Viral invasion and host defense: strategies and counter-strategies.

The outcome of infection of plants by viruses is determined by the net effects of compatibility functions and defense responses. Recent advances reveal that viruses have the capacity to modulate host compatibility and defense functions by a variety of mechanisms.

Structure and assembly of turnip crinkle virus. IV. Analysis of the coat protein gene and implications of the subunit primary structure.

The structure of the turnip crinkle virus (TCV) coat protein and coat protein gene has been examined by cDNA cloning, nucleotide sequencing and high-resolution mRNA mapping. We have cloned a 1450-nucleotide cDNA fragment, representing the 3' end of the TCV genome, using genomic RNA polyadenylated in vitro as the reverse transcriptional template. Nucleic acid sequence analysis reveals the presence of a 1053 nucleotide open reading frame capable of encoding a protein of 38,131 Mr, identified as the coat protein subunit. The 1446 base subgenomic mRNA for the coat protein, mapped using high-resolution primer extension techniques, contains a 137 nucleotide leader sequence upstream from the initiation codon. We have characterized a second subgenomic RNA of approximately 1700 bases, roughly 250 nucleotides longer than the 1446 base species in the 5' direction. No TCV-related RNAs are polyadenylated in vivo. The derived amino acid sequence of the TCV coat protein has been built into the 3.2 A resolution electron density map of TCV reported in paper I of this series. We describe here some of the important features of the structure. Alignment of the three-dimensional structures of tomato bushy stunt virus and southern bean mosaic virus shows significant sequence relationships in the arms and S domains, although the conserved residues do not appear to have any special role in stabilizing the beta-barrel fold or in mediating subunit interactions. The sequences of TCV and carnation mottle virus can be aligned. Comparisons among the four are discussed in terms of the organization of the S domain.

Identification of essential residues in potyvirus proteinase HC-Pro by site-directed mutagenesis.

Two virus-encoded proteinases are responsible for proteolysis of potyvirus polyproteins. One of these, HC-Pro, is a multifunctional protein that autolytically cleaves at its carboxyl-terminus (J.C. Carrington et al., 1989, EMBO J. 8, 365-370). To identify the class of proteinase to which HC-Pro belongs, tobacco etch virus (TEV) HC-Pro mutants containing single amino acid substitutions at serine, cysteine, aspartic acid, and histidine positions were synthesized by in vitro transcription and translation and were tested for autoproteolytic activity. Combinations of these residues are constituents of the active sites of diverse groups of cellular and viral proteinases. Only those positions that were strictly conserved among four potyvirus HC-Pro proteolytic domains (for which sequences have been deduced) were mutagenized. Of the 19 mutant proteinases synthesized and tested, only those with alterations at Cys-649 and His-722 were defective for HC-Pro autolytic activity. Most of the other mutant proteinases exhibited no impairments in processing kinetics experiments. The spectrum of essential residues, as defined by this genetic analysis, supports the hypothesis that HC-Pro most closely resembles members of the cysteine-type family of proteinases.

Complementary DNA cloning and analysis of carnation mottle virus RNA.

A 3850-base pair (bp) cDNA fragment, representing most of the carnation mottle virus (CarMV) genome, was molecularly cloned in Escherichia coli using a modification of the plasmid-primer method of Okayama and Berg (H. Okayama and P. Berg (1982), Mol. Cell. Biol. 2, 161-170). Poly(dT)-tailed pUC8 was used to prime cDNA synthesis from polyadenylated CarMV RNA template. Oligo(dC) was added to the cDNA 3' end, and second-strand synthesis was specifically primed using an oligo(dG)-tailed linker fragment. cDNA inserts of 3250 and 1950 by were identified from a clone bank as apparently overlapping and were fused at a common BglII restriction site to produce pCarMV-1C. Confirmation that this plasmid harbored CarMV cDNA sequences was achieved by Southern blot hybridization with a randomly primed cDNA probe. Nick-translated pCarMV-1C was used to probe virion and total single-stranded RNA extracts from infected Chenopodtium quinoa leaves by "Northern" blot hybridization. Besides the full-length genomic RNA, two additional species of approximately 1600 and 1750 bases were associated with CarMV. Based on a series of Northern blots with nick-translated subclones derived from pCarMV-1C as probes, these subgenomic RNAs were found to originate from the 3' domain of the virus genome.

Characterization of the cell-free translation products of carnation mottle virus genomic and subgenomic RNAs.

