high levels of resistance to them in mosquito Culex quinquefasciatus (Wirth et al., 1997). Structure
The Cyt class of B. thuringiensis endotoxins are 24-28 kDa in mass, and occur primarily in subspecies toxic to nematoceran dipterans. The crystal structure for Cyt2A has been solved, and based on sequence similarities among Cyt proteins, it is assumed all have a similar structure. In contrast to the three-domain structure of activated Cry toxins, the Cyt2A molecule is a single domain, consisting of a/3sheet core wrapped in two outer layers of c~-helix hairpins (Li et al., 1996). Owing to the latitude of correlating structure with function, the structure of Cyt2A can be used to support either the pore-forming or detergent-like mode of action for Cyt proteins.
M A J O R B.T. PATHOTYPES The numerous isolates of B.t. that have been screened for insecticidal activity can be divided among three major pathotypes, those that exhibit toxicity to (1) lepidopterous, (2) dipterous, or (3) coleopterous insects. By far most isolates and subspecies fall into the first category, which from the turn of the century until the mid-1970s was the only pathotype known. The first isolate exhibiting substantial
toxicity to dipterous insects, mainly to the larvae of nematocerous dipterans such as mosquitoes and blackflies, was the ONR 60A isolate of B. thuringiensis subsp, israelensis (H 14) discovered in Israel in 1976 (Goldberg & Margalit, 1977). Several years later, the first B.t. isolate with high toxicity to coleopterous insects, the "tenebrionis" isolate of B. thuringiensis subsp, morrisoni (H 8a8b), was discovered in Germany (Krieg et al., 1983). Since these discoveries, isolates of these and other subspecies have been found that exhibit substantial toxicity to dipterous or coleopterous insects, but isolates with toxicity to lepidopterous insects remain the most common pathotype known. Isolates of B.t. active against nematodes have also been reported (Feitelson et al., 1992).
the smaller Cry2A protein forms the associated cuboidal inclusion. Another important subspecies of B.t. is B. thuringiensis subsp, aizawai (H 7). Typically, isolates of this subspecies produce a single bipyramidal crystal per cell in which CrylAa, CrylAb, CrylC, and CrylD have cocrystallized. The most complicated situation occurs in B. thuringiensis subsp, israelensis, where two Cry4 proteins, Cryl 1A, and two Cyt proteinsmCytlA being the dominant one m crystallize into three different inclusion types that are bound together in a fibrous envelope (Federici et al., 1990).
Parasporal B o d y S h a p e , C o m p l e x i t y , and Pathotype
The first commercial products based on B. thuringiensis consisted of sporulated, lysed cells of the fermented natural isolates described above, formulated as powders or liquid emulsifiable concentrates. Products based on these isolates remain the most common types of products in use today (Table 3). However, in an effort to improve upon these products, studies of the molecular biology and genetics of B. thuringiensis led to the development of several new types of products in which B.t. genes have been manipulated in different ways. These products include bacterial insecticides based on transconjugate or recombinant bacteria, and insect-resistant transgenic plants. In transconjugate bacteria, also referred to as transconjugants, plasmids encoding cry genes that do not occur together naturally have been combined in a comon host cell, usually a derivative of B. thuringiensis subsp, kurstaki. In recombinant bacteria, one or more cloned cry genes are expressed in either B. thuringiensis or another bacterium such as Pseudomonas flourescens, to improve the insecticidal properties of the products. Impovements consist of increased activity and stability, or expanded pest spectra achieved by combining cry genes which encode proteins that target pests such as those belonging to the Spodoptera complex, which are difficult to control with products based on the HD1 isolate of B. thuringiensis subsp, kurstaki. Lastly, in transgenic plants, usually after optimizing gene codon usage for expression in plants, one or more cry genes have introduced into commercial crops such as cotton, corn, soybeans, and potatoes to control certain major lepidopteran and coleopteran pests that feed on these crops.
In general, the shape of the parasporal body is a good but not absolute indicator of an isolate's pathotype. For example, most isolates of B.t. produce a large bipyramidal parasporal crystal (0.5 • 1 ~m) that is almost always only toxic to lepidopterous insects (Heimpel & Angus, 1963; Moar et al., 1989). In isolates of B.t. active against lepidopterous insects, such as the HD1 isolate of B. thuringiensis subsp, kurstaki (H 3a3b), the bipyramidal crystal may be accompanied by a smaller cuboidal crystal toxic to both lepidopterous and dipterous insects (Yamamoto & McLaughlin, 1981). Others, such as the ONR60A isolate of B. thuringiensis subsp, israelensis (H 14) and the PG-14 isolate of B. thuringiensis subsp, morrisoni (H 8a8b), produce spherical parasporal bodies (0.7 to 1 ~m) that are toxic primarily to nematocerous dipterans (e.g., mosquito and blackfly larvae; Federici et al. 1990), whereas the tenebrionis strain of B. thuringiensis subsp, morrisoni (H 8a8b) produces a thin, cuboidal crystal that is toxic only to certain species of coleopterans (Krieg et al., 1987; Keller & Langenbruch, 1993). The degree of protein complexity within the parasporal body of different isolates can vary considerably (Fig. 6). Single crystals can be composed of a single type of protein molecule or a mixture of as many as three. In addition, a single parasporal body may be composed of two (e.g., certain isolates of B. thuringiensis subsp, kurstaki) to four (e.g., B. thuringiensis subsp, israelensis) crystals (Hofte & Whiteley, 1989; Federici et al., 1990). An example of a simple crystal is that of the HD-73 isolate of B. thuringiensis subsp, kurstaki. This isolate only encodes and produces the CrylAc protein that forms a typical bipyramidal crystal during sporulation. The related HD1 isolate of the same subspecies, however, carries at least five cry genes (crylAa, crylAb, crylAc, cry2A, and cry2B) and produces at least four of these encoded proteins (CrylAa, CrylAb, CrylAc, and Cry2A) during sporulation. The three Cry lA proteins cocrystallize, forming a single bipyramidal crystal, whereas
COMMERCIAL PRODUCTS BASED ON B A CIL L LIS THLIRINGIENSIS
Products Based on Natural Isolates of
B. thuringiensis The isolates of the subspecies described above serve as the basis for most of the commerical bacterial insecticides used in pest control throughout the world. Whereas the proteins in the parasporal bodies of these isolates are the principle insecticidal components, the high insecticidal activity and broad insect spectrum of the HD1 isolate of B.
Chapter 21 Bacillusthuringiensis in Biological Control
thuringiensis subsp, kurstaki and the ONR60A isolate of B. thuringiensis subsp, israelensis are not due to the proteins alone, but instead to synergistic interactions among the proteins, proteins and spores, or other molecules produced during fermentation. The isolates and strains used in commercial products from the simplest, B. thuringiensis subsp. morrisoni strain tenebrionis (DSM 2803), to the most complex, B. thuringiensis subsp, israelensis (ONR60A), are described below. It should be realized that in most cases these strains served as the starting materials for the selection of strains that are actually used in commercial products. Once an isolate with good insecticidal properties is identified, it is typically subjected to additional selection and screening in the laboratory to develop strains with consistent commercial properties such as high toxicity, stability, high fermentation biomass yield, and good shelf life. Most of the products based on B.t. isolates and strains are applied at a rate < to 1 lb or 1 quart/acre.
