Extreme Reduction of Disease in Oats Transformed with the 5′ Half of the Barley Yellow Dwarf Virus-PAV Genome

Abstract

Barley yellow dwarf viruses (BYDVs) are the most serious and widespread viruses of oats, barley, and wheat worldwide. Natural resistance is inadequate. Toward overcoming this limitation, we engineered virus-derived transgenic resistance in oat. Oat plants were transformed with the 5' half of the BYDV strain PAV genome, which includes the RNA-dependent RNA polymerase gene. In experiments on T2- and T3-generation plants descended from the same transformation event, all BYDV-inoculated plants containing the transgene showed disease symptoms initially, but recovered, flowered, and produced seed. In contrast, all but one of the BYDV-PAV-inoculated nontransgenic segregants died before reaching 25 cm in height. Although all of the recovered transgenic plants looked similar, the amount of virus and viral RNA ranged from substantial to undetectable levels. Thus, the transgene may act either by restricting virus accumulation or by a novel transgenic tolerance phenomenon. This work demonstrates a strategy for genetically stable transgenic resistance to BYDVs that should apply to all hosts of the virus.

A potential mechanism for selective control of cap-independent translation by a viral RNA sequence in cis and in trans

“…Previously, we found that the 109-nt 39TE, in trans, inhibited translation of a reporter gene carrying the 39TE in cis (Wang et al+, 1997)+ Therefore, we tested the ability of the 109-nt 39TE RNA to inhibit translation of genomic and sgRNA1 in trans+ A 100-fold molar excess of the 39TE RNA inhibited translation of gRNA by 50%, whereas four times as much 39TE RNA was required to inhibit translation of sgRNA1 by 50% (Fig+ 4A)+ The defective 39TE RNA containing the filled-in BamHI 4837 site (39TEBF RNA) was far less inhibitory of either mRNA (Fig+ 4A)+ A 300-fold excess of 39TE RNA reduced translation of the 39-kDa product of ORF 1 from gRNA by sixfold (Fig+ 4B, lanes 2-3), whereas translation of coat protein from sgRNA1 was only halved (Fig+ 4B, lanes 5-6)+ Most strikingly, when equal amounts of genomic and sgRNA1 were present in the same reaction, presence of excess 39TE RNA dropped gRNA translation by 11-fold, whereas translation of sgRNA1 was reduced by only 20% (Fig+ 4B, lanes 8-9)+ In all cases, the defective 39TE had little effect on the translation from genomic RNA or sgRNA1 in trans+ (The apparent inhibition of gRNA by 39TEBF RNA in Fig+ 4A and apparent stimulation in Fig+ 4B, lane 4, reflects experimental variation (632%)+ Inhibition greater than twofold is considered significant+) Thus, the transinhibition requires a functional 39TE sequence and it specifically inhibits gRNA much more than sgRNA1+ sgRNA2 accumulates to a 20-40-fold molar excess over genomic RNA in infected cells (Kelly et al+, 1994;Mohan et al+, 1995;Koev et al+, 1998)+ The ratio of sgRNA2 to translatable gRNA is even greater than that seen on Northern blots, because much of the genomic RNA is encapsidated (Mohan et al+, 1995) and thus sequestered from translation+ Because the 39TE comprises the complete 59 UTR of sgRNA2, it is possible that sgRNA2 inhibits translation of genomic and sgRNA1 in trans+ Because of the preferential inhibition of gRNA versus sgRNA1, we propose that as sgRNA2 accumulates, translation of gRNA is reduced, favoring translation of sgRNA1 late in infection+ To test this hypothesis, the effect of sgRNA2 on translation of gRNA and sgRNA1 was evaluated as in the previous experiment+ As predicted, sgRNA2 inhibited translation of the genomic RNA more effectively than it inhibited translation FIGURE 3. Cap-independent translation of capped (C) and uncapped (U) subgenomic RNAs+ A: Wheat germ translation products of sgRNA1 (map at top), which was transcribed from pSG1 linearized with ScaI (lanes 3, 4), Pst I (lanes 5, 6) or SmaI (lanes 7, 8)+ Proteins were analyzed by 10% polyacrylamide gel electrophoresis+ Lanes 9 and 10 are the products of SmaI-cut pSG1BF transcript that contains the GAUC duplication in the BamHI 4837 site of the 39TE+ Lanes 11 and 12 show translation products of SmaI-cut pSP17 transcript in which the 59-terminal 99 nt of the 188-nt 59 UTR of sgRNA1 were deleted+ Mobilities of products of ORFs 3 (22 kDa), 4 (17 kDa) and 3ϩ5 (72 kDa, made by the in-frame read-through of the ORF 3 stop codon) are indicated at right+ Other bands indicate cleavage products of the labile 72-kDa protein (Filichkin et al+, 1994) and premature termination products within ORF 5 (Brown et al+, 1996)+ Relative moles of translation product (of ORF 3) determined with a Phosphorimager using ImageQuant software are indicated below each lane+ Samples in lanes 9-10 and 11-12 were from different experiments, and the products of the 100% standard (capped SmaI-cut pSG1 transcript) for these are not shown+ B: Products of transcripts from SmaI-cut pSG2 (lanes 2, 3) and pSG2BF (lanes 4, 5), following electrophoresis on a 10% polyacrylamide gel+ Mo...…”

