development of recombinant viral insecticides by expression of … · 2013-10-15 · archives of...

Archives of Insect Biochemistry and Physiology 22:315-344 (1 993)

Development of Recombinant Viral Insecticides by Expression of an Insect-Specific Toxin and Insect-Specific Enzyme in Nuclear Polyhedrosis Viruses Bruce D. Hammock, Billy F. McCutchen, Jeffrey Beetham, Prabhakara V. Choudary, Elizabeth Fowler, Reiji Ichinose, Vernon K. Ward, Joanna M. Vickers, Bryony C. Bonning, Lawrence G. Harshman, David Grant, Tamon Uematsu, and Susumu Maeda DepartmentsofEntomology(B.D.H., B.F.M.,J.B., P.V.C., R.I., V.K.W.,J .M.V. , B.C.B., L.G.H., D.G., T.U., S.M.)andEnvironmental Toxicology(B.D.H., B.F.M.,].B., P.V.C., X.I . , V.K. W . , J .M. V . , B.C.B., L.G.H., D.G., T . U J , Antibody Engineering Facility (P. V.C.), University of California, Davis, California: ABRU, Ciba-Geigy Carp., Research Triangle Park, North Carolina (E.F.)

As supplements to classical chemical insecticides, two approaches to develop recombinant baculovirus insecticides are described. In one approach an insect- specific toxin is expressed leading to a dramatic reduction in time to death. In the second approach an insect juvenile hormone esterase is expressed which leads to a reduction in feeding. Modifications of the wildtype esterase led to viruses which reduced the time to death as efffectively as did the toxin-expressing virus. In both cases existing recombinant viruses are viewed as leads, and approaches to further improvement in the engineered viruses are suggested. Many of these approaches are based on analogy with the development of classical synthetic insecticides. Using these viruses as examples, the potential utility and limitations of recombinant viruses and other biological insecticides are discussed. Q 1993 Wiley-Liss, Inc.

Acknowledgments: The assistance of Dr. Robert Possee (NERC institute of Virology and Environ- mental Microbiology, Oxford, England), Dr. Elizabeth Fowler (ABRU, Ciba Ceigy Corp., Research Triangle Park, North Carolina), and their associates is gratefully acknowledged. ABRU provided the AaHlT gene used in this work. This work was supported in part by grants from the National Science Foundation (DCB-8818875), the U.S. Department of Agriculture (88-37234-4000,91-37302-6186), National Institute of Environmental Health Sciences (ESO-271 O), and the North Atlantic Treaty Organization. B.F. McCutchen is a recipient of a NIEHS Training Grant in Environmental Toxicology awarded to the University of California, Davis. I . Beetham is a recipient of a NIH Training Grant in Biotechnology awarded by the Biotechnology Program at the University of California, Davis. B.D. Hammock is a Burroughs Wellcome Scholar in Toxicology, and P.V. Choudary is a Biotechnology Career Fellow of the Rockefeller Foundation and Director of the University of California, Davis, Antibody Engineering Laboratory. R. lchinose was supported by Sankyo Co., Ltd. T. Uematsu was supported by Sumitomo Chemical Co., Ltd.

Received November 13, 1991 ; accepted July 9, 1992

Address reprint requests to Bruce D. Hammock, Department of Entomology, University of California, Davis, CA 9561 6.

0 1993 Wiley-Liss, Inc.

316 Hammock et al.

Key words: nuclear polyhedrosis virus, scorpion toxin, recombinant development of insecticide

INTRODUCTION

A variety of pressures in the agricultural industry are making alternative agents more attractive as supplements to, or replacements for synthetic pesticides in integrated pest management. However, synthetic chemicals have set a high standard of expectation in terms of ease of use and efficacy. In order to compete with synthetic pesticides in developed countries and to address serious problems in developing countries, biological agents must be more effective. There are several approaches to obtaining improved biological insecticides, but recombinant DNA technology offers many exciting possibilities.

Nuclear polyhedrosis viruses will be among the engineered organisms to be evaluated in agriculture. Because of extensive research on these viruses as expression vectors for proteins of medical importance, the NPVs* are useful systems for testing the utility of peptides and proteins in insect control, in addition to providing insight for peptide expression in other systems.

The NPVs are natural biological control agents already in use as biological pesticides. In spite of many advantages, a variety of disadvantages have limited their use in agriculture (Table 1). However, some of these limitations such as host range could also be interpreted as advantageous. In integrated pest management programs, other problems such as large-scale production and field stability can be overcome by a variety of research strategies if the viruses appear to offer sufficient commercial potential for investment in research by industry (Table 2). Among the attributes of the viruses as natural control agents is their ability to replicate to tremendous numbers while allowing their host to feed for several days or weeks. Although a positive attribute in a natural control strategy, this trait is a severe limitation if the biological pesticide must reduce feeding damage. The problem becomes even worse if the pest quickly bores into the plant where the virus cannot reach. This trait of slow kill also is one which can be addressed by recombinant DNA technology.

There are at least four approaches which can be used alone or in combination to reduce feeding damage by pest species. The approach of viral gene deletion was elegantly applied by O’Reilly and Miller [l] to the ecdysteroid glucosyl transferase. A second approach of expressing a natural insect chemical mediator was illustrated by Maeda [2] with the diuretic hormone. This article will summarize the status of work at University of California, Davis on a third approach of expressing a peptide toxin under a viral promoter [3,4] and a fourth approach of expressing an insect enzyme [5] . The toxin approach

*Abbreviations used: AaHlT = the toxin from the scorpion Androctonus australis (Hector); AcNPV = nuclear polyhedrosis virus of Autographa californica; BmNPV = the nuclear polyhedrosis virus of Bombyx rnori; BSA = bovine serum albumin; DDT = 1,l-(p,p’dichloro)diphenyltrichloroethane; IPM = integrated pest management; JH = juvenile hormone; ]HE = juvenile hormone esterase; kbp = kilo base pair; NPV = nuclear polyhedrosis virus; PIB = polyhedron inclusion body.

Development of Recombinant Viral insecticides 31 7

TABLE 1. Disadvantages of Baculoviruses as Insecticides

Killing time Slow death of host Field stability Degraded by UV light Specificity Narrow host range Expense Not conducive to large scale production Infectivity Oral transmission only; no systemic or contact activity

is attractive in the short term to pesticide scientists due to its dramatic effects and the simplicity of concept and has been used by the laboratory of Robert Possee [6] with the AaHIT toxin discussed below and by the laboratory of Lois Miller [7l with a toxin gene from the straw itch mite, Pyemotes tritici, termed TxP-I for the toxin and tox34 for the gene. The approach of expressing a catalytically active enzyme may be less intimidating to the public, and in the long run it may have more profound effects by disrupting the biology of pest species. Further research on fundamental aspects of the developmental biology of insects is certain to indicate other approaches to practical control of pest species.

EXPRESSION OF THE INSECT-SPECIFIC NEUROTOXIN AaHIT

NPVs are double-stranded DNA viruses with the viral particles embedded in a protein matrix termed the polyhedron in the case of baculoviruses. This polyhedron encapsulation of the viral particles dramatically enhances stability and provides a mechanism of selectivity in being dissolved at the high pH in the insect gut. A second layer of selectivity is provided by the high specificity of the virus for a limited group of insects. In spite of these desirable attributes of selectivity, it is critical that early work on peptide expression involves materials that are extraordinarily specific for insects. The Algerian scorpion Androctonus australis Hector provides such a toxin in the 70 amino acid AaHITl (Fig. 1) [8-111. Toxin names such as AaHITl are derived from the scientific name and author: AaH; insect specificity, I; toxin, T; and a number.

In spite of the apparent safety of this peptide to vertebrate species, our initial studies were carried out using the polyhedrin promoter of the NPV of the silkworm Bornbyx rnori. This expression system is well characterized and offers several technical advantages {12], and in addition, it provides added safety factors. For instance, there are no species related to B. mori that are native to California, and the polyhedrin promoter system used leads to polyhedron- negative viruses that are highly unstable, are easily destroyed in the laboratory, and are only active by injection.

Using a synthetic gene for AaHIT based on published amino acid and nucleotide sequences [13], two transfer vectors containing the AaHIT gene were constructed for BmNPV, one with a signal sequence (Fig. 1) for secretion and one without [3] (Fig. 2).

TABLE 2. Critical Projects to Improve the Field Activity of Biologicals

Formulation Production Biological enhancers Residue analysis Synergists Risk evaluation IPM studies Marketing

318 Hammock et al.

55 60 65 70

Signal Sequence AalT Coding Sequence

Fig. 1 . The amino acid sequence and disulfide bonds of synthesized AaHIT. The toxin produced in Bombyx mori and Heliothis virescens by the engineered viruses has proven indistinguishable from the natural AaHIT.

Fig. 2. Diagrammatic representation of the recombinant baculovirus AcNPV-AaHIT. Relative locations and directions of the AaHIT, p10, and polyhedrin genes in the recombinant baculovirus are shown on the 130 kb circular DNA of AcNPV below. Above is greater detail of the region containing the AaHlT gene under the p10 promoter. Both of these are reinserted into a polyhedron-negative virus during cotransfection. The polyhedron-positive ]HE viruses were prepared using a similar system.

Development of Recombinant Viral Insecticides 31 9

Following transfection and purification of the recombinant virus, the resulting viruses were shown by Western blot analysis to produce AaHIT-like materials in tissue culture. When these viruses were injected into B. mori, speed of kill was dramatically increased in comparison to control viruses engineered to produce other peptides. Since the gene coding for AaHIT with a secretion signal sequence derived from the insect neuropeptide bombyxin was more active, it was used in subsequent studies. With other peptides it may be advantageous to remove the leader sequence to retain the toxin in infected cells.

