Robert E. Johnston

The nucleotide sequence of the coding region of tobacco etch virus genomic RNA: Evidence for the synthesis of a single polyprotein☆

The complete nucleotide sequence of the tobacco etch virus (TEV) RNA genome has been determined excepting only the nucleotide(s) present at the extreme 5 terminus. The assembled TEV genomic sequence is 9496 nucleotides in length followed by a polyadenylated tract ranging from 20 to 140 residues. A computer search of the sequence reveals the following. A 5 untranslated region, rich in adenosine and uridine, is present between nucleotides 1 and 144. A putative initiation codon, at nucleotides 145-147, marks the beginning of a large open-reading frame (ORF) which ends with an opal (UGA) termination codon at positions 9307-9309. A 186-nucleotide untranslated region is present between the termination codon of the ORF and the beginning of the 3 polyadenylated region. The predicted translation product of this ORF is a 3054 amino acid polyprotein with a mol wt of 345,943. A function for the large (54,000 Mr) nuclear inclusion protein is suggested by a comparison of the deduced amino acid sequence with a protein data bank. This protein displays biochemical similarities to other viral RNA-dependent, RNA polymerases.

In vitro synthesis of infectious venezuelan equine encephalitis virus RNA from a cDNA clone: Analysis of a viable deletion mutant☆

A molecular clone of Venezuelan equine encephalitis virus (VEE) was constructed from four cDNAs that were synthesized using the viral RNA genome as template. Together, these cDNAs are believed to represent all but the nine 5-terminal nucleotides of the VEE genome sequence. A T7 promoter, followed by a single intervening G residue, and the exact 5-terminus of VEE were added to the 5-most clone using in vitro mutagenesis. Appropriate restriction fragments isolated from the cloned cDNAs were joined to form a candidate full-length VEE cDNA clone. RNA transcripts synthesized in vitro from the cDNA clone were able to initiate a productive infection in DEAE-dextran-treated chicken embryo fibroblasts (CEF). VEE antigens were demonstrated in RNA-transfected cells, and supernatants from transfected cultures contained infectious virus particles. The candidate full-length cDNA clone lacked 102 nucleotides of the VEE genome sequence. The deletion, which also was present in the genomes of progeny virions derived from the clone, did not appear to affect growth in cultured CEF, baby hamster kidney cells, or Vero cells. The site of the deletion was mapped to the 3-end of the nsP3 gene by comparison to other alphavirus sequences. In this region, the VEE genome sequence includes two tandem 102-nucleotide repeats which can be arranged in a stable stem and loop structure. The sequence remaining in the deleted clone retains one copy of the duplicated sequence and, in addition, faithfully preserves a portion of the predicted stem.

Biochemical analysis of the capsid protein gene and capsid protein of tobacco etch virus: N-terminal amino acids are located on the virion's surface

The sequence of the 1491 nucleotides found at the 3 end of the genome of the highly aphid-transmissible (HAT) isolate of tobacco etch virus (TEV) has been determined. The nucleotide sequence of the capsid protein gene has been identified and compared with the corresponding region of the not-aphid-transmissible (NAT) isolate of TEV and with pepper mottle virus (PeMV). The deduced amino acid sequences of the two TEV capsid proteins displayed 98% homology and a 66% homology with PeMV capsid protein. Three of the six amino acid differences between the capsid proteins of the two TEV isolates occurred near the N terminus of the protein. Biochemical and immunological evidence suggested the N-terminal 29 amino acids of the capsid protein were hydrophilic and were located at or near the virions surface.

Topographic analysis of tobacco etch virus capsid protein epitopes

Monoclonal antibodies have been prepared, which react with capsid protein of an aphid-transmitted isolate of tobacco etch virus (TEV). Ten different monoclonal antibodies were characterized with reference to (1) antibody class, (2) reactivity with different plant virus antigens, (3) the spatial relationship between epitopes, and (4) whether these epitopes were located on the exterior surface of the virion. Three monoclonal antibodies were specific for TEV isolates. These monoclonal antibodies reacted with epitopes exposed on the external surface of the TEV particle. Seven monoclonal antibodies reacted with a variety of different potyviruses including TEV, potato virus Y, tobacco vein mottling virus, pepper mottle virus, watermelon mosaic virus II, and maize dwarf mosaic virus. In general, these seven monoclonal antibodies defined epitopes not readily accessible on the virion surface.

