RNA silencing is an RNA-directed gene regulatory system that is present in a wide range of eukaryotes, and which functions as an antiviral defense in plants. Silencing pathways are complex and partially overlapping, but at least three basic classes can be distinguished: cytoplasmic RNA silencing (or post-transcriptional gene silencing; PTGS) mediated by small interfering RNAs (siRNAs), silencing mediated by microRNAs (miRNAs), and transcriptional gene silencing (TGS) mediated by siRNA-directed methylation of DNA and histone proteins. Recent advances in our understanding of different geminivirus silencing suppressors indicate that they can affect all three pathways, suggesting that multiple aspects of silencing impact geminivirus replication.
Most plant viruses are initiators and targets of RNA silencing and encode proteins that suppress this adaptive host defense. The DNA-containing geminiviruses are no exception, and the AL2 protein (also known as AC2, C2, and transcriptional activator protein) encoded by members of the genus Begomovirus has been shown to act as a silencing suppressor. Here, a three-component, Agrobacterium-mediated transient assay is used to further examine the silencing suppression activity of AL2 from Tomato golden mosaic virus (TGMV, a begomovirus) and to determine if the related L2 protein of Beet curly top virus (BCTV, genus Curtovirus) also has suppression activity. We show that TGMV AL2, AL2(1-100) (lacking the transcriptional activation domain), and BCTV L2 can all suppress RNA silencing directed against a green fluorescent protein (GFP) reporter gene when silencing is induced by a construct expressing an inverted repeat GFP RNA (dsGFP). We previously found that these viral proteins interact with and inactivate adenosine kinase (ADK), a cellular enzyme important for adenosine salvage and methyl cycle maintenance. Using the GFP-dsGFP system, we demonstrate here that codelivery of a construct expressing an inverted repeat ADK RNA (dsADK), or addition of an ADK inhibitor (the adenosine analogue A-134974), suppresses GFP-directed silencing in a manner similar to the geminivirus proteins. In addition, AL2/L2 suppression phenotypes and nucleic acid binding properties are shown to be different from those of the RNA virus suppressors HC-Pro and p19. These findings provide strong evidence that ADK activity is required to support RNA silencing, and indicate that the geminivirus proteins suppress silencing by a novel mechanism that involves ADK inhibition. Further, since AL2(1-100) is as effective a suppressor as the full-length AL2 protein, activation and silencing suppression appear to be independent activities.
Transient expression studies using a Nicotiana benthamiana suspension cell-derived protoplast system have identified a minimal sequence that is necessary and sufficient for activation of the tomato golden mosaic virus coat protein (CP) promoter by the viral TrAP protein (also called AL2). The sequence has a bipartite arrangement in which elements located between −125 to −107 and −96 to −60 from the transcription start site are both required for TrAP-mediated activation. One of the sequences (−96 to −60) also appears to interact with a repressor, as its deletion increases basal promoter activity in the absence of TrAP. That competition experiments using the −107 to −60 sequence to titrate the repressor also resulted in increased basal transcription is consistent with this idea. Thus, in a protoplast system which models mesophyll, regulation of the minimal CP promoter involves both activation and derepression by TrAP.
AL2 and L2 are related proteins encoded by geminiviruses of the Begomovirus and Curtovirus genera, respectively. Both are pathogenicity determinants that cause enhanced susceptibility when expressed in transgenic plants. To understand how geminiviruses defeat host mechanisms that limit infectivity, we searched for cellular proteins that interact with AL2 and L2. Here, we present evidence that the viral proteins interact with and inactivate adenosine kinase (ADK), a nucleoside kinase that catalyzes the salvage synthesis of 5'-AMP from adenosine and ATP. We show that the AL2 and L2 proteins inactivate ADK in vitro and after coexpression in Escherichia coli and yeast. We also demonstrate that ADK activity is reduced in transgenic plants expressing the viral proteins and in geminivirus-infected plant tissues. By contrast, ADK activity is increased after inoculation of plants with diverse RNA viruses or a geminivirus lacking a functional L2 gene. Consistent with its ability to interact with multiple cellular kinases, we also demonstrate that AL2 is present in both the nucleus and the cytoplasm of infected plant cells. These data indicate that ADK is targeted by viral pathogens and provide evidence that this "housekeeping" enzyme might be a part of host defense responses. In previous work, we showed that AL2 and L2 also interact with and inactivate SNF1 kinase, a global regulator of metabolism that is activated by 5'-AMP. Together, these observations suggest that metabolic alterations mediated by SNF1 are an important component of innate antiviral defenses and that the inactivation of ADK and SNF1 by the geminivirus proteins represents a dual strategy to counter this defense. AL2 proteins also have been shown to act as suppressors of RNA silencing, an adaptive host defense response. A possible relationship between ADK inactivation and silencing suppression is discussed.
