Methylation is Involved in Arabidopsis thaliana Recovery from Geminivirus and Prevention of Geminivirus Seed Transmission
Loading...
Date
2008-06
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
The Ohio State University
Abstract
While most plant viruses are RNA viruses that replicate in the cytoplasm of the host cells, members of the Geminiviridae family are unique in that they are DNA plant viruses. Geminiviruses possess either monopartite or bipartite circular single-stranded DNA (ssDNA) genomes of 2.5-3 kb, that replicate in the nucleus through double-stranded DNA (dsDNA) intermediates (Bisaro, 1996). The genome components of bipartite geminiviruses are almost entirely different in sequence except for a 200-250 bp conserved region. The geminiviruses used in my study are the bipartite Cabbage leaf curl virus (CaLCuV) of the genus Begomovirus, and the monopartite Beet curly top virus (BCTV), of the genus Curtovirus (Bisaro, 1996). Geminiviruses are typically transmitted from plant to plant via insect vectors such as the whitefly (Begomovirus) or leafhopper (Curtovirus) (Bisaro, 1996). They infect several plant species and for my studies I focused on the model plant host Arabidopsis thaliana. Geminiviruses replicate predominantly through rolling circle replication and rely on the host replication machinery to accomplish amplification and transcription (Bisaro, 1996). The AL2 protein (Begomoviruses) is a transcription factor that is responsible for the expression of late viral genes (Sunter and Bisaro, 1997). For Curtoviruses, the related L2 protein is not a transcription factor, but acts as a suppressor of plant silencing pathways, as does AL2. One mechanism of silencing involves AL2 and L2 interacting with adenosine kinase (ADK), which is responsible for the phosphorylation of adenosine to AMP (Fig.1). ADK activity is also required for efficient operation of the methyl cycle (Wang et. al., 2003). Plants use two main RNA silencing pathways for defense against geminiviruses. Post transcriptional gene silencing (PTGS) involves small interfering RNAs (siRNAs) to repress gene expression by directing the cleavage of homologous mRNAs in the cytoplasm (Vaucheret, 2008) (Fig. 2). Plants deficient in PTGS pathways are usually more susceptible to infection by plant viruses. PTGS is induced by dsRNA, and two classes of proteins, Argonautes (AGO) and Dicer-like proteins (DCL), are key players in the pathway. In PTGS in Arabidopsis, DCL2 or DCL4 process dsRNA into 21-22 nucleotide (nt) siRNA duplexes, and one strand of the siRNA in incorporated into an RNA induced silencing complex (RISC) which contains an AGO1 protein which is also an endonuclease (slicer). The siRNA programs RISC to cleave mRNAs with complementary sequence (Vaucheret, 2008) (Fig. 2). The second pathway is transcriptional gene-silencing (TGS), which results in the methylation of target DNA sequences in the nucleus. This pathway uses similar RNA silencing machinery, except in this case dsRNA is processed by DCL3 into slighter larger siRNA (~24 nt) which enter a RISC complex that contains AGO4. The function of this RISC complex is to recruit methyltransferase proteins to specific DNA sequences (Qi et. al., 2006). Methylation of DNA and associated histone proteins is strongly correlated with reduced transcription. One of the mutants used in this study is deficient in ago4, which is used in Arabidopsis siRNA production to suppress viral transcription.
Methylation is directed by siRNAs in cooperation with the ADK methyl cycle to methylate histones and viral DNA with a variety of methyltransferases (Wang et. al., 2005) (Fig. 3). The process of viral DNA and histone methylation is carried out by many proteins, but this study will only be testing a handful of them. AGO4 has been described above. PolIV is involved in siRNA directed de novo cytosine methylation in Arabidopsis. Loss of PolIV brings about the absence of most cytosine methylation thus suggesting that it is integral in producing the siRNAs used to carry out genome methylation (Pikaard, et. al., 2005). DRB3 is a dsRNA binding protein, which interacts with DCL proteins and likely facilitates the loading of siRNAs into RISC complexes (Hiraguri, 2005). DRM1/2 and CMT3 are both cytosine methyl transferases and may be partially redundant. Their complete removal with a ddc mutant removes all de novo DNA methylation at non-CG sites while CG sites are controlled by MET1 (methyltransferase1) which is not used in this study. KYP2 (kryptonite2) is a histone 3 lysine 9 (H3K9) methyltransferase and is also involved in heterochromatin methylation.