Bacillus thuringiensis subsp, morrisoni (H8a8b), Isolate DSM 2803 The DSM 2803 isolate of B. thuringiensis subsp, morrisoni, strain tenebrionis, was originally isolated from diseased larvae of beetle Tenebrio molitor (Krieg et al., 1983). Studies of this isolate showed that it produced a single thin rectangular crystal composed of a 67-kDa Cry3A protein toxic to many species of coleopterous insects (Keller & Langenbruch, 1993). Subsequent studies showed that the gene encoding the protein encodes a toxin of 73.1 kDa, but that the protein was processed before being packaged into the crystal. In the 15 years since the discovery of the DSM 2803, numerous other isolates of B.t.-producing Cry3 proteins have been obtained from soil and grain dust samples, and currently four different types of Cry3 proteins (Cry3AD) are recognized. Because of the need to control many coleopterous pests, especially the Colorado potato beetle, Leptinotarsa decemlineata, the DSM 2803 isolate and several others where quickly developed into commercial bacterial insecticides such as MTRAK R, Trident R, and Novodor R (Table 3). In addition, the cry3A gene was used to construct transgenic potatoes (NewLeaf R, Monsanto Company), resistant to the Colorado potato beetle. While several of these products initially sold quite well, the advent of synthetic chemical insecticide Admire R (imidocloprid) has greatly reduced the market, at least at present, for bacterial insecticides based on Cry3 proteins.
Bacillus thuringiensis subsp, kurstaki (H3a3b), Isolates HD 1 and NRD 12 Probably the most common subspecies of B.t. isolated from nature is B. thuringiensis subsp, kurstaki. As noted
earlier, the parasporal body of this subspecies can vary considerably in complexity by having from only one protein to as many as four. The strains used in the first commercial formulations marketed over 30 years ago, and still commonly used today in products such as Dipel R, Foray R, and Thuricide R are those derived from the HD 1 isolate (see Table 3). These products are widely used to control caterpillar pests of vegetable and field crops, ornamentals, and forests (Navon, 1993; van Frankenhuyzen, 1993). HD1 has several properties that contribute to its ongoing success. First, the parasporal body consists of four Cry proteins, CrylAa (133.2 kDa), CrylAb (131 kDa), CrylAc (132.3 kDa), and Cry2Aa (70.9 kDa). Though these vary in their insect spectrum and specific toxicity (Table 4), together they give this isolate a broad spectrum of activity against a wide range of caterpillar species attacking field crops (cotton, corn, and soybeans), vegetables (tomatoes, broccoli, lettuce, and cabbage), fruit (strawberries, grapes, and peaches), and forests (deciduous and fir trees). This protein complexity probably also accounts for the lack of economically important resistance after more than 30 years of use in all but a very few species of lepidopterans, the notable exception being larvae of the diamondback moth, Plutella xylostella (Tabashnik, 1994). Second, in addition to the parasporal body proteins, HD 1 produces several components that synergize or potentiate the toxicity of the Cry proteins. These include the synergist zwittermycin which is an organic molecule, and the spore (Moar et al.; 1989; Miyasono et al., 1994). The mechanism of zwittermycin is unknown. The spore is a synergist in the broad sense of the word, and has its greatest effect against insects with moderate or low sensitivity to the Cry toxins, such as larvae of gypsy moth (Lymantria dispar), the diamondback moth (P. xylostella), and the beet armyworm (Spodoptera exigua). In larvae of these species, after an initial intoxication resulting from activation of the Cry toxins in the midgut, the spore germinates and produces enzymes such as phospholipases and proteases. In addition, insecticidal proteins produced during vegetative growth have been identified (Estruch et al., 1996). These contribute to the permeabilization and lysis of the midgut epithelial cells. Third, many isolates of B. thuringiensis subsp, kurstaki produce the/3-exotoxin, an adenine nucleotide that acts as a competitive inhibitor of DNA-dependent RNA polymerase [see Lecadet and de Barjac (1981), and Sebesta et al. (1981) for reviews]. It occurs widely in many natural isolates of B.t. and in many countries it is not allowed in commercial B.t. preparations because it can be teratogenic for mammals. However, it is permitted in formulations of B.t. in Finland and in certain African countries where it is used to control filth breeding flies, and it is also used in Russia against insects such as the Colorado potato beetle. Though /3-exotoxin has received relatively little attention recently, it does provide another interesting example of
R e p r e s e n t a t i v e E x a m p l e s of C o m m e r c i a l l y Available Microbial Insecticides B a s e d on Bacillus th uringiensis a
B. thuringiensis kurstaki (H3a3b)
Crop or habitat
Product b, c
Producer d, e
Biobit Condorf Cutlassf Dipel Thuricide Javelin Toarow CT Dipel Foray Thuricide Xentari Certan g Clorbac
Abbott Laboratories Ecogen Ecogen Abbott Laboratories Thermo-Trilogy Thermo-Trilogy Toagosei Chemical Abbott Laboratories Abbott Laboratories Thermo-Trilogy Abbott Laboratories Thermo-Trilogy Abbott Laboratories
B. thuringiensis aizawai (H7)
Vegetables Beehives Flowers
B. thuringiensis morrisonih(H8a8b)
Foilf MTRAK Novodor Trident
Ecogen Mycogen Abbott Laboratories Thermo-Trilogy
Mosquitoes and blackflies (Diptera)
B. thuringiensis israelensis (H 14)
Vectobac Teknar Skeetal
Abbott Laboratories Thermo-Trilogy Abbott Laboratories
aFrom Shah and Goettel (1999) and technical bulletins from manufacturers. bMost of these are general trade names that apply to a variety of different types of formulations such as flowable emulsifiable concentrates, wettable powders, and dusts. At present, over 100 different B.t. products are marketed throughout the world, with worldwide sales estimated at $100 million per annum. CThe target species for vegetables crops are numerous lepidopteran and coleopteran pests, among the most important of which are Trichoplusia ni (cabbage looper), Heliocoverpa zea (corn earworm), Spodoptera exigua (beet armyworm), Plutella xylostella (diamondback moth), Ostrinia nubilalis (European corn borer), and Leptinotarsa decemlineata (Colorado potato beetle). Major forest pests include the lepidopterans Lymantria dispar (gypsy moth) and Choristoneura fumiferana (spruce budworm). dAddresses can be found in Shah and Goettel (1999), or the website www.sipwebiorg. eThis list is not meant to be exhaustive, but instead to provide the name of major present and upcoming manufacturers. There are many smaller companies in industrialized and developing countries that also produce a variety of B.t. products. fTransconjugate bacteria. gFor control of larvae of the wax moth, Galleria mellonela. h The name "tenebrionis" has been used to describe pathotypes of this subspecies/serotype toxic to coleopteran insects.
synergism between different potential components of B.t. Because it occurs widely in natural isolates, Dubois (1986) examined its potential to synergize the activity of B.t. spore-crystal mixtures. By using a derivative of the HD1 isolate of B. thuringiensis subsp, kurstaki, he found that the LCs0 of this isolate against the gypsy moth larvae was reduced significantly by the addition of 0.01% fl-exotoxin. Though synergism could be detected earlier, by day 11 postfeeding, larvae treated with only the fl-exotoxin or B.t. preparation had suffered only 20% mortality, whereas mortality was greater than 75% in those treated with a combination of the two. As in the case of the spores, the mechanism of synergism between the fl-exotoxin and other B.t. components is not well defined. Based on its mode of action, one possibility is that the fl-exotoxin interferes with protein synthesis in midgut cells, especially differentiating regenerative cells
that require a high level of protein synthesis, and by doing so it impedes the larva's ability to repair damage to the midgut epithelium. This would hasten the action of endotoxins and spore germination. Another isolate of B. thuringiensis subsp, kurstaki currently used in several commercial preparations is NRD12 isolated by Norman Dubois. This isolate, which has the same Cry toxin composition as the HD1 isolate (Moar et al., 1989) was originally reported to have much better insecticidal activity against species of the genus Spodoptera. However, it was later demonstrated that much of this activity was due to low levels of fl-exotoxin.
