There are currently 14 proposed species of lyssavirus and these are classified into different phylogroups according to genotype and cross reactivity of the G protein.
The phylogroup classification suggests differing degrees of protection afforded by existing rabies vaccines. Current rabies vaccines are based on inactivated classical rabies viruses which belong to phylogroup I and are known to provide protection against other classical rabies viruses although the level of protection afforded against other viruses is ill defined. It is known, however, that protection against phylogroup II viruses with current vaccines is unlikely (Hanlon et al., 2005).
The potential for protection against the newly discovered Bokeloh Bat Lyssavirus (BBLV) and Shimoni Bat Virus (SHIBV) have yet to be determined although from phylogenetic evidence it is thought that protection may be afforded against BBLV, but is unlikely for SHIBV. This variation in protection afforded by current lyssavirus vaccines demonstrates the importance of the development of vaccines against non-rabies lyssaviruses. In order to develop an inter-phylogroup cross protective vaccine, knowledge of the antigenic epitopes and variation within these sites is essential.
It is widely accepted that the lyssavirus G protein is the sole target of neutralising antibodies following vaccination and as such, areas of antigenicity are of great interest. Across the lyssavirus G protein there are four main and one minor antigenic site and these are thought to determine pathogenicity and are essential for the induction of a neutralising immune response following vaccination.
Studies with antigenic site I have suggested that it contains both conformational and linear epitopes as it was originally delineated by mAbs recognising correctly folded G (Lafon et al. 1990) although more recent studies have suggested the presence of a linear epitope within this site (Bakker et al. 2005). Interestingly, antigenic site I is defined as a stretch of 6 amino acids with the general sequence 245/XLCGXX/250 where the LCG motif is conserved across all phylogroup I and II lyssaviruses. Indeed this short motif is also conserved in WCBV with only a minor L246I amino acid difference. Antigenic site II is defined as being a discontinuous conformational epitope having two domains, IIa (217-219) and IIb (53-61), separated by a 156 amino acid stretch. Studies have suggested that rather than being two independent antigenic sites, that sites IIa and IIb form a single antigenic epitope within the mature protein.
Antigenic site III (349-357) is a continuous stretch of 8 amino acids. Interestingly, no site specific mAbs are able to bind immature forms of the G protein indicating that this site forms part of a loop on the protein surface (Benmansour et al. 1991). It may be this tertiary structure which enables binding by antibodies or the interaction of the viral G protein with neuronal receptor molecules. Site III also plays an important role in pathogenicity as it contains the residue at position 352 which tends to be an arginine or lysine in pathogenic viruses but has been substituted in less virulent viruses (Dietzschold 1983). Antigenic site IV consists of only two amino acids and is continuous though contains overlapping linear epitopes. Minor site ‘a’ is located in close proximity to site III but contains no overlapping epitopes and consists of only two amino acids.
As well as conserved antigenic sites there are 13-16 cysteine residues conserved across all phylogroups which form disulphide bonds to provide the mature protein’s structure which as yet has not been fully determined with x-ray crystallography (Walker & Kongsuwan 1999). There are 7 disulphide bridges present in the G protein of the lyssaviruses and the putative tertiary structure is shown in figure 1.
Figure Figure 1: The proposed putative structure of rabies virus glycoprotein, as deduced from the location of conserved cysteines. (Walker & Kongsuwan 1999)
Lyssaviruses remain a significant threat to public and animal health where these viruses remain endemic. As described earlier, current rabies vaccines are all based on classical rabies strains and, since the discovery and documented human infection with non-rabies lyssaviruses, the protection afforded by current vaccines is an area of considerable interest. There have been a number of studies that have attempted to quantify the level of protection afforded by standard vaccines to divergent lyssavirus species and these studies form the basis of phylogroup classification of the different lyssavirus isolates. It is generally understood that the current rabies vaccines are unable to neutralise Lagos Bat Virus (LBV) and Mokola Virus (MOKV) and that there is no neutralisation of WBCV by sera raised against viruses of phylogroups I and II.
Regarding cross phylogroup neutralisation, little is currently known. MOKV (phylogroup II) has been shown to share at least one common epitope with the phylogroup I lyssaviruses RABV Challenge Virus Strain-11 (CVS-11) and Duvenhage Virus (DUVV) as all of these isolates were neutralised by mAb 1049-7 (Dietzschold et al. 1988). However, further characterisation of the antigenicity of different regions of G and their role in the induction of a neutralising response have not been defined and form the core of this project.
Reverse genetics techniques have also been used to investigate the ability of chimeric G proteins to function. In this regard, chimeras of MOKV and RABV G were developed and functionality assessed (Mebatsion et al. 1995). Interestingly, although no sequence homology was found between signal sequences, transmembrane or cytoplasmic domains between these two lyssaviruses, when G protein ectodomains were swapped between these isolates the chimera was able to function in all elements of the viral life cycle. It was also shown that the cytoplasmic domain of the chimera could interact heterotypically with the remaining RABV proteins showing that strict conservation of the whole tail sequence is not essential for formation of infectious lyssavirus particles. This shows promise for use in future strategies. Perhaps if a chimeric G were generated that incorporated the neutralising antibody inducing antigenic features of all the phylogroups it could be rescued into a full length clone and potentially used as a cross protective vaccine.
Whilst the true threat of non-rabies lyssaviruses on human life is unknown, with the continued discovery of antigenically divergent lyssavirus species there is a need to define the limits of protection afforded by current vaccines.