In model plants such as Arabidopsis and rice (Oryza sativa L.), many tools of molecular biology have been developed to greatly improve the process of isolating genes. Two tools that have significantly accelerated the speed with which gene function can be determined in model plants are T-DNA knockout libraries and T-DNA activation libraries (Weigel et al., 2000). Here, large collections of plants have been generated that contain insertions of T-DNAs. In knockout libraries the T-DNAs interrupt and thereby inactivate the genes they have inserted into, while in activation libraries the T-DNAs are engineered to have enhancer elements near their borders, which activate the expression of genes near the T-DNA insertion site. Once plants have been identified with the desired phenotypes, isolation of the relevant gene is readily accomplished by finding the genomic location of the T-DNA that cosegregates with the mutant phenotype. Unfortunately, none of these tools exist for wheat (Triticum aestivum L.) and barley (Hordeum vulgare L.) because these plants can only be transformed with very low efficiency and the genomes are so large that the number of transformants needed is too great. An additional complication for determining gene function through the analysis of loss-of-function mutations in wheat is the fact that all cultivated varieties are polyploid and, therefore in most cases, expression of homeologous genes will mask loss-of-function phenotypes.
Virus-induced gene silencing (VIGS) is a tool for the rapid assessment of gene function that overcomes many of the limitations present in cereal crops. Recently a system for VIGS has been developed for use in wheat and barley. This article will discuss the properties of this system and how it is being used for functional genomics in wheat and barley. It will also discuss the strengths and weaknesses of this approach and what might be done to improve its utility.
WHAT IS VIGS?
Virus-induced gene silencing is a very useful research tool for rapid creation of gene knockdown phenotypes that can be used to assess plant gene function (Baulcombe, 1999; Kumagai et al., 1995; Ratcliff et al., 1997). The biological principles on which VIGS is based were uncovered as molecular biologists studied the consequences of virus infection in plants (Lindbo and Dougherty, 1992). Through this work, it was discovered that many RNA viruses activate a conserved, RNA-based plant antiviral defense response, which targets the RNA produced by infecting viruses for sequence-specific degradation (Ratcliff et al., 1997). This RNA-based plant defense response is triggered when double-stranded RNA (dsRNA) accumulates within cells, as occurs during the replication of RNA viruses. All the sequence within the dsRNA becomes targeted by the host defense system for sequence-specific degradation. This process is exploited in VIGS to permit researchers to down-regulate or “knockdown” the expression of plant genes of their choosing by infection with engineered viruses. By inserting a fragment of transcribed sequence from a plant gene, which the researcher wishes to silence, into the VIGS construct, transcripts of the gene-of-interest are targeted to undergo homology-dependent degradation, thereby causing the gene to be silenced.
Several aspects of VIGS make it a particularly useful tool for plant functional genomics. (i) It is a rapid experimental process. In most instances, the knockdown phenotype of a gene-of-interest is generated within 1 to 2 mo of identifying the candidate sequence. This is far quicker than what is possible through the production and analysis of knockout mutant or stably transformed RNA interference (RNAi) plants. (ii) VIGS does not require full-length cDNA sequences to function, so experiments can be initiated without complete gene sequence information. (iii) VIGS is initiated by infecting plants with a viral construct, so it is possible to observe the effect of transient silencing of genes that would have homozygous lethal phenotypes in conventional mutant analyses. (iv) VIGS can be particularly useful for research in polyploid plants such as wheat because gene silencing occurs through homology-dependent RNA-mediated gene silencing and, therefore, any genes sharing at least ∼85% sequence identity are likely to be down-regulated. In this way, knockdown phenotypes can be observed because the closely related homeologous genes present in polyploids are likely to be silenced as well. (v) VIGS is initiated by virus infection and so can be performed on species that are difficult to transform for stable RNAi studies.