virus -[IC] AF101476, AF101478CdTV-[H8] Cotton leaf crumple virus
The AL2 gene of the geminivirus tomato golden mosaic virus (TGMV) encodes a transcriptional activator protein (TrAP) that is required for efficient expression of the viral coat protein (CP) and BR1 gene promoters. In contrast, L2, the positional homolog of AL2 in the related beet curly top virus (BCTV), is not required for CP expression, raising questions about the functional relationship between the AL2 and L2 gene products. In this study, transgenic Nicotiana benthamiana and N. tabacum var. Samsun plants expressing a truncated AL2 gene (AL2(1-100), lacking the activation domain) or full-length L2 were prepared. These transgenic plants showed a novel enhanced susceptibility (ES) phenotype following inoculation with TGMV, BCTV, or tobacco mosaic virus (TMV), an unrelated RNA virus. ES is characterized by a reduction in the mean latent period (from 1 to 9 days) and by a decrease in the inoculum concentration required to infect transgenic plants (ID50 reduced 6- to 60-fold). However, ES does not result in an enhancement of disease symptoms, and viral nucleic acids do not accumulate to substantially greater levels in infected transgenic plants. That both viral transgenes condition ES suggests that their products share the ability to suppress a host stress or defense response that acts against DNA and RNA viruses. The data further indicate that the transcriptional activation activity of AL2 protein is not required for suppression. The nature of the response targeted by the AL2 and L2 gene products is discussed.
Vascular puncture inoculation (VPI) is an effective technique for transmission of maize viruses without using arthropods or other biological vectors. It involves using a jeweler's engraving tool to push minuten pins through a droplet of virus inoculum toward the major vascular bundle in the scutellum of germinating kernels. Here, VPI is shown to be useful for introducing RNA and DNA viral genomes into maize. Maize dwarf mosaic potyvirus (MDMV) virions, MDMV genomic RNA, foxtail mosaic potexvirus (FoMV) genomic RNA and maize streak geminivirus (MSV) DNA were introduced into kernels by VPI, and infection rates determined. At high concentrations, both MDMV virion and genomic RNA preparations produced 100% infection of susceptible maize. However, MDMV genomic RNA was transmitted with about 100-fold lower efficiency than virions. FoMV genomic RNA and MSV DNA were transmitted at lower efficiency than the MDMV RNA, and the highest transmission rates were about 50%. Ribonuclease A pretreatment prevented genomic MDMV and FoMV RNA transmission, but not MDMV virion transmission indicating the viral RNA was the infectious entity. Proteinase K (ProK) pretreatment reduced transmission of MDMV RNA suggesting that integrity of the viral genomic protein bound covalently to the viral RNA may be important for efficient transmission.
The AL2 gene found in members of the genus Begomovirus of the Geminiviridae encodes a transcriptional activator protein (TrAP; also known as AL2, AC2, or C2 protein). TrAP activates expression from the viral coat protein (CP) and BR1 movement gene promoters in mesophyll cells and protoplasts and acts to derepress the CP promoter in vascular tissue. The experiments presented here were designed to elucidate some of the biochemical properties of this multifunctional regulatory protein and to define its activation domain. The results indicate that TrAP from tomato golden mosaic virus (TGMV) binds single-stranded DNA in a sequence nonspecific manner and only weakly interacts with double-stranded DNA, confirming earlier results obtained with TrAP from other begomoviruses. In addition, evidence is presented that indicates that TrAP binds zinc and that zinc is necessary for optimal interaction with ssDNA. We also show that TrAP is phosphorylated when expressed in insect cells and that it contains a transcriptional activation domain of the acidic type. The minimal activation domain is quite small; the region comprising only the 15 C-terminal amino acids of the protein is capable of activating transcription in mouse fibroblasts (NIH3T3 cells) when fused to a heterologous DNA-binding domain.
