Introduction

Geminiviruses are known to be the cause of several devastating diseases of important crop plants. Papaya and tomato are two such crops throughout the world, which are severely damaged by the destructive diseases caused by geminiviruses [1–3]. Considering the significant impact of these diseases on commercial as well as socio-economic aspect of crop production in case of tomato and papaya, development of an effective control strategy against the geminiviruses is of paramount importance. Many physical, chemical, and biological strategies, including traditional as well as transgenic, are being practiced now-a-days to control the leaf curl disease but most of these management practices, still have several drawbacks [4, 5]. Cultural practices such as rouging, intercropping, avoidance, use of barriers, crop residue disposal, among others, are recommended, but they should be combined with the use of insecticides and/or resistant varieties in order to be effective [6, 7]. Intercropping may provide a useful approach to control geminivirus disease spread [8, 9]. Planting an insusceptible crop such as cucumber or pumpkin with tomato may delay PYMV-TT infection in tomato. Intercropping systems would affect several factors that may be critical to disease management. In the presence of a variety of host plants, whitefly behavior (feeding, rate of movement between plants) becomes more sporadic [8]. Shorter feeding times may lower transmission rates and reduce disease incidence in the affected crop [8]. It has also been shown that intercrops may result in a decrease in whitefly populations in Cassava mosaic virus (CMV) epidemics [9]. Control practices have focused primarily on controlling the vector by contact or systemic high toxicity insecticides with the concomitant problems of development of pesticide resistant forms, low cost benefit ration, and environmental concerns. Insecticides are the most often employed and most expensive approach to geminivirus management. Several insecticides, oils, and soaps are used to reduce whitefly populations and incidence of infected plants both in transplant-production houses and in the field [10]. In some places and seasons, the use of insecticides can reduce incidence of infected plants to economically satisfactory levels. However, under some circumstances, e.g., when there is a large source of viruliferous whiteflies nearby, insecticides have been less successful. Focused cultural methods of disease suppression should be supplemented with regulated chemical use. The difficulty experienced with reducing vector numbers using chemical sprays is explained by the behavior (feeding, ovipositing, and mating) of the adult, nymphal, and larval forms of the whitefly. Nymphs and older larvae are found in the lower regions of the plant canopy and insecticide sprays (including natural or synthetic soaps, oils, and detergents) may not adequately access these areas. Systemic insecticides may have a select advantage by reducing vector numbers irrespective of developmental stage as all regions of the plant are accessed. Systemic insecticides only work to reduce disease if applied before onset of infection. Neonicotinoids (imidacloprid) and nonneurotoxic insect growth regulators (buprofezin and pyriproxyfen) have been used to control Bemisia tabaci in agronomic and horticultural production systems [11]. However, intensive and unregulated use of insecticides has resulted in insecticide resistance or reduced susceptibility of the vector [11] and may suggest that the efficacy of chemical control may be temporary.

Antignus et al. demonstrated that TYLCV incidence in tomato plants grown under the UV-absorbing sheets was only 1% compared with approximately 80% in the uncovered control [12]. The various other traditional methods include the use of sanitation, scouting, as well as biological methods such as cross breeding. Screening of resistant germ plasm has also not given any positive results or resistance to the virus. Thus transgenic approach is come in existence. The development of transformation techniques in the 1980s opened new possibilities for the generation and evaluation of sources of resistance outside of traditional breeding methodologies. In 1986, Powell-Abel et al. demonstrated that tobacco plants transformed with the coat protein of tobacco mosaic virus (TMV) showed resistance to TMV infection. Shortly thereafter, it was shown that the expression of the coat protein genes of alfalfa mosaic, tobacco streak, and tobacco rattle viruses in tobacco conferred resistance to infection by those viruses [13]. Nonstructural genes (replicase and movement protein genes) have also been used successfully to engineer resistance, primarily against RNA viruses [14, 15]. There are fewer reports of engineered resistance to DNA viruses, including the geminiviruses. Two types of resistance have been used previously: first, resistance derived from several species of host, which is mostly multigenic; and second, several different genes from geminiviruses (pathogen-derived resistance, PDR), when transformed into plants, have been shown to be useful sources of resistance [7]. Several approaches have been developed to engineer geminivirus resistance in different plant species. But these approaches do not provide a long-term resistance to the virus. Therefore, far most of the reported transgenic lines failed to confer high level of geminivirus resistant or chosen resistant strategy was effective only in small subset of transgenic lines [16].