The in vitro translation products of carnation mottle virus (CarMV) genomic and subgenomic RNAs were analysed using a rabbit reticulocyte lysate system. Viral RNAs directed synthesis of three main polypeptides, p80, p40, and p34. p40, which was the predominant product using unfractionated virion RNA as template, was identified as coat protein based on electrophoretic mobility through SDS-polyacrylamide gels and immunoprecipitation with anti-CarMV serum. Upon size-fractionation of virion RNAs by sucrose gradient centrifugation and subsequent translational analysis, p40 was found to be encoded by subgenomic RNA, whereas p80 and p34 were synthesized from templates of genome length. p80 and p34 were found to contain common or overlapping amino acid sequences by analysis of cleavage peptides formed during proteolysis with alpha-chymotrypsin. These data support CarMV translational mechanisms other than those proposed in a previous study (R. Saloman, M. Bar-Joseph, H. Soreq, I. Gozes, and U. Z. Littauer, 1978,Virology 90, 288-298).

Complementation of tobacco etch potyvirus mutants by active RNA polymerase expressed in transgenic cells.

A genetic complementation system was developed in which tobacco etch virus (TEV) polymerase (NIb)-expressing transgenic plants or protoplasts were inoculated with NIb-defective TEV mutants. A beta-glucuronidase (GUS) reporter gene integrated into the genomes of parental and four mutant viruses was used to assay RNA amplification. Two mutants (termed VNN and EDE) contained substitutions affecting the conserved "GDD" polymerase motif or a nuclear localization signal sequence, respectively; one (aD/b) contained a mutation debilitating the NIb N-terminal cleavage site, whereas the other (delta b) lacked the entire NIb sequence. Each mutant was unable to amplify in nontransformed tobacco protoplasts. In contrast, the VNN, EDE, and delta b mutants were complemented to various degrees in NIb-expressing cells, whereas the aD/b mutant was not complemented. The VNN mutant was complemented most efficiently, reaching an average of 11-12% the level of parental TEV-GUS, although in some experiments the level was near 100%. This mutant also replicated in, and spread through, whole transgenic plants to the same level as parental virus. The EDE mutant was complemented relatively poorly, reaching 1% or less of the level of parental TEV-GUS. Despite the close proximity of the EDE substitution to the N-terminal cleavage site, proteolytic processing of NIb was unaffected in an in vitro assay. The delta b mutant was complemented to an intermediate degree in protoplasts, reaching 3.5% the level of parental virus, and replicated and moved systemically in transgenic plants. These data indicate that free NIb supplied entirely in trans can provide all NIb functions essential for RNA amplification. The relative inefficient complementation of the EDE mutant suggests that the resulting mutant protein was transinhibitory.

Debilitation of plant potyvirus infectivity by P1 proteinase-inactivating mutations and restoration by second-site modifications.

Tobacco etch virus (TEV) encodes three proteinases that catalyze processing of the genome-encoded polyprotein. The P1 proteinase originates from the N terminus of the polyprotein and catalyzes proteolysis between itself and the helper component proteinase (HC-Pro). Mutations resulting in substitution of a single amino acid, small insertions, or deletions were introduced into the P1 coding sequence of the TEV genome. Deletion of the N-terminal, nonproteolytic domain of P1 had only minor effects on virus infection in protoplasts and whole plants. Insertion mutations that did not impair proteolytic activity had no measurable effects regardless of whether the modification affected the N-terminal nonproteolytic or C-terminal proteolytic domain. In contrast, three mutations (termed S256A, F, and delta 304) that debilitated P1 proteolytic activity rendered the virus nonviable, whereas a fourth proteinase-debilitating mutation (termed C) resulted in a slow-infection phenotype. A strategy was devised to determine whether the defect in the P1 mutants was due to an inactive proteinase domain or due simply to a lack of proteolytic maturation between P1 and HC-Pro. Sequences coding for a surrogate cleavage site recognized by the TEV NIa proteinase were inserted into the genome of each processing-debilitated mutant at positions that resulted in NIa-mediated proteolysis between P1 and HC-Pro. The infectivity of each mutant was restored by these second-site modifications. These data indicate that P1 proteinase activity is not essential for viral infectivity but that separation of P1 and HC-Pro is required. The data also provide evidence that the proteinase domain is involved in additional, nonproteolytic functions.

Evidence that the potyvirus P1 proteinase functions in trans as an accessory factor for genome amplification.