Bacillus thuringiensis s u b s p , aizawai (H7) Although the HD 1 isolate of B. thuringiensis subsp, kurstuki is the most widely used an successful isolate of B.t.,
Bacillus thuringiensis in Biological Control
TABLE 4 Toxicity of B.t. Cry Proteins to First lnstars of Three Lepidopteran Pest Species a LCso in ng/cm 2 of diet Cry protein
Cotton leaf worm
Cry 1Aa CrylAb CrylAc CrylC
5.2 8.6 5.3 > 128
90 10 1.6 > 256
> 1350 > 1350 > 1350 104
aTobacco hornworm (Manduca sexta), tobacco budworm (Heliothis virescens), cotton leaf worm (Spodoptera littoralis). From Hofte and Whiteley (1989).
products based on it are not always effective at economical rates in controlling certain noctuid pests, especially species of Spodoptera such as the beet armyworm (S. exigua), the fall armyworm (S. frugiperda), and cotton and sorghum pests that occur in other regions of the world such as S. litura and S. littoralis. This is due to the relative lack of sensitivity of species in this genus to the Cry lA proteins produced by HD1 (see Table 4). The search for isolates of B.t. that would be more effective against Spodoptera species led to the discovery of several isolates of B. thuringiensis subsp, aizawai that are more effective than HD1. These isolates typically produce a single bipyramidal crystal per cell that contains a complex of four proteins, CrylAa (133.2 kDa), CrylAb (131 kDa), CrylC (134.8 kDa), and CrylD (132.5). These isolates are the basis for the product Xentari R (see Table 3) recommended for use against Spodoptera species (as well as many others, including populations of the diamondback moth, P. xylostella, resistant to HD1 preparations), and Certan R recommended for control of the waxmoth, Galleria mellonella, in beehives (see Table 3). The Cry lC protein is especially important to the increased activity of Bothuringiensis subsp, aizawai against the beet armyworm and diamondback moth.
Bacillus thuringiensis s u b s p , israelensis (H 14), isolate O N R 6 0 A The second most widely used isolate of B.t. used in insect control is the ONR60A isolate of B. thuringiensis subsp, israelensis, which is used in many different regions of the world primarily for the control the larvae of nuisance and vector mosquito and blackfly species (Mulla, 1990; Becker & Ludwig, 1993; Becker & Margalit, 1993). Though used primarily against mosquito and blackfly larvae, this isolate is toxic to all species that have been tested of the dipteran suborder Nematocera (the "long-horned" flies), which includes flies such as the mushrooms flies
(family Sciaridae), crane flies (family Tipulidae), and midges (family Chironomidae). The parasporal body of this isolate, which has an LCs0 in the range of 10 to 15 ng/ml of water against fourth instars of Aedes and Culex mosquitoes, is the most toxic per unit weight of all known B.t. isolates. This high toxicity and broad spectrum among nematocerous dipterans are due a complex of four major proteins, Cry4A (134 kDa), Cry4B (128 kDa), Cryl 1A (72 kDa), and CytlA (27.3 kDa). These are localized within three different inclusion types within the parasporal body, each of which is individually enveloped in a fibrous envelope of unknown composition, several layers of which also surround the entire parasporal body. The apparent function of this envelope is to hold the different toxin inclusions together to increase the probability that all toxins will enter the midgut simultaneously. After ingestion and dissolution in the midgut of sensitive insects, these proteins interact synergistically to bring about the high toxicity characteristic of this isolate. In addition to the major proteins, lesser amounts of related CytlA and Cryl 1 proteins also occur in the parasporal body. All these toxin proteins are encoded on a large plasmid of approximately 125 kb. Although this plasmid was initially identified in the ONR60A isolate of B. thuringiensis subsp, israelensis, it has been reported to occur quite widely in other mosquitocidal subspecies of B.t. including subsp, thompsoni, entomocidus, kenyae, and morrisoni (Ibarra & Federici, 1986; Ragni et al., 1995; Lopez-Meza et al., 1996). An unusual feature of the Cry and Cyt toxins of B. thuringiensis subsp, israelensis, and other subspecies carrying the same 125-kb plasmid, is the ability of these proteins to interact synergistically. Though synergism was initially reported between combinations of the CytlA and Cry4 or C r y l l A proteins (Wu & Chang, 1985; Ibarra & Federici, 1986; Chilcott & Ellar, 1988), it was later shown that combinations of Cry4 and/or Cry 11A proteins are also synergistic (Crickmore et al., 1995; Poncet et al., 1995). As noted earlier, these synergistic interactions, an example of which is provided in Table 5, account for the broad nematocerous dipteran spectrum and high toxicity of B. thuringiensis subsp, israelensis to mosquito and blackfly larvae. In addition, laboratory studies by Georghiou and Wirth (1997) suggest that the CytlA protein delays the development of resistance to the Cry4 and Cry l 1A proteins. In a series of experiments using various combinations of these toxins against larvae of Culex quinquefasciatus, toxin combinations containing CytlA led to only a 3.2-fold level of resistance after 28 generations, whereas the levels of resistance in combinations of Cry toxins lacking CytlA ranged from 90- to 900-fold or more. Based on the high efficacy of B. thuringiensis subsp. israelensis, a range of commercial products based on the ONR60A isolate were quickly developed shortly after its discovery (Mulla, 1990; Becker and Margalit 1993). Rep-
Examples of Cyt 1A: Cryl 1A Synergism against
First Instars of the Mosquito Aecles aegFpti
P r o d u c t s Based o n T r a n s c o n j u g a t e or R e c o m b i n a n t Bacteria
Range (ng/ml) b
CytA Cry11A CytA + Cry11A CytA + CryllA CytA + Cryl 1A CytA + Cryl 1A CytA + Cry11A Parasporal body
m 1:1 3:1 1:3 1 : 10 10 : 1
60a 85b 14.8c 14.8c 13.7c 11.2c 15.2c 2.5d
45-104 60- 200 10-22 10-22 7-22 6-18 12-19 0.5-4.4
1.96 < 0.9 1.54 1.64 1.32 1.44 2.2 2.7
aNumbers followed by the same letter indicate no significant difference at the 95% confidence level. bConfidence limits of 95%.