The symptoms observed in larvae infected with AaHITl were consistent with those expected from a sodium channel blocker and were similar to those resulting from injected nonrecombinant AaHIT peptide purified from scorpion venom by HPLC. The time to death was significantly shortened in all instars of B. mori tested with the recombinant virus. In all cases feeding was dramatically reduced 10-20 h earlier as the larvae showed signs of general irritability, loss of abdominal control, and finally paralysis [3].

Based on these data we moved to the NPV of the alfalfa looper caterpillar, Autograph culifornica (AcNPV). AcNPV represents a well developed expression system [14,15] as well as a virus active on a moderate spectrum of major insect pests. For expression we used the p10 promoter system described by Weyer et al. [16] due to ease of selection, slightly earlier and higher expression levels compared to the polyhedron system [17-191, and because this approach results in viruses that can orally infect target species and are stable under field conditions.

A transfer vector for AcNPV was constructed which was analogous to that made for B. mori. Following cotransfection and purification of the polyhedron- positive recombinant virus, it was found to speed kill relative to wildtype and control engineered virus in several pest species in the insect family Noctuiidae including Heliothis virescens, Spodopteru exigua and Trichoplusiu ni [4]. Time to death also was reduced in every larval stage tested in H . virescens. The presence of an AaHIT-like substance was demonstrated in the hemolymph of larvae infected with the AaHIT virus, but not with the control virus, by both bioassay with blow fly larvae and by Western blot. As in the case of the recombinant NPV tested in B. mori, the neurotoxic effects of AaHIT reduced feeding at least 10 h before actual death [4]. This paralytic effect offers advantages since it allows viral replication to continue for in vivo production of the virus for commercial use.

Similar results with AaHIT have been reported by Stewart et al. [6] and by Tomalski and Miller [7] using a toxin from the straw itch mite. These studies and the work described here illustrate the concept of using a toxic peptide to improve the efficacy of biological pesticides.

IMPROVING THE ENGINEERED TOXIN VIRUS Based only onefficacy, it is likely that even the existing recombinant AcNPV

that expresses AaHIT has commercial potential on a variety of crops. How- ever, it is far better to view these results in the same way as a promising lead with a synthetic chemical in an early screen. There are obvious ways to

320 Hammock et al.

Decision tree for methods to improve efficacy and

speed of kill of recombinant baculoviruses expressing foreign peptides

/ / \ Other Virus

Biological ,, ,, Other AcNPV

/ \ Viruses

Other AaHlT Peptide

lncreased lncreased lncreased Increased Production of Stability of Stability of Potency of

Message Message Peptide Peptide

Improved Leader

I

New P I 0 I I Promoter Polyhedron Existing System Leader

Fig. 3. Outline of approaches for improving recombinant viral insecticides. Properties of the viruses can be improved at many levels and most of these improvements are likely to be additive if not synergistic in increasing the efficacy of the recombinant viruses.

improve the efficacy and host range of the material at the level of the virus, transcription of DNA, translation of RNA, and pharmacokinetics of AaHIT (Fig. 3). Obvious alternative approaches are to use other biologically active peptides or to employ the AaHIT peptide in other expression systems.

There are numerous situations where one or more of the species of noctuiids controlled effectively by AcNPV are the major pests on a crop. The noctuiids alone (with the pest problems resulting from attempted control of noctuiids) account for a massive amount of insecticide use. Recombinant barnloviruses could reduce environmental contamination and pest resurgence, retard or cir- cumvent resistance, and increase agricultural profitability. The advantages of being able to treat crops shortly before harvest and the potential of providing affordable insect control in developing countries are clearly attractive.

Expanding Host Range

A narrow host range clearly is attractive for initial studies on the release of a recombinant pesticide. As discussed by Hammock and Soderlund [20], selectivity can be economically beneficial in a pesticide as well. High selectivity will lead to a reduced rate of the development of resistance since selection only will occur when one is treating for the target pest. In contrast, population

Development of Recombinant Viral Insecticides 321

pressure leading to resistance can occur every time a broad spectrum material is applied even if it is applied to control another species. Selectivity also is likely to reduce problems with pest resurgence due to nonselective destruction of natural enemies. However, the host range of the engineered virus clearly is too narrow for many applications. For instance, even within the noctuiids the poor activity of AcNPV on Helicoverpa zed may limit many cotton applications with current technologies. This limited spectrum of activity will require pest managers to identify correctly the pest in the field to be treated.

The effective host range can be extended in many ways. For instance a cocktail of engineered viruses with different host ranges could be used. Such approaches could be made commercially attractive by having formulations color coded for blending at the farm or to work with farm advisors to formulate blends for specific crops and specific regions. As viruses are engineered to produce more toxin, the effective host range will be extended to insects only marginally sensitive to the wildtype virus. The use of specific factors to increase host range also is promising [21,22]. Possibly the most exciting approach involves investigation of the fundamental mechanisms of specificity at the molecular level and expansion of baculovirus host range by viral selection and recombinant DNA technology [23]. This approach is certain to yield exciting biological data and may allow us to tailor specificity to the level that is desired. Also, as discussed below, a more intimate understanding of viral biology may lead to the production of viruses which target critical tissues for expression of recombinant proteins.

The success of the engineered AcNPVs makes similar approaches with other viruses attractive. Clearly, alternate viruses attacking members of the noctuiid complex such as Helicoverpu sp or Spadopteru sp [24] are attractive targets. From a world perspective the economic impact of the diamondback moth and its propensity for resistance make engineered viruses for this organism attractive. The sensitivity of many forest ecosystems makes engineered baculoviruses attractive for the control of species such as the gypsy moth or spruce budworm. Rapid kill may not be of high priority in a forest ecosystem; however, it is very important in an urban setting. A list of possibilities could be very long. Unfortunately in many cases the systematics and biology of possible viruses are unknown, not to mention cell lines and cloning systems. Many of the potential markets are too small to support even the requisite research much less the production and regulatory costs.

Alternate Promoters and Improvements of the Gene Both the polyhedrin and p10 promoters are only activated very late in viral

replication. There apparently is little expression in the rnidgut tissue first infected. Thus, at the DNA level the use of earlier, more powerful promoter systems or hybrid promoter systems is likely to further increase speed of kill. It recently has been found that the p10 promoter produces recombinant proteins earlier and at a higher level than the polyhedrin promoter [17-191. This may not be significant in the field, but it certainly shows the correct trend. Even more exciting is that the basic protein promoter produces more recombinant protein even earlier than p10 [19]. Tomalski and Miller report that a hybrid promoter containing elements from late and very late promoter results

322 Hammock et al.

in excellent expression of the mite toxin and slightly quicker kill than the polyhedrin promoter [7]. As we learn more about promoter systems it is likely that still higher and earlier production of recombinant proteins can be obtained by using hybrid viral promoters, promoters from other species possibly including target insects, or other approaches.

Gene duplication, careful placement of the foreign gene within the recombinant virus, or even classical breeding to adapt the genetic background of the virus to the engineered gene could increase activity. Since some amplifi- cation of the virus is critical for optimal activity, it is not likely that the engineered viruses will offer the quick kill typical of many synthetic pesticides. However, it is certain that viruses can be engineered to yield kill rates that are far faster than those reported in the four studies to date. Using several technologies in this laboratory, the time needed to kill 50% of orally infected H . virescens has been reduced from approximately 110 h to under 70 h and the time to stop feeding to approximately 35 h following an oral dose of 1,000 PIBs to neonate larvae [Betana and McCutchen, unpublished]. Also, quick kill is not important in all agronomic situations, and even current constructs dramatically reduce feeding [3,4,6,7].

Improvements of the Message

At the level of the message, improvements in the design of ribosome binding site (cap region recognition), codon usage patterns as well as techniques to stabilize the message have the potential of ultimately yielding higher levels of toxin. In the latter case we slowly are developing knowledge of the properties of the primary and secondary structure of both translated and untranslated regions of a message that will increase the stability and rate of translation of a mRNA. Our synthetic gene for AaHIT includes some modifications to bring it in line with the published natural cDNA. However, improvements could be obtained just by using the cDNA from A. australis coding for the natural toxin and its leader [25] .

Improving AaHIT The AaHIT example illustrates several ways to improve the activity of a

recombinant virus. First, the levels of the toxin detected in the hemolymph are quite low [3,4] and one wonders if this is due to low production, suboptimal distribution, or rapid uptake and/or metabolism. We already have demonstrated the dramatic effect of adding a signal sequence for secretion [3,6,7]. It is likely that alternate leaders will yield further improvements in efficacy. Undoubtedly the leader needs to be selected based on the physical properties and mechanism of action of the peptide expressed. With different peptides, one may need to keep the foreign gene product within the cell or to target certain organelles. Perhaps the relatively poor expression of AaHIT is not a hindrance in accelerating paralysis since the toxin may be concentrated in nerve cells. Further knowledge of leader and targeting sequences will expand our predictive abilities in these areas.