During the last decade, our knowledge repertoire of RNA-mediated functions has been greatly increased with the discovery of small noncoding RNAs which play a central part in a process called RNA silencing. Ironically, the very important phenomenon of co-suppression has recently been recognized as a manifestation of RNA interference (RNAi), an endogenous pathway for negative post-transcriptional regulation. RNAi has revolutionized the possibilities for creating custom “knock-downs” of gene activity. RNAi operates in both plants and animals, and uses double stranded RNA (dsRNA) as a trigger that targets homologous mRNAs for degradation or inhibiting its transcription or translation [17]. RNA-mediated resistance can confer protection against a wide variety of RNA and DNA viruses. Compared to coat protein-mediated resistance, RNA-mediated protection provides a higher level of virus resistance [15]. However, this type of resistance appears to be effective only against viruses with closely related sequences [4], which is a shortcoming for virus resistance in crops. The RNA-mediated resistance approach is also effective against DNA viruses, e.g., geminivirus [18]. This type of RNA-mediated resistance is governed by the PTGS phenomenon which is a conserved mechanism for mRNA regulation in plants, animals, and fungi. PTGS controls numerous developmental processes and is required for innate immunity regulating virus accumulation. There are two hallmarks in PTGS. First, silencing of target mRNAs occurs in the cytoplasm. Second, small RNA molecules (~21–25 nt) are generated from the silenced target mRNAs. Based on the differences in their biogenesis, two types of small RNAs, siRNA and miRNA, have been identified.

The increasing impact of geminivirus infection on crop yields has emphasized the importance of developing efficient and sustainable geminivirus resistance in the geminivirus-affected crops. Since attempts to obtain robust PDR to geminiviruses have not been successful, development of geminivirus resistant plants is considered a major challenge [19, 20]. In this context, it is worth considering the merits of the siRNA methodology for achieving desired level of efficient protection from viruses in these crop plants.

RNA interference (RNAi) technology is a powerful methodology recently developed for the specific knockout of targeted genes. It is a process in which the double stranded short interfering RNA (siRNA) induces the post-transcriptional degradation of homologous transcripts [21]. Short interfering RNAs (siRNA) are 21–22 nucleotide long RNA duplexes, known to silence gene expression by specific cleavage of target RNA which it is homologous to [22]. DNA viruses are considered to be good targets of RNA silencing and so it was natural to think of geminiviruses, infecting papaya and tomato crops also as putative targets of RNA silencing using siRNA. RNA silencing or post-transcriptional gene silencing (PTGS) is a highly homology-dependent process that is known to be triggered by double stranded RNA [23]. Studies have revealed that geminiviruses which replicates in nucleus can induce PTGS and become the target for it. RNAi has a strong potential to reduce the infection of geminiviruses [24]. This sequence specific siRNA-mediated gene silencing can be observed by either expressing or introducing 21–25 base transcripts capable of forming duplex directly into the plant [25]. Chemically synthesized siRNA duplexes mediate effective target RNA cleavage when introduced into the plants [26]. However, most literature available so far pertains to gene interference caused by siRNA sequences for silencing of selected or specific single target genes. An RNA interference construct to silence the sequence region of AC1 viral gene was reported to generate highly resistant transgenic common bean plants [27], in which they have used partial sequence from AC1 gene of Bean Golden Mosaic Virus (BGMV) genome. Similarly, Vanderschuren et al. [16] have engineered transgenic cassava with resistance to African cassava mosaic virus (ACMV), by expressing ACMV AC1 homologous hairpin double stranded RNAs. It is also reported that Tomato yellow Leaf Curl Sardinia virus (TYLCSV) could overcome transgene-mediated RNA silencing of replicase and nested C4 [18]. RNAi constructs to silence the AC1 viral genes of BGMV were successfully demonstrated in common bean Phaseolus vulgaris, and the work was reported as first transgenic geminivirus resistant plant in the field [27]. This being a success story, our observation is that in case of BGMV isolates the authors could not find any significant variation in the 421 bp fragment used for intron hairpin construct and have reported that all the isolates were having 100% similarity or variation with one point polymorphism. This supports our strategy which we are using in order to design siRNA as above reports on silencing the geminivirus gave us this clue that in order to have a generic and sustainable resistance against a given virus we have to take into account not only the single gene but also all the given isolates of that given virus. Furthermore, after multiple sequence alignment and finding a region in the gene which is conserved in all the isolates of a given virus if used to design siRNA, may give a broad generic and sustainable resistance. To our knowledge, there is no broad strategy, for geminiviruses in general, and of tomato and papaya in particular, that is based on siRNA technology.

