Charles M. Rice

Association of Sindbis Virion Glycoproteins and their Precursors

We have studied the association of the Sindbis virus glycoproteins in mature virions and infected cells. The glycoproteins were cross-linked with bifunctional amino-reactive reagents (11 A cross-linking distance), some of which could be subsequently cleaved by reduction. Using monospecific rabbit antisera against each viral envelope glycoprotein it was found that >90% of the cross-linked glycoprotein dimers from intact virions or virions solubilized with Triton X100 prior to cross-linking were heterodimers of E1 and E2. The pattern of cross-linked glycoproteins from intact virions as well as infected cells suggested that three E1-E2 dimers may be associated to form a hexameric subunit. Cross-linking of pulselabeled infected monolayers showed that PE2 was cross-linked to E1 less efficiently than was E2, suggesting that if PE2 and E1 are associated as a complex in infected cells then their conformation with respect to reactive amino groups is distinct from that of the mature virion glycoproteins. ts mutants of Sindbis virus in the complementation groups corresponding to E1 and PE2 fail to cleave PE2 at the non-permissive temperature (40 °C). In monolayers infected with these mutants or a heat-resistant variant of Sindbis virus, the glycoprotein precursors synthesized at the elevated temperature were readily cross-linked into large aggregates, indicating a temperature-sensitive tendency for the proteins to aggregate.

Isolation and Characterization of the Hydrophobic COOH-terminal Domains of the Sindbis Virion Glycoproteins

Digestion of intact Sindbis virions with α-chymotrypsin produced a single membrane-associated peptide derived from each of the two virion glycoproteins (referred to as RE1 and RE2, or roots derived from El and E2, respectively). Amino acid composition data and NH_2-terminal sequence analysis established their location at the extreme COOR-terminal end of each glycoprotein. REI and RE2 are rich in hydrophobic amino acids and insoluble in aqueous solutions in the absence of detergents, and show differential solubility in organic solvent systems designed for the extraction of lipids. Essentially all of the covalently attached palmitic acid associated with El and E2 was found to be clustered in their hydrophobic. membrane-associated roots. Beginning six to seven residues from their NH2 termini, RE1 and RE2 contain uninterrupted sequences of hydrophobic amino acids similar in terms of amino acid composition and length to the transmembrane anchors found in other bitopic integral membrane proteins. By comparing the sequence and composition data obtained here with the sequences of E1 and E2 deduced from complementary DNA sequence analysis (Rice & Strauss, 1981) we can make several observations. First, following their uncharged, putative intramembrane segments (33 and 26 amino acids, respectively), El and E2 contain clusters of predominantly basic amino acids. By structural analogy to known transmembrane proteins, El probably spans the bilayer but contains only a few residues exposed on the inner face of the virion envelope. In contrast, E2 and PE2 (the precursor to E2), which have been shown to span the bilayer completely, contain an additional 33 COOR-terminal residues, which could be either exposed on the cytoplasmic face of the lipid bilayer or which could loop back into the membrane. This region at the extreme COOR-terminal end of E2, which is protected by the virion envelope from digestion by α-chymotrypsin, contains a second uncharged domain (23 amino acids in length) whose orientation is unknown, but which may be involved in the highly specific interaction of the transmembrane glycoproteins in the plasma membrane with the cytoplasmic nucleocapsid during budding.

Partial N-terminal amino acid sequences of three nonstructural proteins of two flaviviruses

Partial N-terminal amino acid sequences for the three largest nonstructural proteins of two flaviviruses, yellow fever virus and St. Louis encephalitis virus, have been obtained. The determined sequences of these proteins exhibit significant amino acid sequence homology, and allow the positioning of these three nonstructural proteins in the polyprotein sequence deduced from the nucleotide sequence of yellow fever virus (C. M. Rice, E. M. Lenches, S. R. Eddy, S. J. Shin, R. L. Sheets, and J. H. Strauss, 1985, Science 229, 726-733.) The deduced start points support the hypothesis that the N terminus of nonstructural glycoprotein NS1 results from cleavage by signalase, whereas the N termini of NS3 and NS5 result from cleavages following double basic residues that are flanked by amino acids with short side chains.

Synthesis, Cleavage and Sequence Analysis of DNA Complementary to the 26 S Messenger RNA of Sindbis Virus

Conditions for synthesis of long complementary DNA transcripts of Sindbis virus 26 S and 49 S RNA in high yield have been developed. This single-stranded complementary DNA could be cut with HaeIII, HhaI, RsaI or TaqI to give reproducible patterns of discrete, virus-specific fragments that were suitable for subsequent end-labeling and direct sequence analysis. Using these methods we present the strategy used for obtaining nearly the entire 26 S RNA sequence from complementary DNA synthesized in vitro. This approach should prove useful for sequence analysis of any purified RNA available in microgram quantities.

Biochemical studies of the maturation of the small sindbis virus glycoprotein E3

A small glycoprotein (E3) was purified from the culture fluid of Sindbis virus-infected primary chick embryo fibroblasts. Tryptic peptide mapping and pulse-chase studies verified that this protein was produced as a by-product of the cleavage of the precursor protein PE2 to produce the envelope glycoprotein E2. A 2600-fold purification was achieved via a procedure which used differential ethanol precipitation, gel filtration, ion-exchange chromatography, and affinity chromatography on a lentil lectin column. Amino acid composition analysis, N-terminal microsequencing, and labeling studies yielded information about the fine structure of E3 and its relationship to E2 and virion maturation. The N-terminal sequence of E3 is identical to that of PE2, including the result that 90% of the molecules appear to be blocked. The first 19 amino acids are uncharged and presumably serve as the signal sequence for the insertion of PE2 into the membrane of the endoplasmic reticulum, but this sequence is unusual in that it is not immediately cleaved from PE2 and is glycosylated at the asparagine at position 14. The two residues at the C-terminus of E3, Lys-Arg, are removed during or shortly after cleavage from PE2. Labeling studies imply that, although the PE2----E2 + E3 cleavage is necessary for virion budding, these two events are not closely coupled. E3 is cleaved and released into the culture fluid under conditions where virions do not bud, and the kinetics of the appearance of E3 in the culture fluid and E2 in virions are quite dissimilar. The maturation of E3 is discussed as it relates to the processing of cellular membrane and secretory glycoproteins.

The N-terminus of PE2 in Sindbis virus-infected cells

One of the two envelope glycoproteins of Sindbis virus, E2, and its intracellular precursor PE2 were obtained in chemically pure form from Sindbis infected cells by immunoabsorption with antibody against E2. The N-terminus of PE2, which is probably the signal sequence of this protein, was examined and two N-termini were found. One, the primary product of translation, was sequenced by Edman degradation and the other was found to be blocked and produced by N-acetylation of the first. It was also found that an asparagine residue at position 14 of PE2 is glycosylated. The putative signal sequence is not cleaved by signal peptidase either under normal conditions or when glycosylation is prevented with tunicamycin.