Structure-function relationship of the baculovirus envelope fusion protein F

Baculoviruses are a large group of enveloped, double-stranded DNA viruses exclusively infectious for arthropods, predominantly insects of the order Lepidoptera. Baculovirology has advanced considerably in recent years as a consequence of the successful use of baculoviruses (i) in biological control of insect pests, (ii) as efficient expression vectors for eukaryotic proteins and (iii) as gene delivery vectors into mammalian cells. In the baculovirus infection cycle two phenotypes of virions, occlusion derived virions (ODVs) and budded virions (BVs), are produced. ODVs start the infection in insect larvae by direct fusion with epithelial cells in the midgut, whereas BVs infect cells from other internal tissues and cells in culture through adsorptive endocytosis. Major envelope fusion proteins located in the apical end of BVs play a major role in these processes as they mediate internalization and fusion for successful penetration of cells.In the family Baculoviridae two distinct envelope fusion proteins are identified in BVs. GP64 is the major envelope fusion protein present in the baculovirus Autographa californica multi capsid nucleopolyhedrovirus (AcMNPV) and other group I nucleopolyhedrovirus BVs. In group II NPVs, for example Spodoptera exigua multi-capsid NPV (SeMNPV), Lymantria dispar MNPV (LdMNPV) and Helicoverpa armigera single-capsid NPV (HearNPV) (Chapter 1), a different type of envelope fusion proteins, designated as F, was present in BVs. Viral F proteins in general mediate low-pH-dependent fusion during virus entry (endocytosis) and baculovirus F proteins are functional analogs of GP64. In spite of significant divergence in primary sequence, baculovirus F proteins are classified as class I fusion proteins based on striking similarity in the architecture and functional elements (signal peptide, proteolytic cleavage site, fusion peptide, heptad repeats and transmembrane region) with fusion proteins of enveloped RNA viruses. To date baculovirus F proteins are the first class I virus envelope fusion proteins identified in a DNA virus. Therefore, to enhance our knowledge of class I virus fusion proteins in general and to further understand the functional implications of the specific characters of baculovirus F proteins, the HearNPV F protein was chosen as a model for study.In Chapter 2 the F protein of HearNPV was identified by N-terminal sequencing of the major BV envelope protein (59 kDa). HearNPV F protein was found to be encoded by open reading frame 133 ( Ha133 ) of the HearNPV genome. HearNPV F protein was first synthesized as an 80kDa precursor protein (F 0 ), which was proteolytically cleaved by a furin-like protease, residing in the trans-Golgi network, into two disulfide-linked subunits: F 1 (a subunit of 59 kDa carrying a transmembrane region) and F 2 (a 20 kDa subunit on the extracellular side) both forming a heterodimer. The HearNPV F protein gene successfully rescued the infectivity of an AcMNPV gp64 deletion mutant, confirming that baculovirus F homologous proteins from group II NPVs are functional analogs of GP64. In addition, by chemical cross-linking it appeared that HearNPV F is present in BVs in the form of multimers whereby, unlike in the case of GP64, disulfide-bonds are not involved. Deglycosylation assays showed that both the F 1 and F 2 subunits of HearNPV possess N-linked glycans. However, these glycans did not have a C3 fucose modification when produced in Hz2E5 cells, but did have such a modification when produced in Trichoplusiani BTI-Tn-5B1-4 (High Five) cells. Since C3 fucose is a major inducer of allergic responses in humans, this observation makes the HearNPV- Heliothis cells system an attractive alternative for the production of recombinant glycoproteins for therapeutic applications in animals and humans.Virus envelope fusion proteins are generally heavily N-linked glycosylated. N-linked glycosylation is a modification step in protein biosynthesis occurring in the endoplasmatic reticulum. It is important for proper folding and intracellular trafficking, but also for the functionality such as virus attachment, fusogenicity and antigenicity. Baculovirus F proteins are N-linked glycosylated. However the number and location of predicted N-linked glycosylation sites vary from different baculovirus F proteins. One or more N-linked glycosylation sites are present in F 2 subunits of baculovirus F proteins and one site is conserved suggesting a possible role in F functionality. In Chapter 3 the only asparagine (N104) of the only predicted N-linked glycosylation site NXS/T in F 2 of HearNPV F was substituted by a glutamine (Q 104 ) and as a consequence F 2 was not glycosylated. Absence of N-linked glycans on the F 2 subunit of HearNPV F protein did not affect F protein synthesis and incorporation of F into BVs. Plaques produced by recombinant HearNPV expressing a mutant F protein (N104Q) are even slightly larger than those produced by recombinant HearNPV expressing wild-type F protein. The recombinant baculoviruses expressing N104Q also induced much more efficient low-pH activated syncytium formation and produced more BVs at the early stage of infection. These observations suggest that N-linked glycans attached on F 2 of baculovirus F proteins are involved in virus infectivity and F fusogenicity. Total inhibition of N-linked glycosylation resulted in low cellular expression levels of the HearNPV F protein and