Characterization of Sindbis virus epitopes important for penetration in cell culture and pathogenesis in animals

Two anti-glycoprotein E2 monoclonal (MC) antibodies, designated R6 and R13, prepared against an attenuated mutant (SB-RL) of Sindbis virus, preferentially neutralize attenuated, rapidly penetrating strains of Sindbis as opposed to virulent, more slowly penetrating strains. An antigenic variant of SB-RL which displayed reduced reactivity with R6 and R13, demonstrated virulence in suckling mice and slow penetration in baby hamster kidney cells. Anti-E2 MC antibodies to SB-RL and to an independently propagated wild-type Sindbis strain, SIN, were used to compare the topographical relationship of the R6 and R13 epitopes to other E2 antigenic sites. The R6 and R13 epitopes constituted a previously undescribed E2 site, E2-c. Antibody competition experiments using virions in solid phase and in suspension, as well as binding of MC antibody to antigenic variants, demonstrated that E2-c was a spatially distinct and independently mutable site at the surface of Sindbis virions. Of three E2 antigenic sites, E2-a and E2-c were conserved among the Sindbis strains examined, while E2-b was strain specific. Although attenuated strains were preferentially neutralized by E2-c specific MC antibodies, the critical mutation in these strains did not alter their ability to bind E2-c MC antibody. Rather, the mutation was responsible for the altered biological effect of antibody binding at E2-c; the in vitro rapid penetration phenotype; and the in vivo phenotype of attenuation in suckling mice.

Selection for accelerated penetration in cell culture coselects for attenuated mutants of Venezuelan equine encephalitis virus.

Previous studies with Sindbis virus (SB) suggested that a single point mutation in glycoprotein E2 (serine 114 to arginine 114) conferred three phenotypic alterations: attenuation in neonatal mice, accelerated penetration of cultured cells, and efficient neutralization by two E2-specific monoclonal antibodies (Davis, Fuller, Dougherty, Olmsted, and Johnston (1986) Proc. Natl. Acad. Sci. USA 83, 6771-6775). Moreover, selection for rapidly penetrating mutants of SB coselected for attenuation in vivo, indicating that a domain of SB E2 which influences penetration in culture overlaps an E2 domain which influences pathogenesis (Olmsted, Meyer, and Johnston (1986) Virology 148, 245-254). To test the possibility that overlapping penetration and pathogenesis domains exist in other alphaviruses, the virulent Trinidad donkey strain of Venezuelan equine encephalitis virus (TRD-VEE) was serially passed in baby hamster kidney (BHK) cells under a stringent selective pressure for accelerated penetration. Isolates were biologically cloned from the first through the fourth passages and were characterized as to penetration time course in BHK cells and virulence in adult mice following intraperitoneal inoculation. Twenty-two of the 27 isolates segregated into two major categories: slowly penetrating and virulent (like the TRD-VEE parent) and rapidly penetrating and avirulent. Mice which received the avirulent mutants were positive for anti-VEE neutralizing antibody and were refractory to challenge with TRD-VEE. Of the seven mouse avirulent mutants, two also were attenuated in hamsters, indicating the presence of at least two genetic loci at which mutations may influence both pathogenesis and penetration.

Alternative forms of a strain-specific neutralizing antigenic site on the Sindbis virus E2 glycoprotein

Experiments with monoclonal antibodies raised against two laboratory strains of Sindbis virus, SB and SIN, suggested the existence of a strain-specific neutralizing antigenic site (E2-b) on the E2 glycoprotein. A comparison of monoclonal antibody binding patterns and E2 glycoprotein gene sequences of six laboratory strains distinguished three different configurations of E2-b that correlated with specific amino acid substitutions at position 216 of the E2 glycoprotein. Further study of neutralization escape mutants selected with E2-b-specific antibodies confirmed that amino acid 216 is a major determinant of the E2-b antigenic site. Eight of nine mutants showed a coding change at position 216. One neutralization escape mutation created a new glycosylation site at position 213 and resulted in an E2 protein with an altered migration rate in SDS-PAGE. The neutralization escape mutants studied included amino acid substitutions not found in the laboratory strains that revealed differing binding requirements for two E2-b-specific monoclonal antibodies. The E2-b site is contrasted with the E2-c neutralizing antigenic site described previously (R.A. Olmsted, W.J. Meyer, and R.E. Johnston, 1986, Virology 148, 245-254).

