Structure-Function Relationships in the Structural Proteins and in the RNAs of Alphaviruses and Flaviviruses
<p>The RNA virus families <i>Togaviridae</i> and <i>Flaviviridae</i> were considered one family as recently as 1983. These two families contain more than 100 members, many of which are important pathogens for humans and domestic animals. Studies on members of both famil...
| Summary: | <p>The RNA virus families <i>Togaviridae</i> and <i>Flaviviridae</i> were considered one family as recently as 1983. These two families contain more than 100 members, many of which are important pathogens for humans and domestic animals. Studies on members of both families which were undertaken to increase our understanding of the functions of the virus structural proteins and of RNA sequence elements that interact with virus proteins, and of the evolution of these two families of RNA viruses, are presented in this thesis. These investigations include two on the nature and function of a virus-encoded self-protease that functions in the processing of the structural proteins, several studies on the role of the virus structural glycoproteins in assembly of progeny virions and viral virulence, and studies on the evolution of these viruses, including the demonstration that recombination has occurred in the <i>Togaviridae</i> to produce an important new pathogen, and that RNA sequence elements have been conserved during the evolution of the <i>Flaviviridae</i>.</p>
<p>Alphavirus structural proteins are translated from a subgenomic messenger RNA as a polyprotein, which is cleaved to the final products by proteolytic processing. This processing was studied by comparative sequence analysis of three temperature sensitive mutants of Sindbis virus (the type alphavirus) which have a defect in processing of the polyprotein at the nonpermissive temperature. These mutations were localized in the C-terminal region of the capsid protein. From the position of these mutations and from sequence similarities between the alphavirus capsid proteins and animal serine proteses, we hypothesized that the capsid protein was a serine autoprotease whose active site is formed by His-141, Asp- 147 and Ser-215. To study this capsid protein protease activity in more detail, we have altered the proposed catalytic triad of the protease by site-directed mutagenesis. We have assayed the protease activity in the mutagenized capsid proteins by <i>in vitro</i> transcription and translation, and attempted to rescue virus from mutagenized full-length "infectious" clones. The results supported our hypothesis.</p>
<p>Sindbis virus matures when preformed nucleocapsids acquire their envelopes by budding through virus-modified areas of the cell surface membrane. ts103 is a mutant of Sindbis virus which has a defect in this late maturation step such that it generates multicored particles, and it has provided a good system for studying structure-function relationships during viral assembly and maturation. Hybrid genomes were constructed that were formed from a full-length cDNA clone of wild type Sindbis in which restriction fragments were replaced with cDNA from ts103. Virus rescued from these constructs were used to determine the protein responsible for the multicored phenotype and to map the mutation. tsl03 was found to have a single amino acid substitution in glycoprotein E2. The implications of this mutation for our understanding of virus assembly are
discussed.</p>
<p>Virus surface glycoproteins are believed to be important determinants of virulence and tissue tropism. Ne urovirulence of Sindbis virus for mice has been used as an animal model system in which to explore the effects of each individual protein or of a particular domain of a given protein, or even of a single amino acid residue, on neurovirulence. By constructing hybrid genomes among various strains of Sindbis virus at the cDNA level and rescuing virus <i>in vitro</i> using <i>in vitro</i> transcription and transfection, it was possible to evaluate the effect of each protein on neurovirulence. From these studies, we concluded that Sindbis virus glycoproteins are important determinants of neurovirulence, but not the sole determinants. The virulence phenotypes of various recombinant viruses in both weanling mice and suckling mice are discussed, with reference to the role of particular residues in producing the neurovirulent phenotype.</p>
<p>We have also undertaken a study of virulence in flaviviruses. The 17D vaccine strain of yellow fever virus, the type flavivirus, is one of the most reliable and stable live virus vaccines ever developed. By comparison of the nucleotide sequences of the 17D vaccine strain and its parental virulent Asibi virus we have located all of the changes which occurred during the attentuation of yellow fever to produce the 17D vaccine. This comparison led us to the conclusion that changes in the viral envelope protein play an important role in attenuation.</p>
<p>The 25 members of the genus <i>Alphavirus</i> have for the most part diverged by linear descent from a common ancestor. We have now found that Western equine encephalitis virus is an exception to this. Western equine encephalitis virus is a close relative of Sindbis virus as determined by immunological cross-reaction, but it is a New World virus that causes encephalitis in humans and horses, whereas Sindbis virus is an Old World virus not normally associated with encephalitis. The nucleotide sequence and deduced amino acid sequence of the structural proteins of Western equine encephalitis virus reveal that it arose by recombination between Eastern equine encephalitis virus (or a recent ancestor of it) and a virus closely related to Sindbis virus. The importance of recombination in the evolution of RNA viruses and in the generation of new potentially pathogenic virus strains, as well as the implications of the amino acid changes which have occurred in Western equine encephalitis virus (subsequent to the initial recombination event) for our understanding of the interaction between the structural proteins of alphaviruses, are discussed.</p>
<p>To study evolution in flaviviruses, sequences at the 5'and 3' ends of several flaviviruses have been compared. Conserved structures or sequence elements have been identified, one pair of which could result in cyclization of flavivirus RNA. The significance of these sequences is discussed.</p> |
|---|