in poor proteolytic cleavage, indicating that N-linked glycosylation is indeed a prerequisite for proper F protein biosynthesis and F function. It is proposed that the N-linked glycosylation of F 1 is an evolutionary adaptation to modulate the virulence of the virus.Enveloped viruses enter target cells through envelope fusion mediated by virus envelope fusion proteins. This fusion step occurs either on the surface of the cell membrane at neutral pH or in the late endosome activated by a low pH. Entry of baculovirus BVs into insect and mammalian cells is generally thought to occur through a low-pH-dependent endocytosis pathway, possibly through clathrin-coated pits. This insight, primarily based on (immuno-) electron microscopy studies, lacks biochemical and experimental support. Moreover, other entry mechanisms for virus exist and the relevance of these for baculovirus entry needs to be investigated. In Chapter 4, using specific inhibitors of endocytosis, it was demonstrated that both group I (GP64) and group II (F) NPVs baculoviruses enter insect cells primarily through clathrin-mediated endocytosis. Functional entry of baculoviruses (AcMNPV) into mammalian cells (BKH21) is dependent on the same endocytic pathway. Additionally, using a caveolar endocytosis inhibitor, genistein, the significantly enhanced baculovirus transduction in BHK21 suggests that caveolae are somehow involved in baculovirus entry in mammalian cells. This is not the case in insect cells, where the clathrin pathway seems to be the only one. These observations might promote the engineering of novel baculovirus gene transfer vectors and support a new strategy to enhance baculovirus transduction efficiency in mammalian cells by locking up of the caveolar pathway.F proteins from different baculovirus species have a cytoplasmic tail domain (CTD), ranging from 48 ( Spodoptera litura multicapsid NPV [MNPV]) to 78 ( Adoxophyes honmai NPV) amino acid (aa) residues. This is much longer than the CTD of GP64-like envelope fusion proteins (7aa) in group I NPV, the latter CTD being non-essential for infectious BV production. In Chapter 5, the functional role of the CTD of HearNPV F was studied using truncation mutagenesis and F swapping approaches. A HearNPV F gene without CTD was not able to rescueanHearNPV virus lacking F, but was able to rescue the infectivity of an AcMNPV lacking GP64. Combining these two lines of evidences, it was demonstrated that the CTD of HearNPV F is not essential for F biosynthesis, processing, trafficking and fusogenicity, but essential for infectiousHearNPV BVproduction. This finding further suggests that the F CTD is important for F protein incorporation during BV assembly and budding. In group I NPVs such as AcMNPV, F homologs with a CTD are not required for BV production, which might suggest that the assembly of group I and group II NPV BVs occurs differently. It is hypothesized that BV formation of group II NPVs requires a minimal CTD, probably to interact with nucleocapsid proteins.Heptad repeats (HRs), a conserved structural motif of class I viral fusion proteins, are held responsible for the formation of a six-helix bundle structure of the envelope fusion proteins. Peptides derived fromtheseHRs can be potent and specific inhibitors of membrane fusion and viral infection. Similar to many viral fusion proteins, a conserved leucine zipper motif was predicted downstream of the fusion peptide proximal to the HR region, HR1, in baculovirus F proteins. To study the functional role of the HR1 domain in baculovirus F protein, site-directed mutagenesis was used and mutant F was allowed to rescue the infectivity of an AcMNPV lacking GP64 (pseudotyping) (Chapter 6). The HearNPV F mutant proteins with a single leucine to alanine substitution showed lower rescuing capacity than the wild-type F. Thus the conserved leucine residues were critical for F protein function. The F L216R mutant was not able to rescue the AcMNPV GP64-minus virus for reasons yet unknown. Using AcMNPV as expression vector, it was further shown that all F proteins mutated in HR1 were incorporated in BVs, indicating that substitutions of the leucine residues did not affect intercellular trafficking of F protein, although F L216R showed a lower viral incorporation rate. Proteolysis was not affected by any of the substitutions of the leucine residues. However, folding of F was significantly affected. A large number of malformed multimers of F was detected in F L216A , F L216R and F L223A mutants and F L216R showed predominantly malformed oligomers. In addition to the expected paucimannose N-linked glycans, heavier N-linked glycans were found attached to the F 2 subunits of the F L216R mutant. Collectively these results indicate that the HR1 domain within baculovirus F proteins is important to secure the conformational properties of the F protein and that the HR1 domain is critical more specifically for proper folding and post-translational modification.The research described in this thesis unraveled some of the important characteristics of baculovirus F proteins in relation to their function in the infection process. These findings further enhanced our knowledge on the role of baculovirus F protein in the uptake in insect cells versus mammalian cells, in the fusion process and in BV maturation and assembly. Finally, in Chapter 7 the data obtained from this study are discussed in the light of the current understanding of the function of baculoviruses and other F proteins.