Background and objectives

Baculoviruses are insect pathogens that are used as biological control agents of insect pests as alternatives to chemical insecticides. Safety testing confirms that baculoviruses are insect specific and cause no hazards to beneficial insects, other animals or to plants. The major drawback for the commercial use of baculoviruses is their slow action, and this is especially relevant for crops which have low damage thresholds. They also have a limited host range and are costly to produce. These problems called for the improvement of the insecticidal ability of baculovirus strains by genetic modification.

This study aimed to test the biosafety of genetically modified baculovirus so that their behaviour in the environment can be predicted.

Approach and methodology

The behaviour of genetically-modified baculoviruses was assessed by the construction of baculoviruses with increased virulence, but reduced persistence and survival in the environment. The baculovirus Autographa californica nuclear polyhedrosis virus (AcNPV) provided a suitable model. AcNPV was marked, and either a suicide mechanism incorporated or deletions introduced to reduce its persistence and survival in the field. The biosafety of the recombinant AcNPV baculovirus was compared to that of the wild type.

Subsequently, we designed and tested a microcosm system and carried out limited field release of the genetically modified AcNPV.
Top

Main findings and outcome

The deletion of six AcNPV genes, believed to affect the persistence, survival and spread of AcNPV in the environment, did not affect the replication of AcNPV or its efficiency as an insecticide. Thus, deletion mutants in these genes may have a reduced ability to spread in the environment. Single, double and triple mutants were constructed, with and without the lacZ reporter gene. The promoters of the deleted genes were left intact to avoid disturbing the overall transcription during virus replication.

Attempts were made to construct a suicide recombinant by placing the Escherichia coli lac operator (lacO) up- or downstream of the polyhedrin (ph) gene transcriptional start site and the lac repressor (lacI) under the control of the Drosophila heat shock promoter hsp70. Thus, in theory, the expression of ph and the production of polyhedra is blocked in the field and unblocked by the addition of IPTG. In practice, the introduction of lacO was insufficient to block the expression of ph.

To study UV-persistence, dry virus deposits were inactivated by direct exposure to artificial UV-sunlight. There was no detectable difference in half-life between the wild-type AcNPV and an AcNPVpe- mutant (resulting in envelope-less polyhedra). The loss of the polyhedron envelope does not appear to increase UV sensitivity.

Comparison of the six AcNPV deletion mutants in a sensitive host showed that the infectivity of single deletion mutants does not differ significantly from the wild type. However, double deletions may be less infective.
Top

The survival and spread of genetically modified viruses was compared with that of non-modified viruses after their release into the environment. Mixed infections of wild type and recombinant AcNPV were passaged in cohorts of insects. After each passage the proportion of each virus type in the mixture was estimated and the remainder re-passaged. After several passages the recombinant was lost from the system, suggesting reduced biological fitness. Similar experiments were run in which one virus was introduced a set time after the other. This showed that the first virus does not prevent infection by a second virus, but does interfere with its replication, significantly reducing its titre. Consequently, the opportunities for recombination between genetically modified and non-modified baculoviruses are considerably reduced when infection with another virus is asynchronous.

Due to concerns that baculoviruses may interact with non-target Lepidoptera we examined the extent of infection in permissive, semi-permissive and poorly permissive hosts. In the semi-permissive host the level of infection was determined by the dose/host ratio, mortality only occurring at a certain dose threshold. In the poorly permissive host no mortality occurred regardless of the dose.

Genetically modified AcNPV were tested in a contained microcosm prior to field release to obtain data about their yield and dispersal. Two mutants both carrying a LacZ reporter gene were extensively tested. The microcosm was separated into two compartments by an insect screen that prevented larval migration but allowed air circulation. Each compartment contained soil, sugar beet plants and second instar larvae of S. exigua. Wild-type or recombinant AcNPVs were applied to one compartment. After one larval generation the quantity of virus in the soil, plants and water was determined. There was no difference between the infectivity of the wild type and that of the recombinant AcNPV. The spread of wild type AcNPV was however five times greater than the recombinant. Equal quantities of non-modified and genetically modified viruses resulted in only small differences in larval mortality (39 and 48% respectively). The recombinants were better retained in the upper layers of the soil whereas the wild types penetrated the throughout the soil and contaminated infiltration water.
Top

Conclusions

We successfully generated stable genetically modified baculoviruses with multiple deletions whilst maintaining their specificity and biological activity. We tested the behaviour of single deletion mutants in a microcosm and demonstrated that they have a reduced ability to spread. Genetically modified baculoviruses may be less competitive than the wild type during asynchronous mixed infections.

The trachea was identified as the main conduit for dissemination of infection through the insect host’s body.