Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

16.1 Introduction

There are at least 12 viral families associated with insects and other arthropods (Erlandson 2008). The Baculoviridae is the most commonly investigated with regard to its development as a microbial insecticide due to its favorable characteristics such as safety to the environment, humans, other vertebrates, plants, and natural enemies of pests (particularly predators and parasitoids). These viruses are generally highly selective, not affecting other insect species including those that are pests. Consequently, the baculoviruses are ideal control agents to be used in integrated pest management (IPM) programs in agriculture, forests, and pastures. Use of these agents as microbial insecticides was advocated in the 1960s and 1970s (e.g., Ignoffo and Couch 1981; Tanada and Kaya 1993; Cunningham 1995; Moscardi 1999; Szewczyk et al. 2006, 2009).

Baculoviruses have also proven to be extremely valuable tools in biotechnology. The baculovirus–insect cell expression system has become one of the most widely used systems for routine production of recombinant proteins. More recently, baculoviruses have demonstrated the ability to make ideal vectors for a variety of mammalian cell lines and are potential candidates in gene therapy (Kost et al. 2005; Hitchman et al. 2009).

16.2 State of Taxonomy and Biology of Baculoviruses

16.2.1 Taxonomy

Baculoviruses are a large and diverse group of viruses pathogenic to arthropods, primarily insects from the orders Lepidoptera, Hymenoptera, and Diptera. More than 700 baculoviruses have been isolated from invertebrates and reported in the literature (Moscardi 1999; Herniou and Jehle 2007). These viruses occur naturally in insect populations and are normally named for the initial host from which they were isolated. Owing to their high virulence, specificity to insects, and environmental stability, they have been widely used as bioinsecticides for the control of numerous agricultural and forest pests. A number of these viruses have been used to control insects as biological alternatives to chemical pesticides (Moscardi 1999; Szewczyk et al. 2009).

Baculoviruses replicate in the nuclei of infected host cells and possess circular, covalently closed, double-stranded DNA genomes ranging from 80 to 180 kbp in length, encoding for 100–180 proteins (Theilmann et al. 2005). Genomes of more than 50 baculoviruses have been sequenced (NCBI databases) and many have been analyzed and published (van Oers and Vlak 2007; Rohrmann 2008a). These viruses belong to the family Baculoviridae, which is currently subdivided on the basis of phylogenetic evidence and molecular characteristics into four genera: Alphabaculovirus (lepidopteran nucleopolyhedrovirus), Betabaculovirus (lepidopteran granulovirus), Gammabaculovirus (hymenopteran nucleopolyhedrovirus), and Deltabaculovirus (dipteran nucleopolyhedrovirus). This classification of baculoviruses (Jehle et al. 2006) has been proposed for the 9th International Committee on Taxonomy of Viruses Report (www.ictvonline.org). Lepidopteran NPVs can be further classified into two groups, i.e., I and II. This subdivision has been correlated with the presence of unique envelope fusion proteins, GP64 (Group I) and F (Group II), encoded by viruses from each group (Zanotto et al. 1993; Pearson et al. 2000; Ijkel et al. 2000; Herniou et al. 2001, 2003). Virions of Alphabaculoviruses are designated single (S) or multiple (M) depending on the number of nucleocapsids per ODV (occlusion-derived virus), whereas delta- and gammabaculoviruses normally contain a single nucleocapsid per ODV (Volkman et al. 1995; Theilmann et al. 2005).

Baculoviruses exist as two phenotypes, i.e., occlusion-derived virus (ODV) and budded virus (BV), which have a common nucleocapsid structure and carry the same genetic information (Blissard 1996). These virions are produced at different cell locations and times in the infection cycle. Also, they differ with relation to some of their virus-derived proteins, in the composition of their viral membranes, and in their mechanisms of entry into the host cell. BVs are produced in the late phase of ­infection, obtain their envelope from the cell membrane, and require the fusion protein GP64 (Monsma et al. 1996; Hefferon et al. 1999) or another unrelated protein termed the F protein (Lung et al. 2002; Westenberg et al. 2004) that facilitates systemic infection. This protein forms structures called peplomers at one end of the budded virus particle, but they are not present in ODVs (Monsma et al. 1996), although a number of other proteins are only associated with ODV. Several ODV envelope proteins have been identified as essential for primary infection of midgut cells of insect larvae and others as ODV components whose specific location and function have not yet been determined (Kuzio et al. 1989; Faulkner et al. 1997; Kikhno et al. 2002; Pijlman et al. 2003; Ohkawa et al. 2005; Slavicek and Popham 2005; Fang et al. 2007, 2009; Li et al. 2007). ODVs are produced in the very late phase of the infection when nucleocapsids become enveloped within the nucleus and are subsequently occluded in a protein crystal structure forming the occlusion bodies (OBs).

16.2.2 Viral Life Cycle

In the baculovirus life cycle, ODVs establish primary infection in the midgut and are required for horizontal transmission of baculoviruses between insect hosts. These virions are derived from the nuclear membrane of the insect cell and at a very late time, postinfection, become occluded in a protein matrix, forming paracrystalline structures termed occlusion bodies (OBs). The occlusion bodies are composed mainly of a protein called polyhedrin in NPVs and granulin in GVs that are highly stable and facilitate virus survival and dispersal in the environment (Olszewski and Miller 1997). BVs are highly infectious for insect cells and are capable of spreading infection from cell to cell both within the insect and in cell culture. These virions have an envelope distinct from ODV that facilitates systemic infection. They acquire their envelopes by budding through the plasma membrane.

The viral life cycle begins when a susceptible host ingests OBs that have been deposited on foliage by a previously infected host, resulting in the release of hundreds of ODVs in the gut. In the host midgut, crystalline polyhedron matrix surrounding the ODVs is dissolved by the alkaline environment. The released ODVs then pass through the peritrophic membrane, attach to the microvilli, and subsequently initiate primary infection of mature columnar epithelial cells within the midgut. Budded virus (BV) produced in these cells initiates secondary infections, spreading throughout the host. The nucleocapsids are released from the endosomes and are transported to the nucleus, where viral transcription, DNA replication, and assembly of progeny nucleocapsids occur, resulting in the production of BV and ODV. In the final stage of infection, most of the nucleocapsids remain in the nucleus and become occluded in a protein matrix to form OBs. Progeny OBs are released upon death and disintegration or liquefaction of the infected insect and subsequently initiate a new round of infection to other hosts. The terminally infected insect can migrate to a higher elevation on the branch of a plant, facilitating dispersal of the occlusion bodies (Kamita et al. 2005a, b; Rohrmann 2008b). The consecutive steps of this complex process of infection are shown in Fig. 16.1.

