Background
Viruses of the family
Geminiviridae have circular, single-stranded (ss) DNA genomes and are divided into four genera based upon genome arrangement, host range and insect vectors. The most numerous, and economically the most destructive, are the whitefly-transmitted geminiviruses that are included in the genus
Begomovirus[
1,
2]. Begomoviruses are transmitted by the whitefly
Bemisia tabaci and exclusively infect dicotyledonous plants. They have emerged everywhere in the world where environmental conditions support large whitefly populations, and have become a major constraint in the production of food and fiber crops such as cassava, tomato, cucurbits, pepper, beans and cotton [
3‐
5].
Chilli leaf curl disease (ChLCD) is an important factor limiting chilli production on the Indian subcontinent and is caused by begomoviruses [
6‐
8]. Symptoms of the disease are severe leaf curl with cup-shaped, upward curling leaves, yellowing, and stunted plant growth. Previously chilli leaf curl betasatellite (ChLCB) has been identified in a large collection of chilli samples with leaf curl symptoms from all over the Pakistan [
9]. A single species of betasatellite (ChLCB) was found associated with isolates showing geographical segregation and are similar to that reported earlier [
6]. Chilli peppers often show symptoms similar to tomato leaf curl disease, such as yellowing, leaf curling, a reduction in leaf size and stunting. Since chilli and tomato crops overlap in the field, it is likely that chilli peppers may become infected with tomato begomoviruses. The disease was experimentally transmitted from infected to healthy chilli and tomato seedlings by the whitefly
Bemisia tabaci[
10,
11]. Inoculated chilli plants developed typical symptoms of the disease. However, the inoculated tomato plants developed severe leaf curl symptoms similar to those of leaf curl disease of tomato caused by
Tomato leaf curl New Delhi virus (ToLCNDV) [
11,
12]. Analysis of a large collection of chilli samples from Pakistan showed that diverse begomoviruses may infect chillies [
7]. Recently another distinct begomovirus,
Pepper leaf curl Lahore (PepLCLV), has been identified in chilli in Pakistan although the infectivity of the virus to chillies was not established experimentally [
13].
Here we have characterised a further isolate of PepLCLV from chilli and have investigated its interaction with betasatellites and the DNA B component of ToLCNDV. We show that the virus requires the DNA B of ToLCNDV to infect plants and induce disease symptoms but its interaction with ChLCB is poor.
Discussion
The geminiviruses are a rapidly emerging group of plant viruses, which can be attributed to various factors, including increased insect vector populations and the presence of alternative hosts. Though it was speculated that geminiviruses had the capacity to evolve rapidly in response to changes in their environment (such as alterations in cropping systems and/or population dynamics of the insect vector), there are few studies documenting geminivirus evolution. The success in the Old World of begomoviruses that associate with betasatellites appears to be due to the ability of betasatellites to be replicated by several distinct begomoviruses [
27]. Thus, the ability of begomoviruses to interact with diverse betasatellites, the mobilization of begomovirus components from alternate hosts and recombination among begomoviruses or associated satellites has been documented as the major driving force in rapid emergence and resistance breakdown by begomovirus-betasatellite complexes [
5]. There are numerous reports documenting mixed infection of geminiviruses. For example, in cassava, infection by two distinct begomovirus species resulted in a severe disease due to synergism [
28]. Interestingly, ToLCNDV, a bipartite begomovirus has been consistently detected in several hosts in the Indian subcontinent and suggest that the virus has flexibility in its interaction with other begomovirus components that may help virus to expand host range. We have previously shown that ToLCNDV interacts with ChLCB under field conditions that result in severe symptoms [
12]. Another study showed that
Tomato leaf curl Gujarat virus captured the DNA B component of ToLCNDV, resulting in a virus capable of inducing more severe disease symptoms [
29,
30]. Our recent analysis has shown that ToLCGV may exist without a DNA B in some weeds (M. Mubin, manuscript in preparation) which suggests that ToLCGV was mobilized from a weed into tomato upon its interaction with the DNA B of ToLCNDV. Thus, it appears that component capture during mixed infection, probably in weed hosts, may result in viruses with enhanced virulence to crop plants.
