Transgenic organisms expressing genes from Bacillus thuringiensis to combat insect pests

Abstract

Various subspecies (ssp.) of Bacillus thuringiensis (Bt) are considered the best agents known so far to control insects, being highly specific and safe, easily mass produced and with long shelf life.¹ The para-crystalline body that is produced during sporulation in the exosporium includes polypeptides named δ-endotoxins, each killing a specific set of insects. The different entomopathogenic toxins of various Bt ssp. can be manipulated genetically in an educated way to construct more efficient transgenic bacteria or plants that express combinations of toxin genes to control pests.² Joint research projects in our respective laboratories during the last decade demonstrate what can be done by implementing certain ideas using molecular biology with Bt ssp. israelensis (Bti) as a model system. Here, we describe our progress achieved with Gram-negative bacterial species, including cyanobacteria, and some preliminary experiments to form transgenic plants, mainly to control mosquitoes (Diptera), but also a particular Lepidopteran and Coleopteran pest species. In addition, a system is described by which environment-damaging genes can be removed from the recombinants thus alleviating procedures for obtaining permits to release them in nature.

Synergy between Cyt1Aa and Cry Toxins

Bti is highly efficient and specific against mosquito and black fly larvae.³ Most importantly, no resistance has been observed in nature after about 30 years of extensive use worldwide.⁴ Different activities and modes of action of its four major toxins form a lethal combination against larvae of all mosquito species tested.⁵ Resistance is not selected due to synergy among Bti components, mostly the low-toxic, non-specific Cyt1Aa. High synergy levels affected by Cyt1Aa were observed by Crickmore⁶ and Wirth,⁷ the latter also demonstrated that Cyt1Aa prevented selection of resistant mosquitoes.⁸

The question raised was whether a combination of anti-Lepidopteran toxins with Cyt1Aa imitates this rare advantage of Bti. To partially answer this question, two genes were cloned for expression in Escherichia coli, cry1Ac (from Bt ssp. kurstaki) and cry1Ca (from Bt ssp. aizawai), with and without cyt1Aa, and tested against three pests, Helicoverpa armigera, Pectinophora gossypiella and Spodoptera littoralis.⁹ Co-expression of all three genes, and with p20 encoding an accessory protein (for reasons beyond the scope of this report), indeed synergized toxicity against H. armigera but antagonized it against P. gossypiella. Moreover, very high toxicity against S. littoralis and huge synergy value between the two tested Cry's were found without Cyt1Aa and P20. Thus, one cannot predict which gene combination would be useful in pest control, and each idea must experimentally be tested separately.

Cyanobacteria to Deliver Bti Toxins against Mosquitoes

Several disadvantages hamper the use of Bti: in nature, it does not proliferate whereas the toxins disappear by sinking and adsorption to silt particles and are inactivated by sunlight.³ To achieve a real biological control, the vector must multiply in the same niche as its target. Photo-synthetic cyanobacteria have several additional features that render them excellent candidates to control mosquito larvae:¹⁰ their floating capacity avoids sinking and adsorption to silt hence keeps them in the same zone (upper level of water bodies) and available to the larvae. Cyanobacteria are amenable to recombinant DNA technology, and their pigments¹¹ are likely to protect the toxins from sunlight inactivation.

Our attempts to circumvent the disadvantages of Bti by using the nitrogen-fixing, filamentous cyanobacterium Anabaena PCC 7120 were quite successful.^12–14 The first condition for cyanobacteria to control mosquito larvae, feeding them, was demonstrated (Fig. 1): the closely related species Anabaena siamensis is ingested and digested by larvae of Aedes aegypti, similarly to Anabaena PCC 7120 (not shown).

Figure 1.

Anabaena siamensis is ingested and digested by Aedes aegypti larvae. A second instar larva of A. aegypti ingests (A) Anabaena siamensis (B), excreting from its anus, (C) leaving behind digested filaments and intact heterocysts (D).

In cooperation with the Sanger Institute, we sequenced pBtoxis,¹⁵ the 128 kb plasmid of Bti that harbors all the genetic information necessary for mosquito larvicidal activity.¹⁶ All 15 possible combinations (= 2⁴ − 1) of four genes, three encoding the toxins Cry4Aa, Cry11Aa and Cyt1Aa and the accessory P20, were cloned for expression in E. coli.¹⁷ This protocol of cloning gene combinations under identical promoters allowed comparisons of toxicities and synergy levels among the toxins in vivo. Cyt1Aa was indeed found to synergize Cry4Aa and Cry11Aa hence is anticipated to reduce the likelihood of selection for resistance in the target organisms.⁸

The most toxic combinations were appropriately moved into Anabaena PCC 7120 and confirmed our working hypotheses: Cyt1Aa synergizes the Cry's in Anabaena as well,¹⁴ and the toxic activity is protected from silt in the laboratory and from sunlight inactivation in semi-field conditions. They were about seven-fold more effective than a commercial preparation of Bti itself.¹⁸ One of our future plans, adding the last major Cry gene cry4Ba to this battery, would improve this bio-control agent.