Fig. 16.1
figure 1_16

Natural life cycle of baculovirus AcMNPV. Polyhedra are taken orally by the larvae along with plant material (a) and are dissolved in the alkaline environment of the midgut (b). ODVs are liberated and infect epithelial midgut cells (c). The virus replicates, and budded viruses (BVs) are produced (d), and they infect other tissues (e). After secondary infection (f  ), polyhedra are accumulated (g). Finally, the larval body disintegrates, and millions of new polyhedra are released to the environment (h). BV – budded form of the virus. ODV – occlusion derived form of the virus

Full size image

16.2.3 Molecular Biology of Baculoviruses

Baculoviruses are a large group of double-stranded DNA viruses. They infect arthropods and do not replicate in vertebrates, plants, or microorganisms. Though they do not replicate, they may, under special conditions, enter animal cells. This unexpected property has made baculoviruses a valuable tool for studies of transient expression of foreign genes under vertebrate promoters introduced into the baculovirus genome (Boyce and Bucher 1996; Kost et al. 2005).

The baculoviruses have gained immense attention in molecular biology laboratories because they are one among the most versatile genetic engineering tools (for a review see van Oers 2006). The most widely studied baculovirus is the Autographa californica nucleopolyhedrovirus (AcMNPV). Our current knowledge about the biology of AcMNPV is, to a large extent, a consequence of the developments of baculovirus-based expression vectors. This system of foreign gene expression has many advantages over other systems, which are as follows:

  • A high level of foreign gene expression is usually achieved compared to other eukaryotic expression systems.

  • It is possible to express more than one foreign gene.

  • The baculovirus genome can accommodate large pieces (around 20 kbp) of foreign DNA.

  • Insertion of specific signal sequences in front of a foreign gene often leads to export of the gene product outside of the infected cell.

The circular DNA genome of AcMNPV is surrounded by a small basic protein that neutralizes the negative charge of the DNA. This structure is protected by proteins forming a nucleocapsid. Virions consist of one or more nucleocapsids embedded in a membranous envelope. The genomic circular DNA is infectious in the naked form. As mentioned above, the two morphologically distinct, but genetically identical, viral forms (ODV and BV) are produced at different periods after infection. The occlusion bodies (polyhedra) contain many occlusion-derived virions (ODV) surrounded by a matrix composed mainly of polyhedrin, a major structural protein (Braunagel et al. 2003). It should be stressed here that polyhedrin is produced in large quantities (approx. 30% of total protein mass at the time of host death) but is not needed for transmission of the virus from cell to cell. Polyhedra (OBs) are relatively stable, and protected virions under favorable conditions can survive in the environment for decades. They are large enough to be seen under a light microscope. Under magnification of 1,000×, polyhedra resemble clear, irregular salt crystals.

Recombinant baculoviruses are usually constructed in two steps. Initially, a heterologous gene is introduced into a baculovirus transfer vector. The vector consists of a bacterial replicon of a multicopy plasmid, a selection marker gene, promoter and terminator regions along with flanking baculovirus sequences from a nonessential locus, and a multiple cloning site (or a single unique restriction site) downstream from a viral promoter. When commercial production of a recombinant protein is required, the promoters and the flanking DNA usually originate from one of the very late genes, either polyhedrin or p10. The latter is another viral gene coding for a protein produced in large quantities very late in the infection. It is the main component of the fibrillar structures, which accumulate in the nucleus and in the cytoplasm of infected cells. For some purposes, e.g., for earlier enhancing of the insecticidal properties of a baculovirus, weaker late but not very late promoters [e.g., basic protein promoter (p6.9)] or early promoters (e.g., ie1, p35 or DA26) are sometimes preferred.

Classical methods of recombinant construction are based on the homologous recombination in insect cells as the second step of engineering the recombinant. The baculovirus transfer vector containing foreign DNA and genomic viral DNA are introduced into insect cells where they recombine yielding recombinant virus with an integrated heterologous gene. Many improvements over classical methods (Summers and Smith 1987) for recombinant selection have been made in recent years. Linearization of the baculovirus genome at one or more locations simplifies the construction of recombinant baculoviruses. Linear baculovirus DNA exhibits a greatly reduced infectivity compared to preparations of circular DNA. When a unique restriction site was introduced into the AcMNPV genome, which allows for linearization in the vicinity of the polyhedrin gene, recombinant viruses were obtained at a frequency of about 30% (Kitts et al. 1990). It should be pointed out that recombination between linear genomic DNA and a transfer vector results in circularization of the genome. Therefore, even though the titer of recombinants per transfection is similar to that of the normal cotransfections with circular genomic DNA, the percentage of recombinants is greatly increased because the background of nonrecombinants originating from linear DNA is greatly reduced. Further developments of the above method increased the percentage of recombinant viruses to almost 100% (Kitts and Possee 1993).

Many laboratories specializing in the production of recombinant proteins routinely use the Bac-to-Bac expression system for constructing baculovirus recombinants (Luckov et al. 1993). The diagram shown in Fig. 16.2 outlines the key steps of recombinant construction. A bacmid (baculovirus shuttle vector) is an engineered low-copy bacterial plasmid (F1 derivative) containing the complete baculovirus genome. The gene of interest is cloned into another small plasmid (e.g., pFastBac) downstream of the polyhedrin promoter. This plasmid also contains two transposable elements flanking the gene of interest and a gentamycin-resistance gene. The donor plasmid is used to transform special bacterial strains containing the baculovirus genome. These bacteria also contain a plasmid coding for the enzyme transposase that catalyzes transposition between the transposable elements engineered in the donor plasmid and those engineered in the viral genome. As a result, the bacmid containing the gene of interest is obtained and can be visually selected because of the presence of an additional LacZ marker gene within the viral genome. After verifying the presence of the gene of interest in the baculovirus genome, recombinant bacmid preparations from bacteria are used to transfect insect cells. Following transfection, viable recombinant ­baculovirus should be budding into the culture medium within 2–3 days posttransfection.

Fig. 16.2
figure 2_16

Schematic diagram of Bac-to-Bac expression system. The foreign gene is inserted into pFastBac vector, and the plasmid is introduced into specially engineered Escherichia coli strain containing Bacmid – large low-copy plasmid and a helper plasmid allowing for site-specific transposition. The recombinant Bacmid is isolated and transfected into cultured insect cells. The transfected cells are then lysed to yield recombinant baculovirus used later for large-scale preparation of the baculovirus and for the production of the recombinant protein

Full size image

A few hundred insect cell lines that can potentially be used for in vitro propagation of baculoviruses are known. A few that support the growth of AcMNPV were obtained from two parental organisms, Spodoptera frugiperda and Trichoplusia ni (Lepidoptera: Noctuidae). The most widely used is Sf9, which grows well in suspension (Summers and Smith 1987). BTI-Tn5B1-4 derived from T. ni, known as High Five cells, has also been used for viral growth (Granados et al. 1994). Cell lines that can be used for propagation of Lymantria dispar nucleopolyhedrosis virus (LdMNPV), Helicoverpa zea nucleopolyhedrosis virus (HzSNPV), Bombyx mori nucleopolyhedrovirus (BmSNPV), Anticarsia gemmatalis nucleopolyhedrovirus (AgMNPV), and a few other baculoviruses are also currently available.