resentative examples products still marketed in the United States, Europe, and other countries are shown in Table 3. In the United States and Europe, these products are used primarily for the control of nuisance mosquitoes and blackflies, and in some situations, chironomid midges. However, the most important use of B. thuringiensis subsp, israelensis from a medical perspective has been in the World Health Organization's Onchocerciasis Control Program in West Africa. Formulations of B. thuringiensis subsp, israelensis, mainly Teknar R and Vectobac R, have been used in the dry season since the mid-1980s to control blackfly larvae that had begun to develop resistance to the chemical insecticide Abate (Guillet et al., 1990). The use of products based on B. thuringiensis subsp, israelensis preserved the Onchocerciasis Control Program, allowing the return of farmers to many river valleys that had become largely uninhabitable due to the prevalence of onchocerciasis. Products based on B. thuringiensis subsp, israelensis have been in use now for well over a decade, and no significant resistance has yet been reported in field populations of mosquitoes, blackflies, or chironomids. In the Rhine Valley in Germany, where B. thuringiensis subsp. israelensis is used against the floodwater mosquito (Aedes vexans) for more than 10 years, specific tests for the development of resistance have indicated no decrease in the sensitivity of mosquito populations (Becker & Ludwig, 1993). This lack of resistance is probably due to two major factors, the complexity of the B. thuringiensis subsp, israelensis parasporal body as described earlier, and the treatment of only selected areas along the Rhine. Because A. vexans is capable of moving several miles from its origin, the it latter permits the seasonal and year-to-year mixing of treated and untreated mosquito populations (natural refugia), thereby reducing the buildup of resistance genes in the treated populations.
T r a n s c o n j u g a t e Bacteria
Transconjugate bacteria used in commercial products are based on various strains of B. thuringiensis as the host cell. Typically, the cry gene complement of a recipient strain is increased by transforming the strain with a plasmid carrying a cryl or cry3 gene. The introduced plasmid is used to amplify copies of genes already in the strain, for example crylAc, to increase the yield of a particular protein, or to introduce new cry genes to expand the spectrum of activity. Examples of products on the market based on transconjugate strains include Cutlass | , Condor | , and Foil | . In the first two, a strain of B. thuringiensisis subsp, kurstaki has been enhanced by transformation with a plasmid from B. thuringiensis subsp, aizawai that contains a crylA gene to increase toxicity against certain lepidoptera. In Foil | a plasmid bearing a cry3A gene was introduced into a strain of B. thuringiensis subsp, kurstaki that produces Cry lAc, yielding a strain toxic to both lepidopteran and coleopteran insects (Baum et al., 1998).
R e c o m b i n a n t Bacteria
During this decade, several bacterial strains constructed using recombinant DNA technology have been registered in the United States. The first products registered were developed by Mycogen (San Diego, CA), and used Pseudomonas fluorescens as the host strain. The bacterium was engineered to produce large amounts of wild-type or recombinant Cry proteins, after which the cell wall was chemically fixed around the crystal to provide protection from ultraviolet light. Examples of products and the protein(s) they contain include MVP | which contains a CrylAc-CrylAb chimera toxic to lepidopteran insects; MTRAK | which contains Cry3A toxic to coleopteran insects; and MATTCH | which contains CrylAc and CrylC, toxic to lepidopteran insects (this product is a physical mixture of two P. flourscens strains, each producing one of the toxins). Ecogen (Langhorne, PA) pioneered a strategy in which strains of B. thuringiensis were engineered to produce more complex mixtures of toxins using recombinant plasmids. Current products on the market are CRYMAX | which contains CrylAc, and Cry2A, and is toxic to lepidopteran insects; Lepinox | which contains CrylAa, CrylAc, a CrylAc-CrylF chimera, and Cry2A, toxic to lepidopteran insects and designed especially for the Spodoptera complex; and Raven | which contains CrylAa, Cry3A and
Bacillus thuringiensis in Biological Control
Cry3Ba, and is toxic to both lepidopteran and coleopteran insects (Baum et al., 1998).
in the field is a significant threat when pest populations are placed under intensive selection pressure.
RESISTANCE TO BACILLUS THURINGIE.NSIS INSECTICIDES
TRANSGENIC BACILLUS THURINGIENSIS CROPS The ability to clone B.t. endotoxin genes led quickly to the development of the first transgenic B.t. plants in the mid-1980s. Since then, many major crops that suffer substantial economic damage from caterpillar and beetle pests have been genetically engineered to produce Cry proteins to control these insect pests, though only a few of these are available commercially (see Jenkins, 1998; Schulet et al., 1998 for recent reviews). In the United States, the available crops are cotton, corn, and potatoes (Table 6). Numerous others under development include rice, soybeans, broccoli, lettuce, walnuts, apples, and alfalfa. During the 1998 growing season, approximately 12 million acres of B.t.-corn were grown in the United States and 2.8 acres of B.t.cotton. Several tens of thousands of acres of B.t.-potatoes were also grown. These amounts are expected to grow considerably over the next few years as more companies commercialize B.t.-crops. Moreover, if these initial crops prove to be economic successes, most minor crops will eventually be engineered to produce Cry and other proteins for control of their major invertebrate pests. A list of B.t.transgenic crops currently under development can be found on the United States Department of Agriculture's website, hhtp ://www. aphis, usda. gov/bbep/bp/.
Though resistance to B.t. products among insect species under field conditions has been rare, laboratory studies show that insects are capable of developing high levels of resistance to one or more Cry proteins. Under laboratory selection, for example, populations of the Indian meal moth (Plodia interpunctella) developed levels of resistance ranging from 75- to 250-fold to CrylAa, CrylAb, CrylAc, Cry2A, and Cry 1C. In addition, under heavy selection pressure in the laboratory, populations of mosquitoes (C. quinquefasciatus), beetles (the Colorado potato beetle and the cottonwood leaf beetle, Chrysomela scripta), and tobacco budworm (Heliothis virescens) all developed levels of resistance ranging from several 100- to several 1000-fold to the Cry toxins against which they were selected (see Tabashnik, 1994; Bauer, 1995 for reviews). Rotation of Cry proteins in bacterial insecticides is a potentially useful tactic for managing resistance to individual Cry proteins. However, because most Cry proteins are related, the potential for cross-resistance remains a major problem. In fact, high levels of cross-resistance among Cry proteins has already been demonstrated in the laboratory in populations of the tobacco budworm. Whereas the preceding results were obtained using laboratory models, resistance in the field has become a major problem in the diamondback moth in Hawaii, Japan, the Philippines, and Florida, where populations were treated heavily and frequently (weekly or more) with products based on the HD1 isolate of B. thuringiensis subsp, kurstaki. In Hawaiian populations, because of cross-resistance, resistance extended to include CrylAa, CrylAb, CrylAc, CrylF, and CrylJ, whereas certain Floridian populations were resistant to CrylAa, CrylAb, and CrylAc. These results make it clear that even with complex products containing mixtures of Cry endotoxins and synergists, resistance
Resistance D e v e l o p m e n t C o n c e r n s There has been concern about the use of B.t. transgenic crops ever since the concept was first developed. Initially, environmentalists raised concerns over the safety to vertebrates of eating products containing Cry proteins. However, because safety studies demonstrated that vertebrate stomach juices rapidly inactivate Cry proteins, the emphasis has shifted to concern over the development of resistance to Cry proteins in insect populations. Were this to occur, the value of microbial insecticides based on B.t. Cry proteins
Insecticidal Protein Composition of B.t. Bacterial Insecticides versus Transgenic Plants
Cry and Cyt proteins B.t. subspecies (isolate)
Bacillus thuringiensis kurstaki (HD 1)+ B. thuringiensis israelensis (60A)
B.t. cotton B.t. corn B.t. corn