Environmental Considerations

For various reasons, field tests to release living genetically engineered microorganisms are not yet allowed worldwide. One justifiable reason that has been demanded by The European Council Directive,¹⁹ namely the use of markers that confer resistance to “clinically used” antibiotics must be phased out! Drug resistance markers must be removed from transgenic clones before they are even to be considered for release in nature. Release to the environment of Anabaena transgenic clones such as ours that were derived by selection of antibiotic resistance markers requires marker-free strains. The most elegant way to achieve this goal is by site-specific recombination.

The site-specific recombination system of the λl-like coliphage HK022, which has been implemented in Arapidopsis plants²⁰ and in human cells,²¹ is designed to remove these genes from the Anabaena genome. The responsible enzyme, Integrase (Int) catalyzes site-specific integration and excision of DNA provided that the recombination target sites attP + attB or attR + attL, respectively, are available. In human cells Int is active on the extra-chromosomal level with plasmids²² as well as on the chromosomal level,²¹ in both cis and trans orientations.

Expression of lacZ in Anabaena PCC 7120 was designed to demonstrate the Int-catalyzed excisive recombination reaction, whether located on a plasmid or on the chromosome.²³ A plasmid pMVO carrying the four Bti toxin genes was constructed such that its antibiotic resistance marker nptII can be excised with Int24 (Fig. 2). This plasmid was introduced into the Anabaena chromosome by homologous recombination after conjugation using neomycin selection.²⁴ The excision of nptII along with additional unnecessary DNA (luxAB) out of the resultant mosquito larvicidal transgenic Anabaena is underway.

Figure 2.

Plasmid pMVO, designed for excision by Int via site-specific recombination between the attR and attL sites. The 23.4 kb pMVO carries the four Bti toxin genes (cry4A, cry11A, p20 [twice] and cyt1A) and antibiotic resistance marker nptII. Two promoters were introduced, P_psbA (of the photosystem II's D1) and P_A1 (T7 phage early promoter) at the denoted positions. nptII, neomycin/kanamycin resistance gene; ORF all3924, a PCR amplified sequence²⁴ encoding a probable penicillin amidase (see in http://bacteria.kazusa.or.jp/cyanobase/index.html).

How Can Maize Control Mosquitoes?

Pollen of maize (Zea mays) provide complete food source for Anopheles arabiensis larvae, which is the reason for the sharp rise in malaria prevalence in Africa during blooming seasons.²⁵ Vast quantities of maize pollen accumulate on the surface of nearby puddles, enhancing development of mosquito larvae in breeding sites that lie within 50–60 meters range.²⁶ Moreover, maize pollen is phagostimulant for mosquito larvae.²⁷ Maize is therefore engineered to express combination of genes for mosquito larvicidal toxins in the pollen. The combinations to be exploited are of Bti toxin genes together with tmfA, encoding the peptide hormone TMOF (Trypsin Modulating Oostatic Hormone) of A. aegypti.²⁸ This hormone, an unblocked decapeptide (YDPAPPPPPP) that exerts its effects against a relatively narrow range of targets,²⁹ starves the larvae to death by blocking translation of trypsin-like mRNA in the midgut.²⁸ Since starved larvae are 6–35-fold more sensitive to Bti toxins than are fed larvae,³⁰ TMOF is anticipated to synergize the Bti toxins in suppressing larval densities around fields of maize genetically modified appropriately. Continuous anti-vector coverage for an entire village is likely to be achieved with a few patches of transgenic maize producing larvicidal pollen. Transformed plantlets with cry11Aa-tmfA, cry4Aa-tmfA and cyt1Aa-tmfA have been generated, moved to the greenhouse (to be published elsewhere), and additional gene combinations are currently being prepared.

Anti-Coleopteran Active Genes to Control Capnodis ssp.

Our recently-embarked project is to discover Bt genes that will control a pest prevalent in countries surrounding the Mediterranean. Three ssp. of the flat-headed borer Capnodis, Capnodis tenebrionis, C. carbonaria and C. cariosa, kill trees of cultivated stone-fruits.³¹ The larvae destroy the root systems of almond, apricot, cherry, nectarine, peach, plum and pistachio. Tree mortality and economic losses are reported from all Southern-European and Mediterranean countries.³² Since natural occurring arthropod enemies of Capnodis are rare,³³ growers use intensively organophosphates or carbamates onto the foliage or the stem and the surrounding soil.^34,35 The populations of Capnodis increase in areas where they had been considered minor pests few decades ago. Development of environmentally friendly measures to control them is thus highly important.

As the first step to achieve this goal, an artificial diet was recently developed,³⁶ and is currently exploited to screen Bt strains for toxicity against them. Of a battery of 215 field isolates, 38 that were found to include at least one gene encoding anti-Coleopteran Cry toxin⁷ are being bioassayed. The genes from the best isolates will be cloned for expression in the roots of the target trees.

Concluding Remarks

Use of environment friendly and cost effective alternatives to chemical pesticides improves health and safety, enhances crop output and lowers levels of pollution. Toxins of entomopathogenic bacteria have become leading bio-pesticides to control populations of insect pests and vectors transmitting severe human diseases. Implementation of innovative ideas to exploit molecular methods and interactions between organisms, together with considering various ecological aspects, is likely to become hallmark of future generations.

Abbreviations

Bt

Bacillus thuringiensis

Bti

Bacillus thuringiensis subsp. israelensis

TMOF

trypsin modulating oostatic factor

ssp.

subspecies