Tomato golden mosaic virus (TGMV) is a bipartite member of the subgroup III Geminiviridae. Like all geminiviruses, TGMV replicates in the nucleus of susceptible cells by rolling circle replication (RCR). Double-stranded replicative form DNA generated during RCR serves as template for the transcription of viral genes by RNA polymerase II and the associated cellular transcription machinery. Previous studies in tobacco protoplasts and Nicotiana benthamiana leaf discs have shown that the viral AL2 gene product transactivates expression of the coat protein (CP) and BR1 movement protein genes, and that activation occurs at the level of transcription. Because of its function and properties, we propose the name TrAP, transcriptional activator protein, for the AL2 gene product. Using transgenes consisting of complete and truncated versions of the CP promoter fused to the GUS reporter gene, we show in the studies presented here that TrAP is required for CP gene expression in both mesophyll and phloem tissues. Surprisingly, TrAP appears to induce CP expression by different mechanisms in different cell types: it may activate the CP promoter in mesophyll cells, and acts to derepress the promoter in phloem tissue. In addition, TrAP is clearly capable of inducing the expression of responsive chromosomal promoters and could, in principle, activate host genes. Distinct viral sequence elements mediate expression and derepression in phloem and activation in mesophyll, suggesting that TrAP interacts with different components of the cellular transcription machinery to accomplish CP gene expression in different cell types, and underscoring the intricacy and complexity of virus-host interactions. (C) 1997 Academic Press.
The monopartite DNA genome of beet curly top geminivirus (BCTV, strain Logan) contains four leftward (complementary sense) open reading frames (ORFs) designated L1, L2, L3, and L4. We investigated the functions of the L2 and L3 ORFs by mutational analysis. We found that in Nicotiana benthamiana and sugarbeet plants, neither a functional L2 nor a functional L3 gene is required for infectivity. Double mutants were also infectious, and no evidence for a synergistic effect of these genes was evident. However, while sugarbeet plants inoculated with L2 or L3 mutants showed symptoms that were indistinguishable from those elicited by wild type virus, mutant-inoculated N. benthamiana plants displayed a novel phenotype in which recovery of the plant from initially severe disease symptoms was greatly enhanced. Enhanced recovery was associated with a large reduction in viral DNA levels. Our studies did not provide evidence for functional homology between the BCTV L2 gene and its presumed homologue (AL2) in the bipartite geminiviruses. In contrast, mutants with lesions in the L3 ORF accumulated three- to five-fold less DNA than wild type virus in a protoplast replication assay, consistent with the interpretation that the BCTV L3 gene is a homologue of the bipartite geminivirus AL3 gene which is known to function as a replication enhancer. Functional homology was directly confirmed in experiments which demonstrated that the BCTV L3 gene can complement a tomato golden mosaic virus AL3 mutant, and vice versa.
The geminivirus group is diverse and contains viruses which can be placed into three distinct subgroups on the basis of their genome organization and biological properties. However, most dicot-infecting geminiviruses possess AL1, AL2, and AL3 open reading frames. AL1 encodes the only vital protein that is absolutely required for replication, AL2 codes for a protein that transactivates the expression of virion sense promoters, and AL3 specifies a protein that enhances vital DNA replication. In the studies presented here, we examined the functional specificity of the AL1, AL2, and AL3 gene products of subgroup II and subgroup III geminiviruses. Surprisingly, we found that all viruses tested were able to produce a gene product which complemented the reduced DNA replication phenotype of a tomato golden mosaic virus (TGMV) AL3 mutant. We also found that all bipartite subgroup III viruses tested produced a protein that could transactivate the virion sense promoters of a TGMV AL2 mutant, although a subgroup II virus did not. In contrast, the replication activity of AL1 protein proved to be virus specific. The data are discussed with regard to the functions of these proteins in viral replication and their practical significance for the development of crop protection strategies.
Recombinant genomes derived from the Logan and CFH strains of the geminivirus beet curly top virus (BCTV) have been constructed and analyzed for pathogenicity on Nicotiana benthamiana and sugar beet (Beta vulgaris L.). Infectivity assays indicated that the latent period on N. benthamiana was primarily determined by a DNA fragment bearing the leftward open reading frames (ORFs) L1, L2, L3, and L4. Recombinants bearing leftward ORFs from the CFH strain were characterized as having a short latent period (mean = 6-11 days), while the reciprocal recombinants bearing leftward ORFs from the Logan strain had latent periods defined as long (mean = 16-22 days). Infectivity assays on sugar beet indicated that certain recombinant BCTV genomes exhibited novel pathogenic properties not common to either wild type strain, including the loss of systemic movement and replication competency, or asymptomatic systemic infection of sugar beet. The results indicate that N. benthamiana is a more permissive host than sugar beet with respect to heterologous combinations of BCTV genes, and that pathogenicity and virulence of BCTV in sugar beet requires the interaction of certain viral gene products and/or cis-elements that have coevolved in the same strain.
