Flavivirus Envelope Protein Research Articles

The Fusion Glycoprotein Shell of Semliki Forest Virus: An Icosahedral Assembly Primed for Fusogenic Activation at Endosomal pH

Semliki Forest virus (SFV) has been extensively studied as a model for analyzing entry of enveloped viruses into target cells. Here we describe the trace of the polypeptide chain of the SFV fusion glycoprotein, E1, derived from an electron density map at 3.5 Å resolution and describe its interactions at the surface of the virus. E1 is unexpectedly similar to the flavivirus envelope protein, with three structural domains disposed in the same primary sequence arrangement. These results introduce a new class of membrane fusion proteins which display lateral interactions to induce the necessary curvature and direct budding of closed particles. The resulting surface protein lattice is primed to cause membrane fusion when exposed to the acidic environment of the endosome.

Evaluation of monoclonal antibodies for generic detection of flaviviruses by ELISA.

Three monoclonal antibodies (Mabs) specific for the envelope (E) protein of flaviviruses were evaluated for use in an antigen capture ELISA. Three combinations of Mabs and a combination of polyclonal antibodies (Pabs) were evaluated in antigen capture ELISAs for their ability to detect 18 flaviviruses. The Mab ELISAs detected 50% of flavivirus antigens with a sensitivity between 1 and 9 x 10(4)/ng viral protein/ml, however, none of the ELISAs evaluated proved to be useful for generic detection of flaviviruses, being unable to detect tick-borne flaviviruses and some mosquito-borne flaviviruses. The inability of the ELISAs to detect tick-borne flaviviruses is thought to be due to the conformation of surface epitopes, which the Mabs were unable to recognise. This was again observed using recombinant TBE virus prM/E protein as antigen in direct and antigen capture ELISAs. The Mabs reacted with the prM/E protein when it was denatured by binding directly onto the solid phase, but the antibodies were unable to detect the native protein in antigen capture ELISAs. The antigen capture ELISAs evaluated in this study were considered to be unsuitable for the generic detection of flaviviruses, but may provide a sensitive diagnostic assay for specific flavivirus infection.

Analysis of flavivirus envelope proteins reveals variable domains that reflect their antigenicity and may determine their pathogenesis

Studies on the molecular basis of flavivirus neutralisation, attenuation and tropism indicate that amino acid substitutions, in different parts of the envelope gene, may be responsible for the altered phenotypes. However, the association of particular substitutions with individual characteristics has proven difficult. Comparative analysis of all known tick-borne flavivirus envelope proteins through sequence alignment and a sliding window, reveals clusters of amino acid variation distributed throughout the envelope protein coding region. Further comparison with mosquito-borne flaviviruses reveals essentially the same profile of variability throughout the envelope protein sequence although there is a major difference within the postulated B domain of these viruses which may reflect their different evolutionary development. Most phenotypically variant properties, such as serotypic differences, variants characteristic of vaccine strains, altered tropisms and neutralisation escape mutants, map within the variable clusters. Thus, we propose that natural mutagenesis and selection may occur at specific sites that do not destroy the secondary and tertiary E protein structure and that the variable clusters represent the exposed surface amino acids of the envelope protein defining antigenicity, tropicity and pathogenesis.

Oligomeric rearrangement of tick-borne encephalitis virus envelope proteins induced by an acidic pH.

The flavivirus envelope protein E undergoes irreversible conformational changes at a mildly acidic pH which are believed to be necessary for membrane fusion in endosomes. In this study we used a combination of chemical cross-linking and sedimentation analysis to show that the envelope proteins of the flavivirus tick-borne encephalitis virus also change their oligomeric structure when exposed to a mildly acidic environment. Under neutral or slightly alkaline conditions, protein E on the surface of native virions exists as a homodimer which can be isolated by solubilization with the nonionic detergent Triton X-100. Solubilization with the same detergent after pretreatment at an acidic pH, however, yielded homotrimers rather than homodimers, suggesting that exposure to an acidic pH had induced a simultaneous weakening of dimeric contacts and a strengthening of trimeric ones. The pH threshold for the dimer-to-trimer transition was found to be 6.5. Because the pH dependence of this transition parallels that of previously observed changes in the conformation and hydrophobicity of protein E and that of virus-induced membrane fusion, it appears likely that the mechanism of fusion with endosomal membranes involves a specific rearrangement of the proteins in the viral envelope. Immature virions in which protein E is associated with the uncleaved precursor (prM) of the membrane protein M did not undergo a low-pH-induced rearrangement. This is consistent with a protective role of protein prM for protein E during intracellular transport of immature virions through acidic compartments of the trans-Golgi network.