In principle VIGS should be a very useful tool for research in any plant species. However, a major limitation to its widespread adoption is the lack of viral vectors that generate useful gene silencing in different plant species. Initially, VIGS was almost exclusively performed in Nicotiana benthamiana (Karel Domin) using vectors derived from tobacco mosaic virus (Kumagai et al., 1995), potato virus X (Ratcliff et al., 1997), and tobacco rattle virus (Liu et al., 2002; Ratcliff et al., 2001). In recent years, new protocols and vectors have expanded the list of dicotylendonous plants in which VIGS can be employed, for example, Arabidopsis (Burch-Smith et al., 2006) and potato (Solanum tuberosum L.) (Brigneti et al., 2004; Faivre-Rampant et al., 2004), but it was not until recent experimentation with barley stripe mosaic virus (BSMV)–based vectors that VIGS became an option for functional genomics research in wheat (Scofield et al., 2005) and barley (Hein et al., 2005; Holzberg et al., 2002).
A VIGS SYSTEM BASED ON BARLEY STRIPE MOSAIC VIRUS
Barley stripe mosaic virus is a single-stranded RNA virus of the Hordeivirus genus. Its genome is tripartite, comprised of the α, β, and γ RNAs. Petty et al. (1989) synthesized cDNAs of the three RNAs and cloned them each into DNA plasmids such that infectious BSMV RNAs could be synthesized from the three plasmids by in vitro transcription of capped RNAs. Barley stripe mosaic virus infection is initiated by mixing together in vitro transcripts from the α, β, and γ DNA plasmids and rub-inoculating them onto susceptible host plants.
Holzberg et al. (2002) first demonstrated that BSMV could trigger useful VIGS in barley plants. In this initial study, it was demonstrated that BSMV constructs engineered to carry 0.19- to 1.4-kb inserts of the barley phytoene desaturase (PDS) gene would cause down-regulation of PDS expression. Phytoene desaturase expression is essential for the synthesis of carotenoid pigments, which protect chlorophyll from photolysis. Tissue in which PDS had been down-regulated can therefore be visualized by the appearance of photobleaching where chlorophyll had undergone photolysis (see Fig. 1 ). Typically in BSMV-VIGS studies, a 120- to 500-bp fragment, representing a portion of a transcribed sequence from a plant gene, is inserted into the γ RNA plasmid at restriction sites immediately 3′ to the stop codon of the γb gene (Holzberg et al., 2002). The 120-bp minimum size for the plant gene fragment is based on the observation that host insert sequences <120 bp are significantly less effective in BSMV-VIGS (Bruun-Rasmussen et al., 2007; Scofield et al., 2005). The upper size limit of 500 bp is less well defined, but reflects the fact that all sequences inserted into plant viral vectors are unstable as the virus replicates (Pogue et al., 2002), and larger fragments may be lost with greater frequency (Bruun-Rasmussen et al., 2007; Cakir and Scofield, 2008).
Unlike genetic mutations abolishing gene function or gene silencing in transgenic plants expressing RNAi constructs, silencing in VIGS occurs transiently. The temporal and spatial patterns of gene silencing have been analyzed for BSMV-VIGS in wheat and barley seedlings. When BSMV infection is initiated on the second leaf, BSMV moves systemically into the third leaf and significant silencing can be detected there 3 d postinoculation (dpi) and will persist until at least 21 dpi (Hein et al., 2005; Scofield et al., 2005).
THE UTILITY OF BSMV-VIGS FOR FUNCTIONAL GENOMICS IN WHEAT AND BARLEY
Much of the initial use of the BSMV-VIGS system has been in the functional analysis of disease resistance pathways in wheat and barley. The general design of these experiments is to initiate silencing of a candidate gene in a wheat genotype that is normally resistant to a pathogen of interest, challenge the silenced plants with this pathogen, and then assess whether or not the plants are still resistant to the pathogen. Conversion to susceptibility of plants infected with the viral constructs silencing the candidate gene, while plants infected with control viral constructs remain resistant, is strong evidence for the candidate gene having an essential function in the resistance pathway.