Here, we present the steps involved in this strategy i.e., in silico search for siRNA sequence highly homologous to the most conserved region of the genomes of geminiviruses infecting papaya and tomato, to counter disease caused by them. Since, the strategy is highly homology dependent, we are emphasizing on the conserved regions of the geminiviruses genome, as genetic variability has been reported in case of several geminiviruses [28], and there may be more than one isolate/strain of a given geminivirus infecting papaya and tomato having genetic variability [29] so that designing siRNA against one isolate may not work against another infecting the same crop. The siRNA should ideally be designed from the conserved regions of geminiviruses. In this study, we have taken into account the designing of siRNA from two essential viral genes, namely, the coat protein gene (AV1) and replicase gene (AC1) of papaya leaf curl virus (PLCV) and tomato leaf curl virus (TLCV) isolates. Our preliminary sequence analysis of TLCV genomes of various isolates revealed that TLCV isolates from Northern India share a tight homology and can be grouped in a separate subsection in order to design tailor made siRNA against TLCV Northern India. To have a stable and foolproof resistance against the virus, this exercise of searching conserved regions in PLCV, TLCV and TLCV (Northern India) were necessary to design siRNA against the respective viruses.

Materials and methods

Identification/retrieval of target viral coat protein and replicase gene sequences

Plant Virus Online database i.e., Description of Plant Virus (DPV; www.dpvweb.net) and NCBI (www.ncbi.nlm.nih.gov) nucleotide database were used to retrieve sequences of different isolates of geminiviruses infecting papaya (PLCV) and tomato (TLCV) from all over the world including isolates of tomato leaf curl viruses from northern region of India (TLCV, Northern India).

Alignment of sequences

The retrieved sequences were aligned and a phylogenetic tree was created for each using CLCBiO sequence viewer version 5.1.1 (Free Workbench bioinformatics software; www.clcbio.com). Multiple sequence alignment was done to choose most conserved short stretches from coat protein gene or replicase gene sequences among viral isolates. The most conserved region of 21–25 nt from coat protein gene (AV1) and replicase gene (AC1) of PLCV, TLCV, and TLCV (Northern India) were selected to further design putative siRNA sequences to be used for gene silencing of geminiviruses, infecting papaya and tomato.

Selection of siRNA

Selection of siRNA sequences depends on several criteria [30, 31], which we have taken into account while selecting the putative siRNAs. In general, we have targeted the most conserved regions of the coding sequence of both, coat protein and replicase genes, of PLCV, TLCV, and TLCV (Northern India), and designed putative siRNAs following the guidelines given at Ambion siRNA design guidelines (www.ambion.com/techlib/tb/tb_506). Thus, designed siRNAs were further cross checked by subjecting the complete AV1 and AC1 gene sequence to siRNA target finder (Ambion; www.ambion.com/techlib/misc/siRNA_finder.html) and then best ones were chosen accordingly.

Analyzing putative siRNAs by BLAST

Only those conserved regions were selected to design putative siRNAs, whose sequences are not even remotely similar to other plant gene sequence. A cross homology search using NCBI, BLASTn (www.ncbi.nlm.nih.gov/BLAST) was also made on the selected 21–25 nt sequence to screen them (if at all, by chance the selected siRNA bears homology to any plant gene) to avoid gene silencing of any plant gene, i.e. off target cleavage.

Results and discussion

The sequences of different isolates of geminiviruses infecting papaya (PLCV) and tomato (TLCV & TLCV, Northern India) used for alignment and designing of siRNA were retrieved from the databases. Details of these sequences including their accession numbers are given in Table 1. These sequences were aligned using the CLCBiO sequence viewer version 5.1.1 and phylogenetic trees were generated from these alignments. The alignments and phylogenetic tree in case of PLCV coat protein and replicase genes (AV1, AC1) are given in Figs. 1a, b and 2a, b, respectively. Similarly, the alignment and phylogenetic tree were constructed for corresponding genes of TLCV and TLCV (Northern India) are given in Figs. 3a, b, 4a, b, 5a, b, and 6a, b, respectively. The multiple sequence alignment revealed a number of well-conserved region of 21–25 nt from AV1 (coat protein gene) and AC1 (replicase gene) of all three geminiviral groups.