could be greatly diminished because of the documented capacity, even under field conditions, of insect populations to develop cross-resistance. The potential for the development of resistance to Cry proteins is of concern not only to environmentalists, but also to most scientists in academia, government, and B.t. insecticide and transgenic plant industries. With respect to the latter, failure of B.t. crops in the field after more than a decade of development would result in significant economic losses to these companies. The threat of resistance is real and easy to understand. Basically, the first generation of transgenic crops is based on plants that produce only a single Cry protein, and thus these lack the complexity described earlier of conventional B.t.-based bacterial insecticides. For example, current lines of B.t. cotton produce the Cry lAc protein, and are targeted to control the most important cotton pest, the tobacco budworm, in the southeastern United States. This protein was selected for engineering into cotton because it is the most toxic to the tobacco budworm (see Table 4). Similarly, in the case of corn, most lines have been engineered to produce the CrylAb toxin to control the European corn borer (Ostrinia nubilalis), and in potatoes, to produce the Cry3A toxin to control the Colorado potato beetle. In addition to the lack of complexity in comparison to bacterial insecticides, each of these toxins is produced continually by the plant. While perhaps an advantage from the standpoint of saving application costs, continuous production of toxin places the insect population under heavy selection pressure. Actually, over the past two growing seasons, studies of B.t. cotton in the United States and Australia have shown that Cry lAc production in the plant decreases over the growing season, in some cases leading to sublethal doses for larvae near the end of the season. This is of particular concern because it is known from experience with conventional chemical insecticides that exposure of populations to sublethal doses of a toxin permits the survival of individuals that are heterozygous for toxin resistance genes, allowing these genes to build up in the population. Because first generation B.t. crops generally target only a single pest species, another problem they present is lack of adequate control of insects not very sensitive to the toxin produced by the crop. It is important to realize that lack of or low sensitivity is not resistance, but nevertheless can lead to control problems and resistance to other toxins. For example, as can be seen from Table 4, the CrylAc protein is not very toxic to species of armyworms (Spodoptera species) or bollworms (Helicoverpa species). Already problems have been encountered in the United States and Australia in B.t. cotton with bollworms. In the United States, corn earworm (H. zea) populations invaded limited acreages of B.t. cotton in Texas, virtually destroying the crop because of their high tolerance to CrylAc. In 1998 in Australia, where the principal cotton pests are bollworms (H. armigera and H. punctigera), which are also quite tolerant
to Cry lAc, the efficacy of the B.t. cotton lasted for only approximately half the season. Applications of other pesticides, including B.t.-based insecticides, had to be made to control these pests. The lack of control resulted not only from the lower sensitivity of the bollworms to Cry lAc, but from the levels of toxin decreasing in the cotton plants as the growing season progresses. A danger of decreasing toxin levels is that this could prime target populations for the development of resistance to other Cry proteins to which these populations may initially be quite sensitive.
Bacillus thuringiensis Resistance Management in Transgenic Crops The possibility of resistance to transgenic crops has prompted the development of a variety of conceptual strategies for managing resistance (McGaughey & Whalon, 1992; Tabashnik, 1994; Gould, 1998). The most prominent of these are listed in Table 7. Most of these, for example, pyramiding various toxin genes within plants, use of tissuespecific toxin production, and induced toxin synthesis in which toxin is only synthesized after an insect begins to feed, are years away from field deployment or commercial availability. In the meantime, resistance management relies primarily on the refuge strategy. In this strategy, some percentage of the crop, usually 4 to 30%, must consist of non-B.t, plants. The value of the non-B.t, plants is to maintain a high percentage of susceptible insects (i.e., frequency of susceptible genes) in the target population. When moths lay eggs on B.t. plants, a high percentage of the first instars that feed on these plants will die, because this is the stage most sensitive to the toxin. However, a high percentage of the larvae that emerge and feed on non-B.t, plants will survive, and theoretically mate with adults that survived growth on B.t. plants. The latter survivors presumably survived because they were heterozygous or homozygous for resistance. They should constitute a low percentage of the mating population, and by mating with insects not selected for
Key Strategies for Managing Resistance to
B.t. Crops a Protein pyramiding Protein synthesis
Multiple Cry genes Cry plus other insecticidal genes Consitutive Tissue specific Chloroplast specific Inducible Refuges spatial, temporal Cry gene crop rotation
aModified from McGaughey, W. H., & Whalon, M. E. (1992). Sci-
ence, 258, 1451-1455.
Bacillus thuringiensis in Biological Control
resistance, the percentage of resistance genes is diluted and remains low in the target population. To delay resistance, the refuge strategy is being used in B.t. cotton and B.t. corn in the United States. B.t. cotton has been planted for 3 years, but it is not yet possible to assess the success of this strategy. During the first 3 years, there has been no confirmed evidence of resistance. However, aside from the refuges planted in the crops, many surrounding non-B.t, crops and noncrop plants provide refuges for susceptible insects, which then contribute to the dilution of resistance genes in the target population gene pool. The true test of this strategy will come when large contiguous areas, comprising square miles of B.t. crops producing Cry genes, are planted successively for several years. While these experiments in nature are ongoing, more sophisticated plant engineering strategies are being used to engineer resistance management strategies directly into the plants. In addition to pyramiding Cry genes, mixtures of insecticidal proteins with different modes of action will be engineered into plants. Already several types of proteins to meet these needs have been identified, including non-Cry proteins from B.t. insecticidal non-B.t, toxins, lectins, and enzymes selectively toxic to specific insects (Jenkins, 1998; Schuler et al., 1998).
Other Concerns about Bacillus th uringiensis Crops Concerns have been raised about the safety of B.t. proteins, and therefore B.t. crops, to nontarget insects, especially the predators and parasites used as biological control agents. It must be realized that B.t. kills the target pest, but even B.t. insecticides also cause mortality in certain nontarget insects. B.t. products used to control caterpillar pests such as gypsy moth and spruce budworm larvae in forests, for example, will also kill certain species of nontarget lepidopteran larvae in these habitats. The mortality caused in the nontarget populations must be kept in perspective and viewed in the context of the relative risk of using B.t. in comparison with using available synthetic chemical insecticides. The latter typically have a much broader spectrum of toxicity, and will kill pest and nontarget insects belonging to a wide range of insect orders. Because all B.t. proteins have a very restricted spectrum of activity, each usually only exhibiting toxicity to insects of a single order (Lepidoptera, Diptera, or Coleoptera), the use of B.t. proteins is much more environmentally compatible than is the use of chemical insecticides. B.t. proteins may persist in the environment, when used in either insecticides or B.t. crops, but their toxic effects on nontarget predators and parasites are low and temporary, being reduced even further after the crop is harvested. It should also be obvious that as a control agent targeted to kill insect pests, B.t. crops will also reduce the predator and parasite populations, particularly the latter, that depend
on the target insect for their reproduction. Again, the mortality caused by B.t. in these nontarget populations must be kept in perspective. Most crop production, especially that of field, vegetable, and fruit crops, occurs in monocultures that are not natural ecological habitats. The crop uniformity characteristic of these monocultures permits pest insects as well as their predators and parasites to build up into populations that are much larger than those that would occur in more diverse natural habitats. In this context, the use of B.t. crops has a neutral effect in that it reduces predators and parasite populations that would not occur if it were not for the unnatural presence of a large crop habitat and concomitantly large host pest populations that the predators and parasites use as a resource. Moreover, even in the unnatural ecological habitat of a crop monoculture, the use of B.t. because of its greater specificity, will have less impact overall on the predator and parasite populations than will the use of broad-spectrum chemical insecticides.