With classical insecticides, it has been argued that there are four approaches to the development of insect control chemicals [26,27]. First generation approaches are based on folk remedies and chance observations. Second


generation approaches are based on a random screening process followed by bioactivity-directed synthesis. Third generation approaches are based on rational synthesis based on a knowledge of the comparative physiology and biochemistry of the target and nontarget species. Finally, fourth generation approaches are based on molecular biology [28]. DDT is an example of a compound developed exclusively by a second generation approach, while the juvenoids clearly were developed with a combination of second and third generation approaches. The highly successful pyrethroids are based on first generation natural pyrethrins whose structures were optimized with both second and third generation approaches, since the wide ranging screening efforts from industry were based in part on both an appreciation of the site of action and the metabolism of these compounds. The viruses expressing the mite toxin and AaHIT represent initial leads. These leads can be exploited with both second and third generation approaches. Certainly one of the roles of current researchers is to illustrate how the research paradigms developed with classical pesticides can be applied to optimize these fourth generation insect control agents.

Based on analogy with classical insecticides, a metabolism and pharmacokinetic study with AaHIT and the mite toxin clearly are warranted. It is unlikely that a lepidopteran target drove the evolution of AaHIT since these insects are not common prey of the scorpion [29]. Thus, it is possible that the toxin is rapidly degraded, possibly through uptake by certain cells followed by proteolytic cleavage. Random or site-directed mutagenesis, development of chimeric toxins, or even truncated toxins could yield recombinant peptides which are more stable or are directed away from tissues involved in catabolism and toward target tissues. As with classical pesticides, pharmacokinetic and metabolism studies on prototype toxins will provide the scientific basis for rational modifications in the toxin structure. Alternatively, molecular techniques offer the possibilities of generating large numbers of random mutants. A combination of rapid in vitro assays to insure that action on the sodium channel is not lost followed by rapid in vivo assays could lead to an improvement in action of the recombinant viruses. This would represent a second generation or screening approach to optimization of the structure of the biologically active peptide.

A program leading to rational modification of toxic peptides such as AaHIT can be conceptually parallel to similar programs with classical insecticides. Certainly rational optimization of even a small peptide will be difficult; however, such improvements can be based on computer-aided design with information coming from a combination of technologies including X-ray crystallography of the toxin and target, energy minimization or quantum mechanical calculations of toxin structure, NMR structures, circular dichroism, antigenic relationships, and a knowledge of the evolutionary relationships among toxins in a group [25,30-39].

Injected AaHIT shows surprisingly low activity in lepidopterous larvae, when compared with other insect species [29]. Thus, its great effectiveness when expressed at low levels in NPV-infected larvae presents an anomaly. Possibly continuous exposure to low levels of toxin is more effective than a single large dose from an injection or scorpion sting. It also is possible that

324 Hammock et al.

direct viral infection of nervous or adjacent tissue leads to a locally high concentration of the toxin in the well insulated lepidopteran nervous system. Removal of the sheath around the central nervous system of lepidopterous larvae dramatically increases the in vitro toxicity of many toxins including AaHIT. This anomaly raises the possibility of using pharmacodynamic studies to improve the virus further.

Alternate Toxins

The results with AaIT and TxP-I are very important in that they demonstrate that natural toxins which are highly specific for insects can be used to reduce feeding of pest insects infected with a recombinant virus at levels which are certain to be important in the field. They also provide an indication of the potency of the toxin used to obtain an effect. Certainly more stable toxins, earlier and stronger promoters and other approaches discussed above could lead to equal effects using a weaker toxin. However, if a toxin is too weak to reach pharmacologically active concentrations before the virus kills the insect, it will have little effect. Thus, the trend will be to engineer more potent toxins. Excellent leads for such toxins exist as discussed below [32,4043].

EXPRESSION OF THE JUVENILE HORMONE ESTERASE FROM HELIOTHIS VIRESCENS

Metabolism of Juvenile Hormone

The success of the juvenile hormone mimics (juvenoids, insect growth regulators) in insect control illustrates that the process of metamorphosis can be exploited commercially. However, it would be very attractive in agriculture to have materials that give an anti-juvenile hormone effect and reduce feeding [44]. In holometabolous insects and specifically the Lepiodoptera (butterflies and moths), a reduction in the titer of the terpenoid juvenile hormone initiates a sequence of events leading ultimately to pupation. One of the early events in this process is an initiation of wandering behavior and a cessation of feeding .

Several factors lead to this reduction in JH titer but metabolism clearly plays a role (Fig. 4). In insects from several orders and at some stages in lepidopterous larvae, hydration of the JH epoxide is important in degradation [28]. The alternate hydrolytic pathway involves cleavage of the highly stable conjugated methyl ester. In all lepidopterous species examined a highly specific enzyme known as JH esterase has been shown to hydrolyze JH. A dramatic rise in the titer of JHE has been shown by a variety of independent methods to be a key factor in reduction of the hormone. For instance, selective inhibition of the enzyme by transition state mimics or organophosphates leads to higher than normal titers of JH, continuation of feeding behavior, and giant larvae [45,46]. Alternatively, injection of the affinity-purified enzyme into larvae of the tobacco hornworm, Munduca sextu, causes the insects to turn black. This dose-dependent blackening is associated with anti-JH effects in M . sexta and can be reversed by topical application of JH or its mimics [47]. For a variety of reasons, including its usually soluble nature and the availability of


BIOSYNTHESIS OF JH

t + REGULATION

OCH, JH (active). keep6 Inaeci In fsedlng M e t o

I DEGRADATION OF JH

Epoxlde hydroiese(s)

OCHJ OH

JH dlol (Inactive) JH acld (inactive)

Fig. 4. Regulation of juvenile hormone titers. JH synthesized within the corpora allata is exported to the hernolyrnph. JH catabolism involves epoxide hydrolases and esterases (JHE) that hydrolyze JH into the biologically inactive JH acid and JH diol metabolites. In vivo regulation of JH levels occurs via modulation of i ts rate of synthesis and degradation. These sites of regulation are logical targets for insect control agents.

rapid assays, JH esterase has been much better studied than the alternate pathway involving epoxide hydration. In many insects, epoxide hydration appears to be the dominant pathway, and for reasons discussed below, it ultimately may offer advantages over JHE for insect control.

Cloning of JH Esterase

Based on knowledge of the temporal expression of JHE, using a degenerate DNA oligonucleotide with deduced coding for the N-terminal sequence of JHE and using an antibody specific to THE, three 3 kbp cDNAs were isolated from an expression library made to mRNA isolated from fatbodies of the tobacco budworm, H . virescens. Sequence data showed the clones code for the N-terminal peptide of JHE and have the expected esterase consensus sequences [48].

Advantages of JH Esterase in Insect Control

In theory, expression of JHE should reduce JH titer, reduce iqsect feeding and possibly lead to precocious metamorphosis. One way to obtain such precocious expression involves the use of the NPV expression system described above for AaHIT. JHE offers several advantages over AaHIT in such a system. JHE should reduce or eliminate feeding while the virus replicates and ultimately kills the insects. Thus, viral-encoded JHE may disturb the biology of the viral infection cycle less than some toxins. Since JHE is an enzyme natural to the insect, it is hoped that public acceptance will be easily

326 Hammock et al.

obtained. Conventional wisdom suggests that a decreased JH titer should reduce insect feeding and induce behavioral changes without killing the insect. This attribute could be useful in production. Because NPVs currently are commercially produced in insects, one could use inexpensive JH mimics as antidotes for the recombinant virus, allowing it to be produced in normal- sized or even giant insects. Finally, many of the desirable characteristics of a reporter protein for molecular biology [49] and for microbial ecology are characteristics of JHE. These attributes outlined below will make the fulfilling of regulatory agency requirements of field monitoring economically feasible.

JH Esterase as a Reporter Enzyme JHE has many attributes ideal for a reporter enzyme. The coding region for

JHE is small (1.7 kbp) and the 60,000 molecular weight enzyme is stable even on extended storage and to repeated freeze thaw cycles. For analysis, there is a very sensitive and highly specific radiochemical assay using commercial reagents [50,51] as well as a sensitive colorimetric assay (McCutchen, unreported data). This catalytic activity is not dependent on cofactors and is tolerant of a variety of organic solvents, salts, and pH conditions. Finally, there is a sensitive and specific ELISA for the enzyme, and we recently have developed an affinity-amplified ELISA of greater sensitivity [52] and specificity.

Expression of JH Esterase For NPV expression, the coding region of one of the 3 cDNAs (termed

3Hv16) was isolated in a Bluescript plasmid and has been expressed under the polyhedrin promoter in BmNPV (Hanzlik et al., unreported data) and under the polyhedrin and p10 promoters in AcNPV [5,17,18,53]. High levels of in vitro expression were obtained in each system with over 95% of the protein being exported into the media in a glycosylated form. Expression was significantly earlier and reproducibly higher under the p10 promoter than under the polyhedrin promoter using the promoter system of Weyer et al. [16,17,54] and earlier and slightly lower than with the polyhedrin promoter using the p10 promoter system developed by Vlak et al. [18,55]. A joint study using three different culture systems recently confirmed the above results in independent trials [19]. Such work indicates that further improvement of viral expression is possible by varying the promoters used and possibly the arrangement of genes on the viral chromosome. These systems are now being used to produce JHE and site-specific modifications of JHE for catalytic site studies, pharmacokinetics, and biological testing [56].

When recombinant JHE viruses were tested in vivo in T. ni, JHE was detected in tissues and hemolymph at levels corresponding to but not greatly exceeding the levels normally observed in the last larval instar [5,17]. No physiological or behavioral effects were seen that could be attributed to the recombinant virus when older larvae were treated. When younger larvae were treated with the JHE virus, most larvae grew at a rate indistinguishable from the wildtype or a control engineered virus. However, with expression under both the polyhedrin and p10 promoters and in both T. ni and in H . virescens, stunted larvae were found which grew at a much slower rate than control or


infected control larvae. All of these stunted larvae showed high levels of JHE in the hemolymph, while such developmental effects were never seen in control populations [5,17] (McCutchen et al., unreported data). The results of this work were encouraging, but the effects were too irregular and infrequent to exploit for practical insect control.