Geminivirus genomes replicate by a rolling circle mechanism which is initiated by the virus-encoded replication-associated protein (Rep) [
31]. Rep is a sequence specific DNA binding protein which recognises and binds to repeated sequences, known as iterons, in the intergenic region immediately upstream of a hairpin structure that contains the ubiquitous (for geminiviruses) nonanucleotide sequence (TAATATTAC). Rep then initiates replication by nicking in the nonanucleotide sequence. The DNA A and DNA B components of bipartite begomoviruses have the same iteron sequences, thereby ensuring that the DNA A-encoded Rep may initiate replication of both components; maintaining the integrity of the split genome. However, mutational analyses and sequencing of field isolates suggests that begomoviruses may tolerate some sequence variation in iteron sequences without deleterious effects on Rep recognition. Here we show that the predicted iteron sequence of PepLCLV (GGGGAC) differs at two base positions from the iteron sequence of ToLCNDV (GGTGTC; Figure
5). Thus, it appears that the first two or three bases may be more important in iteron recognition by Rep. An interesting finding of the study is that PepLCLV isolate examined here has no, or at least only limited, ability to trans-replicate betasatellites. This contrasts with the results of Tahir
et al. [
13], who showed the association of ChLCB with another isolate of PepLCLV from Pakistan. Our analysis suggests that the putative Rep protein of the virus has an N-terminal leader sequence (Figure
2) which may be important in the inability of the virus to trans-replicate betasatellites. However, in the absence of any evidence this remains a hypothesis that requires confirmation by mutagenesis. The other possibility may be that there are natural variants of PepLCLV that lack leader peptide in the Rep protein that may be able interact with betasatellites. Since our efforts to characterize begomoviruses in recent samples is not exhaustive, it will not be surprising if another begomovirus capable of interaction with betasatellite is present in naturally infected chillies.
Chilli leaf curl disease (ChLCD) is an important factor limiting chilli production across Pakistan and India. The genomes of begomoviruses are either bipartite (with two genomic components known as DNA A and DNA B) monopartite (with a genome consisting of only a homolog of the DNA A component) or monopartite associated with a symptom determining satellite (collectively known as betasatellites). All three types have been previously identified in chillies [
9,
10,
12]. The full-length genome of a begomovirus associated with ChLCD originating from the Punjab (Pakistan), PGL1 was cloned and shown to consist of 2747 nucleotides. Sequence comparisons showed that the genome had the highest sequence identity (99%) with Pepper leaf curl Lahore virus (PepLCLV-[PK:Lah:04]) indicating that it represents an isolate of PepLCLV based on the 89% species demarcation threshold for begomoviruses [
23].
Agrobacterium-mediated inoculation of the clone to
N. benthamiana induced only very mild symptoms. Inoculation of the clone with a betasatellite previously isolated from a ChLCD affected plant gave similarly mild symptoms and virus levels were not detectable by Southern hybridization. However, inoculation with the DNA B component of ToLCNDV induced symptoms typical of ChLCD in
N. benthamiana,
N. tabacum and
C. annum. These results suggest that the virus characterised here may be bipartite. A surprising finding is that the levels of viral DNA were lower in plants inoculated with PepLCBDV, ToLCNDV DNA B and ChLCB in comparison to plants inoculated in the absence of ChLCB (Figure
5). This is the first experimental demonstration of infectivity for a bipartite begomovirus causing ChLCD.
These results presented here demonstrate that ChLCD in Pakistan may be caused by a bipartite variant of PepLCLV, which is associated with a DNA B component related to ToLCNDV DNA B, in addition to a monopartite variant of PepLCLV which associates with a betasatellite (ChLCB) [
13]. The difference in the ability of the two PepLCLV isolates to interact with ChLCB may be due to the presence, in the isolate characterised here, of additional N-terminal amino acid sequences of Rep. However, since we as yet do not fully understand the mechanism of interaction of begomovirus-encoded Rep with betasatellites to initiate satellite replication (betasatellites lack the iteron sequences encoded by their helper viruses [
27]), this will require experimental confirmation. Despite the differences in the predicted iteron sequences of PepLCLV and ToLCNDV, PepLCLV has the ability to trans-replicate ToLCNDV DNA B and induce ChLCD in experimental hosts and chilli. This is the first experimental demonstration of Koch's postulates using cloned viral DNA components for a bipartite begomovirus causing ChLCD. The bipartite PepLCV has some residual ability to interact with betasatellites although the presence in plants of both a DNA B and a betasatellites appears to reduce virus titre and symptom severity, suggestive of interference. The nature of this interference will be the focus of our future studies since this may provide a novel mechanism of obtaining resistance to the viruses causing ChLCD. The complex nature of ChLCD across the Indian sub-continent, which has been shown to be caused by several bipartite and monopartite, betasatellite-associated begomoviruses will be a challenge for the development of resistant varieties either by conventional or non-conventional means. The high yield losses resulting from ChLCD are threatening chilli cultivation and are forcing farmers in some areas to grow other crops.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
MS performed the experiments. MS, SA, YS, RWB and SM were involved in data analysis. SA, YS, RWB and SM provided overall direction and experimental design. RWB and SM wrote the manuscript. All authors read and approved the final manuscript.