The baculovirus expression system is widely used for production of glycoproteins with therapeutic potential for humans and animals. Most posttranslational modifications of these proteins are the same as in mammalian cells. However, N-glycosylation of proteins in mammalian cells is more complex than that in insect cells. In the latter, N-glycans with terminal mannose residues are produced, in contrast to sialic acid-terminated glycans in vertebrate cells. In most cases, the extent and quality of glycosylation in insect cells are sufficient for preservation of biological activities of glycoproteins, and such insect-derived glycoproteins fulfill the requirements for a potential therapeutic agent. In rare cases, when the role of glycan chains in preservation of biological activity is very high, it is possible to use “humanized” insect cell lines (Harrison and Jarvis 2006), which are genetically engineered to produce the required vertebrate-type complex N-glycans with terminal sialic acids.

16.3 Baculovirus Production Technology

16.3.1 In Vivo Production

At present, commercial production of baculoviruses has been carried out only in vivo, either by applying the virus against the host insect in the field and collecting diseased or dead larvae or by producing the target insect in the laboratory on an artificial diet and contaminating the diet with a baculovirus for further collection of virus-killed insects. The latter is the most commonly used method for producing baculoviruses in different countries. Both methods have been used successfully for commercial production of the Anticarsia gemmatalis Alphabaculovirus (AgMNPV) in Brazil (Moscardi 1999, 2007). For some insects, there are no available artificial diets, and therefore, commercial production of baculoviruses of these insects is generally too difficult or impossible under laboratory conditions due to dependency of host plant leaves for viral inoculation. On the other hand, field production of baculovirus agents is viable, resulting in a product of lower cost (Moscardi 1999). However, field production is difficult when liquefaction of the insect body is very intense, as, for instance, in larvae infected by Spodoptera spp. baculoviruses, making it almost impossible to collect dead larvae. In this case, live larvae must be collected close to death when the body has not yet ruptured. These larvae may, however, contain less virus than would dead larvae. It is known that two viral enzymes, chitinase and cathepsin, are important in the process of cuticle disintegration and liquefaction of the insect body, which are common among species of Lepidoptera. Among natural isolates of a same baculovirus, it is possible to find a few which lack these enzymes, thus facilitating field and laboratory production. The commercial field and laboratory production of the AgMNPV are discussed in Sects. 16.4.1.2 and 16.4.1.3, to exemplify details of both production methods.

16.3.2 In Vitro Production

Baculovirus production in insect cell cultures offers advantages over in vivo multiplication for being a controllable, sterile, highly pure product yield process, besides the fact that hundreds of cell lines have already been established. The process of baculovirus production for agricultural pest control needs to be efficient, with competitive cost, leading to a final product that is highly pathogenic to the target pest. There is a strong limitation for in vitro production, however, since successive passages of the virus in cell culture result in genetic alterations, leading to loss of virulence (Krell 1996; Rhodes 1996). In laboratory culture, production of occlusion-derived virions (ODV) is not necessary for survival of the virus. The budded virus (BV) particle is the form used for cell-to-cell transmission in cell culture. The main protein of the BV particle is the GP64 (Blissard 1996). During infection, this glycoprotein is essential for virus budding and is responsible for entrance of the virus into the next host cell (Monsma et al. 1996). Various culture conditions are known to influence infection of lepidopteran cells by baculoviruses and include temperature, pH, dissolved oxygen concentration, osmolality, and nutrient composition of the culture medium. Most lepidopteran cells proliferate optimally at temperatures between 25 and 28°C with an optimum pH of 6.2. Insect cells present several comparative advantages to mammalian cells such as ease of culture, higher tolerance to osmolality and by-product concentration, and higher expression levels when infected with a recombinant baculovirus (Agathos 1996; Ikonomou et al. 2003).

In vitro production remains an important requirement from a commercial perspective for the use of baculoviruses as insecticides. One of the most important effects of the viral passage is the change from the parental, many polyhedra per cell (MP) phenotype, to the few polyhedra per cell (FP) phenotype (Fraser and Hink 1982; Fraser et al. 1983; Pedrini et al. 2004; Rezende et al. 2009; Slavicek et al. 1996). A key problem associated with the passage effect is the reduced occlusion and loss of virulence of the occluded virus. Frequent mutations have been identified within a specific region in the Few Polyhedra mutants (FP) that contains the 25k fp locus. This gene encodes a 25-kDa protein that is essential for virion occlusion and polyhedron formation (Chakraborty and Reid 1999; Harrison and Summers 1995; Lua et al. 2002; Pedrini et al. 2004; Slavicek et al. 1996). Another type of mutant generated during serial passage of baculovirus is the formation of Defective Interfering Particles (DIPs). These mutants have lost the ability to be replicated in the host cell without the aid of a helper virus, and large sizes portions of their genome are usually deleted (Bangham and Kirkwood 1990; Kool et al. 1991; Pijlman et al. 2001).

Another challenge for in vitro production of baculovirus is the requirement for a highly productive insect cell line (Jem et al. 1997) and a highly productive culture medium (Chakraborty et al. 1999). Many cell lines are available for production purposes and are derived from various sources, thus exhibiting a wide variety of growth and production characteristics. Careful screening or formulation of media must be performed for a particular virus isolate–cell line combination, as different media can greatly affect polyhedra yields (Pedrini et al. 2006). Recently, a new strategy for in vitro production has been proposed based on Many Polyhedra (MP) variants. These are clones selected using the plaque assay technique after several passages of the virus in cell culture. MPs maintain the wild-type features such as formation of many polyhedra in the cell nucleus and Budded Virus high titer (Slavicek et al. 2001; Pedrini et al. 2005), which allow them, in principle, to be competitive with the population of Few Polyhedra mutants accumulated in cell culture. The investigation of factors associated with loss of genetic stability and the use of new strategies such as isolation of more stable variants (MP), as well as the reduction of cost of cell culture medium components, is an important requirement for process optimization of in vitro baculovirus production.

16.4 Use of Baculoviruses for Pest Control

Since the comprehensive review by Moscardi (1999) on use of baculoviruses for control of Lepidoptera, other works have been published on the state of virus utilization against insect pests of agricultural, forest, and vegetable production systems (e.g., Copping and Menn 2000; Szewczyk et al. 2006, 2009; Souza et al. 2007; Erlandson 2008). In this chapter, we focus on the most important programs worldwide, with emphasis on those aspects that benefit or limit use of these agents in IPM programs. The use of the AgMNPV in Brazil is presented as a case study to discuss how a very successful program (the most important one worldwide) experienced a serious setback over the past 6 years.