When wildtype JHE was placed under promoters which gave polyhedrin-positive viruses [17-191 the engineered virus gave kill limes similar to or even slower than the nonengineered virus. When JHE was expressed in later larval instars, high levels of the enzyme were seen in the hemolymph with levels exceeding 1 g per liter in last instar larvae of B. mori, but no dramatic behavioral changes were observed. When the JHE was tested in late larval instars using viruses lacking the ecdysone glucose transferase gene, the virus expressing wildtype JHE also failed to cause precocious development or mortality [l].

Distribution and Pharmacokinetics A partial solution to this dilemma has been found by expanding on

unpublished observations of Sparks and Philpott. When expressed JHE was injected into M. sextu, the half-life of the protein in the hemolymph was found to be short relative to BSA, mammalian epoxide hydrolase or the hemolymph carrier protein of M. sextu. However, affinity-purified THE was stable in the presence of a variety of tissues held in vitro or in the presence of tissue homogenates [57,58] (Roe et al., unreported data).

The volume of distribution of injected JHE was very close to the blood volume, indicating that it did not penetrate rapidly into major tissues. Half-life measurements of injected JHE (1.2 h) were similar over a wide range of amounts, but the half-life could be extended by injection of large amounts of active (data not shown) or catalytically inactive JHE (Fig. 5). The half-life of injected JHE was not dramatically changed by the coinjection of large amounts of BSA (Fig. 5). These data suggest that a saturable and presumably specific receptor-mediated endocytosis is involved in the rapid removal of injected JHE, and these observations may be associated with the normal rapid removal of hemolymph JHE at critical times in insect metamorphosis [58]. A variety of experiments including reisolation of JHE and irnrnunohistochemistry strongly suggest that the pericardial cells are responsible for this rapid uptake of the enzyme from the hemolymph [58,59].

Activity of JH Esterase on Diptera Since the JHE of H. virescens is the most active enzyme isolated and since

it degrades all known homologs of JH with a high kcATto KM ratio, the enzyme will act as an efficient scavenger for JH in any insect where there is an efficient system for delivering the enzyme to the hernolymph or target tissue. The anti-juvenile hormone activity of JHE is not limited in application to agricultural pests, and could be used to control insects which transmit human disease. We have recently investigated the effects of recombinant JHE on Aedes aegypti [60]. Injection of JHE into mosquito larvae caused mortality in a dose-dependent manner. At several treatment levels, high mortality was observed within two days. The results of injecting larvae or pupae with JHE extended to the adult stage and included retardation of ovariole maturation.

328 Hammock et al.

100

m c .- .5 l o

P E

w 5

1

"\

0 1 2 3 4 5 Time after injection (hr)

Fig. 5. Pharmacokinetics of JHE showing disappearance from hemolymph is a phenomenon saturable by inactive JHE but not BSA. Proteins (0 JHE; 0 ]HE 4- BSA; WJHE + inactivated JHE) were injected into the hernocoel of second instar larvae of Manduca sexta. Determination of remaining hemolymph JHE activity was by standard partition assay [301. Injected amounts of protein were as follows: JHE, 1 nmol/min activity, 0.26 rng as protein; inactivated JHE, <0.01 nmol/rnin activity, 10.6 mg as protein; BSA, 10.6 mg. Each datum point represents four larvae assayed independently.

Thus, the impact on mosquito populations may be compounded through several life stages. Recombinant JHE, targeted to disease vectors by an appropriate delivery system, could prove to be a safe alternative for controlling vectors of human disease.

IMPROVING THE ENGINEERED JHE VIRUS

Tissue Uptake of JH Esterase In addition to the approaches outlined for the AaHIT virus which are

generally applicable to chemical mediators and enzymes, there are at least three additional ways to improve the JHE virus. First, Ichinose et al. [57,58] have shown that this chemically stable protein rapidly disappears in vivo. Since JHE is very stable in the hemolymph of normal and viral infected larvae of a number of stages, this degradation was thought to occur after uptake by specific tissues. It now is clear that THE is rapidly sequestered from the hemolymph by the pericardial cells and possibly other tissues where it presumably is degraded by one or more pathways. There is evidence from the derived amino acid sequence of JHE that a ubiquitin-mediated pathway could be involved in degradation of the enzyme [61]. Such an intracellular process involves the attachment of ubiquitin molecules via an isopeptide bond to the epsilon amino group of surface lysines [62-641. The ubiquitinated protein is then recognized and degraded by an ATP-dependent protease and the ubiquitins are recycled by ubiquitin hydrolases. By removing ubiquitin


SITE DIRECTED wH2--iu; MUTAGENESIS

CONJUGATION

NO r UBQ CONJUQATION

x ubiquitin

1 Ublqulbn’ I Ubiquitin

NH2

dFl Ubiquibn x

R I

DEGRADATION STABLE JHE

Fig. 6 . Proposed ubiquitin (USQ) degradation pathway for juvenile hormone esterase. There are two proposed recognition systems for the degradation of proteins by the ubiquitin pathway. The first involves the recognition of surface lysine residues (K) which are within an amino-acid environment rich in proline, glutamate, serine, and threonine (PEST) 1391, or show a consensus sequence of Lys-Phe-Leu-Cln [42l. Lysine residue 522 in JHE fulfills both of these requirements and is thus a potential candidate site for ubiquitin attachment via enzyme E2 and subsequent degradation of JHE. The second proposed mechanism of ubiquitination involves the recognition of a susceptible N-terminal amino acid (tryptophan in JHE) and an adjacent lysine residue by factor E3, then attachment of ubiquitin by E2 to this lysine (K29) which signals JHE for protein degradation [41]. By removing the two lysine residues implicating ubiquitin degradation of JHE, JHE should survive much longer in the insect, subsequently accumulate to much higher levels, and have a greater biological activity.

attachment and/or recognition sites, one should be able to stabilize the enzyme against degradation by this pathway (Fig. 6). In addition, there may be other proteolytic cleavage sites or targeting sequences which can be identified by metabolism studies and removed.

Ichinose et al. [58] have used several techniques including immunohis- tochemistry to demonstrate the selective uptake of JHE. The presence of recombinant JHE in pericardial cells now has been demonstrated by Booth et al. [59] at the ultrastructural level. The receptor-mediated process suggested

330 Hammock et al.

by pharmacokinetics and the specificity of uptake indicate that there are specific receptor binding sequences on JHE. Possibly these can be removed from the sequence to stabilize the protein in the blood. As we develop an understanding of specific uptake systems in insect tissues, these sequences can be used to target recombinant proteins to individual tissues. Initially, workers in the field worked with the hypothesis that the greatest biological activity would be obtained from biological molecules most closely related to the pest species to be controlled. In retrospect, it is reasonable that a species will have efficient degradation systems to remove its own regulatory molecules. Thus, peptide hormones and certainly JHE isolated from species other than H . virescens may be very attractive for insect control.

Modified JH Esterase A series of viruses have been prepared producing modified JHEs under the

control of the p10 promoter as described for AaHIT and wildtype JHE. Two of these viruses lead to a shorter time to death than the AaHIT virus when tested with identical oral doses in both H . virescens and T. ni. These data suggest that additional changes can be made in JHE and possibly other natural insect proteins to increase speed of kill further and that other insect proteins expressed either as wildtype or modified proteins can lead to insect death.

Synergism of JH Esterase It has been argued for many years that the hemolymph binding protein

clearly acts to keep JH out of lipophilic depots [65,66] and possibly to protect it from degradation by enzymes with a low affinity [67,68]. However, it can accelerate the degradation of JH in the presence of THE by making it available to the enzyme [28,68]. Expression of high levels of the JH hemolymph binding protein in the presence of high levels of JHE may act synergistically to reduce effective JH titers.

A second possibility is the use of JH epoxide hydrolase (or another epoxide hydrolase acting on JH) either independently or in combination with JH esterase. There is the possibility that JH acid produced by JHE can be recycled in the tissues by remethylation to JH by methyl transferases. Such a process would reduce the effectiveness of JHE in reducing JH titers. There is no known biochemical process to recycle a diol back to an epoxide. Since JH diol has greatly reduced biological activity, this could be an effective way to reduce JH titers. There also is indication in some species that the JH acid is an even better substrate for JH epoxide hydrolase than is JH; thus the enzymes could work synergistically. In most insects, JHE appears largely soluble and JH epoxide hydrolase largely membrane bound. Thus the enzymes also could be comple- mentary by reducing extracellular and intracellular JH.

In the insect, the reduction in JH titer occurs, due to a reduction in biosynthesis, as well as an increase in degradation. As we develop an understanding of peptides that regulate JH biosynthesis, there is the hope that these peptides (such as allatostatin) can be coexpressed with JHE to obtain a synergistic reduction in JH. The observations on AaHIT support the concept that even peptides involved in neurosecretory innervation can be used in such an approach. Just as the most active JH mimics and anti-JH chemicals have


imperfect action leading to morphogenetic effects in larvae, it may be possible to use an understanding of tissue targeting to effect local reduction of hormone titers.

THE JHE GENE: A LEAD TO NEW TARGETS

The regulation of insect metamorphosis undoubtedly is far more complex than even the most intricate scenario yet published. It seems clear that there is at least one temporal factor controlling the tissue-selective expression and export of JHE [28]. The observation that parasitization of T. rii and S. exigua by hymenopterous parasitoids in the genus Chelonus can lead to almost perfect precocious pupae in early instars with both a reduction in JH and an increase in JHE [69] suggests that both of these processes have a common regulatory factor [7O-73]. Such regulatory factors are conceptually even more attractive than JHE and allatostatin for recombinant approaches to insect control. Several approaches are possible in trying to identify these factors. One approach involves the development of appropriate bioassays and classical purification. We currently are emphasizing the approach of isolating the JHE gene and then looking for possible regulatory regions.

