Iterons Homologous to Helper Geminiviruses Are Essential for Efficient Replication of Betasatellites

Abstract

Betasatellites associated with geminiviruses can be replicated promiscuously by distinct geminiviruses but exhibit a preference for cognate helper viruses. However, the cis elements responsible for betasatellite origin recognition have not been characterized. In this study, we identified an iteron-like repeated sequence motif, 5'-GAGGACC-3', in a tobacco curly shoot betasatellite (TbCSB) associated with tobacco curly shoot virus (TbCSV). Competitive DNA binding assays revealed that two core repeats (5'-GGACC-3') are required for specific binding to TbCSV Rep; TbCSB iteron mutants accumulated to greatly reduced levels and lost the cognate helper-mediated replication preference. Interestingly, TbCSV also contains identical repeated sequences that are essential for specific Rep binding and in vivo replication. In order to gain insight into the mechanism by which TbCSB has acquired the cognate iterons, we performed a SELEX (systematic evolution of ligands by exponential enrichment) assay to identify the high-affinity Rep binding ligands from a large pool of randomized sequences. Analysis of SELEX winners showed that all of the sequences contained at least one core iteron-like motif, suggesting that TbCSB has evolved to contain cognate iterons for high-affinity Rep binding. Further analyses of various betasatellite sequences revealed a region upstream of the satellite conserved region replete with iterative sequence motifs, including species-specific repeats and a general repeat (5'-GGTAAAT-3'). Remarkably, the species-specific repeats in many betasatellites are homologous to those in their respective cognate helper begomoviruses, whereas the general repeat is widespread in most of the betasatellite molecules analyzed. These data, taken together, suggest that many betasatellites have evolved to acquire homologous iteron-like sequences for efficient replication mediated by cognate helper viruses.IMPORTANCE The geminivirus-encoded replication initiator protein (Rep) binds to repeated sequence elements (also known as iterons) in the origin of replication that serve as essential cis elements for specific viral replication. Betasatellites associated with begomoviruses can be replicated by cognate or noncognate helper viruses, but the cis elements responsible for betasatellite origin recognition have not been characterized. Using a betasatellite (TbCSB) associated with tobacco curly shoot virus (TbCSV) as a model, we identify two tandem repeats (iterons) in the Rep-binding motif (RBM) that are required for specific Rep binding and efficient replication, and we show that identical iteron sequences present in TbCSV are also necessary for Rep binding and the replication of helper viruses. Extensive analysis of begomovirus/betasatellite sequences shows that many betasatellites contain iteron-like elements homologous to those of their respective cognate helper begomoviruses. Our data suggest that many betasatellites have evolved to acquire homologous iteron-like sequences for efficient replication mediated by cognate helper viruses.

Keywords: betasatellite; geminivirus; iterons; replication; tobacco curly shoot virus.

FIG 1

TbCSB iteron deletion mutants are defective in replication and specific Rep binding. (A) Schematic representation of the genome organization of TbCSB isolate Y35. The positions of the adenine-rich (A-rich) region and the βC1 gene are indicated, and the stem-loop structure within the satellite conserved region (SCR) is diagrammed. The putative iteron sequences in the RBM are shown in boldface, and the spacer sequence is italicized. (B and C) DNA gel blot analyses of viral and satellite DNAs extracted from N. benthamiana plant leaf tissues (B) or leaf discs (C) agroinoculated with the helper virus (Y35) either alone or in the presence of the wild-type (WT) betasatellite (Y35β) or the indicated mutated betasatellite at 15 dpi (B) or 6 dpi (C). Total-DNA samples were blotted with helper- and betasatellite-specific probes and were also stained with ethidium bromide to show loading controls. The levels of betasatellite DNA accumulation in two independent plants were quantified and normalized against the value for the WT, which was arbitrarily set at 100%. (D) Symptoms of infected N. benthamiana plants and a mock-treated control plant at 15 dpi. Images from a close-up view of the upper surfaces (center panels) and lower surfaces (right panels) of infected leaves are shown to indicate the curled-leaf symptoms. (E) EMSA of TbCSV-[Y35] Rep–Y35β RBM binding. Binding reaction mixtures contained recombinant MBP or MBP-Y35 Rep protein purified from E. coli cells and an Alexa Fluor 680-labeled Y35β RBM probe in the absence of a competitor or in the presence of 10-, 100-, and 500-fold molar excesses of competitor DNAs, as indicated above the panels. The binding reaction mixtures were resolved on 1% agarose gels and the DNA bands were visualized using an Odyssey infrared imaging system to compare relative binding efficiencies. The shifted bands were quantified and normalized against the background level for the reaction without protein (lane 1; percentage of binding set at 0%) and the binding band for the reaction without any competitors (lane 3; percentage of binding set at 100%). The experiment was repeated three times, and similar data were obtained. One representative gel with associated quantification values is shown.