Table 1 Gene sequences of different isolates of geminiviruses infecting papaya and tomato that were retrieved from the databases
Full size table
Fig. 1
figure 1figure 1

a Papaya leaf curl virus AV1 sequence alignment. b Papaya leaf curl virus AV1 sequence alignment tree

Full size image
Fig. 2
figure 2figure 2

a Papaya leaf curl virus AC1 sequence alignment. b Papaya leaf curl virus AC1 sequence alignment tree

Full size image
Fig. 3
figure 3figure 3figure 3figure 3figure 3

a Tomato leaf curl virus AV1 sequence alignment. b Tomato leaf curl virus AV1 sequence alignment tree

Full size image
Fig. 4
figure 4figure 4figure 4figure 4figure 4figure 4figure 4

a Tomato leaf curl virus AC1 Sequence alignment. b Tomato leaf curl virus AC1 sequence alignment tree

Full size image
Fig. 5
figure 5figure 5

a Tomato leaf curl virus (Northern India) AV1 alignment. b Tomato leaf curl virus (Northern India) AV1 alignment tree

Full size image
Fig. 6
figure 6figure 6figure 6

a Tomato leaf curl virus (Northern India) AC1 alignment. b Tomato leaf curl virus (Northern India) AC1 alignment tree

Full size image

In the context of genetic variability among the various virus isolates reported so far, the siRNA approach, which is a highly homology-dependent methodology, may not function against an isolate or a strain of it when designed for another isolate or strain. Hence, it is necessary to first perform multiple sequence alignment to identify the conserved genomic regions among the different viral isolates and then select the most appropriate conserved sequence(s) to further design siRNAs. Such siRNA will not only result in a broad spectrum resistance against many isolates/strains of the same virus but also be useful in the prediction of stability of resistance.

Sequences selected as putative siRNA

Conserved regions of the coding AV1 and AC1 genes obtained by multiple sequence alignment were critically evaluated as effective targets for siRNAs, fulfilling the criteria suggested by various groups [30, 31]. A number of conserved stretches with 5′ AA dinucleotide or triplet AAG/C with 3′ adjacent 18/19 nucleotides, respectively, were scanned as target sequence for siRNA. Stretches of >4 T’s or A’s in the target sequence were avoided as it may act as a termination signal for RNA pol III. Of the several oligonucleotides regions putatively identified, ~21 bp long oligonucleotides were manually selected as candidate siRNA (sense strands) using the criteria that these oligonucleotides have 30–50% GC content, have no G residues in the overhang and having extra UU or dTdT dinucleotide (the presence of an overhang on the antisense strand has been demonstrated to be important for some sequences) added at the 3′-ends. To crosscheck these manually selected oligonucleotides, they were subjected to siRNA target finder analysis that resulted in a number of putative siRNA validating the manually selected siRNAs also. The three best sequences finally selected from each of PLCV, TLCV, TLCV (Northern India) coat protein (AV1), and replicase genes (AC1), are summarized in Table 2.

Table 2 Putative siRNA sequences selected by using siRNA target finder
Full size table

Analyzing putative siRNAs by BLAST

A cross homology search using NCBI, BLAST was also made on the above-selected sequences to avoid silencing of any host plant gene (if at all, by chance the selected siRNA bears any homology to any plant gene) i.e. off target cleavage. The results showed a similarity of 95–100% to various geminiviruses but not to any gene of the plant genome, thus assuring the silencing of only geminiviral genes. It is very important to compare the potential target sites to appropriate genome database (human, Arabidopsis, etc.) using BLAST (www.ncbi.nlm.nih.gov/BLAST) and eliminate any sequence showing more than 16–17 bp contiguous base pairs of homology to other coding sequence.

Conclusions

Geminiviruses are a major threat to world agriculture, and breeding crops resistant to these DNA viruses is one of the major challenges faced by plant virologists and biotechnologists [32]. DNA geminiviruses infecting Papaya and Tomato can be targets of RNA silencing. It is reported that siRNA sequences with predicted functionality and specificity characteristics are expected to deliver 75% or greater gene silencing/knockdown in 80% or more of cases [30]. Thus, we expect that the siRNA thus selected shall be able to do gene silencing of a number of isolates of geminiviruses infecting papaya and tomato.

We have selected those putative siRNA for engineering resistance those are fulfilling the general guidelines of siRNA. Thus, we have done multiple sequence alignment and further selected the most conserved region which are fulfilling all the criteria of a putative siRNA also based on various software/algorithms. Also, only those siRNA from conserved region were selected which did not show any homology to plant genes by BLAST search. This is important as we wish to avoid cross silencing of the plant gene while attempting silencing of viral genes. The efficiency of the siRNA sequence thus selected may be checked by further synthesizing, construction/chemical synthesis, transformation/introduction into papaya and tomato and finally analysis of resistant plants by challenge viral inoculation. Such a study can lead to a novel and an efficient siRNA-based strategy against this virus to counter the leaf curl disease and its management in case of both tomato and papaya crops in India.