For this work, an 851 bp fragment corresponding to the N-terminal portion of JHE was used to screen a lambda genomic library constructed in EMBL3 from embryonic DNA of an outbred line of H. virescens. Several strongly positive clones were identified and shown to contain sequences hybridizing to all or portions of the cDNA from JHE. Following analysis using a battery of restriction enzymes and oligonucleotide probes, four of the clones were selected for further investigation (Fig. 7). Subcloning and further analysis, including partial sequencing, indicated that all of the coding region for JHE as well as substantial portions of the upstream and downstream regions were contained in two largely overlapping clones (Fig. 7). Further sequencing of these clones is underway. It is hoped that the resulting information will provide insight into the regulation of JHE, and thus to JH. Biological data indicate integrated regulation of JH biosynthesis and degradation. It is possible that factors involved in such regulation could be effective in the baculovirus or other vector systems.

POTENTIAL OF RECOMBINANT BACULOVIRUSES IN AGRICULTURE

Results both with JHE and with AaHIT illustrate that the timeline between fundamental research and practical application is becoming shorter. Ap- proaches to improve the activity of each virus clearly rely on fundamental knowledge of both the target insect and the virus. As will be discussed below, such information is needed for wider practical exploitation of both the virus and recombinant protein and to address regulatory questions. In addition, the process of solving many of the practical problems of engineered viruses will help us to address fundamental questions in regulatory and developmental biology.

332 Hammock et al.

574-851 2479-2986

1 1123f-17f4 I .I,. 1-227

JHEcDNA 5 , ~ I

Probe- 1 2 3 4 2986 bp v v v v

GENOMIC

CLONES .F . + -

Deduced Alignment of Clones __ B I

DT El F L

5'- JHEGene -3'

Fig. 7. Characterization of JHE genomic clones. Cloned genomic DNA (cDNA) was dot blotted onto nylon membranes, then probed with radiolabeled DNA fragments (probes 1-4) derived from ]HE cDNAclone 3Hvl6 (top). Hybridization of probes 1 4 to the genomic clones (B,D,E,F, bottom) was used to indicate if the cloned genomic DNA corresponded to the JHE message. Hybridization was performed in 7% sodium dodecyl sulfate, 1 mM EDTA, 0.5M NaP04 at 60°C. After hybridization, membranes were washed twice in 2 x SSC at 65"C, once in 2 x SSC plus 0.1% sodium dodecyl sulfate at 6S"C, then once in 0.1 x SSC at 65°C.

Exploiting Current Knowledge

There are numerous environmental reasons to think of the recombinant viruses as insecticides, but there are commercial reasons as well. If we develop an analogy between the engineered viruses expressing AaHlT or straw itch mite toxin versus a lead structure in an early insecticide screen, it becomes obvious, as shown in Figure 3, that there are numerous, systematic ways to improve the properties of the engineered viruses. It also is likely that improvements in the virus, the toxin, delivery systems, and other factors will be additive if not synergistic. As we compare the potential of these engineered viruses to classical insecticides, it is important to consider that insecticide development is based on the accumulated knowledge of decades of research in both the agricultural and pharmaceutical industries. In contrast, the concept of using biologically active peptides commercially, much less an engineered virus as a delivery system, is very new. Progress is certain to be exciting, but it is likely to be slow as well until a critical mass of scientists and industries are involved.


There are many differences between the engineered viruses and classical insecticides. For instance, we have the opportunity to improve the effectiveness of a toxin expressed in a bacterium or virus at the DNA, RNA, or protein level in contrast to a single structure with classical compounds. The low levels of expression of AaHIT, Txp-I, and JHE in early instars relative to other proteins studied certainly leave room for improvement. It could well be that the biologically active materials discussed here are just lead structures that illustrate a concept, and the materials actually used in the field will be very different. If this is the case, the paradigms developed for improving engineered organisms and integrating them into a pest management framework are likely to be far more important than the early products themselves.

Use of Other Chemical Mediators The effectiveness of AaHIT when expressed in AcNPV brings up several

possibilities. First, by illustrating the concept of an engineered viral insecticide, it provides a test system for a wide variety of chemical mediators which may be useful in insect control. Maybe of greatest importance is that it provides a commercial justification for the investigation of the biochemical basis of arthropod-arthropod interactions. Structures of many peptides biologically active on insects have been published as side projects of studies on human toxicity or as interesting probes for fundamental studies of the nervous system [30,31,74]. Now there is a clear practical outlet for research on peptides which influence insects. Bionomics provide us with a plethora of elegant examples of potential leads that can be exploited, utilizing this system with arachnids [32,41,75,76] and parasitoids being most obvious [40,42]. Clearly the first targets for expression must be peptides like AaHIT which are highly specific for insects. However, the specificity of the NPV system itself probably provides sufficient safeguards for more general toxins to be explored in the future. Possibly even peptides with toxicity to vertebrates will serve as useful leads, since their toxicity may be modified by mutagenesis to change receptor binding or pharmacokinetic properties.

Use of Baculoviruses to Test Peptides Scientists have long speculated on the utility of an insect-specific vector

system for the delivery of peptides and proteins for insect control [77,78]. The efficacy of this concept now is clearly proven. Just as the viral vector system can be used for many peptides, AaHIT now is of proven utility and can be applied to other biological systems. In this work, the baculovirus system may play a role in producing and testing constructs and mutants to insure that they maintain their effects on insects before more complex engineering in plants or other expression systems is attempted [13]. There are a variety of possible applications of these peptides, many leading possibly to subtle effects on behavior.

Dangers of Over Exploitation Many of the problems that we now face with current insecticides do not

stem from the products themselves. The problems stem from capitalistic factors which force the overuse of insecticides and from the tendency to use

334 Hammock et ai.

insecticides to absolve our lack of understanding of ecosystems. If the engineered viruses meet the multiple criteria needed for successful development, there is the danger that they will face the same problems of misuse as classical synthetic compounds [79].

Potential for Misuse of AaHIT

As recombinant biological approaches are considered with AaHIT and similar peptides, it is important to move cautiously with an appreciation for the sensitivities of the public. The concept of avoiding the image of a biotechnology barbarian is important. We must avoid squandering the AaHIT gene in the fashion that the genes for the Bacillus thuringiensis endotoxins are being exploited [80,81]. Inappropriate exploitation clearly will lead to resistance. With the advent of every new insect control strategy, some scientists always claim that resistance will be impossible. Modern experience with insecticides demonstrates that the only compounds for which resistance does not occur are those which are worthless. This is a blunt way of saying that successful materials always cause a selection pressure. If properly used, the engineered baculoviruses have properties that should overcome much cross resistance and retard resistance development. They are very promising tools for resistance management. However, if the AaHIT gene or JHE are overused, resistance will be accelerated. If used properly, wildtype and engineered NPVs could be very valuable in resistance management programs [82].

Along these lines, the success of one approach in biotechnology encourages analogous approaches and raises difficult ethical questions. For instance, AaHIT could be expressed easily in cyanobacteria (blue green algae) for the control of aquatic Diptera such as mosquitoes. There clearly are mosquito control problems where there is little risk of ecological disruption with such an organism. In the paragraphs below some of the questions that this exercise illustrates are presented.

There are reasons for serious concern about expression of toxins in algae, bacteria, or other reproducing components of an ecosystem. Even engineered baculoviruses that could compete in nature raise concern [83]. The rapid evolution of resistance could be a problem when there is significant survival of aquatic Diptera if organisms producing a toxin are present for extended periods. A second concern is ecological. The toxin-secreting organisms could cause extensive mortality to aquatic insect guilds, with ecological damage as a possible consequence. The third concern is that, if released from biological control, the engineered cyanobacteria or other organisms could become an environmental problem. Although many mosquito outbreaks occur in ecolog- ically disturbed areas where recombinant cyanobacteria would not create a problem, the cyanobacteria and algae are notorious in their ability to distribute quickly. Finally, it is uncertain how recombinant organisms will be regulated. When considering recombinant DNA systems for pest control, it may be best to pave the way with an approach that is perceived to be the safest and the most innocuous so that early experiments do not lead to public reactions or a stifling regulatory climate. With this example one can see that expressing AaHIT in an important component of an aquatic ecosystem clearly would not be considered innocuous by most ecologists.


Risk Assessment, National vs. Global

A complicating factor is that the ethics of biotechnology, like the ethics or risk benefit analysis of classical pesticides, will vary depending upon when decisions are being made and where they are made. Deciding on the use of recombinant biologicals against insect-vectored diseases is easy in a rich nation with relatively minor problems with arthropod-borne disease. However, given that millions of people suffer from such disease, is it ethical to retard the technology based on a small or incalculable chance of undesirable effects? The technology involved in genetic engineering is not the exclusive property of developed countries. Even if a national risk benefit analysis of a construct shows it to be undesirable in a developed country, that construct probably can be made and tested in a country where the risk benefit analysis is different. As shown with classical pesticides like DDT, there are no borders to ecological effects. Thus, would a developed country be better off to make and carefully test a questionable construct like cyanobacteria-expressing AaHIT, rather than relying on countries lacking the resources for proper evaluation?