References Angus, T. A. (1954). A bacterial toxin paralyzing silkworm larvae. Nature (London), 173, 545-546. Aronson, A. I. (1993). The two faces of Bacillus thuringiensis: insecticidal proteins and post-exponential survival. Molecular Microbiology, 7, 489-496. Ash, C., Farrow, J. A. E., Wallbanks, S., & Collins, M. D. (1991). Phylogenetic heterogeneity of the genus Bacillus revealed by comparative analysis of small-subunit-ribosomal RNA sequences. Letters in Applied Microbiology, 13, 202- 206. Bauer, L. S. (1990). Resistance: a threat to the insecticidal crystal proteins of Bacillus thuringiensis. Florida Entomologist, 78, 414-443. Baum, J. A., Johnson, T. B., & Carlton, B. C. (1998). Bacillus thuringiensis: natural and recombinant bioinsecticide products. In F. R. Hall & J. J. Menn (Eds.), Biopesticides: use and delivery (pp. 189-209). Totowa: Humana Press. Baumann, L., Okamoto, K., Unterman, B. M., Lynch, M. J., & Baumann, P. (1984). Phenotypic characterization of Bacillus thuringiensis (Berliner) and B. cereus (Frankland & Frankland). Journal of Invertebrate Pathology, 44, 329- 341. Becker, N., & Ludwig, M. (1993). Investigations on possible resistance in Aedes vexan after a 10-year application of Bacillus thuringiensis israelensis. Joumal of the American Mosquito Control Association, 9, 221224. Becker, N., & Margalit, J. (1993). Use of Bacillus thuringiensis israelensis against mosquitoes and blackflies. In P. F. Entwistle, J. S. Cory, M. J. Bailey, & S. Higgs (Eds.), Bacillus thuringiensis, an environmental biopesticide: theory and practice (pp. 147-170). London: J. Wiley & Sons. Berliner, E. (1915). Ueber die schlaffsucht der Ephestia kuhniella und Bac. thuringiensis n. sp. Z. Allg. Entomologie, 2, 21-56. Butko, P., Huang, F., Pusztai-Carey, M., & Surewicz, W. K. (1996). Membrane permeabilization induced by cytolytic delta-endotoxin CytA from Bacillus thuringiensis var. israelensis. Biochemistry, 35, 11355-11360. Butko, P., Huang, F., Pusztai-Carey, M., & Surewicz, W. K. (1997). Interaction of the delta-endotoxin CytA from Bacillus thuringiensis var. israelensis with lipid membranes. Biochemistry, 36, 1286212868. Carlson, C. R., & Kolsto, A.-B. (1993). A complete physical map of a Bacillus thuringiensis chromosome. Journal of Bacteriology, 175, 1053-1060.
Chen, X. J., Lee, M. K., & Dean, D. H. (1993). Site-directed mutations in a highly conserved region of Bacillus thuringiensis 6-endotoxin affect inhibition of short circuit current across Bombyx mori midguts. Proceedings of the National Academy of Sciences, USA, 90, 9041-9045. Chilcott, C. N., & Ellar, D. J. (1988). Comparative toxicity of Bacillus thuringiensis var. israelensis crystal proteins in vivo and in vitro. Journal of General Microbiology, 134, 2551-2558. Crickmore, N., Bone, E. J., Williams, J. A., & Ellar, D. J. (1995). Contribution of the individual components of the delta-endotoxin crystal to the mosquitocidal activity of Bacillus thuringiensis subsp, israelensis. FEMS Microbiology Letters, 131,249-254. Crickmore, N., Zeigler, D. R., Feitelson, J., Schnepf, E., Van Rie, J., & Lereclus, D., (1998). Revision of the nomenclature for the Bacillus thuringiensis pesticidal crystal proteins. Microbiology and Molecular Biology Reviews, 62, 807-813. de Barjac, H., & Franchon, E. (1990). Classification of Bacillus thuringiensis strains. Entomophaga, 35, 233-240. De Maagd, R. A., Kwa, M. S. G., van der Klei, H., Yamamoto, T., Schipper, B., Vlak, J. M., Stiekema, W. J., & Bosch, D..(1996). Domain Ill substitution in Bacillus thuringiensis CrylA(b) results in superior toxicity for Spodoptera exigua and altered membrane protein recognition. Applied and Environmental Microbiology, 62, 1537-1543. Dubois, N. R. (1986). Synergism between/3-exotoxin and Bacillus thuringiensis subspecies kurstaki (HD-1) in gypsy moth, Lymantria dispar, larvae. Journal of Invertebrate Pathology, 48, 146-151. Estruch, J. J., Warren, G. W., Mullins, M. A., Nye, G. J., Craig, J. A., & Koziel, M. G. (1996). Vip3A, a novel Bacillus thuringiensis vegetative insecticidal protein with a wide spectrum of activities against lepidopteran insects. Proceedings of the National Academy of Sciences USA, 93, 5389-5304. Federici, B. A. (1994). Bacillus thuringiensis: Biology, application, and prospects for further development. In R. J. Akhurst (Ed.), Proceedings of the Second Canberra Meeting on Bacillus thuringiensis (pp. 1-15). Canberra: CPN Publications Pty, Ltd. Federici, B. A., & Bauer, L. S. (1998). CytlAa protein of Bacillus thuringiensis is toxic to the cottonwood leaf beetle, Chrysomela scripta, and suppresses high levels of resistance to Cry3Aa. Applied and Environmental Microbiology, 64, 4368-4371. Federici, B. A., Luthy, P., & Ibarra, J. E. (1990). The parasporal body of BTI: structure, protein composition, and toxicity. In H. de Barjac & D. Sutherland (Eds.), Bacterial control of mosquitoes and blackflies; biochemistry, genetics, and applications of Bacillus thuringiensis and Bacillus sphaericus (pp. 16-44). New Brunswick: Rutgers University Press. Francis, B. R., & Bulla, Jr., L. A. (1997). Further characterization of BTR1, the cadherin-like receptor for CrylAb toxin in tobacco hornworm (Manduca sexta) midguts. Insect Biochemistry and Molecular Biology, 27, 541- 550. Feitelson, J. S., Payne, J., & Kim, L. (1992). Bacillus thuringiensis: Insects and beyond. Bio/Technology, 10, 271-276. Garczynski, S. F., Crim, J. W., & Adang, M. J. (1991). Identification of putative brush border membrane-binding molecules specific to Bacillus thuringiensis 6-endotoxin by protein blot analysis. Applied and Environmental Microbiology, 57, 2816-2820. Ge, A. Z., Shivarova, N. I., & Dean, D. H. (1989). Location of the Bombyx mori specificity domain on a Bacillus thuringiensis 6-endotoxin protein. Proceedings of the National Academy of Sciences, USA, 86, 4037-4041. Georghiou, G. P., & Wirth, M. C. (1997). Influence of exposure to single versus multiple toxins of Bacillus thuringiensis subsp, israelensis in development of resistance in the mosquito Culex quinquefasciatus (Diptera: Culicidae). Applied and Environmental Microbiology, 63, 1095-1101. Gill, S. S., Cowles, E. A., & Pietrantonio. (1992). The mode of action of Bacillus thuringiensis endotoxins. Annual Review of Entomology, 37, 615-636. Goldberg, L. J., & Margalit, J. (1977). A bacterial spore demonstrating