The first published studies employing BSMV-VIGS demonstrated the system's utility in the functional dissection of two gene-for-gene disease resistance pathways from wheat (Scofield et al., 2005) and barley (Hein et al., 2005). Lr21-mediated resistance to leaf rust in wheat and the Mla13-mediated resistance to powdery mildew in barley are each controlled by nucleotide binding–leucine-rich repeat (NB-LRR) class resistance genes (R-genes). The NB-LRR class of R-genes is the most extensively studied class of R-genes. The NB-LRR proteins are known to either directly or indirectly detect the presence of elicitors expressed by avirulent pathogens and initiate a signal-transduction process that results in the hypersensitive response (HR) and activation of pathogen defense responses. These BSMV-VIGS studies each demonstrated that resistant genotypes became susceptible when infected with BSMV-VIGS constructs designed to silence the wheat Lr21 and barley Mla13 R-genes (see Fig. 2 ). Resistant genotypes infected with control constructs remained resistant, indicating that BSMV infection did not disturb plant physiology sufficiently to interfere with these R-gene responses. Both studies also tested whether the Lr21- and Mla13-resistant responses were dependent on the expression of the RAR1, SGT1, or HSP90 genes, which have been found to be essential in other NB-LRR resistance pathways of Arabidopsis, tobacco (Nicotiana tabacum L.), tomato (Lycopersicon esculentum Mill.), and barley. The BSMV-VIGS experiments showed that Lr21- and Mla13-mediated resistance were abrogated when RAR1, SGT1, and HSP90 were silenced, indicating their essential role in these resistance pathways.
Two published studies have employed BSMV-VIGS in the final stages of map-based cloning to functionally confirm that the correct gene had been isolated. BSMV-VIGS was used to functionally confirm that the wheat Lr1 leaf rust R-gene (Cloutier et al., 2007) and the barley Rpg5 stem rust R-gene (Brueggeman et al., 2008) had been isolated after chromosome-walking procedures. Before the availability of a suitable VIGS system, confirmation would have been performed by the time-consuming step of transforming a susceptible genotype with a construct that would express the candidate R-gene. In fact, transformation was required to unambiguously identify Lr1 because it was found that several highly homologous NB-LRR genes were present at the Lr1 locus. The VIGS analyses performed in this study were unable to discriminate between the highly homologous R-gene-like sequences at this locus (Cloutier et al., 2007).
Microarray studies have generated large lists of differentially expressed genes during compatible and incompatible interactions. The great challenge now is to sort through these lists to determine which genes have causal roles in the outcome of the plant–pathogen interactions. Several recent studies have employed BSMV-VIGS to test if genes implicated in disease resistance by induction of transcription during resistance may have bona fide functions in resistance. Zhou et al. (2007) took this approach to test if three wheat receptor-like kinase genes (TaRLK 1, 2, and 3), which were up-regulated during incompatible reactions with stripe rust, had functional roles in resistance. Using VIGS constructs designed to specifically silence each TaRLK gene or constructs that target the silencing of all three genes, they observed that all three TaRLKs contribute to stem rust resistance (Zhou et al., 2007).
Two recent publications from the Wise laboratory have also used BSMV-VIGS to test for functional roles in resistance of genes first implicated through differential expression detected in microarray analysis of barley–powdery mildew interactions (Hu et al., 2009; Meng et al., 2008). Both of these studies are focused on identifying the genetic pathways of basal defense, which functions independently of R-genes. Transcription of a small family of genes encoding barley-specific peptides was found to be up-regulated during interactions with the powdery mildew fungus, Blumeria graminis f. sp. hordei Speer (Bgh). Silencing this gene family, called Blufensin1 (BLN1), by BSMV-VIGS results in enhanced resistance that is independent of Mla R-genes. When BLN1 was silenced in plants expressing the Mlo susceptibility factor, they also observed increased resistance. Taken together, they conclude that BLN1 functions in basal defense as a negative regulator of resistance to penetration by Bgh The second study examines three genes, chorismate synthase, anthranilate synthase, and chorismate mutase, which function in the aromatic amino acid biosynthetic pathway and are coordinately induced from 0 to 16 h after infection in compatible and incompatible interactions with Bgh (Meng et al., 2008). Silencing each of these genes by BSMV-VIGS results in increased penetration by Bgh, while overexpression causes increased resistance, strongly suggesting a function for these genes in resistance to penetration by Bgh
When an entire species is resistant to a pathogen, it is termed nonhost resistance This is one of the least understood areas of plant disease resistance and likely results from a wide variety of mechanisms. Johal and Briggs (1992) isolated the first nonhost resistance gene from plants, the HM1 gene of maize (Zea mays L.). All maize plants are resistant to the pathogen Cochliobolus carbonum Nelson race 1 (CCR1), except mutants in which the HM1 and HM2 loci are nonfunctional. The hm1/hm2 mutants are highly susceptible to CCR1, which kills the mutant plants at any stage of their growth. Johal and Briggs (1992) found that HM1 encodes a reductase enzyme, which detoxifies the toxin produced by CCR1 that is absolutely required for pathogenesis. They found that HM1 homologs are only found in grass species, and as no other grasses are hosts to CCR1, it begged the question: Do HM1 genes present in other grasses provide nonhost resistance to CCR1? This hypothesis was tested by employing BSMV-VIGS to silence the six copies of HM1 present in barley (Sindhu et al., 2008). It was found that silencing the barley HM1 genes conferred very strong susceptibility to CCR1. This result is quite sobering to contemplate. The grass family provides the most important crops for human survival, and they are all being protected against this devastating pathogen by the HM1 gene, which arose uniquely in the grass lineage approximately 40 million yr ago (Sindhu et al., 2008). Understanding how this system has provided such durable resistance will be very important for future efforts to engineer disease resistance.