Epitope analysis of tick-borne encephalitis (TBE) complex viruses using monoclonal antibodies to envelope glycoprotein of TBE virus (persulcatus subtype)

The arrangement of envelope protein epitopes of tick-borne encephalitis viruses (TBEV) (persulcatus or eastern subtype, Sofjin strain and ricinus or western subtype, Minsk-256 strain) and Kyasanur Forest disease virus (KFDV) was investigated using competitive binding of monoclonal antibodies against the Sofjin E protein. The E protein of TBEV Sofjin strain forms three antigenic domains: E1, E2 and E3, represented by 12, 9 and 2 epitopes respectively; two additional epitopes stand alone. Domains E1 and E2 are heterogeneous. On the epitope map of the Minsk-256 strain domain E3 remains intact, domains E1 and E2 overlap and the relative arrangement of virus-neutralizing epitopes from E1 and E2 domains is changed. The epitope map of KFDV is significantly dissimilar to TBEV. The viruses can be distinguished by epitopes with identical serological reactivity. A satisfactory agreement between our epitope maps and previously published antigenic models of flavivirus envelope protein (Guirakhoo et al., 1989; Mandl et al., 1989a) was observed. The main difference of our map is that domains corresponding to domains B and C (Sofjin strain) and A, B and C (Minsk-256 strain) in Heinz's model are overlapping. The results of competition analysis depend on the nature of the antigen (virion or purified protein) and the immunoassay technique.

Envelope protein sequences of dengue virus isolates TH-36 and TH-Sman, and identification of a type-specific genetic marker for dengue and tick-borne flaviviruses.

Complementary DNAs were synthesized from the envelope protein genes of two isolates of dengue virus (TH-36 and TH-Sman, previously suggested as possible dengue virus type 5 and dengue virus type 6 respectively) and amplified by the polymerase chain reaction using sense and antisense primers designed from conserved dengue virus gene sequences. The amplified cDNA clones were sequenced in both directions by double-stranded dideoxynucleotide sequencing. Alignment with published dengue virus sequences enabled us to assign these viruses accurately to classified serotypes, confirming that TH-36 and TH-Sman are strains of dengue virus type 2 and dengue virus type 1 respectively. Amino acid changes between the proteins encoded by these two isolates and strains of their respective serotypes may account for the significant antigenic differences observed during previous serological typing of these viruses. Moreover, sequence alignment of flavivirus envelope proteins revealed a hypervariable region, within which members of the dengue and tick-borne virus antigenic complexes show unique peptide sequences. This type-specific hypervariable domain may be useful as a genetic marker for typing dengue and tick-borne flaviviruses.

The flavivirus envelope protein E: isolation of a soluble form from tick-borne encephalitis virus and its crystallization

By the use of limited trypsin digestion of purified virions, we generated a membrane anchor-free and crystallizable form of the tick-borne encephalitis virus envelope glycoprotein E. It retained its reactivity with a panel of monoclonal antibodies, and only subtle structural differences from the native protein E were recognized. Treatment with the bifunctional cross-linker dimethylsuberimidate resulted in the formation of a dimer. Crystallization experiments yielded hexagonal rod-shaped crystals suitable for X-ray diffraction analysis.