It is well-established that the PTGS pathway acts to degrade RNA virus transcripts and genomes, as well as the transcripts of geminiviruses. However, the Bisaro laboratory has recently shown that methylation of the geminivirus genomes, leading to transcriptional gene silencing, is an important defense against these DNA viruses (Bisaro et al., 2008). Methylation-deficient mutant plants show enhanced susceptibility to geminivirus infection and, and viral genomes obtained from these mutants show reduced methylation (Bisaro, 2008). Furthermore, that geminivirus AL2 and L2 proteins can block the the methyl cycle suggests that heterochromatin methylation is integral in the suppression of the viral genome.
We wished to ask further questions about methylation defense in Arabidopsis. One hypothesis tested is that the methylation pathway is required for plants to recover from geminivirus infection. Previous work has shown that Nicotiana benthamiana plants can recover from infection with BCTV L2 mutant viruses. Recovered tissue is nearly symptom-free and contains low levels of virus (Hormuzdi and Bisaro, 1995). Moreover, L2 mutant-infected plants show enhanced ADK activity, suggesting the involvement of L2 in the suppression of ADK and implicating the methylation pathway in recovery (Wang et. al., 2003). Thus one goal of this study was to observe the behavior of two BCTV L2 mutant viruses on various methylation mutant Arabidopsis plants. It was predicted that wild-type plants would recover because of their ability to properly suppress the virus. However, methylation deficient plants (ago4, cmt3, kyp2, and ddc) were predicted to be unable to suppress the virus and would therefore still exhibit symptomatic tissue with the new plant growth (i.e. no recovery). We also predicted that virus extracted from recovered tissue would be hypermethylated. Conversely, viral DNA from tissue showing symptoms was predicted to have considerably lower methylation.
A second hypothesis is that plant methylation pathways are needed to prevent geminiviruses from invading the meristem. Most viruses are excluded from the meristem, and it has been shown that in the case of certain RNA viruses, meristem exclusion requires the PTGS pathway (Baulcombe and Martin-Hernandez, 2008). It is possible that for a DNA virus, the methylation pathway might be involved in this type of defense.
It is advantageous for plants to keep viral genomes out of the meristem so that there can be no potential infection of its progeny by virus, which might otherwise invade the seed stock. It must also be noted that most viruses are not seed transmitted because hosts typically have strong defense pathways involved in preventing any form of meristem invasion. Meristem cultures are even used to produce virus free plants (Manganaris, 2003). Typically, vegetative meristems prevent the virus from entering into its undifferentiated tissue and infection only spreads to developmentally older areas. The removal of these infected areas causes new growth to occur out of the vegetative meristem resulting in non-symptomatic tissue, or recovery. While this tissue is by no means virus free, the viral levels are low enough as to reduce symptoms. The same applies to floral meristems. Methylation mutants infected with geminiviruses exhibit stunting as a symptom. Stunting is a result of the virus getting close to the meristem so it can be thought that the virus in these plants might also be getting into the floral meristems and thus into the seeds. It has been discovered that the Tobbaco rattle virus encodes a silencing suppressor, 16K, which suppresses plant silencing enough to allow invasion of the meristem in N. benthamiana (Baulcombe and Martin-Hernandez, 2008). This study will assess the ability of BCTV and CaLCuV to invade the meristem of Arabidopsis by checking for viral DNA in the progeny of infected plants. If the methylation pathway is involved, it is expected that wild-type plants will be able to suppress any attempts at seed transmission/meristem invasion, but methylation mutants should allow some of the virus to make its way into the seed stock of infected plants. Several mutants will be analyzed in this context, including mutants deficient in ago4, polymerase IV (polIV), or double-stranded RNA binding protein 3 (drb3).
Description
Keywords
Virology, Plant Pathology