Functional analysis of a novel baculovirus envelope fusion protein

Baculoviridae are a family of large double stranded DNA viruses that are exclusively pathogenic to arthropods. Baculoviruses have been studied i) with the aim to develop alternatives to chemical pest control, ii) for their application as<span class=GramE>an</span>eukaryotic expression system to express heterologous proteins, and recently iii) as gene delivery vehicle in gene therapy. Baculoviruses cluster into two distinct genera on the basis of the occlusion body (OB) morphology: Nucleopolyhedrovirus (NPV) and Granulovirus (GV). The NPVs are further subdivided into a group I and a group II, based on the phylogeny of some of their genes. A feature, shared by all baculoviruses, is the production of two viral phenotypes during the infection cycle: the occlusion derived virus (ODV), which is responsible for the horizontal transmission of the virus in insect populations, and the budded virus (BV), which is responsible for the spread of the virus within the host and in cultured insect cells. A NPV infection starts with the uptake of OBs, which dissolve in the midgut, liberating numerous ODV particles. The nucleocapsid (NC) enters the midgut epithelial cells after fusion of the ODV envelope with the cellular membrane. Then the NCs are transported to the nucleus, where transcription, replication and assembly of progeny NCs occur. The progeny NCs are translocated to the plasma membrane, where they acquire an envelope containing, at least in group I NPVs, the<span class=GramE>virus</span>encoded major envelope glycoprotein GP64. After budding from the plasma membrane the progeny BV is responsible for the systemic infection of the insect beyond the initially infected midgut epithelium and enters other insect cells by the mechanism of receptor mediated endocytosis. The acidification of the endosome triggers the GP64-mediated fusion of the viral and endosomal membrane. The NCs are thereby released in the cytoplasm and transported to the nucleus. Determination of the genome sequence of the group II NPVs Lymantria dispar (Ld<span class=GramE>)MNPV</span>and Spodoptera exigua (Se)MNPV led to new insights with regard to BV envelope structure. Neither LdMNPV nor SeMNPV contains a gp64 gene. Instead, group II NPVs, or at least LdMNPV and SeMNPV, possess another BV envelope (glyco<span class=GramE>)protein</span>that has a function analogous to GP64.<o:p></o:p></span>First it was investigated whether the SeMNPV BV enters the cell by direct membrane fusion or by the endocytotic pathway ( <strong>chapter 2</strong> ). The use of a lysosomothrophic reagent inhibited SeMNPV BVs infection, indicating that the virus enters by the endocytotic pathway. The major envelope protein of the SeMNPV BVs was sequenced and this appeared to be a C-terminal cleavage fragment of the primary translation product of ORF 8 in the SeMNPV genome. The Se8 gene product as well as its homolog (Ld130) was shown to be localized at the plasma membrane. Syncytium formation assays showed that these proteins alone were sufficient to mediate pH-dependent membrane fusion and were therefore named the fusion (F) protein. The fact that F was detected in BVs as a smaller fragment derived from a larger proprotein, suggested that upon maturation F was posttranslationally cleaved. Such cleavage is a general mechanism, found among vertebrate viruses to activate their fusion proteins. The occurrence of a conserved proprotein convertase cleavage site in the primary translation product of SeMNPV F as well as LdMNPV F suggested the possible involvement of a cellular furin in the cleavage reaction. The use of a specific inhibitor indicated that furin is indeed involved in the cleavage of SeMNPV F and that this cleavage is required for the fusiogenic activity of the protein ( <strong>chapter 3</strong> ). This was further confirmed by demonstrating that a mutant F protein with a dysfunctional cleavage site did not mediate membrane fusion in a syncytium formation assay. By means of Western analysis it was shown that F is a BV specific protein and that both its N-terminal (F ₂ ) and C-terminal (F ₁ ) cleavage product is present in BVs and presumably covalently linked by a disulfide-bridge.<o:p></o:p></span>As far as its function as fusion protein, all results indicated that F was functionally analogous to GP64. To investigate whether this analogy also held with respect to receptor binding and virion budding, an AcMNPV pseudotyping system was developed ( <strong>chapter 4</strong> ). A gp64 -deficient AcMNPV bacmid was successfully pseudotyped with SeMNPV and LdMNPV F proteins, as infectious BVs were obtained upon transfection of the bacmids into insect cells. This pseudotyping also demonstrated once more that F has to be cleaved to release its fusiogenic capabilities, since an F protein with a dysfunctional cleavage site was not able to rescue the loss of the gp64 gene. Western analysis showed that the mutated protein was incorporated in BVs, but in its uncleaved (F ₀ ) conformation. The meantime available genome sequences of Xestia c-nigrum (Xecn<span class=GramE>)GV</span>,</span>Plutella xylostella</span> (Plxy)GV and ; mso-fareast-language:EN-US;layout-grid-mode:line'>Cydia pomonella</span> (Cp)GV revealed that<span style='mso-spacerun:yes'> </span>also these granuloviruses do not contain a gp64 gene. Therefore it was also examined whether the F protein homolog of PlxyGV was capable to pseudotype the gp64 -null AcMNPV mutant. However, no infectious BVs were produced upon transfection of the bacmid in insect cells, although the PlxyGV F protein was demonstrated in the produced BVs.<span style='mso-spacerun:yes'> </span><o:p></o:p></span>In addition to a gp64 gene group I NPVs have a remnant f gene in their genome. GP64 homologs are also found on two tick-transmitted Orthomyxoviruses, Thogoto virus and Dhori virus. This suggests that an ancestral group I NPV may have acquired the gp64 gene either from their host or from another virus, thereby acquiring a selective advantage in replication and obviating the need for a functional F protein. To experimentally address this hypothesis a SeMNPV f -null mutant was pseudotyped with GP64 ( <strong>chapter 5</strong> ). However, no infectious BVs could be obtained in this experiment, indicating that the f gene is essential for viral propagation of SeMNPV. Reinsertion of the f gene of SeMNPV restored the wild-type phenotype of SeMNPV. These results suggest that F from group II NPVs might contain additional essential functions in producing infectious BVs. These functions are essential in group II NPVs, but not in group I NPVs.<span style='mso-spacerun:yes'> </span><o:p></o:p></span>The biological significance of the posttranslational cleavage of SeMNPV F was further investigated ( <strong>chapter 6</strong> ). In fusion proteins of vertebrate viruses this cleavage occurs just in front of a hydrophobic, called "fusion peptide". As the N-terminus of SeMNPV F ₁ possesses common features to these vertebrate viral fusion peptides this sequence was studied by mutational analysis. SeMNPV F proteins with single point mutations at conserved positions as well as with a complete deletion of the fusion peptide domain were analyzed in the baculovirus pseudotyping system. Two of the F mutants, including the deletion mutant, were unable to rescue the propagation defect of the gp64 -null bacmid, while other mutants were significantly reduced in their viral infectivity or viral propagation in cell culture. These results indicate that N-terminus of SeMNPV F ₁ is most likely the "fusion peptide" involved in the viral entry mechanism.<o:p></o:p></span>Finally the results obtained during the reported PhD study are discussed ( <strong>chapter 7</strong> ) with respect to baculovirus entry, evolution and taxonomy