Generation and characterization of monoclonal antibodies reactive with the 49-kDa proteinase of tobacco etch virus.

Monoclonal antibodies (McAbs) were generated against two tobacco etch virus (TEV)-encoded nonstructural proteins, the 49-kilodalton (kDa) proteinase and the 58-kDa putative RNA-dependent RNA polymerase. This process was facilitated by the fact that these two TEV nonstructural proteins cocrystallize in the nuclei of virus-infected cells to form nuclear inclusion (NI) bodies which can be purified readily. The anti-NI McAbs were shown by Western blot analysis to be specific for either the TEV 49-kDa or the 58-kDa protein. Those McAbs reactive with the 49-kDa proteinase were characterized further with respect to the 49-kDa domain with which they reacted and with respect to their ability to inhibit the autocatalytic or self-processing activity of the 49-kDa proteinase. The 49-kDa antigens were synthesized from a TEV cDNA sequence using cell-free transcription and translation systems. Each anti-49-kDa McAb was used in immunoprecipitation studies with a series of 49-kDa antigens which represented a nested set of 49-kDa proteins with common amino termini but varying in length. Immunoprecipitation results showed that all of the anti-49-kDa proteinase McAbs reacted with one of five binding regions, designated A through E from the carboxy terminus of the proteinase, which were 77, 38, 81, 18, and 61 amino acids long, respectively. The 38-amino-acid binding region B contained the proposed catalytic cysteine 339 residue and was recognized by only one McAb, 4911. McAb 4911 was the only anti-49-kDa McAb capable of inhibiting the self-processing reaction in which the 49-kDa proteinase is released from its 75-kDa polyprotein precursor.

A sindbis virus variant with a cell-determined latent period☆

Abstract A rapidly growing variant of Sindbis virus was isolated by repeated selection for virions released from cells at 5.5 hr postinfection. The variant, SB-RL, was cloned from the eleventh 5.5-hr passage. SB-RL was characterized by a significant reduction in latent period and an increased rate of penetration. These characteristics were expressed only in the baby hamster kidney cell line used for the original selection. In other cell lines, the latent period and rate of penetration were identical to wild-type Sindbis. SB-RL also produced higher virus titers and exhibited an increased rate of RNA synthesis. However, the expression of these characteristics was independent of cell type. SB-RL could be distinguished from the parental Sindbis strain on the basis of neutralization of infectivity, suggesting an alteration in the E2 glycoprotein of SB-RL.

Differences in virion stability among Sindbis virus pathogenesis mutants

The structure of closely related Sindbis virus strains, which differ in their virulence for neonatal mice, has been probed by measuring the sensitivity of virions to heat, varying concentrations of dithiothreitol (DTT), phenylglyoxal or low pH. Attenuated mutants (SB-RL and SB-FP) were much more sensitive to loss of infectivity after heat or DTT treatment than either the prototype virulent strain (SB) or same-site virulent revertants of SB-RL. Incubation of SB-RL virions in the presence of DTT increased their density and exposed sulfhydryl groups on both E1 and E2 glycoproteins as determined by [14C]iodoacetamide labeling. Incubation of SB-RL at pH 6.1 for 1 h at 37 degrees C resulted in a 50% decrease in titer, whereas an equivalent decrease in SB titer was obtained at pH 6.0. Attenuated (SB-RL) and virulent (SB) strains could not be distinguished on the basis of phenylglyoxal sensitivity. Comparison of the nucleotide sequences of these strains and analysis of strains derived from recombinant full-length cDNA clones demonstrated that reduced virion stability under these conditions was a consequence of an arginine for serine substitution at position 114 of the surface glycoprotein E2. This same amino acid substitution is responsible for attenuation in neonatal mice. However, experiments with a second-site virulent revertant of SB-RL, which retained E2 arginine 114 and the reduced stability phenotype, indicated that virion instability in itself was not directly responsible for reduced virulence in vivo.