The tobacco etch potyvirus (TEV) polyprotein is proteolytically processed by three viral proteinases (NIa, HC-Pro, and P1). While the NIa and HC-Pro proteinases each provide multiple functions essential for viral infectivity, the role of the P1 proteinase beyond its autoproteolytic activity is understood poorly. To determine if P1 is necessary for genome amplification and/or virus movement from cell to cell, a mutant lacking the entire P1 coding region (delta P1 mutant) was produced with a modified TEV strain (TEV-GUS) expressing beta-glucuronidase (GUS) as a reporter, and its replication and movement phenotypes were assayed in tobacco protoplasts and plants. The delta P1 mutant accumulated in protoplasts to approximately 2 to 3% the level of parental TEV-GUS, indicating that the P1 protein may contribute to but is not strictly required for viral RNA amplification. The delta P1 mutant was capable of cell-to-cell and systemic (leaf-to-leaf) movement in plants but at reduced rates compared with parental virus. This is in contrast to the S256A mutant, which encodes a processing-defective P1 proteinase and which was nonviable in plants. Both delta P1 and S256A mutants were complemented by P1 proteinase expressed in a transgenic host. In transgenic protoplasts, genome amplification of the delta P1 mutant relative to parental virus was stimulated five- to sixfold. In transgenic plants, the level of accumulation of the delta P1 mutant was stimulated, although the rate of cell-to-cell movement was the same as in nontransgenic plants. Also, the S256A mutant was capable of replication and systemic infection in P1-expressing transgenic plants. These data suggest that, in addition to providing essential processing activity, the P1 proteinase functions in trans to stimulate genome amplification.

Requirement for HC-Pro processing during genome amplification of tobacco etch potyvirus.

The helper component-proteinase (HC-Pro) of tobacco etch potyvirus (TEV) is a multifunctional protein with several known activities. The N-terminal region is required for aphid transmission and efficient genome amplification, the central region is required for long-distance movement in plants, and the C-terminal domain is a cysteine-type proteinase that autocatalytically cleaves between itself and the P3 protein. To investigate the requirement for HC-Pro-mediated proteolysis during viral replication, a variety of mutations resulting in amino acid substitutions and insertions in the proteolytic domain and cleavage site motif were introduced into the viral genome. Mutations affecting the active site residues, His722 and Cys649, or the cleavage site P1' residue, Gly764, inhibited proteolytic activity of HC-Pro. Mutant genomes containing these modifications were amplification-defective in protoplasts and plants. Mutants with substitutions affecting several conserved, but non-active site, residues (Ser610, Cys694, and Asp715) within the proteinase domain encoded active HC-Pro proteinases and were similar to parental virus in protoplasts and plants. To determine if the replication defect of the proteinase-debilitated mutants was due to inactivation of HC-Pro proteolytic activity or simply to the inability of HC-Pro and P3 protein to separate, a sequence coding for a heterologous cleavage site recognized by the TEV NIa proteinase was inserted between the HC-Pro and P3 coding regions of an active site mutant. This cleavage site was functional in vitro using purified NIa proteinase. However, this modification was insufficient to restore amplification activity to the mutant. In addition, the active site mutant was not complemented by wild-type HC-Pro supplied in trans by transgenic plants. These results suggest that an active HC-Pro proteinase is required in cis for TEV genome amplification.

RNA binding activity of NIa proteinase of tobacco etch potyvirus.

The C-terminal domain of NIa protein (NIaPro) from tobacco etch potyvirus (TEV) is a sequence-specific proteinase required for processing of the viral polyprotein. This proteinase also interacts with NIb, the TEV RNA-dependent RNA polymerase. NIaPro and two NIaPro-containing polyproteins (NIa and 6/NIa) were analyzed from extracts of recombinant Escherichia coli. Using RNA-protein blot and UV-crosslinking assays, NIaPro and the NIaPro-containing polyproteins were shown to possess RNA-binding activity. NIaPro bound nonspecifically to several RNAs, including plus- and minus-strands of the TEV 5' and 3' noncoding regions. Saturation binding data obtained using the UV-crosslinking assay were consistent with a possible cooperative RNA-binding activity of NIaPro. In addition, the RNA-binding activities of NIaPro and full-length NIa protein were similar. Based on its RNA-binding activity and other known functions, NIaPro or a NIaPro-containing polyprotein is proposed to serve one or more direct roles during TEV RNA synthesis.

A counterdefensive strategy of plant viruses: suppression of posttranscriptional gene silencing.

Posttranscriptional gene silencing (PTGS) in plants inactivates some aberrant or highly expressed RNAs in a sequence-specific manner in the cytoplasm. A silencing mechanism similar to PTGS appears to function as an adaptive antiviral response. We demonstrate that the P1/HC-Pro polyprotein encoded by tobacco etch virus functions as a suppressor of PTGS. A locus comprised of a highly expressed beta-glucuronidase (GUS) transgene was shown to exhibit PTGS. Genetic crosses and segregation analyses revealed that a P1/ HC-Pro transgene suppressed PTGS of the GUS sequence. Nuclear transcription assays indicated that the silencing suppression activity of P1/HC-Pro was at the posttranscriptional level. These data reveal that plant viruses can condition enhanced susceptibility within a host through interdiction of a potent defense response.