FIG 2

Identification of core iteron sequences required for specific Rep binding and betasatellite replication. (A) Sequences of the TbCSB wild-type (WT) iteron and mutant derivatives. The mutated nucleotides are underlined, and the long dash indicates the deletion of five nucleotides. (B and C) Analysis of the binding affinities of TbCSB (Y35β) RBM mutants by competitive EMSA. The indicated competitor DNAs were added in 10-, 100-, and 500-fold molar excesses relative to the probe, and the bound probe were quantified as described for Fig. 1E. The experiment was repeated three times, and one representative gel with associated quantification values is shown. (D) DNA gel blot analysis of satellite DNAs. Total-DNA samples were extracted at 15 dpi from N. benthamiana plants agroinoculated with the helper virus together with the WT betasatellite or the indicated mutant and were hybridized to a betasatellite-specific probe. Ethidium bromide-stained total DNAs are shown to indicate equal loading. The relative accumulation levels of betasatellite or mutant DNAs are shown.

FIG 3

Rep binds to cognate iterons with higher affinity than to noncognate iterons. (A) Comparison of RBM sequences in TbCSB-[Y35] and TYLCCNB-[Y10]. (B and C) EMSA analyses of the binding affinities of Rep to cognate and noncognate RBMs, as well as to hybrid RBM mutants containing heterologous iteron sequences. The binding reaction mixtures contained either the MBP-Y35 Rep and the Y35β RBM probe (B) or the MBP-Y10 Rep and the Y10β RBM probe (C) in the presence of the indicated unlabeled competitor dsDNAs added in 10-, 100-, and 500-fold molar excesses relative to the probe. The experiment was repeated three times, and one representative gel with associated quantification values is shown.

FIG 4

Homologous iterons in TbCSV-[Y35] are required for specific Rep binding and virus replication. (A) Iteron organization in TbCSV-[Y35] and TbCSB-[Y35]. Iteron sequences are shown in boldface, and the relative orientations are indicated by arrows. The putative TATA box of the leftward promoter is boxed. (B) Identification of specific binding sites in the TbCSV-[Y35] RBM by EMSA. The binding reaction mixtures contained the MBP-Y35 Rep and the labeled TbCSV-[Y35] RBM in the presence of excess amounts of competitor dsDNAs corresponding to the WT RBM or the mutant with the indicated iteron(s) deleted. The bound probes were quantified as described for Fig. 1E. (C) DNA gel blot analysis of viral DNA extracted at 15 dpi from N. benthamiana plants agroinoculated with WT TbCSV-[Y35] or iteron deletion mutants. Total DNAs were stained with ethidium bromide to serve as loading controls. The accumulation levels of viral DNA were quantified and are expressed as percentages of the level for the WT virus. (D) Symptoms of infected N. benthamiana plants and a mock-treated control plant at 30 dpi.

FIG 5

Rep binding affinity and replication of TbCSB-[Y35] SELEX winners. (A) Analysis of the Rep binding affinities of the Y35β RBM SELEX winners by competitive EMSA. The competitor DNAs corresponding to each of the SELEX winner sequences were added in 10-, 100-, and 500-fold molar excesses relative to the labeled Y35β RBM probe, and the bound probes were quantified as described for Fig. 1E. (B and C) DNA gel blot analyses of viral and satellite DNAs extracted from N. benthamiana plant tissues (B) or leaf discs (C) agroinoculated with a helper virus either alone or in the presence of the wild-type (WT) betasatellite or the indicated SELEX mutant. Total-DNA samples were isolated at 15 dpi (B) or 6 dpi (C) and were stained with ethidium bromide to indicate equal loading. The accumulation levels of betasatellite DNA were quantified and are expressed as percentages of the level for the WT satellite. (D) Symptoms of infected N. benthamiana plants and a mock-treated control plant at 15 dpi.

FIG 6

Pairwise comparison of the organization of repeated sequence motifs in helper begomoviruses and their associated betasatellites. Representative begomoviruses with repetitive sequence motifs homologous to (A) or distinct from (B) those of their betasatellites are shown. The viral isolate designation is followed by the GenBank accession number and then the consensus repeated sequence. The number in parentheses preceding each sequence designates the nucleotide coordinate for the starting repeat; the number of spacing bases, set off by periods, is given within the sequence. Repeated elements are shown in boldface, and lowercase letters represent bases that do not match the consensus. The positions and orientations of the species-specific repeats and the general repeats are indicated by arrows under the sequences and thick black arrows within the sequences, respectively. The putative TATA box sequence of the leftward promoter in each begomovirus, and in some cases also in betasatellites, is italicized. (C) Genome structure of betasatellites. The positions of the IRS, the A-rich region, and the βC1 gene are indicated, and the putative hairpin structure within the SCR is diagrammed. (D and E) Schematic illustrations of the specific repeated sequences in begomoviruses (D) and betasatellites (E). The arrows show the orientations and positions of repeated sequences. The positions of the putative TATA box in helpers are indicated by solid boxes, and the less-conserved TATA-like elements in betasatellites are indicated by dashed boxes.