While most of the published applications of BSMV-VIGS have addressed the functional genomics of disease resistance pathways, there are a few studies investigating other biological systems. Aside from the studies characterizing BSMV-VIGS, where plant genes are silenced which give photobleaching phenotypes, there are two publications investigating the silencing of genes involved in cell wall biosynthesis. Oikawa et al. (2007) employed BSMV-VIGS to characterize the function of the P23k gene that is unique to monocots. Silencing P23k in barley resulted in asymmetrically shaped leaves with frequent cracks along the margins. These results, together with the finding that P23k expression is induced in the vascular bundles, supports their assertion that this gene is involved in the synthesis of cell wall polysaccharides and secondary wall formation. Held et al. (2008) utilized BSMV-VIGS to observe the consequences of silencing the barley cellulose synthase, HvCesA. Surprisingly, the HvCesA VIGS experiment resulted in not only down-regulation of HvCesA but also a number of nontarget cellulose synthase–like (Csl) genes. Investigation of the basis of this effect identified the existence of naturally occurring antisense transcripts of HvCesA. The antisense and sense transcripts form dsRNAs, which are then processed by a dicer enzyme to form small interfering RNAs (siRNAs). These siRNAs accumulate late in leaf development as cellulose synthesis decreases, consistent with their acting to coordinately down-regulate other Csl genes during cell wall biosynthesis. This result points not only to the power of VIGS, but also to the complexities of plant genomes that will be encountered as this technique is applied.
STRENGTHS, WEAKNESSES, AND FUTURE DIRECTIONS
Genetic improvement of cereal crops is crucial to meeting the rapidly increasing production requirements for world food supply. Genetic experimentation in wheat and barley has been greatly impeded by the size and complexity of their genomes and also aspects of their biology that prevent the easy implementation of the many advanced technologies developed in model plants. The examples discussed here make it clear that BSMV-VIGS is opening many new avenues for functional genomics in wheat and barley. The ability to generate knockdown phenotypes without having to perform the difficult and time-consuming process of transformation and regeneration is a highly significant advantage, as is the ability to silence all copies of a gene present in complex, polyploid genomes.
However, BSMV-VIGS is certainly not without limitations and experimental complications. Perhaps the biggest difficulty in employing any VIGS systems is variation in the extent with which silencing phenotypes develop. In the context of performing functional genomics assays, an optimal VIGS system would generate silencing over a predictable and sufficiently large area of the plant so that the expected phenotype can be easily recognized. VIGS results from a complex interaction between the plant and virus in which the size of the area exhibiting silencing is determined by the balance between the replication, movement, and pathogenicity of the virus and the strength of the silencing response mounted by the host. Movement of the virus and spreading of the silencing signals is driven by the source–sink relationships within the plant, so careful attention to plant growth is crucial for obtaining uniform results. The addition of a pathogen assay to the VIGS experimental system adds a third organism and an additional source of variation.