The main baculoviruses that have been or are currently being utilized are depicted in Table 16.1. In Latin America, the AgMNPV is the most commonly used biological product to control A. gemmatalis in soybean (Glycine max). This virus was used in about 2.0 million hectares during the 2003/2004 growing season in Brazil, representing approximately 10% of the soybean cultivated area in the country. It has also been used in Argentina, Colombia, Bolivia, Paraguay, and Mexico (Moscardi 1999, 2007; Sosa-Gómez et al. 2008). Another virus that is presently used in Brazil is the nucleopolyhedrosis of the poplar moth, Condylorrhiza vestigialis. This virus has been produced on insects reared on an artificial diet. The primary objective of its application is the treatment of 2,000 ha/year, which represents the infested area among 5,500 ha of poplar (Populus spp.) plantations in south Brazil (Sosa-Gómez et al. 2008). In Peru, a granulovirus has been developed to control larval populations of the potato tuber moth, Phthorimaea operculella, in field and stored potatoes, by the initiative of the International Potato Center (CIP) (Raman et al. 1992). This virus has also been used in Bolivia, Colombia, and Ecuador (Moscardi 1999; Sosa-Gómez et al. 2008). Presently, another baculovirus used in Latin America is the Erinnyis ello GV in Colombia, which has replaced chemical insecticides in sites of endemic occurrence of the insect (Bellotti 1999, Bellotti, pers. communication). Apparently, there are no significant programs using entomopathogenic viruses in Cuba, since there are no reports in the literature, and contacts with Cuban researchers on use of baculoviruses for pest control have not been acknowledged.

Table 16.1 Main examples of baculoviruses developed as microbial insecticides
Full size table

The genera Heliothis and Helicoverpa represent key pests of several annual crops and vegetables worldwide (Ignoffo and Couch 1981; Cunningham 1995), responsible for losses of millions of dollars annually. An NPV of H. zea was developed in the 1960s and registered in 1975 in the USA (Ignoffo and Couch 1981), representing an important breakthrough in virus use. Elcar™, developed by Sandoz, was the first viral insecticide registered in the USA for use in cotton. The HzSNPV has a relatively broad range, infecting other species belonging to the genera Helicoverpa and Heliothis. An HzSNPV formulated product registered as GemStar™ has been used to control Helicoverpa armigera on cotton in Australia. Locally obtained isolates of H. armigera SNPV have also been produced and applied to cotton, soybean, pigeon pea, maize, and tomato crops in China, India, and Australia (Zhang et al. 1995; Sun and Peng 2007; Erlandson 2008; Srinivasa et al. 2008). The potential application of the NPVs of H. zea and H. armigera is enormous, as H. zea and H. virescens in the Americas and H. armigera in Africa, Asia, and Australasia cause severe losses to several crops and vegetables. In China, NPVs of H. armigera are used on over 100,000 ha annually involving at least 12 HaSNPV producers (Sun and Peng 2007).

Another insect genus that causes a severe economic impact on food production is the Spodoptera complex, including S. frugiperda, S. exigua, S. littoralis, and S. litura. In Brazil, an indigenous isolate of S. frugiperda NPV (SfNPV) was used to control the insect in maize and was applied to 20,000 ha/year (Valicente and Cruz 1991; Moscardi 1999). Owing to difficulties and high cost of SfNPV production by the Brazilian Organization of Agricultural Research (Embrapa), a government research institution, this program has been discontinued temporarily – at present, no SfNPV product is available to maize farmers. Presently, a S. exigua NPV, under different trade names, has been used to control this species on vegetable crops in the USA, Europe, China, and Thailand. Also, an NPV of S. litura is used in China, India, and Thailand (Sun and Peng 2007; Erlandson 2008; Kumari and Singh 2009; Szewczyk et al. 2009).

One of the most important successes in commercial production and use of a baculovirus in Europe may be the coddling moth, Cydia pomonella, GV (CpGV) for use in orchards, particularly apples and pears. The CpGV has been produced under different trade names (Table 16.1) and has been used in Argentina, Canada, France, Germany, Russia, and Switzerland, among other countries (Moscardi 1999; Arthurs and Lacey 2004; Vincent et al. 2007; Erlandson 2008; Kutinkova et al. 2008). The product Madex® (Andermatt BIOCONTROL), initially developed to support organic fruit producers in Europe, is now produced for application on over 250,000 ha units/year (Vincent et al. 2007). Considering application of other trade names of the CpGV, this may be the most important worldwide viral insecticide currently applied in terms of treated area.

Other important viruses that are currently employed to control insects include the tea tortricids Adoxophyes honmai and Homona magnanima granuloviruses (GV) in Japan. From 1990 to 1993, five GV production facilities were established in Kagoshima County. These facilities were government-subsidized but operated by a tea growers cooperative. The area sprayed with GVs comprised 5,850 ha in Kagoshima in 1995, equivalent to 80% of all the tea fields in the prefecture (Nishi and Nonaka 1996). The GVs of H. magnanima and A. honmai were registered in 2003 and produced by Arysta LifeScience Corporation (M. Nakai, pers. communication); however, the use of GVs has declined. One reason for the reduction in use of GVs in Japanese tea fields is the changing pattern of occurrence of other pests. Mulberry scale, for example, has been increasing recently, and chemical treatment is required to control this insect at the same time GVs are sprayed. However, the spray also kills H. magnanima and A. honmai. Furthermore, GVs have been applied in Kagoshima for more than 10 years, and the populations of H. magnanima and A. honmai have been reduced (Nakamura 2003). In China, approximately 12 baculoviruses have been authorized as commercial insecticides, including H. armigera NPV (cotton, pepper, tobacco) (which is the most heavily used virus in the country), S. litura NPV (vegetables), S. exigua NPV (vegetables), Buzura suppressaria NPV (tea), and Pieris rapae GV, and Plutella xylostella GV (vegetables) (Sun and Peng 2007). Use of baculoviruses in China is the greatest worldwide, regarding the number of viruses being registered for insect control. Sun and Peng (2007) also report a Cypovirus (CPV) produced in China for control of Dendrolimus punctatus, an insect pest of pine forests.

In forest systems, especially in temperate regions, defoliating larvae of Lepidoptera and Hymenoptera are often significant pests. A Lymantria dispar (Lep.: Lymantriidae) MNPV has been developed since the 1980s as a viral insecticide under the trade names Gypcheck and Disparvirus, among others (Moscardi 1999; Reardon et al. 1996; Erlandson 2008; Szewczyk et al. 2009). Also, NPVs of hymenopterans such as Neodiprion sertifer, N. abietis, and Diprion pini (Diprionidae) have been developed as bioinsecticides (Lucarotti et al. 2007; Erlandson 2008; Szewczyk et al. 2009). Forest ecosystems tend to be more stable than agricultural systems, allowing for natural or applied baculoviruses to remain in the environment for long periods of time.

16.4.1 Use of the Alphabaculovirus of Anticarsia gemmatalis (AgMNPV) in Brazil and Latin America: A Case Study

The virus AgMNPV serves as the most important testament that baculoviruses are a viable insect control strategy in the context of an IPM program. Conversely, when an IPM program is not available or not functioning adequately, use of a baculovirus may not succeed. The evolution of the AgMNPV use may serve as an example (Moscardi 1999, 2007), as its applications in soybean in Brazil reached approximately 2.0 million hectares in 2003/2004 season. However, due to changes in agricultural practices by soybean growers, use of the AgMNPV experienced a sharp decline in the last 7 years. This program is summarized and discussed in the sequence.