Variable and hypervariable domains are found in the regions of HCV corresponding to the flavivirus envelope and NS1 proteins and the pestivirus envelope glycoproteins

Based on the flavi- and pestivirus model of genome organization for the hepatitis C virus (HCV) (1–5), the nucleotide and deduced amino acid sequences of the putative envelope (E1) and the junction between the El and NS1 /envelope 2 (E2) region from six different human isolates of HCV were compared with the nucleotide and predicted amino acid sequences of the prototype hepatitis C virus (HCV-1) ( 5). The overall percentage of nucleotide and amino acid changes among all six isolates, including HCV-1, from nucleotide 713 to 1630 (amino acid 129 to 437) was between 3 and 7%, which is comparable to that seen in some flaviviruses (6–8). An analysis of the number of nucleotide and deduced amino acid sequence changes among all six isolates and HCV-1 revealed a moderately variable domain of approximately 40 amino acids in the E1 region and a hypervariable domain (Region V) of approximately 28 amino acids, which is directly downstream from a putative signal peptide sequence, in the junction between E1 and NS1/E2. A similar hypervariable domain is not found in the C-terminus of the envelope polypeptide or in the N-terminus of the NS1 polypeptide domain of the flaviviruses. These findings suggest that the mature NS1 /E2 polypeptide starts about amino acid 380 and that the NS1/E2 domain may correspond to a second envelope glycoprotein as in the case of the pestivirus. The observed heterogeneity in the putative structural proteins of HCV may have important ramifications for future vaccine development.

An Analysis of the Antibody Response against West Nile Virus E Protein Purified by SDS-PAGE Indicates that This Protein Does Not Contain Sequential Epitopes for Efficient Induction of Neutralizing Antibodies

The large envelope (E) protein of flaviruses is the viral surface protein that contains neutralizing epitopes. We have analysed the E protein of the West Nile (WN) flavivirus for neutralizing epitopes generated from linear segments of the protein; if effective, these might allow the synthesis of peptides suitable for vaccination. For this study we used the E protein and defined fragments thereof as antigens in rabbits. The sera thus obtained, containing antibodies to E protein as shown by Western blot analyses, were tested for neutralizing activity by the plaque reduction neutralization test. If the E protein used as antigen was reduced (the native E protein contains six disulphide bridges) and denatured the resulting antibodies did not consistently demonstrate neutralizing activity. These results show that the E protein of WN virus does not contain a linear segment that is able to induce neutralizing antibodies efficiently. Studies using denatured E protein fragments containing subsets of the intact disulphide bridges showed that the local covalent primary structure of the protein involved in each of the six bridges was also insufficient for inducing the synthesis of neutralizing antibodies. The complete E protein, with all six disulphide bridges intact, purified by preparative SDS-PAGE under the conditions used in our experiments could, however, induce the synthesis of neutralizing antibodies in rabbits. These data indicate that at least one complex epitope which is able to induce neutralizing antibodies is not completely denatured or can be reformed to some extent if the complete E protein has been subjected to SDS-PAGE without prior destruction of the disulphide bridges.

Antigenic structure of the flavivirus envelope protein E at the molecular level, using tick-borne encephalitis virus as a model

A model of the tick-borne encephalitis virus envelope protein E is presented that contains information on the structural organization of this flavivirus protein and correlates epitopes and antigenic domains to defined sequence elements. It thus reveals details of the structural and functional characteristics of the corresponding protein domains. The localization of three antigenic domains (composed of 16 distinct epitopes) within the primary structure was performed by (i) amino-terminal sequencing of three immunoreactive fragments of protein E and (ii) sequencing the protein E-coding regions of seven antigenic variants of tick-borne encephalitis virus that had been selected in the presence of neutralizing monoclonal antibodies directed against the E protein. Further information about variable and conserved regions was obtained by a comparative computer analysis of flavivirus E protein amino acid sequences. The search for potential T-cell determinants revealed at least one sequence compatible with an amphipathic alpha-helix which is conserved in all flaviviruses sequenced so far. By combining these data with those on the location of disulfide bridges (T. Nowak and G. Wengler, Virology 156:127-137, 1987) and the structural characteristics of epitopes, such as dependency on conformation or on intact disulfide bridges or both, a model was established that goes beyond the location of epitopes in the primary sequence and reveals features of the folding of the polypeptide chain, including the generation of discontinuous protein domains.