The context of issues in biotechnology can be complex, as in this case involving human health and the environment. Moreover, these issues are embedded in an uncertain regulatory structure and an unpredictable public.

Use of Recombinant Viruses in Developing Countries

The above arguments may, in fact, be immediately relevant to the viruses already described in the literature [3,4,6,7]. Pest control in many developing countries is limited due to the very high cost of classical insecticides even by Western standards. The industrial base to produce these materials locally often is lacking. Disasters with classical materials such as Leptophos have demonstrated that shipment of materials banned in developed countries to developing countries is dangerous. The infrastructure to establish effective pest management systems often is lacking, and the impact of problems associated with overuse of classical pesticides such as resistance, resurgence, and toxicity may be even greater than in developed countries.

The engineered baculoviruses offer the potential for production in cottage industries. It would be administratively simple for international organizations to provide periodic shipments of innocula of either wildtype or engineered organisms for local production. The labor costs which make in vivo virus production of lower profitability in developed countries will be less of a problem in developing areas. Several successful control programs are in use in developing countries based on the local production of viruses. The same technology could be used to produce recombinant viruses if they offer advantages in pest management. With recombinant viruses potentially useful in developing countries, international organizations must develop guidelines for their testing and use based on local as well as international risk/benefit evaluation.

336 Hammock et al.

IMPACT ON TOXICOLOGY AND REGULATION

Impact on Toxicology Simply expressing materials to reduce feeding or enhance speed of kill does

not solve all of the problems of NPVs as biological pesticides (Table 1). However, it should provide the profit motive to support the allocation of resources to solve these problems along the lines shown in Table 2. A major worry involves agency and public acceptance. How the public will accept these materials is uncertain and depends in part upon how the topic is presented. Possibly the recent clinical successes with recombinant approaches will lead the public to look with more favor upon both genetic engineering and viruses. We must take an active role in working with the public, the press and regulators to establish effective standards.

There are clear approaches that toxicologists and environmental microbiol- ogists can take to facilitate acceptance. First, the recombinant pesticides must be viewed by industry as valuable commodities worthy of research in their own right rather than a cheap biological used to avoid registration costs. Thus, we need high standards of production and quality control as well as the analytical technology to support it [84]. Unfortunately only a few markets will support such work. Second, the development of these biological insecticides must be supported by high quality science. Metabolism, environmental analysis, and risk assessment on biopolymers will present new challenges to toxicologists [57]. This work is critical for the rational design of materials with improved efficacy as well as for insuring that biological pesticides are safe for both man and the environment.

Semantic Problems As recombinant pesticides become commercial, we must be very careful in

our use of terms. Although semantics may help us to gain regulatory ap- proval, we must not use these terms to delude ourselves or the public. Historically, Alexander Fleming’s use of the term antibiotic (anti-life) to describe diverse chemicals isolated from organisms perceived as distasteful by the general public has not prevented wide acceptance of these often highly toxic chemicals [85]. Many antibiotics have such a great toxicity and such an unfavorable therapeutic index that they could not be registered as agricultural chemicals today.

A caution regarding the use of terms comes from the early development of juvenile hormone mimics where the US. Environmental Protection Agency objected to the term ”hormone.” However, the agency was comfortable with the term “insect growth regulator,” which on analogy with plant growth regulator, suggests a chemical which will make insects grow better. Hope- fully, recombinant biological pesticides will be accepted based on their relative benefit and risk rather than semantics.

It is possible that the term ”chemical” now has such unfavorable connota- tions that it will soon be lost to the English language. This situation is not helped by the implication by some scientists that in contrast to insecticides, biologicals are composed of some unknown matter which is not chemical. It also must be remembered that these recombinant viruses as well as wildtype


viruses are proposed insecticides. Sparks and Hammock [27] suggested that pesticide generations be defined based on the method of discovery rather than the final compound. By analogy, a return to the early concepts of biocontrol where we define an agent or organism based on how it is used is valuable. Natural biological control [86,87] and the classical biological control procedure of importing natural enemies or conserving natural enemies of pests [88,89] clearly are distinct from a pesticide philosophy, yet the same organism can be used for each. Ehler [90] provides a concise summary of these traditional definitions of categories of biocontrol. Clearly, some augmentative approaches utilize a pesticide approach regardless of whether the organism is natural or engineered [79]. Many of the ethical and regulatory considerations suggested by Hoy et al. [91] for commercial releases of natural enemies apply equally to recombinant organisms. The term “biological insecticide” [79] seems to describe these agents that will be used for inundative releases, as defined by Cate [92]. The term “biological insecticide” or the broader term “biological pesticide” for those materials that will be used in a manner similar to classical synthetic insecticides gives the positive connotation of the origin of the material and cautions against careless use without giving too many negative impressions [79].

As emphasized by Strong [80], transgenic techniques have a role in biological control, but they are not per se biocontrol. Neither synthetic pesticides nor biological pesticides are inherently good or evil; rather, they have properties that may lend themselves to proper use or misuse. Syn- thetic pesticides have set a very high standard of efficacy, sometimes at a high cost to the environment. Hopefully the expensive lessons learned with the use of classical synthetic pesticides will not be forgotten as we develop biological pesticides. Many of the biological pesticides, including the recombinant baculoviruses described here, have numerous desirable properties which can be exploited in pest management programs designed with an appreciation for the ecosystem.

A Dilemma of Terms

If the engineered viruses described here are to be used in insect control, they need to be carefully tested. Superficially it is difficult to imagine a biological insecticide with more desirable environmental properties for pest management or greater safety to man than the AcNPV that expresses AaHITl . However, from a semantic standpoint, the engineered viruses expressing AaHIT present a dilemma. Although apparently safe, a genetically engineered virus expressing a toxin from a scorpion is labeled with at least four terms viewed as alarming by the general public. In discussions with R. Possee and his research team, it was agreed that we use the term ”toxin” for peptides which have been described in the literature as toxins; however, the more general term ”peptide” or ”protein” is to be used for materials not previously described in the literature as toxins. One can hope that the public will develop a more mature understanding of genetic engineering and viruses, if not an appreciation of the beauty of scorpions.

338 Hammock et al.

Regulation

Safety tests should not be ignored simply because a material is a biological rather than a synthetic pesticide. However, we can hope that regulatory agencies will suggest rational tests based upon the properties of materials and their proposed uses with both biological and synthetic pesticides. A poor regulatory climate has stifled the development of the selective classical insecticides that would be most valuable for integrated pest management [20]. The same problem could occur with biological pesticides.

An excellent methodical approach in the evaluation of environmental impact of recombinant NPVs has been taken in Britain [93]. Such work has encompassed environmental microbiology and rigorous testing on nontarget organisms. Williamson [83] brings up that for initial releases, a narrow spectrum of activity probably will be seen as beneficial by regulatory agencies. Although the major problems caused by the use of classical synthetic pesticides have been environmental, regulation and testing in this country have been driven by human health concerns. The British studies have been pleasing in that they have concentrated on the ecosystem questions where deleterious impact is most likely to be seen.

As we look at potential risks of engineered viruses, it is important that we consider both the recombinant material being produced and the genetic environment into which it is inserted. In the case of JHE, AaHIT, and the mite toxin inserted into AcNPV, both the expressed materials and vectors appear safe. The products encoded by the genes carried by these recombinant viruses play a role in accelerating the killing rate of the insect pest, but do not confer any selective ecological advantage on the recombinants over the wildtype NPVs. In fact, there is a selective disadvantage of all these recombinants, since they produce less viral progeny. This offers an advantage to commercial producers, since the recombinants will not recycle extensively in the environment. It also allays the fears expressed by Williamson [83] that a recombinant virus could become established to the detriment of natural fauna.

While it is conceivable that recombinant baculoviruses may acquire altered host range capabilities, the available data suggest that the host range deter- minant(s) is localized to a rather specific locus in the baculovirus genome. There is no evidence to suggest that this locus is more mobile than others in the baculovirus genome. Thus, the probability for horizontal transmission of the engineered genes to nontarget insect populations would be negligible (Maeda, unpublished). The sequencing efforts in progress in the laboratories of Possee and Miller on AcNPV, and Maeda with BmNPV, certainly will shed more light on the structure-function relationships of various genes which will enable us to construct still safer disabled baculoviruses for use as insecticides. For example, the p10 minus constructs of Vlak and associates are likely to be even less competitive in the field with wildtype NPV than the p10 positive constructs [18,55].

Now that the commercial potential of recombinant NPVs has been clearly demonstrated, we need to perform those experiments that will test if the recombinant organisms offer significant risk under the most rigorous conditions. The scientific method of employing experiments designed to demon-


strate that biological pesticides are not safe, will provide the most rigorous tests of their properties. By such an approach, we can speed the acceptance of a valuable new set of tools for integrated pest management, if they are indeed safe, Our obIigation is not only to provide the data that will satisfy agencies and the public that recombinant NPVs are relatively innocuous, but to educate the public on riskhenefit analysis and to be educated by the public on their concerns regarding these recombinant pesticides.

LITERATURE CITED

1. OReilly DR, Miller LK: A baculovirus blocks insect molting by producing ecdysteroid UDP-glucosyl transferase. Science 245, 1110 (1989).

2. Maeda S : Increased insecticidal effect by a recombinant baculovirus carrying a synthetic diuretic hormone gene. Biochem Biophys Res Commun 165, 1177 (1989).

3. Maeda S, Volrath SL, Hanzlik TN, Harper SA, Maddox DW, Hammock BD, Fowler E: Insecticidal effects of an insect-specific neurotoxin expressed by a recombinant baculovirus. Virology 284, 777 (1991).