16.4.1.1 Historical Perspective

The AgMNPV program was established after a pilot phase, conducted during 1980–1982 in farmers’ fields in various regions in south Brazil. The virus was found to be efficient for control of A. gemmatalis with only one application, compared to 1.8 insecticide applications in areas conducted according to farmer’s perceptions (Moscardi 1999, 2007). Implementation of the program for AgMNPV use began in the 1982/1983 soybean season, when approx. 2,000 ha of soybean were treated. Initially, small amounts of the virus were produced in A. gemmatalis larvae reared on an artificial diet at Embrapa Soybean facilities (Londrina, PR). At that time, frozen killed larvae were distributed to extension officers for treatment of demonstration plots and virus production in the field, which provided inocula to treat other areas in the same season or to collect and store dead larvae for the subsequent season. AgMNPV usage gained momentum with the development of a wettable powder formulation of the virus in 1986 (Moscardi 1989, 1999).

Another important step for consolidation of the AgMNPV program was the legal agreement between Embrapa and five private companies in 1990. Through these agreements Embrapa would transfer all the technology for AgMNPV production, formulation, and quality control of production batches. The products based on the AgMNPV of each company were registered according to Brazilian policies for registration of plant protection insecticides (Moscardi and Sosa-Gómez 1996). With the commercialization of the AgMNPV by private companies, the use of this virus increased from about one million hectares in 1990 to approximately 1.5 million hectares in 1995, with most of the production occurring in the field during each soybean season. Peak AgMNPV use occurred in the 2003/2004 season, when approximately two million hectares of soybeans were treated with the virus. Afterward, the use of this agent declined sharply because of changes in farmers’ procedures to control pests in soybean (Fig. 16.3), which are discussed below. In addition to the early efforts by Embrapa to develop and improve, both technically and economically, in vivo procedures for AgMNPV production under controlled laboratory conditions, two of the companies also attempted to develop production methodologies. One of them (Geratec) produced about 150,000 ha-equivalent of the virus per year in the early 1990s. However, owing to the high cost of labor, disposable rearing containers, and components of the insect artificial diet, laboratory production of the virus was discontinued by both companies. On the other hand, AgMNPV field production became widely adopted by all participating companies as the best available method to obtain large quantities of virus-killed larvae at low cost (Moscardi 1999, 2007). Both methods (field and laboratory) of AgMNPV production are discussed in this sequence. This virus was also used in Argentina and is currently being used in Paraguay and Mexico (Sosa-Gómez et al. 2008).

Fig. 16.3
figure 3_16

Evolution of the use of the Alphabaculovius of Anticarsia gemmatalis (AgMNPV). In Brazil. Y-axis  =  Number of treated hectares (×1,000); X-axis  =  Soybean seasons since 1982/1983. The arrow indicates when the soybean integrated program in Brazil started to have a setback, making difficult the use of a very specific baculovirus, as discussed in the paper

Full size image

16.4.1.2 AgMNPV Field Production

Field production of the virus became a major enterprise during the 1980s and 1990s, involving small companies that specialized in marketing AgMNPV-killed caterpillars to private companies that registered the virus for commercialization (Moscardi 1999, 2007). Growers’ fields were contracted and pest control in their fields was implemented by the AgMNPV producers. Usually, about three fields are sprayed per day during the prevalence of A. gemmatalis larvae in soybeans (December and January). Before collection, fields that are sprayed each day with the AgMNPV are inspected at the 6th or 7th day postapplication for selection of those that would yield the highest number of dead larvae per hectare. Peak ­collection occurs from the 8th to the 10th day after virus application and may involve 200–300 “larval pickers” per day, requiring ten buses to transport them to the fields. In a single day, production at one collection site could reach 600 kg of AgMNPV-killed larvae, enough for treatment of 30,000 ha. To emphasize the importance of production in the field, during the 2002/2003 season, approximately 45 metric tons of AgMNPV-killed caterpillars were collected and sold to the private companies at about US$ 10–12/kg, representing about two million hectare-equivalents of the biological insecticide to be applied in the subsequent soybean season.

Despite its value for producing high quantities of AgMNPV at low cost, field production presented problems that restricted the expansion of its use or affected the quality of the end-product, for example, (1) yearly production was too dependent on natural incidence of the host insect, which may occur in low numbers in certain seasons, thus reducing AgMNPV yield and resulting in variable quantities of the biological insecticide to growers from season to season; (2) quality of field-collected AgMNPV-killed larvae decreased owing to change in collection procedures to attend the high demand by the private companies that registered their AgMNPV commercial products. A key problem was that collection of dead larvae in the field shifted from handpicking to shaking plants over pieces of cloth placed over the ground in between soybean rows. This shift resulted in collection of dead larvae, live host larvae (containing low amounts of virus), larvae from other lepidopteran species, other insects (stink bugs, beetles, etc.), and leaves, which resulted in material with higher amount of extraneous organic matter other than the AgMNPV-killed larvae. While the handpicking method resulted in an average 50 ha equivalent of the virus per kg, the newer procedure resulted in an average 30–35 ha equivalent of the virus per kg. Because of the higher amount of extraneous organic matter, standard procedures for homogenization and formulation had to be modified. Additionally, higher amounts of organic matter in the final product led to nozzle clogging and decreased efficiency of the product in the field. Therefore, commercial laboratory production of the AgMNPV became a requirement to improve quality. To achieve this goal, research was conducted at Embrapa Soybean to carry out the necessary improvements to make commercial laboratory production viable.

16.4.1.3 AgMNPV Commercial Laboratory Production: A Breakthrough

Improvements in the AgMNPV laboratory production procedures up to 1997 (Moscardi et al. 1997) served as a starting point for a PhD study (Santos 2003) aimed at removing the most important bottlenecks related to commercial production of AgMNPV. Among various aspects studied (ingredients of the insect diet, rearing conditions, containers, virus dosage, larval size at inoculation, and number of larvae per container), significant progress in AgMNPV production was attained as a result of these studies. Cost for the artificial diet was reduced by approximately 85% through substitution of agar by another jellifying agent, and through the reduction of casein content by 50%. With these new procedures, the cost of AgMNPV-killed larvae was approximately US$ 0.42 to treat one ha, as compared with US$ 0.30 for those collected in the field (Santos 2003). Considering the much higher quality of laboratory-produced AgMNPV plus the cost involved, the product generated from laboratory production could be offered at a lower cost than that of chemical insecticides to control A. gemmatalis.