Very careful attention to controls is essential to successful VIGS experiments. Without question, infection by the virus has great effects on the physiological state of the host. Therefore, it is essential to run ample numbers of controls in which plants are infected with viral constructs that do not target plant genes for silencing. The control plants must be carefully assessed to ensure that there is no perturbation of the experimental phenotype to be scored in the experiment. Dealing with this variation and the large number of control and experimental plants required to manage the variation are the greatest factors limiting throughput in this experimental system.
Another essential element of VIGS analysis is confirmation that the target gene has been silenced. This is typically done by quantitative reverse transcription polymerase chain reaction (Bustin, 2005). This determination is particularly important when a VIGS experiment does not yield the expected phenotype. Finally, the possibility of “off-target” silencing has been reported in VIGS and RNAi experiments. Here a gene other than the intended target becomes silenced and is the true cause of the knockdown phenotype (Xu et al., 2006). A very strong way to refute this possibility is to perform another round of VIGS in which the same gene is targeted by VIGS constructs carrying fragments from the same gene, but which have no homology to the previous VIGS targeting sequence (Scofield et al., 2005). Generation of a similar knockdown phenotype using fragments of the same gene, but which have no sequence overlap, provides very strong evidence against “off-target” silencing.
All of the BSMV-VIGS studies to date have screened a small number of candidate sequences that were selected based on the results of other experimentation. The ultimate value of VIGS, however, would come when screens can be performed with sufficient throughput so that the function of genes without a priori experimental support could be identified. There is one published account of a high-throughput VIGS assay in which 4992 unique sequences from a normalized tobacco cDNA library were screened for function in the hypersensitive response mediated by the Pto R-gene (Lu et al., 2003). This screen led to identification of 79 cDNAs whose expression was required for the Pto-mediated HR. Of these, six were then demonstrated to be essential for disease resistance. Screening this large number of cDNAs was only possible because the potato virus X (PVX) VIGS system had been engineered so that VIGS could be initiated by injecting cultures of Agrobacterium-carrying T-DNA constructs that then expressed infectious PVX RNA. Additionally, the screening for Pto-mediated HR was performed by transiently infecting the silenced plants with Agrobacterium-carrying T-DNA constructs that transiently expressed the AvrPto elicitor in planta, thereby activating the Pto-mediated HR.
Efforts to adapt the BSMV system for more high-throughput applications are just beginning. The first-generation BSMV-VIGS system requires synthesis of capped in vitro transcripts from the three plasmids carrying the BSMV genomic RNAs. This is both time consuming and expensive. A recent improvement that eliminates the need for in vitro transcription was developed in the Wise laboratory. In this system, plasmids have been constructed so that each BSMV RNA can be transcribed in planta from the cauliflower mosaic virus 35S promoter (CaMV35S), ribozyme processing sequences are cloned at the 3′ end of the viral RNAs to generate the correct 3′ termini. In this system, infection is initiated by mixing the three BSMV plasmids together and then biolistically bombarding the mixture into plants (Hu et al., 2009; Meng et al., 2008). Very recently, progress toward a T-DNA–based BSMV-VIGS system was reported by Jackson et al. (2009) Clearly these improvements should greatly increase the utility of BSMV-VIGS.
The majority of published studies employing BSMV-VIGS, and other VIGS systems as well, are investigating disease resistance pathways. It is not entirely clear why this is the case, although it should be noted that most of VIGS systems have been developed by research groups already studying plant–pathogen interactions. However, activation of disease resistance responses happens very quickly and utilizes signaling components that tend to be of low abundance in plant cells. Such pathways may be particularly well suited for analysis by the transient silencing characteristic of VIGS. Transient silencing may be less effective in interrupting the function of pathways involving proteins that are expressed at high levels or have long half-lives.
A substantial amount of experience has been gained using BSMV-VIGS to dissect disease resistance pathways of wheat and barley. Cereal research groups around the world are rapidly adopting this technique. The increasing user base should lead to improvements in this technology and should increase its effectiveness for the analysis of genetic traits.
This manuscript is dedicated to the memory of Professor Mike Gale FRS, who enriched plant genetics with his many significant scientific contributions, enthusiasm, generosity, and keen sense of humor.