4. McCutchen BF, Choudary PV, Crenshaw R, Maddox D, Kamita SG, Palekar N, Volrath S, Fowler E, Hammock BD, Maeda S : Development of a recombinant baculovirus expressing an insect-selective neurotoxin: Potential for pest control. Biotechnology 9, 848 (1991).

5. Hammock BD, Bonning B, Possee RD, Hanzlik TN, Maeda S: Expression and effects of the juvenile hormone esterase in a baculovirus vector. Nature 344,458 (1990).

6. Stewart LMD, Hirst M, Ferber ML, Merryweather AT, Cayley PJ, Possee RD: Construction of an improved baculovirus insecticide containing an insect-specific toxin gene. Nature 352, 85 (1991).

7. Tomalski MD, Miller LK: Insect paralysis by baculovirus-mediated expression of a mite neurotoxin gene. Nature 352, 82 (1991).

8. Zlotkin E, Fraenkel G, Miranda F, Lissitzky 5: The effect of scorpion venom on blow fly larvae; a new method for the evaluation of scorpion venom potency. Toxicon 8, 1 (1971).

9. Walther C, Zlotkin E, Rathmayer W: Action of different toxins from the scorpion Androctonus australis on a locust nerve-muscle preparation. J Insect Physiol22, 1187 (1990).

10. Teitelbaum Z, Lazarovici P, Zlotkin E: Selective binding of the scorpion venom insect toxin to insect nervous tissue. Insect Biochem 9, 343 (1979).

11. Gordon D, Jover E, Couraud F, Zlotkin E: The binding of the insect selective neurotoxin (AaIT) from scorpion venom to locust synaptosomal membranes. Biochem Biophys Acta 778, 349 (1984).

12. Maeda S: Expression of foreign genes in insects using baculovirus vectors. Annu Rev Entomol34, 351 (1989).

13. Dee A, Belagaje RM, Ward K, Chi0 E, Lai M-HT: Expression and secretion of a functional scorpion insecticidal toxin in cultured mouse cells. Biotechnology 8, 339 (1990).

14. Miller LK: Baculoviruses as gene expression vectors. Annu Rev Microbiol42, 177 (1988).

340 Hammock et al.

15. Luckow VA: Cloning and expression of heterologous genes in insect cells with baculovirus vectors. In: Recombinant DNA Technology and Applications. Prokob A, Bajpai RK, Ho C, eds. McGraw-Hill, New York, pp 97-152 (1991).

16. Weyer U, Knight S, Possee RD: Analysis of very late gene expression by Autograph californica nuclear polyhedrosis virus and the further development of multiple expression vectors. 1 Gen Virol71, 1525 (1990).

17. Bonning BC, Hirst M, Possee RD, Hammock BD: Further development of a recombinant baculovirus insecticide expressing the enzyme juvenile hormone esterase from Heliothis virescens. Insect Biochem Molec Biol22, 453 (1992).

18. Roelvink PW, van Meer MMM, de Kort CAD, Possee RD, Hammock BD, Vlak JM: Temporal expression of Autographa californica nuclear polyhedrosis virus polyhedrin and p10 gene. J Gen Virol73, 1481 (1992).

19. Bonning BC, Roelvink PW, Vlak JM, Possee RD, Hammock BD: Comparison of expression characteristics of various promoters in the Autograph californica nuclear polyhedrosis virus using juvenile hormone esterase as a reporter enzyme. J Gen Virol, Submitted (unreported data).

20. Hammock BD, Soderlund DM: Chemical strategies for resistance management. In: Pesticide Resistance: Strategies and Tactics for Management. Glass E, ed. National Academy Press, Washington D.C., pp 111-129 (1986).

21. Granados RR, Federici BA: The biology of baculoviruses. Biological properties and molecular biology. CRC Press, Boca Raton, vol. I, p 275 (1986).

22. Granados RR, Federici BA: The biology of baculoviruses. Practical application in insect controls. CRC Press, Boca Raton, vol. 11, p 276 (1986).

23. Kondo A, Maeda S: Host range expansion by recombination of the baculoviruses Bornbyx mori nuclear polyhedrosis virus and Autograph californica nuclear polyhedrosis virus. J Virol 65, 3625 (1991).

24. Maeda S, Mukohara Y, Kondo A: Characteristically distinct isolates of the nuclear polyhedrosis virus from Spodopteru lifura. J Gen Virol72, 2631 (1990).

25. Bougis PE, Rochat H, Smith LA: Precursors of Androctonus australis scorpion neurotoxins. J Biol Chem 264, 19259 (1989).

26. Williams CM: Third-generation pesticides. Sci Amer 217, 13 (1967).

27. Sparks TC, Hammock BD: Insect growth regulators: Resistance and the future. In: Pest Resistance to Pesticides: Challenges and Prospects. Georghiou GP, Satio T, eds. Plenum Press, New York, pp 615-668 (1983).

28. Hammock BD: Regulation of juvenile hormone titer: Degradation. In: Comprehensive Insect Physiology, Biochemistry, and Pharmacology. Kerkut GA, Gilbert LI, eds. Pergamon Press, New York, pp 431472 (1985).

29. Herrman R, Fishman L, Zlotkin E: The tolerance of lepidopterous larvae to an insect-selective neurotoxin. Insect Biochem 20, 625 (1990).

30. Loret El’, Mansuelle P, Rochat H, Granier C: Neurotoxins active on insects: Amino acid sequences, chemical modifications, and secondary structure estimation by circular dichroism of toxins from the scorpion Androctonus australis Hector. Biochemistry29, 1492 (1990).


31. Loret El', Martin-Eauclaire M-F, Mansuelle P, Sampieri F, Granier C, Rochat H: An anti-insect toxin purified from the scorpion Androctonus australis Hector also acts on the a- and p-sites of the mammalian sodium channel: Sequence and circular dichroism study. Biochemistry 30, 633 (1991).

32. Zlotkin E, Miranda F, Rochat H: Venoms of Buthinae. In: Arthropod Venoms. Bettini S, ed. Springer-Verlag, New YorklBerlin, pp 317369 (1978).

33. El Ayeb M, Bahraoui EM, Granier C, Rochat H: Use of antibodies specific to defined regions of scorpion a-toxin to study its interaction with its receptor site on the sodium channel. Biochemistry 25, 6671 (1986).

34. Fontecilla-Camps JC, Habersetzer-Rochat C, Rochat H: Orthorhombic crystals and three dimensional structure of the potent toxin I1 from the scorpion Androctonus australis Hector. Proc Natl Acad Sci USA 85, 7443 (1988).

35. Fontecilla-Camps JC: Three-dimensional model of the insect-directed scorpion toxin from Androctonus australis Hector and its implication for the evolution of scorpion toxins in general. J Mol Evol29, 63 (1989).

36. Granier C, Novotny J, Fontecilla-Camps JC, Fourquet P, El Ayeb M, Bahraoui E: The antigenic structure of a scorpion toxin. Molec Immun 26, 503 (1989).

37. Amato I: Scorpion toxin tells an evolutionary tale. Sci News 239, 85 (1991).

38. Bontems F, Roumestand C, Gilquin B, Menez A, Toma F: Refined structure of chary- bdotoxin: Common motifs in scorpion toxins and insect defensins. Science 254, 1521 (1991).

39. Loret EP, Sampieri F, Granier C, Miranda F, Rochat H: Scorpion toxins affecting insects. Meth Neurosciences 8, 381 (1992).

40. Piek T Venoms of the Hymenoptera: Biochemical, pharmacological and behavioural aspects. Academic Press, Orlando, p 570 (1986).

41. Zlotkin E: Toxins derived from arthropod venoms specifically affecting insects. In: Com- prehensive Insect Physiology, Biochemistry and Pharmacology. Kerkut GA, Gilbert LI, eds. Pergamon Press, Oxford, UK, Vol. 10, pp 499-546 (1985).

42. Coudron TA: Nonparalytic venoms of the parasitic Hymenoptera. In: Naturally Occurring Pest Bioregulators. Hedin P, ed. ACS Symposium Series No. 449, pp 4145 (1991).

43. Zilberberg N, Zlotkin E, Gurevitz M: Molecular analysis of cDNA and the transcript encoding the depressant insect selective neurotoxin of the scorpion Leiurus quinquesfriafus hebraeus. Insect Biochem Molec Biol22, 199 (1992).

44. Staal GB: Anti juvenile hormone agents. Annu Rev Entomol31, 391 (1986),

45. Sparks TC, Hammock BD: Comparative inhibition of the juvenile hormone esterases from Trichoplusia ni, Tenebrio rnolitor and Musca domestica. Pestic Biochem Physiol 14, 290 (1980).

46. Abdel-Aal YAI, Hammock BD: Transition state analogs as ligands for affinity purification of juvenile hormone esterase. Science 233, 1073 (1986).

47. Philpott ML, Hammock BD: Juvenile hormone esterase is a biochemical anti-juvenile hormone agent. Insect Biochem 20, 451 (1990).

342 Hammock et al.

48. Hanzlik TN, Abdel-Aal YAI, Harshman LG, Hammock BD: Isolation and sequencing of cDNA clones coding for juvenile hormone esterase from Heliothis virescens: Evidence for a charge relay network of the serine esterases different from the serine proteases. J Biol Chem 264, 12419 (1989).

49. Pugsley AP: Protein targeting. Academic Press, San Diego, p 279 (1989).

50. Hammock BD, Sparks TC: A rapid assay for insect juvenile hormone esterase activity. Anal Biochem 82, 573 (1977).

51. Hammock BD, Roe RM: Analysis of juvenile hormone esterase activity. In: Methods in Enzymology. Law JH, Rilling HC, eds. Academic Press, Orlando, vol 111, pp 487494 (1985).