In May 2003, a private company (Coodetec) established a pilot laboratory for virus production in Cascavel, PR, Brazil and by the end of that year was inoculating 100,000 A. gemmatalis larvae per day, employing 14 people. After processing 1,000 kg of dead larvae, viral yield was 65–72 ha-equivalent/kg. Coodetec subsequently built large laboratory facilities in 2004, consisting of two independent laboratories of 750 m2 each: one for insect production and the other for virus production, with another facility (500 m2) for virus storage, processing, and formulation. In the first laboratory, eggs are collected daily in adult oviposition rooms, and larvae are reared in separate rooms up to the 4th instar in 500-ml cardboard cups containing insect diet. Daily, 4–5% of the 4th instar larvae are transferred to plastic trays with diet and vermiculite to obtain pupae and maintain the insect colony. The remaining larvae (95–96%) are taken to the virus production laboratory where they are transferred from the 500-ml cups to plastic trays containing AgMNPV-treated diet. Seven days later, dead larvae are collected in plastic bags and stored at −4°C for further processing and formulation of the biopesticide. The laboratories implemented at Coodetec were implemented to employ 45 people to inoculate 800,000–1,000,000 larvae per day, resulting in a quantity of AgMNPV to treat 1.8–2.0 million ha/year. However, these laboratories discontinued their production as soybean pest control strategy changed in the last 7 years, drastically reducing the demand for the AgMNPV, which is discussed below.

16.4.1.4 Why Did the AgMNPV Program Experience a Setback in Brazil?

With the implementation of no-till agricultural systems in Brazil, soybean farmers had to apply herbicides as desiccants prior to soybean sowing. In this operation, growers began to mix broad-spectrum insecticides (such as pyrethroids) with the herbicides, “just to kill any insects that were present in the weeds” (Corrêa-Ferreira et al. 2010). At 15–20 days after soybean emergence, when growers apply postemergence herbicides, most also mixed pyrethroids, and in most cases, these two insecticide applications before and early in crop development were found to be detrimental to natural enemies (predators and parasitoids), thus disturbing the equilibrium in soybean systems. Other insects/organisms such as the soybean looper, white flies, Spodoptera spp., and mites, which were considered as secondary pests 7 years ago then became important pests (Bueno et al. 2007). Farmers, therefore, went into a treadmill, as broad-spectrum chemicals had to be applied against these “pests,” and a highly specific biological product such as AgMNPV could no longer be used. Coodetec ceased production of the virus in the laboratory, but the ­developments regarding commercial AgMNPV production under laboratory conditions were proved to be viable and cost-competitive with chemical insecticides available on the market. Presently, this virus is used on about 300,000 ha yearly, compared to about two million hectares 7 years ago (Moscardi 2007). Lastly, with the introduction of soybean rust in Brazil, at least two applications of fungicides are made that may reduce the incidence of important natural entomopathogenic fungi such as Nomuraea rileyi (Sosa-Gómez et al. 2003) and Entomophthorales, which used to hold down populations of various caterpillar species before fungicide applications on the crop.

16.5 Factors Limiting Baculovirus Use

A successful program for use of a baculovirus depends upon a combination of factors (see Moscardi 1999 and references cited in pgs. 274–277), including selection of the most virulent isolate, application timing (as larvae may take a week or more to die), application technology, and plant substrate. However, solar radiation is the major factor affecting field persistence of baculoviruses. Viral activity can be completely lost in less than 24 h, but mean half-life generally has varied from 2–5 days. Ultraviolet radiation in region B (UV-B) (280–310 nm) inactivates baculoviruses. However, UV-A (320–400 nm) may also be critical in baculovirus deactivation. Many substances have been tested as sunscreening agents in formulations of these biological products, with many promoting protection to baculoviruses against UV radiation, such as fluorescent brighteners of the stilbene group. Besides protection against UV, the stilbenes also enhance viral activity (Shapiro 1995, Morales et al. 2001 and literature cited therein).

Baculoviruses, because of their high specificity, are most suited for use in agriculture, forestry, and fruit crop systems where there are no concurrent important insect pests, as was the case of the application of AgMNPV against A. gemmatalis in soybean in Brazil (Moscardi 1999). Also, if an IPM program is not adopted by farmers, it is difficult to succeed in using a baculovirus, since the target insect must be monitored frequently (at least once per week) to time applications against the most susceptible larval instars (i.e., young ones). Farmers may find the need to sample their fields every week troublesome and may prefer the rapid killing by chemical insecticides. A critical issue in baculovirus use is the time necessary to kill the insect host (Moscardi 1999; Szewczyk et al. 2006, 2009; Souza et al. 2007; Erlandson 2008). Farmers may not initially be prepared to observe no obvious control results for 4–5 days following application of a baculovirus insecticide (Moscardi 2007). In Brazil, in the beginning of the AgMNPV program, farmers were not accustomed to wait long for A. gemmatalis larval mortality after virus application. Many would return to the fields within 2 or 3 days and apply chemical insecticides, not waiting for the virus to act on the larval populations. Because of this limitation, research has been directed at developing genetically modified baculoviruses with shorter times to kill their host larva.

16.6 Genetically Modified Baculoviruses to Control Insects

In the past, the practical application of baculoviruses as commercial insecticides was hampered by their relatively slow killing action and technical difficulties for in vitro commercial production. Due to the slow killing action, primary users (used to fast-killing chemical insecticides) regarded baculoviruses as ineffective. With advances in genetic-engineering technologies, many successes have been made in improving the timing of the killing action. Two broad strategies have been pursued in laboratories worldwide to achieve this goal: interference with host physiology and introduction of an insect-specific toxin (Bonning and Hammock 1996; Mishra 1998; Inceoglu et al. 2001).

The first strategy involves introducing genes coding for some insect hormones or enzymes into the baculovirus genome. Alternatively, the deletion of some nonessential baculovirus genes provides a beneficial effect for the speed of kill of a virus, as was found in the case of viral ecdysteroid UDP-glucosyltransferase (egt) gene. The product of this gene catalyzes the conjugation of sugar molecules to ecdysteroids (Tumilasci et al. 2003), thus preventing the ecdysteroid from crossing cellular membranes. Maeda (1989) was the first to introduce a diuretic hormone gene into Bombyx mori baculovirus genome to cause insects to lose water. Modified BmNPV killed larvae about 20% faster than wild-type BmNPV. Ma et al. (1998) expressed pheromone biosynthesis activating neuropeptide (PBAN) fused to the bombyxin signal sequence for secretion using AcMNPV. The recombinant baculovirus reduced survival time of Trichoplusia ni larvae by more than 20% in comparison to larvae infected with a control virus. Two other insect ­hormone genes (eclosion hormone and prothoracicotropic hormone genes) were also studied as potential factors for modification of baculovirus; however, no significant improvement over wild-type virus was observed (Eldridge et al. 1991; O’Reilly et al. 1995). Another strategy for improving the timing of the killing action was based on control of the juvenile hormone, which in lepidopteran larvae regulates the onset of metamorphosis at the final molt. The juvenile hormone is regulated by juvenile hormone esterase which when overexpressed decreases concentration of the hormone. This, in turn, is a signal to stop feeding and to pupate. This elegant hypothesis for improvement of baculovirus action encountered many difficulties in practice but is being pursued to make it more efficient under natural conditions (Hammock et al. 1990; van Meer et al. 2000; Hinton and Hammock 2003; Inceoglu et al. 2001).

