The structure and replication of some insect viruses

Insect pathogenic viruses can be divided into five groups. Nuclear polyhedrosis viruses, cytoplasmic polyhedrosis viruses, granulosis viruses, and 'pox-like' viruses are viruses of the occluded type in which the virus particles are embedded in a large crystalline inclusion which may be either polyhedral or capsular in shape. Other viruses do not occur in such 'inclusion-bodies' and are often referred to as non-occluded viruses. Nuclear polyhedrosis viruses have been extensively investigated but knowledge of many of their properties is ill-defined or lacking and assumptions have been made about their structure and multiplication which are contrary to accepted concepts of virus assembly. In this work two nuclear polyhedrosis viruses have been investigated, largely by electron microscopic methods, in order to determine how the structural and dynamic properties of the viruses can be correlated. Consideration has also been given to the compliance of these properties with current thinking on virus structure and replication. The viruses employed were those of the small tortoise-shell butterfly, Aglais urticae, and the gipsy moth, Lymantria dispar. Polyhedra were extracted from infected cadavers and purified by density gradient centrifugation. Suspensions of polyhedra were subjected to various alkaline treatments to solubilize the crystalline matrix of protein and to liberate the occluded virus particles. These virus particle and an aggregated form of the polyhedron protein were separated and purified most effectively by density gradient centrifugation. Samples of virus particles and aggregated polyhedron protein were examined electron microscopically and used for the production of antisera. Antisera were also produced to polyhedron protein solutions obtained, by precipitation, from dissolutions of polyhedra from which virus particles and aggregated polyhedron protein had been removed by centrifugation. Serological tests in gel and amine acid analysis showed that the different types of polyhedron protein were related, but some probably lacked certain antigens. Serological tests showed no relationship between virus particles and polyhedron protein but the polyhedron proteins from the two viruses were related but not identical. Aggregated polyhedron proteins often exhibited a lattice-like arrangement when examined electron microscopically in negatively stained preparations and from a comparison of such lattices and the lattice structure of the polyhedron protein seen in ultra-thin sections of polyhedra a cubic arrangement has been proposed which is built from inter-connecting six-armed nodal units. The polyhedron is bounded by a membrane and preparations containing such membranes could be harvested from solutions of dissolved polyhedra by centrifugation. When examined electron microscopically the membranes appeared to possess a cubic lattice structure of similar dimensions to polyhedron protein. Closer examination however showed them to be composed of holes or hollows arranged hexagonally each hole having a central core. Polyhedron membranes did not have the structure of a unit membrane when seen in section. Virus particles extracted from polyhedra were either enveloped, naked, or empty. In A. urticae polyhedra the virus envelope (or outer membrane) surrounds a single virus particle whereas in L. dispar polyhedra bundles of virus particles are surrounded by the envelope. The envelope appears to be a three-layered structure with an outer surface layer in which no detailed substructure can be resolved, a layer of hexagonally packed subunits or peplomers 20 mandmu; in diameter arid a flexible membrane composed of 40andAring; subunits, hexagonally packed, which resembles a unit membrane when seen in section. It is likely that the surface layer and the peplomers are virus-coded structures which become attached to a host cell membrane. The virus particle consists of an internal component which is probably coiled, surrounded by a virus capsid (or inner membrane) which is apparently composed of 30andAring; subunits arranged in a diamond-shaped lattice network. The virus capsid appears to connect those virus particles (or nucleocapsids) enveloped in bundles one to another as a result of 'pinching' of the capsid. However it is likely that each individual virus particle is capable of establishing virus replication as the number of virus particles within an envelope varies considerably. The dynamic properties of the viruses were investigated by dissecting larvae at intervals after infection and embedding certain tissues for ultra-thin sectioning and electron microscopy. In A. urticae larvae enveloped virus particles were found adjacent to the microvilli of midgut columnar cells and naked virus particles were found within the microvilli. The nuclei of these cells were usually virus-infected but although some crystalline polyhedron protein was present in the nuclei no polyhedra were formed. Enveloped virus particles were found both in the cell cytoplasm and in the underlying basal lamina. The tracheal epithelium cells beneath the basal lamina were normally virus-infected and polyhedron formation occurred in the usual manner. Such infected columnar cells were not observed in L. dispar larvae but enveloped virus particles were found in the basal lamina. It is suggested that the pathway of infection of the virus is initiated by infection of the gut columnar cells as a result of attachment to the plasma membrane by the virus envelope. Many enveloped virus particles are then produced in the nuclei of these cells without the formation of polyhedra. The enveloped virus particles are therefore available for the infection of cells and tissues in the host larva and the virus spreads largely as a result of infection of the epithelial cells of the tracheal system. Typical development of the viruses was studied in fat body cells. A large densely staining network, the virogenic stroma, is clearly visible in the enlarged nuclei of the cells and naked rod-shaped virus particles appear to be produced in association with it. The virus particles acquire an envelope in spaces within the virogenic stroma or in the area between the virogenic stroma and the nuclear membrane. In L. dispar virus-infected cells the virus particles are often criss-crossed within the envelope. Polyhedron protein crystallines in a lattice arrangement between the enveloped virus particles and compression of the envelope against the virus particles takes place as a result. The criss-cross formation of the virus particles within the envelope in L. dispar virus-infected cells becomes compressed into a side-by-side arrangement because of this process. The polyhedron membrane is visible as a densely stained periphery around mature polyhedra. Similar infected tissue was embedded in water-soluble glycol methacrylate and the sections were treated with enzyme solutions. The virus particles were digested both by DNase and pronase but the virogenic stroma was mostly affected by pronase. Polyhedron protein was somewhat resistant to digestion with pronase. Phospholipase D appeared to cause partial digestion of the virus envelope. Nuclear polyhedrosis viruses are extremely complex structurally but conform to the general concepts of virus assembly. The dynamic properties of the viruses cause gross disturbance of the architecture of the normal host cell nucleus and many of their structural components can be observed during the replicative process. Nuclear polyhedrosis viruses raise problems of nomenclature and classification which can only be resolved by further work on their properties. Such work should help to establish these viruses as useful and stimulating entities for study.