52. SzCkhcs A, Gee SJ, Jung F, McCutchen BF, Hammock BD: Unreported data.

53. Hammock BD, Szekacs A, Hanzlik T, Maeda S, Philpott M, Bonning B, Possee R: Use of transition state theory in the design of chemical and molecular agents for insect control. In: Recent Advances in the Chemistry of Insect Control 11. Crombie L, ed. Royal Society of Chemistry, Cambridge, U.K. pp 256-277 (1990).

54. Bonning BC, Merryweather AT, Possee RD: Genetically engineered baculovirus insecticides. Agric Biotech News Info 3, 209 (1991).

55. Vlak JM, Schouten A, Usmany M, Belsham GJ, Klinge-Roode EC, Maule A, van Lent JWM, Zuidema D: Expression of a cauliflower mosaic virus gene I using a baculovirus vector based on the p10 gene and a novel selection method. Virology 178, 312 (1990).

56. Ward VK, Bonning BC, Huang T, Shiotsuki T, Griffeth VN, Hammock BD: Analysis of the catalytic mechanism of juvenile hormone esterase by site-directed mutagenesis. Int J Biochem 24,1933 (1992).

57. Ichinose R, Harshman LG, Ward V, McCutchen BF, Uematsu T, Huang TL, Beetham JK, Maeda S, Hammock BD: The role of a baculovirus expression system in xenobiotic metabolism: Insecticidal activity, pharmacokinetics in insects and catalytic mechanism of expressed esterases. Xenobiotica In Press.

58. Ichinose R, Kamita SG, Maeda S, Hammock BD: Pharmacokinetic studies of therecombinant juvenile hormone esterase in Manduca sexta. Pestic Biochem Physiol 42, 13 (1992).

59. Booth TF, Bonning BC, Hammock BD: Localization of juvenile hormone esterase during development in normal and in recombinant baculovirus infected Trichoplusia ni larvae. Tissue Cell 24, 267 (1992).

60. Harshman LG, Vickers JM, Ichinose R, Grant DF, Ward VK, Eldridge BF, Hammock BD: Effect of recombinant juvenile hormone esterase on Aedes aegypti. Proc Calif Vector Control Assn 59, 77 (1991).

61. Rogers S, Wells R, Rechsteiner M: Amino acid sequences common to rapidly degraded proteins: The PEST hypothesis. Science 234, 364 (1986).

62. Bachmair A, Finley A, Varshavsky A: In viva half-life of a protein is a function of its amino-terminal residue. Science 234, 179 (1986).

63. Bachmair A, Varshavsky A: The degradation signal in a short-lived protein. Cell 56,1019 (1989).

64. Dice JF, Chiang H-L, Spencer EP, Backer JM: Regulation of catabolism of microinjected ribonuclease. J Biol Chem 261, 6853 (1986).

Development of Recombinant Viral insecticides 343

65. Hammock B, Nowock J, Goodman W, Stamoudis V, Gilbert LI: The influence of hemolymph-binding protein on juvenile hormone stability and distribution in Munducu sextu fat body and imaginal discs in nitro. Mol Cell Endocrin 3, 167 (1975).

66. Akamatsu Y, Dunn PE, Kezdy FJ, Kramer KJ, Law JH, Reibstein D, Sanburg LL: Biochemical aspects of juvenile hormone action in insects. In: Control Mechanisms in Development. Meints RH, Davies E, eds. Plenum Press, New York, pp 125149 (1975).

67. Kramer KJ, Sanburg LL, Kezdy FJ, Law JH: The juvenile hormone binding protein in the hemolymph of Munducu sextu Johannson (Lepidoptera: Sphingidae). Proc Natl Acad Sci USA 71, 493 (1974).

68. Goodman WG, Chang ES: Juvenile hormone cellular and hemolymph binding proteins. In: Comprehensive Insect Physiology, Biochemistry and Pharmacology. Kerkut GA, Gilbert LI, eds. Pergamon Press, Oxford, vol7 Endocrinology, pp 491-510 (1985).

69. Biihler A, Hanzlik TN, Hammock BD: Effects of parasitization of Trichoplusiu ni by Chelonus sp. Physiol Entomol 10,383 (1985).

70. Jones D, Jones G, Hammock BD: Developmental and behavioral responses of larval Trichoplusiu ni to parasitization by an imported braconid parasite Chelonus sp. Physiol Entomol6, 387 (1981).

71. Grossniklaus-Burgin C, Lanz F, Lanzrein B: Influence of the parasitoid Chelonus sp. on the endocrinology of its host, Trichoplusiu ni. In: International Conference of Endocrinological Frontiers in Physiological Insect Ecology. Sehnal F, Zabza A, Denlinger DL, eds. Wroclaw Technical University Press, Wroclaw, Poland, pp 437441 (1988).

72. Grossniklaus-Burgin C, Connat JL, Lanzrein B: Ecdysone metabolism in the hosvparasitoid- system Trichoplusiu nilchelonus sp. Arch Insect Biochem Physiol11, 79 (1989).

73. Kunkel JG, Grossniklaus-Burgin C, Karpells ST, Lanzrein B: Arylphorin of Trichoplusiu ni: Characterization and parasite induced precocious increase in titer. Arch Insect Biochem Physiol 13, 117 (1990).

74. Kharrat R, Darbon H, Granier C, Rochat H: Structure-activity relationships of scorpion a-neurotoxins: Contribution of arginine residues. Toxicon 28, 509 (1990).

75. Eitan M, Fowler E, Herrmann R, Duval A, Pelhate M, Zlotkin E: A scorpion venom neurotoxin paralytic to insects that affects sodium current inactivation: Purification, primary structure, and mode of action. Biochemistry 29, 5941 (1990).

76. Zlotkin E, Eitan M, Bindokas VP, Adams ME, Moyer M, Burkhart W, Fowler E: Functional duality and structural uniqueness of depressant insect-selective neurotoxins. Biochemistry 30, 4814(1991).

77. Miller LK, Lingg AJ, Bulla LA, Jr: Bacterial, viral, and fungal insecticides. Science 219, 715 (1983).

78. Kirschbaum JB: Potential implication of genetic engineering and other biotechnologies to insect control. Annu Rev Entomol30, 51 (1985).

79. Hammock BD: Virus release evaluation. Nature 355, 119 (1992).

80. Strong DR: Interface of the natural enemy and environment. In: New Directions in Biological Control: Alternatives for Suppressing Agricultural Pests and Diseases. Baker RR, Dunn PE, eds. Alan R. Liss, Inc., New York, pp 57-64 (1990).

344 Hammock et al.

81. Williams S, Friedrich L, Dincher S, Carozzi N, Kessmann H, Ward E, Ryals J: Chemical regulation of Bacillus thuringiensis &endotoxin expression in transgenic plants. Biotechnol- ogy 10, 540 (1992).

82. Ignoffo CM, Roush RT: Susceptibility of permethrin- and methomyl-resistant strains of Heliothis virescens (Lepidoptera: Noctuidae) to representative species of entomopathogens. J Econ Entomol 79, 334 (1986).

83. Williamson M: Biocontrol risks. Nature 353, 394 (1991).

84. Hammock BD: Applications of immunochemistry in crop protection and biotechnology, an overview. In: Biotechnology for Crop Protection. American Chemical Society Symposium Series. Hedin PA, Menn JJ, Hollingworth Rh4, eds. American Chemical Society, Washing- ton, DC, vol379, pp 298-305 (1988).

85. Maurois A: The life of Sir Alexander Fleming. E.P. Dutton & Co., New York, p 293 (1959).

86. Doutt RL, DeBach P: Some biological control concepts and questions. In: Biological Control of Insect Pests and Weeds. DeBach P, ed. Reinhold Publishing Co., New York, pp 11S142 (1964).

87. Huffaker CB, Messenger PS, DeBach P: The natural enemy component in natural control and the theory of biological control. In: Biological Control. Huffaker CB, ed. Plenum Press, New York, pp 16-67 (1971).

88. DeBach P: The scope of biological control. In: Biological Control of Insect Pests and Weeds. DeBach P, Schlinger EI, eds. Reinhold Publishing Corp., New York, pp 3-20 (1965).

89. Parker FD: Management of pest populations by manipulating densities of both hosts and parasites through periodic releases. In: Biological Control. Huffaker CB, ed. Plenum Press, New York, pp 365-376 (1971).

90. Ehler LE: Planned introductions in biological control. In: Assessing Ecological Risks of Biotechnology. Ginzburg LR, ed. Butterworth-Heinemann, Boston, pp 21-39 (1991).

91. Hoy MA, Nowierski RM, Johnson MW, Flexner JL; Issues and ethics in commercial releases of arthropod natural enemies. Amer Entomol37, 74 (1991).

92. Cate J R Biological control of pests and diseases: Integrating a diverse heritage. In: New Directions in Biological Control: Alternatives for Suppressing Agricultural Pests and Dis- eases. Baker RR, Schlinger EI, eds. Alan R. Liss, Inc., New York, pp 23-43 (1990).

93. Bishop DHL, Entwistle PF, Cameron IR, Allen IR, Possee R D Field trials of genetically-engineered baculovirus and Noctuide. In: The Release of Genetically-Engineered Microorga- nisms. Sussman M, Collins CH, Skinner FA, Stewart-Tull DE, eds. Academic Press, New York, pp 143-179 (1988).

development of recombinant viral insecticides by expression of … · 2013-10-15 · archives of...

Documents