Another approach to reduce killing time was used by O’Reilly and Miller (1991), who deleted the baculovirus-encoded ecdysteroid glucosyltransferase gene. The product of the egt gene normally prevents larval molting during infection and indirectly increases feeding activity of infected caterpillars. The infection with recombinant virus resulted in 30% faster killing of larvae and significant reduction in food consumption. The egt enzyme is responsible for rendering the hormone ecdysone inactive. Inactivation of ecdysone results in prolongation of the larval stage and increased plant consumption. When larvae are infected with an egt-minus virus, molting proceeds normally, and consequently, the larvae eat less food. The egt gene is not essential for viral replication and can be replaced with an exogenous gene, e.g., with a toxin gene, which may further enhance the insecticidal activity of the recombinant virus (Popham et al. 1997; Sun et al. 2004).

Enhancins are baculovirus-encoded proteins that can increase the oral infectivity of a heterologous or homologous baculovirus. Their infection-enhancing effects are probably due to the degrading action on mucins and to the improved fusion of the virus to the midgut epithelium cells (Wang et al. 1994; Wang and Granados 1997). Enhancin genes have been expressed by recombinant AcMNPVs and subjected to dose-mortality studies (Hayakawa et al. 2000; Li et al. 2003). LD50 values were significantly lower for the recombinant virus in comparison to the wild-type virus (from 4.4- to 21-fold lower). Harrison and Bonning (2001) have constructed a recombinant AcMNPV producing three different proteases from the flesh fly Sarcophaga peregrina, which are known to degrade basement membrane proteins. One of the recombinants expressing cathepsin L under baculovirus promoter of p6.9 gene generated a 51% faster speed of kill in comparison to the wild-type virus. Chitinases are enzymes that degrade chitin into low-molecular-weight oligosaccharides. Baculovirus chitinases are likely to be involved in the degradation of exoskeletons and gut linings of insects. A recombinant AcMNPV expressing the chitinase gene of Manduca sexta was constructed by Gopalakrishnan et al. (1993). When fourth instar Spodoptera frugiperda larvae were infected with the recombinant, their survival time was reduced by approximately 1 day in comparison to the wild-type AcMNPV.

Modification of the baculovirus genome by introduction of specific toxin genes has been much more widely exploited than methods based on interference with host physiology. Most reported research has focused on arthropod toxin genes isolated from mites, spiders, or scorpions (reviewed by Inceoglu et al. 2001; Kamita et al. 2005a, b). This line of research proved to be highly successful, but the reluctant attitude of policy makers in many countries toward genetically engineered products has hampered their introduction. The first reports on successful construction of baculovirus genome containing insect-specific toxin genes were published about 20 years ago (Carbonell et al. 1988; Tomalski and Miller 1991). The most promising insect-specific toxin gene used for construction of baculovirus recombinants is probably the gene coding for AaIT toxin originating from the scorpion Androctonus australis. The reported speed of kill by this baculovirus recombinant was increased up to about 40%, and the feeding damage was also reduced by about 40% (Inceoglu et al. 2001). The AaIT toxin gene was introduced into different baculovirus vectors including NPVs of Bombyx mori (Maeda et al. 1991), Autographa californica (Stewart et al. 1991), mint looper Rachiplusia ou (Harrison and Bonning 2000), cotton bollworm Helicoverpa zea (Treacey et al. 2000), and Helicoverpa armigera (Sun et al. 2004). Baculovirus expression of AaIT provides a continuous supply of freshly produced toxin; therefore, a low level of constant toxin production, even when driven by an early promoter, may be sufficient to elicit a paralytic response. In accordance with this hypothesis, Elazar et al. (2001) found that the concentration of AaIT in the hemolymph of paralyzed Bombyx mori is about 50 times lower when the toxin is delivered by a recombinant baculovirus in comparison to the dose delivered by direct injection of the same toxin. Toxin genes isolated from other scorpions, e.g., Leiurus quinquestriatus hebraeus (Chejanovsky et al. 1995; Gershburg et al. 1998; Froy et al. 2000), straw itch mite Pyemotes tritici (Burden et al. 2000), ants (Szolajska et al. 2004), or spiders Diguetia canities and Tegenaria agrestis (Hughes et al. 1997) and introduced into baculovirus genomes were highly active against lepidopteran larvae and are also under intensive study as potential biopesticides. Most of these toxins attack insect sodium channels, so their target is similar to chemical pesticides belonging to the pyrethroid group (Bloomquist 1996; Cestele and Catterall 2000). However, their specific site of action within sodium channels is different, so they may impart a synergistic effect when used in conjunction with baculovirus recombinants carrying toxin genes (McCutchen et al. 1997). Another promising approach for improvement of baculovirus insecticidal efficacy was suggested by Herrmann et al. (1995), who demonstrated that when excitatory and depressant toxins are simultaneously injected into insect larvae, they may exert a synergistic effect. Regev et al. (2003) have shown that in the case of a recombinant AcMNPV expressing toxin pairs (a combination of excitory and depressant scorpion toxins) used against H. virescens, H. armigera, and Spodoptera littoralis larvae, a cooperative insecticidal effect is observed. The recombinant producing excitory toxin LqhIT1 and depressant toxin LqhIT2 from Leiurus quinquestriatus hebraeus provided an improvement of 40% in effective time to paralysis when compared to wild-type AcMNPV and an improvement of approximately 20% when compared to recombinants producing each toxin separately. Chang et al. (2003) have elaborated a novel and highly successful method for the improvement of recombinant baculoviruses; they generated a baculovirus that produced occlusion bodies incorporating Bt toxin. The recombinant baculovirus genome coded for native polyhedrin and a fusion protein in which polyhedrin is fused to the Bt toxin. The speed of action and pathogenicity of the recombinant were greatly enhanced compared to wild-type virus, thus yielding a biopesticide combining the positive properties of the virus and the bacterial toxin and minimizing the probability of evolution of insect resistance to these two killing factors.