The infection of susceptible plant hosts by single-stranded DNA viruses in the geminivirus group depends on the interaction of host and viral factors for the replication of viral DNA, the expression of viral genes, and the movement of virus throughout the plant. This paper reports that two strains of the geminivirus, beet curly top virus (BCTV) differ in their ability to infect certain ecotypes of Arabidopsis thaliana. Symptoms appeared on susceptible plants approximately 2-3 weeks after inoculation with BCTV-Logan and after 10-15 days with BCTV-CFH. Symptoms were more severe in BCTV-CFH-infected plants and included leaf curling, the formation of stunted, deformed inflorescence structures and the accumulation of anthocyanin pigments in symptomatic tissues. Analysis of viral DNA accumulation indicated that symptom development and severity were correlated with the amount of viral DNA present in the plants. Viral DNA was undetectable in two ecotypes that were phenotypically resistant to BCTV-Logan. Studies of viral DNA replication in excised inflorescence pieces demonstrated that BCTV-Logan could replicate in tissues from these resistant ecotypes, suggesting that resistance was due to a block in viral movement. Genetic studies of these two ecotypes indicate that resistance is due to a single, recessive locus. This is the first example of a single resistance locus to any geminivirus. The identification of resistant and susceptible interactions between different BCTV strains and A. thaliana ecotypes provides an excellent model system for the genetic and molecular analysis of the interaction of a plant host with this important group of plant pathogens.
The monopartite DNA genome of beet curly top geminivirus (BCTV, strain Logan) contains four leftward, complementary sense open reading frames (ORFs) designated L1, L2, L3, and L4 and three rightward, virion sense ORFs designated R1, R2, and R3 (R1 encodes the coat protein). The R3 ORF has not been reported previously in the BCTV genome, and evidence for three functional virion sense genes on one genome component has not been presented before for any geminivirus. We investigated the functions of the virion sense ORFs by introducing mutations into each of them. We found that in Nicotiana benthamiana plants, BCTV genomes containing mutations in ORF R1 were not infectious, whereas an R3- mutant was very weakly infectious. The small proportion of plants infected by the R3- mutant remained asymptomatic and contained greatly reduced amounts of viral DNA. An R2- mutant was highly infectious but asymptomatic, and in infected plants it accumulated mostly the double-stranded DNA (dsDNA) replicative form in nearly wild-type amounts. All of the mutants replicated in tobacco protoplasts, although R1- and R2- mutants accumulated reduced amounts of genomic single-stranded DNA (ssDNA) relative to wild-type virus. In the case of R2- mutants, the reduction was large (approx. ninefold) and was accompanied by a similar increase in dsDNA levels. The results suggest that the R1 and R3 gene products are required for efficient movement of the virus in the infected plant, whereas the R2 gene product may be involved in the regulation of ssDNA vs dsDNA levels.
The genome of the geminivirus tomato golden mosaic virus (TGMV) consists of two DNA components, designated DNA A and DNA B. DNA A encodes AL1, the only viral protein required for DNA replication. AL1 protein interacts specifically with sequences in the common region that is conserved between the two genome components, near sequences involved in the transcription of complementary sense genes encoding BL1 protein and the AL1 protein itseft. In the experiments described here, we replaced the AL1 and BL1 open reading frames with the β-glacuronidase (GUS) reporter gene and used the gone replacement constructs to examine AL1 and BL1 gone expression in tobacco protoplasts. We found that expression of the GUS reporter in the AL1 replacement construct was reduced to background levels when transfactions included a plasmid expressing AL1 protein from the cauliflower mosaic virus 35S promoter, indicating that AL1 gene expression is autoregulated. Surprisingly, a similar repression of BL1 gene expression by AL1 protein was not observed. Plasmids expressing the TGMV AL2 or AL3 proteins had no significant effect on AL1 or BL1 gone expression. In the course of these studies, we showed for the first time that the product of the AL3 ORF alone is sufficient to complement the replication-deficient phenotype of a TGMV AL3 mutant. The results are discussed in light of the multiple activities of AL1 protein.
In addition to ss and ds genomic DNA, agroinoculation of Nicotiana benthamiana plants with the Logan strain of the geminivirus beet curly top virus (BCTV) consistently resulted in de novo production of subgenomic DNAs on initial passage. Single-stranded and dsDNA forms representing at least seven size classes (0.8 to 1.8 kb) of subgenomic DNA were observed in total DNA extracts from inoculated plants. Extracts from infected sugar beet and tomato contained variable but usually smaller amounts of subgenomic DNAs, suggesting that their production may be influenced by the host species. Restriction endonuclease mapping and partial nucleotide sequencing of three independent clones of a 1.5 kb size class indicated that this subgenomic DNA is produced from the standard viral genome by two separate deletion events. One deletion of 941 bp includes portions of the leftward open reading frames (ORFs) L1, L2 and L3, while the other deletion of 579 bp encompasses portions of the intergenic region and the rightward ORFs R1, R2 and R3. The data indicate that the 1.5 kb BCTV subgenomic DNA is a defective DNA that has retained cis-elements essential for replication.