Cap-independent enhancement of translation by a plant potyvirus 5' nontranslated region.

The RNA genome of tobacco etch virus (TEV), a plant potyvirus, functions as an mRNA for synthesis of a 346-kilodalton polyprotein that undergoes extensive proteolytic processing. The RNA lacks a normal 5' cap structure at its terminus, which suggests that the mechanism of translational initiation differs from that of a normal cellular mRNA. We have identified a translation-enhancing activity associated with the 144-nucleotide, 5' nontranslated region (NTR) of the TEV genome. When fused to a reporter gene encoding beta-glucuronidase (GUS), the 5' NTR results in an 8- to 21-fold enhancement over a synthetic 5' NTR in a transient-expression assay in protoplasts. A similar effect was observed when the 5' NTR-GUS fusions were expressed in transgenic plants. By using a cell-free translation system, the translation enhancement activity of the TEV 5' NTR was shown to be cap independent, whereas translation of GUS mRNA containing an artificial 5' NTR required the presence of a cap structure. Translation of GUS transcripts containing the TEV 5' NTR was relatively insensitive to the cap analog m7GTP, whereas translation of transcripts containing the artificial 5' NTR was highly sensitive. The 144-nucleotide TEV 5' NTR synthesized in vitro was shown to compete for factors that are required for protein synthesis in the cell-free translation reaction mix. Competition was not observed when a transcript representing the initial 81 nucleotides of the TEV 5' NTR was tested. These results support the hypothesis that the TEV 5' NTR promotes translation in a cap-independent manner that may involve the binding of proteins and/or ribosomes to internal sites within the NTR.

A viral cleavage site cassette: identification of amino acid sequences required for tobacco etch virus polyprotein processing.

Mature viral-encoded proteins of tobacco etch virus (TEV) arise by proteolytic processing of a large precursor. The proteinase responsible for most of these cleavages is a viral-encoded 49-kDa protein. All known or predicted cleavage sites in the TEV polyprotein are flanked by the conserved sequence motif Glu-Xaa-Xaa-Tyr-Xaa-Gln-Ser or Gly, with the scissile bond located between the Gln-Ser or Gly dipeptide. By using cell-free systems to manipulate and express cloned cDNA sequences, a 25-amino acid segment containing a putative proteolytic cleavage site of the TEV polyprotein has been introduced into the TEV capsid protein sequence. This recombinant protein is cleaved by the 49-kDa proteinase at the introduced cleavage site, thus demonstrating portability of a functional cleavage site. The role of the conserved amino acid sequence in determining substrate activity was tested by construction of engineered proteins that contained part or all of this motif. A protein that harbored an insertion of the conserved 7-amino acid segment was cleaved by the 49-kDa TEV proteinase. Cleavage of the synthetic precursor was shown to occur accurately between the expected Gln-Ser dipeptide by microsequence analysis. Proteins containing insertions that generated only the Gln-Ser, or only the serine moiety of the conserved sequence, were insensitive to the 49-kDa proteinase.

Suppression of long-distance movement of tobacco etch virus in a nonsusceptible host.

To investigate host functions involved in the tobacco etch potyvirus (TEV) infection process, a tobacco line (V20) with a strain-specific defect in supporting systemic infection was analyzed. Using a modified TEV encoding a reporter protein, beta-glucuronidase (GUS), genome amplification, cell-to-cell movement, and long-distance movement were measured in V20 and a susceptible line, Havana425. Comparable levels of TEV-GUS genome amplification were measured in inoculated protoplasts from both tobacco lines. The rates of cell-to-cell movement of virus in inoculated leaves were nearly identical in V20 and Havana425 between 48 and 72 h postinoculation. In contrast, long-distance movement from leaf to leaf was markedly restricted in V20 relative to Havana425. In situ histochemical analysis of inoculated leaves revealed that infection foci expanded radially over time, providing the potential for contact of virus with veins. Immunocytochemical analysis of V20 tissue from infection foci indicated that TEV-GUS entered the phloem parenchyma or companion cells adjacent to the sieve elements, suggesting that the block in long-distance movement was associated with entry into, or exit from, sieve elements. The genetic basis for the V20 restriction was characterized in a segregation analysis of a cross between V20 and Havana425. The heterozygous F1 progeny displayed the susceptible phenotype, indicating that the V20 restriction was a recessive trait. Segregation in the F2 progeny indicated that the restriction was likely due to the interaction of recessive genes at two nonlinked loci. These data support the hypothesis that long-distance movement requires a set of host functions that are distinct from those involved in cell-to-cell movement.