Numerous studies have investigated the effectiveness of factors such as gene promoters and signal sequences in front of cloned genes on the efficiency of production and biological quality of expressed toxins. Historically, initial laboratory studies with recombinant baculoviruses were carried out by infecting caterpillars through ingestion of occlusion bodies or by injecting the budded virus into the hemocoel (O’Reilly et al. 1992). The infection by ingestion of occlusion bodies can be used for recombinants with healthy polyhedrin gene, so in the past, the toxin gene was usually introduced into the p10 locus, while the latter method was employed for recombinants with foreign genes in the polyhedrin locus. As an alternative to larval injections, the recombinant occlusion-negative viruses were packaged into polyhedra by cells infected with a second, occlusion-positive virus (e.g., wild-type virus) (Wood et al. 1993). A breakthrough in the construction of viral recombinants was the elaboration of the method of duplication of a viral promoter (Roy 1992). This procedure allowed for the expression of foreign genes under different promoters, e.g., under a basic protein gene promoter (Bonning et al. 1994) because none of the viral genes are lost. The level of recombinant gene expression in the baculovirus system is promoter-dependent, but factors other than the quantity of the product must also be taken into account. The argument for use of late or very late promoters in recombinant baculoviruses is the reduction of risk that a toxin gene could be expressed in nontarget insects because these promoters are not active in beneficial insects (McNitt et al. 1995). Rachiplusia ou MNPV (RoMNPV) expressing a gene coding for either scorpion Androctonus australis toxin (AaIT) or Leiurus quinquestriatus hebraeus toxin (LqhIT2) killed larvae of corn borer Ostrinia nubilalis most effectively when the gene was cloned behind a late p6.9 promoter. When p10 promoter was used, the level of polyhedra production was reduced in some cases, and virions were not occluded efficiently (Harrison and Bonning 2000). Recombinant AcMNPV expressing cathepsin L of the flesh fly through ie-1 promoter killed H. virescens larvae only slightly faster than wild-type AMNPV, but when the gene was expressed from the p6.9 promoter, the recombinant virus killed the host about 50% faster than did the wild-type baculovirus (Harrison and Bonning 2001). On the other hand, Tuan et al. (2005) showed that the early p-PCm promoter was superior to the very late p10 for controlling insect pests when LqhIT2 scorpion depressant toxin gene was introduced into AcMNPV genome, which may indicate higher susceptibility of earlier instars of these larvae to baculovirus infection. Sun et al. (2004) constructed a chimeric promoter by insertion of a p6.9 promoter downstream of the polyhedrin promoter and used this dual promoter for the expression of AaIT scorpion toxin gene in egt locus of HaSNPV. This HaSNPV-AaIT recombinant was found to be a much more effective biocontrol agent than the wild-type virus or egt-deleted virus.

Speed of action of genetically modified baculoviruses can be also enhanced by signal sequences in front of cloned genes. van Beek et al. (2003) constructed a series of AcMNPV recombinants expressing LqhIT2 scorpion toxin gene with different signal sequences, including signal sequences of AcMNPV GP64, cuticle protein II of Drosophila melanogaster, bombyxin of B. mori, dipteran chymotrypsin, and some scorpion toxins. Bombyxin signal sequence proved to be the most effective for enhancing insecticidal efficacy. Further searches for new promoters and for more effective signal sequences in transporting a toxin outside of the expressing cell are being carried out in many laboratories, and it is expected that many more natural and synthetic promoters and signal sequences will improve the speed of kill and safety of recombinant baculoviruses.

Biosafety of a biopesticide is an important problem, which requires special consideration. Biosafety can never be assured with absolute confidence, but a number of studies indicate that baculoviruses pose no hazard to animals other than their hosts. Though baculoviruses can enter mammalian cells, productive viral infection does not occur even at very high multiplicity of infection (Kost et al. 2005). Additionally, the foreign gene to be expressed after baculovirus infection must be placed under specific mammalian promoters; the expression from the baculoviral promoter has never been observed. Recombinant HaSNPV expressing AaIT scorpion toxin gene was not pathogenic to bees, birds, fish, and other vertebrates (Sun et al. 2002). Genetically modified AcMNPV did not affect the aquatic microbial community in any respect (Kreutzweiser et al. 2001). Natural enemies of larvae such as parasitoids and predators were not adversely affected by preying upon larvae infected with recombinant viruses (Li et al. 1999; Smith et al. 2000; Boughton et al. 2003). Also, it has not been proven thus far that the foreign gene can be transferred from donor recombinant baculovirus to another organism (Inceoglu et al. 2001, 2007). On the basis of these reports, it can be concluded that there is no evidence that recombinant baculoviruses pose greater threats to the animal world and the biosphere than the parental baculoviruses. However, in spite of this fact, field trials of genetically modified baculoviruses have instigated massive public protests, which put further trials on hold. The slow progress in application of genetically modified baculoviruses as pesticides may be, in part, due to the choice of “exotic” toxin genes used for modifications of the baculovirus genome. Taking into account the origin of these social conflicts, the choice of toxins used for this purpose should be reexamined, and baculoviruses should be modified with genes coding for more “natural” insect toxins, e.g., with genes coding for toxic polypeptides of parasitoid wasps occurring in regions infested by a particular pest.

16.7 Final Considerations and Further Prospects on Use of Baculoviruses as Biopesticides

Baculovirus insecticides have not met their full potential to control pest insects worldwide. In his review, Moscardi (1999) previewed the following: (1) The expansion of baculovirus use, in the following 5 years, i.e., up to 2004, would depend on new developments in the areas of recombinant baculoviruses and in the in vitro commercial production of these agents. The development of recombinant baculovirus was efficiently completed by researchers in several countries, but the in vitro commercial technology still lags behind today due to technical problems; (2) The use of baculoviruses would increase substantially in 10 years (i.e., up to 2009). However, this did not occur; (3) The AgMNPV program in Brazil could reach about four million hectares of soybean. This did not happen either. In reality, the use of the AgMNPV declined from two million hectares to about 300,000 ha over the past 7 years due to reasons discussed above (Sect. 16.4.1.4). In spite of this reduction in AgMNPV usage, this program can be considered an example regarding the viability of baculoviruses as insecticides. A current program for revival of the integrated pest management of soybean insect pests in Brazil will help to increase the use of AgMPV.

Despite the low use of viral insecticides worldwide (ca. 0.5%) as compared to biopesticides based on the bacterium Bacillus thuringiensis, total use of microbial insecticides worldwide is only about 2.0–2.5% of the total market of insecticides. Despite the low market influence of baculovirus insecticides, there are over 50 registered products in different countries, including the same product under different trade names. In the future, genetically modified baculoviruses will contribute to the expansion of baculovirus use worldwide, as these GMOs are considered safe through extensive research conducted over many years (Szewczyk et al. 2009). The most important issue for baculovirus use will be public perception regarding the benefits of baculovirus GMOs to control insects, including low impact on the environment. Also, regardless of whether a program is based on a wild-type or a genetically engineered baculovirus, global farmer education toward general use of biological pest control agents will be a key feature for expansion of baculovirus use worldwide. Unfortunately, pest control programs in most countries are directed toward the use of chemical insecticides, as in Brazil, where the official extension services have been “demolished” in 90% of the states over the past 10 years, leaving farmers to the control recommendations of professionals related to agrochemical companies. The use of baculoviruses as very specific bioinsecticides will depend on sound IPM programs, where integration of available techniques to control insects are used to reduce the number of chemical insecticide applications on a given crop and minimize the environmental impact of pest control. In systems where no IPM programs exist, there is little chance of success of use of a very specific baculovirus, especially in crop production systems where the one to be controlled with a baculovirus occurs with other concurrent insect pests. Adoption of the IPM approach by farmers is important for use of baculovirus pesticides for successful sustainable agriculture.