rapid larvicidal activity against Anopheles sergentii, Uranotaenia unguiculata, Culex univitattus, Aedes aegypti, and Culex pipiens. Mosquito News, 37, 355-358. Gould, F. (1998). Sustainability of transgenic insecticidal cultivars: integrating pest genetics and ecology. Annual Review of Entomology, 43, 701-726. Gould, F., Martinez-Ramirez, A., Anderson, A., Ferre, J., Silva, F. J., & Moar, W. J. (1992). Broad-spectrum resistance to Bacillus thuringiensis toxins in Heliothis virescens. Proceedings of the National Academy of Sciences, USA, 89, 7986-7990. Guillet, P., Kurtak, D. C., Philippon, B., & Meyer, R. (1990). Use of Bacillus thuringiensis israelensis for Onchocerciasis control in West Africa. In H. de Barjac & D. Sutherland (Eds.), Bacterial control of mosquitoes and blackflies; biochemistry, genetics, and applications of Bacillus thuringiensis and Bacillus sphaericus (pp. 187-201). New Brunswick: Rutgers University Press. Hannay, C. L. (1953). Crystalline inclusions in aerobic spore-forming bacteria. Nature (London), 172, 1004-1006. Heimpel, A. M. (1967). A critical review of Bacillus thuringiensis var. thuringiensis Befliner and other crystalliferous bacteria. Annual Review of Entomology, 12, 287-322. Heimpel, A. M., & Angus, T. A. (1963). Diseases caused by certain sporeforming bacteria. In E. A. Steinhaus (Ed.), Insect pathology: an advanced treatise, (Vol. 2, pp. 21-73). New York: Academic Press. Hofmann, C., Vanderbruggen, H., Hofte, H., Van Rie, J., Jansen, S., & Van Mallaert, H. (1988). Specificity of Bacillus thuringiensis 6-endotoxins is correlated with the presence of high affinity binding sites in the brush border membrane of target insect midguts. Proceedings of the National Academy of Sciences, USA, 85, 7844-7888. Hofte, H., & Whiteley, H. R. (1989). Insecticidal crystal proteins of Bacillus thuringiensis. Microbiological Reviews, 53, 242-255. Huber, H. E., & Luthy, P. (1981). Bacillus thuringiensis delta-endotoxin: composition and activation. In E. W. Davidson (Ed.), Pathogenesis of invertebrate microbial diseases (pp. 209-234). Totowa: Allanheld, Osman & Co. Ibarra, J. E., & Federici, B. A. (1986). Isolation of a relatively non-toxic 65kilodalton protein inclusion from the parasporal body of Bacillus thuringiensis subsp, israelensis. Journal of Bacteriology, 165,527-533. Ishawata, S. (1901). On a type of severe flacherie (sotto disease). Dainihon Sanshi Kaiho, 114, 1- 5. Jenkins, J. J. (1998). Transgenic plants expressing toxins from Bacillus thuringiensis. In F. R. Hall & J. J. Menn (Eds.), Biopesticides: use and delivery (pp. 211-232). Totowa: Humana Press. Keller, B., & Langenbruch, G. A. (1993). Control of coleopteran pests by Bacillus thuringiensis. In P. F. Entwistle, J. S. Cory, M. J. Bailey, & S. Higgs (Eds.), Bacillus thuringiensis, an environmental biopesticide: theory and practice (pp. 171-191). London: J. Wiley & Sons. Knowles, B. H., & Dow, J. A. T. (1994). The crystal 6-endotoxins of Bacillus thuringiensis: Models for their mechanism of action on the insect gut. BioEssays, 15, 469-476. Knowles, B. H., & Ellar, D. J. (1987). Colloid-osmotic lysis is a general feature of the mechanism of action of Bacillus thuringiensis 6-endotoxins with different insect specificity. Biochemica Biophysica Acta, 924, 509- 518. Koni, P. A., & Ellar, D. J. (1993). Cloning and characterization of a novel Bacillus thuringiensis cytolytic toxin 6-endotoxin. Journal of Molecular Biology, 229, 319-327. Krieg, A., Huger, A., Langenbruch, G., & Schnetter, W. (1983). Bacillus thuringiensis var. tenebrionis: a new pathotype effective against larvae of Coleoptera. Journal of Applied Entomology, 96, 500-508. Krieg, A., Schnetter, W., Huger, A. M., & Langenbruch, G. A. (1987). Bacillus thuringiensis subsp, tenebrionis, Strain BI 256-82: a third pathotype within the H-Serotype 8a:8b. Systematic and Applied Microbiology, 9, 138-141. Lecadet, M.-M., & Debarjac, H. (1981). Bacillus thuringiensis beta-exotoxin. In E. W. Davidson (Ed.), Pathogenesis of invertebrate microbial diseases (pp. 293-321). Totowa: Allanheld, Osman & Co.
Bacillus thuringiensis in Biological Control
Lee, M. L., Milne, R. E., Ge, A., & Dean, D. H. (1992). Location of a Bombyx mori receptor binding region on a Bacillus thuringiensis 8endotoxin. Journal of Biological Chemistry, 267, 3115-3121. Li, J., Carroll, J., & Ellar, D. J. (1991). Crystal structure of insecticical 6endotoxin from Bacillus thuringiensis at 2.5 * resolution. Nature (London), 353, 815-821. Li, J., Koni, P. A., & Ellar, D. J. (1996). Structure of the mosquitocidal 8endotoxin CytB from Bacillus thuringiensis p. kyushuensis and implications for membrane pore formation. Journal of Molecular Biology, 257, 129-152. Lopez-Meza, J. E., Federici, B. A., Poehner, W. J., Martinez-Castillo, M., & Ibarra, J. E. (1995). Highly mosquitocidal isolates of Bacillus thuringiensis subsp, kenyae and entomocidus from Mexico. Biochemical Systematics and Ecology, 23, 461-468. Luthy, P., & Ebersold, H. R. (1981). Bacillus thuringiensis delta-endotoxin: histopathology and molecular mode of action. In E. W. Davidson (Ed.), Pathogenesis of invertebrate microbial diseases (pp. 235267). Totowa: Allanheld, Osman & Co. Meadows, M. P. (1993). Bacillus thuringiensis in the environment: ecology and risk assessment. In P. F. Entwistle, J. S. Cory, M. J. Bailey, & S. Higgs (Eds.), Bacillus thuringiensis, an environmental biopesticide: theory and practice (pp. 193-220). London: J. Wiley & Sons. McGaughey, W. H., & Whalon, M. E. (1992). Managing insect resistance to Bacillus thuringiensis toxins. Science, 258, 1451-1455. Miyasono, M., Inagaki, S., Yamamoto, M., Ohba, K., Ishiguro, T., Takeda, R., & Hayashi, Y. (1994). Enhancement of ~endotoxin activity by toxin-free spore of Bacillus thuringiensis against the diamondback moth, Plutella xylostella. Journal of Invertebrate Pathology, 63, 111-112. Moar, W. J., Trumble, J. T., & Federici, B. A. (1989). Comparative toxicity of spores and crystals from the NRD-12 and HD-1 strains of Bacillus thuringiensis subsp, kurstaki to neonate beet armyworm (Lepidoptera: Noctuidae). Journal of Economic Entomology, 82, 15931603. Mulla, M. S. (1990). Activity, field efficacy, and use of Bacillus thuringiensis israelensis against mosquitoes. In H. de Barjac & D. Sutherland (Eds.), Bacterial control of mosquitoes and blackflies; biochemistry, genetics, and applications of Bacillus thuringiensis and Bacillus sphaericus (pp. 134-160). New Brunswick: Rutgers University Press. Navon, A. (1993). Control of lepidopteran pests with Bacillus thuringiensis. In P. F. Entwistle, J. S. Cory, M. J. Bailey, & S. Higgs (Eds.), Bacillus thuringiensis, an environmental biopesticide: theory and practice (pp. 125-146). London: J. Wiley & Sons. Poncet, S., Delecluse, A., Klier, A., & Rapoport, G. (1995). Evaluation of synergistic interactions among CrylVA, CrylVB and CrylVD toxic components of Bacillus thuringiensis subsp, israelensis crystals. Journal of Invertebrate Pathology, 66, 131-133. Priest, F. G., Kaji, D. A., Rosato, Y. B., & Canhos, V. P. (1994). Characterization of Bacillus thuringiensis and related bacteria by robosomal RNA gene restriction fragment length polymorphisms. Microbiology, 140, 1015-1022. Ragni, A., Thiery, I., & Delecluse, A. (1995). Characterization of six highly mosquitocidal Bacillus thuringiesis strains that do not belong to H-14 serotype. Current Microbiology, 31, 1-7. Sangadala, S., Waiters, F. W., English, L. H., & Adang, M. J. (1994). A mixture of Manduca sexta aminopeptidase and alkaline phosphatase enhances Bacillus thuringiensis insecticidal CrylA(c) toxin binding and 86RblpLK+ efflux in vitro. Journal of Biological Chemistry, 269, 10088-10092. Schnepf, E., Crickmore, N., Van Rie, J., Lereclus, D., Baum, J., & Feitelson, J. (1998). Bacillus thuringiensis and its pesticidal proteins. Microbiology and Molecular Biology Reviews, 62, 775-806. Schnepf, H. E., & Whiteley, H. R. (1981). Cloning and expression of the Bacillus thuringiensis crystal protein gene in Escherichia coli. Proceedings of the National Academy of Sciences USA, 78, 2893-2897. Schuler, T. H., Poppy, G. M., Kerry, B. R., & Denholm, I. (1998). Insectresistant transgenic plants. Trends in Biotechnology, 16, 168-175.
Schwartz, J.-L., Gameau, Savaria, D., Masson, L., Brousseau, R., & Rousseau, E. (1993). Lepidopteran-specific crystal toxins from Bacillus thuringiensis form cation- and anion-selective channels in planar lipid bilayers. Journal of Membrane Biology, 132, 53-62. Sebesta, K., Farkas, J., Horska, K., & Vankova, J. (1981). Thuringiensis, the beta-exotoxin of Bacillus thuringiensis. In H. D. Burges (Ed.), Microbial control of pests and plant diseases, 1970-1980 (pp. 249281). London: Academic Press. Shah, P. A., & Goettel, M. S. (1999). Directory of microbial control products and services. Gainesville: Society for Invertebrate Pathology Press. Smith, R. A., & Couche, G. A. (1991). The phylloplane as a source of Bacillus thuringiensis variants. Applied and Environmental Microbiology, 57, 311-315. Tabashnik, B. E. (1994). Evolution of resistance to Bacillus thuringiensis. Annual Review of Entomology, 39, 47-79. Thiery, I., & Franchon, E. (1997). Bacteria: identification, isolation, culture and preservation of entomopathogenic bacteria. In L. A. Lacey (Ed.), Manual of techniques in insect pathology (pp. 55-77). San Diego: Academic Press. Thomas, W. E., & Ellar, D. J. (1983). Bacillus thuringiensis var. israelensis crystal delta-endotoxin: effects on insects and mammalian cells in vitro and in vivo. Journal of Cell Science, 60, 181-197. Thomas, W. E., & Ellar, D. J. (1983). Mechanism of action of Bacillus thuringiensis var. israelensis insecticidal 6-endotoxin. FEBS Letters, 154, 362-368. van Frankenhuyzen, K. (1993). The challenge of Bacillus thuringiensis. In P. F. Entwistle, J. S. Cory, M. J. Bailey, & S. Higgs (Eds.), Bacillus thuringiensis, an environmental biopesticide: theory and practice (pp. 1-35). London: J. Wiley & Sons. van Frankenhuyzen, K. (1993). The challenge of Bacillus thuringiensis. In P. F. Entwistle, J. S. Cory, M. J. Bailey, & S. Higgs (Eds.), Bacillus thuringiensis, an environmental biopesticide; theory and practice (pp. 1-35). Chichester: J. Wiley & Sons. Van Rie, J., Jansens, S., Hofte, H., Degheele, D., & Van Mellaert, H. (1989). Specificity of Bacillus thuringiensis 8-endotoxins: Importance of specific receptors on the brush border membrane of the midgut of target insects. European Journal of Biochemistry, 186, 239-247. Van Rie, J., Jansens, S., Hofte, H., Degheele, D., & Van Mallaert, H. (1990). Receptors on the brush border membrane of the insect midgut as determinants of the specificity of Bacillus thuringiensis delta-endotoxins. Applied and Environmental Microbiology, 56, 1378-1385. Wirth, M. C., Georghiou, G. P., & Federici, B. A. (1997). CytA enables CryIV endotoxins of Bacillus thuringiensis to overcome high levels of CryIV resistance in the mosquito, Culex quinquefaciatus. Proceedings of the National Academy of Sciences USA, 94, 10536-10540. Wolfersberger, M. G. (1990). The toxicity of two Bacillus thuringiensis delta 8-endotoxins to gypsy moth larvae is inversely related to the affinity binding cites on the brush border membranes for toxins. Experientia, 46, 475-477. Wu, D., & Aronson, A. I. (1992). Localized mutagenesis defines regions of the Bacillus thuringiensis 8-endotoxin involved in toxicity and specificity. Journal of Biological Chemistry, 267, 2311-2317. Wu, D., & Chang, F. N. (1985). Synergism in mosquitocidal activity of 26 and 65 kDa protein from Bacillus thuringiensis subsp, israelensis crystal. FEBS Letters, 190, 232-236. Wu, D., & Federici, B. A. (1993). A 20-kilodalton protein preserves cell viability and enhances CytA crystal formation during sporulation in Bacillus thuringiensis. Journal of Bacteriology, 175, 5276-5280. Wu, D., Johnson, J. J., & Federici, B. A. (1994). Synergism of mosquitocidal toxicity between CytA and CryIVD proteins using inclusions produced from cloned genes of Bacillus thuringiensis. Molecular Microbiology, 13, 965- 972. Yamamoto, T., & McLaughlin, R. E. (1981). Isolation of a protein from the parasporal crystal of Bacillus thuringiensis var. kurstaki toxic to the mosquito larva, Aedes taeniorhynchus. Biochemical and Biophysical Research Communications, 103,414-421.