The Baculoviridae is a large family of viruses that infect arthropods, particularly insects of the order Lepidoptera, Diptera, and Hymenoptera. Baculoviruses contain circular double-stranded DNA genomes of 80 to 180 kb, which are contained within an enveloped, rod-shaped virion. The family Baculoviridae is divided into four genera based on the infected host larva: the alphabaculovirus are lepidopteran-specific nucleopolyhedroviruses (NPVs), the betabaculovirus are lepidopteran-specific granuloviruses, the gam-mabaculovirus are hymenopteran-specific NPVs and the deltabaculovirus are dipteran-specific NPVs The lepidopteran NPVs can be further subdivided into groups Ⅰ and Ⅱ based on phylogenetic studies (15). The group Ⅰ NPVs include Autographa californica multiple NPV (AcMNPV) and Bombyx mori NPV (BmNPV). Both viruses are models for studies on basic baculovirology and application of baculoviruses as gene expression vectors and insecticide.
Since the genome sequence of AcMNPV was reported for the first among baculoviruses, the number of complete sequences has grown rapidly and current database contains the sequences of 48 baculovirus genomes including 12 group Ⅰ NPVs. These demonstrated that there are 30 conserved core genes present in all baculoviruses and 17 genes present only in group Ⅰ NPVs (16). A signature gene among these unique to group Ⅰ NPV is gp64/gp67. A distinguishing difference between group Ⅰ and Ⅱ is that group Ⅱ NPVs utilize an ancient fusion homologue for cell-to-cell spread and this fusion activity is replaced by gp64 in group Ⅰ NPVs. Therefore, it has been sug-gested that acquisition of these unique genes promoted the diversification of group Ⅰ NPVs and may contri-bute to baculovirus speciation by causing alterations in host range (16). The orf8 gene (Bm8) of BmNPV is one of such gene found in only and all group Ⅰ NPVs sequenced to date. In this review, Ⅰ summarize current knowledge of this gene family that was obtained mostly from AcMNPV and BmNPV.
Table 1 summarizes the information on Bm8 and its homologues from group Ⅰ NPVs. The deduced amino acid sequence of Bm8 showed significantly high identity (93 to 96 %) to that of AcMNPV, Rachiplusia ou (Ro) MNPV, and Plutella xylostella (Plxy) NPV homologues. On the other hand, it showed lower identity (approximately 30 %) to the rest. A phylogenetic tree of deduced amino acid sequences of this gene family also showed that Bm8 was closely related to the homologues from AcMNPV, BmNPV, RoMNPV, Maruca vitrata (Mavi) MNPV, and to lesser extent to those from the other viruses (Fig. 1). This phylogenetic tree showed two monophyletic clades; Bm8 was found in a clade B including well-characterized species such as AcMNPV and RoMNPV. The genes in this clade seem to be very closely related to each other except for the homologue from Mavi-MNPV. The other clade A included Epiphyas postvittana (Eppo) NPV, Choristoneura fumiferana (Cf) MNPV, C. fumiferana DEF (CfDEF) MNPV, and Orgyia pseudotsugata (Op) MNPV as prominent species. The amino acid identity between the genes in this clade was not so high (50 to 70 %), suggesting that the genes in clade A are more diverse than the genes in clade B.
Table 1. List of BmNPV orf8 and its homologues
Most of Bm8 homologues were predicted to encode proteins of approximately 200 amino acids (aa) except for Plxy16 and Hycu137 (Table 1). The Plxy16 ORF contains only 74 aa and was approximately 150 aa smaller than others. Harrison and Lynn (14) suggested that this truncation was due to a single base mutation and that the virus with this truncation is maintained in larvae. In addition, the predicted amino acid sequences of flaking small ORFs did not exhibit any similarity to other known proteins including Bm8 (14). On the other hand, Hycu137 includes 127 aa, which are approxi-mately 100 aa shorter than the other homologues. This truncation was due to a single base insertion that was confirmed by re-sequencing of amplified Hycu137 region from larval stock (personal communication with M. Ikeda). However, the flanking ORF (Hycu136 containing 72 aa) also showed significant similarity to the C-terminal region of Bm8 (18). Therefore, there is a possibility that Hycu136 corresponds to the trun-cated C-terminus of Hycu137.
It has been shown that N-terminal region of Bm8 is important for its function and contains a coiled coil domain (19, 20). Prediction programs of coiled coil domains revealed that most of Bm8 homologues contain a coiled coil domain in N-terminal region (between 60 to 110 aa) except for Plxy16 (Table 1). The Plxy16 ORF may be too short to contain this motif since it encodes only 74 aa. The Cf15 ORF exhibited a low level of probabilities for coiled coil prediction, however, the protein encoded by Cf15 may contain a coiled coil domain because the sequence of around this region in Cf15 was aligned well with others.