“…The more efficient trans-inhibition of translation by full-length sgRNA2 than the 109-nt 39TE is not due to the active translation of sgRNA2, because mutation of the ORF 6 start codon had no effect on trans-inhibition (Fig+ 6B)+ Furthermore, translatable (capped) sgRNA2 with a defective 39TE (Fig+ 3B) did not inhibit in trans (Fig+ 6B)+ Thus, like the 109-nt 39TE alone, sgRNA2 inhibits via the 39TE-mediated mechanism+ We speculate the sgRNA2 inhibits more efficiently because it may have a higher binding affinity for protein factors that mediate cap-independent translation+ This could also explain the need for the sgRNA2 sequence in cis for cap-independent translation in vivo+ sgRNA2 may facilitate a switch from early to late gene expression Gene expression of many viruses is divided into temporal stages with nonstructural replication proteins expressed early and structural proteins expressed late+ Synthesis of BYDV subgenomic RNAs requires replication, so the structural genes they encode are not translated until after RNA replication has commenced+ Thus, RNA-templated transcription (subgenomic RNA synthesis) alone can account for turning on late gene expression+ However, the data presented here suggest an additional level of control mediated by viral RNA in trans that may act to shut off expression of early genes+ We propose a model of trans-regulation of translation by the 39TE in which accumulation of sgRNA2 at high levels preferentially inhibits translation of genomic RNA over sgRNA1+ Early in infection, genomic RNA from the invading virion is the only message (Early, Fig+ 7)+ This allows cap-independent translation of ORFs 1 and 1ϩ2 (replicase) facilitated by the 39TE in cis+ The replicase then replicates gRNA and transcribes sgRNAs+ As large amounts of sgRNA2 accumulate (Late, Fig+ 7), it strongly inhibits translation of gRNA, shutting off translation of replication genes (ORFs 1 and 2), while only weakly inhibiting translation of sgRNA1, permitting translation of structural and movement protein genes (ORFs 3, 4, and 5)+ This model is supported by the following observations+ (1) The 39TE is required in cis for translation (Allen et al+, 1999) of the only two genes (ORFs 1 and 2) required for RNA replication (Mohan et al+, 1995)+ (2) Thus, intact 39TE is required for replication in vivo (Allen et al+, 1999)+ (3) Only ORFs 1 and 2 are translated from gRNA (Di et al+, 1993;Mohan et al+, 1995;Allen et al+, 1999)+ (4) The 59 end of the in vivo-defined 39TE sequence that gives cap-independent translation in cis coincides precisely with the 59 end of sgRNA2 (Fig+ 2)+ (5) sgRNA2 inhibits translation of gRNA in trans far more efficiently than it inhibits translation of sgRNA1 (Fig+ 5A)+ (6) When gRNA and sgRNA1 are competing with each other in the presence of sgRNA2 at ratios similar to those in infected cells, only the products of sgRNA1 are translated significantly, and gRNA is virtually shut off (Fig+ 5B)+ sgRNA2 accumulates to at least 20-to 40-fold molar excess to gRNA (Kelly et al+, 1994;Mohan et al+, 1995;Koev et al+, 1998) and probably to a higher ratio when compared to translatable (non-encapsidated) gRNA+…”

Viruses of Poaceae: a case history in plant pathology

“…While conventional breeding will no doubt continue, alternative approaches based on recombinant technology are now becoming important but have yet to reach commercial exploitation. Oats and barley (McGrath et al ., 1997; Koev et al ., 1998) have been transformed with constructs containing the coat protein gene of BYDV‐PAV, ‐MAV and ‐RPV and decreased virus titre was associated with the presence of the coat protein gene in these plants. Wang et al .…”

Resistance and Tolerance to Barley Yellow Dwarf in Cereals

“…So, in order to provide resistance to all strains, a battery of coat protein genes from different strains must be introduced into transgenic plants, or a chimeric protein can be designed with features that confer resistance to each of the strains (Miller and Young, 1995). Koev et al (1998) transformed successfully oat plants with the 5i half of the BYDV strain PAV genome, including the RNA-dependent RNA polymerase gene. These oats reduced dramatically disease symptoms when inoculated with BYDV-PAV.…”