Background
Geminiviruses are plant pathogens that profoundly affect diverse plant crops in tropical and subtropical countries [
1‐
3]. These are emerging class of viruses with new strains still evolving, thereby making them more virulent with wide host range specificity [
4,
5].
Tomato leaf curl New Delhi virus (ToLCNDV) is a member of
begomovirus genus infecting tomato crop and it causes severe yield loss. This group of viruses may have monopartite (DNA A) or bipartite (DNA A and DNA B) circular ssDNA genomes. The DNA A component encodes six Open Reading Frames (ORFs) namely AC1, AC2, AC3, AC4, AV1 and AV2 while only two proteins (BC1 and BV1) are encoded by DNA B. These ORFs are encoded either in the virion or complementary-sense orientations. Most of these proteins have been implicated in virus multiplication and pathogenesis. One of the apparent symptoms associated with ToLCNDV infection is upward leaf curling in tomato leaves.
MicroRNAs (miRNAs) have recently emerged as the key regulatory molecules in diverse biologically relevant processes, both in plants and animals [
6,
7]. The miRNAs are transcribed from their own promoters by RNA pol II activity and have characteristic 5' cap and 3' poly-A tail [
8,
9]. These pri-miRNA transcripts form hairpin like structure and are sequentially processed by the action of RNase III-like proteins, namely HYL1/SER1 and DCL1 in
Arabidopsis, to generate miRNA duplexes [
6,
10]. The mature miRNA enters into a multi-protein complex termed RNA-induced silencing complex (mi-RISC) and guides it to the target mRNAs with complementary sequences. This consequently leads to the target cleavage [
8,
11] and/or inhibits translation of the targets [
12]. In plants, miRNAs have been demonstrated to participate in leaf morphogenesis, phase transition, flower development and root and shoot development [
13‐
18]. It is thus apparent that ToLCNDV induced leaf curling in tomato can be utilized as a model system to study the influence of miRNA-mediated biological actions on leaf deformations.
In
Arabidopsis, few miRs have been demonstrated to critically regulate leaf development
viz., miR165/166, miR164 and miR319/159 [
19‐
21]. For instance, miR165/166 targeted HD-ZIP III transcription factors (TFs) are involved in determining adaxial and abaxial pattern formation [
20] while, miR159 and miR319 play important roles in maintaining leaf phenotype by regulating members of MYB transcription factors and TCP transcription factors, respectively [
19]. Similarly, miR164 that targets CUC2 also takes care of leaf patterning by controlling serration of leaf margins [
21]. The involvement of these miRNAs in leaf morphology has been demonstrated by raising
Arabidopsis transgenic over-expressing miRNAs or targets with mutated miRNA binding sites and these transgenic plants revealed clear leaf development associated defects. Moreover, evidences support the involvement of miRNAs in biotic and abiotic stresses. For instance, miR393 expression is induced under bacterial infection [
22]. The F-box auxin receptor proteins targeted by miR393 are consequently down-regulated, thereby suppressing auxin signaling pathways and probably conferring resistance against pathogens. On the other hand, miR395, miR399, miR398, etc., have been associated with specific abiotic stresses [
7,
23,
24].
Viral encoded proteins interfere with host RNAi pathways and thus these proteins distort the normal cellular activities [
25‐
27]. The Viral Suppressors of RNA silencing (VSRs) are crucial in disease severity and developmental abnormalities [
25]. ToLCNDV encodes three VSRs
viz., AC2, AC4 and the pre-coat protein AV2 [
26,
27]. AC2 is a crucial VSR that disturbs post-transcriptional gene silencing and mutation in AC2 leads to reduced pathogenesis [
28], Karjee et al. Unpublished data]. Further, over-expression of AC2 induces expression of several host genes [
29] and impacts severely on plant architecture.
African cassava mosaic virus (ACMV) AC4, on the other hand, has been shown to bind directly to miRNAs, thereby making mi-RISC non-functional [
27].
ToLCNDV is also accompanied by a satellite DNA (β DNA) that encodes βC1 protein [
30]. Recently, it has been demonstrated that βC1 of
Tomato yellow leaf curl China New virus (TYLCCNV) interacts with host ASYMMETRIC LEAVES 1 (AS1) to alter leaf phenotype [
31]. All these studies suggest that ToLCNDV encodes enough proteins to help virus achieve suitable environment for its survival and cause pathogenesis. The VSRs influence miRNA biogenesis of host, leading to developmental abnormalities [
32]. Although viral proteins, VSRs, have been shown to hamper the development of transgenic plants [Karjee et al. Unpublished data], not much has been studied about which miRNAs are responsible for ToLCNDV induced changes in leaf phenotype. We attempted to understand how ToLCNDV utilizes host miRNAs to bring about curl phenotype in the leaves. Since the involvement of miRNA in biotic responses and leaf patterning is now well recognized, we wanted to explore the molecular principles behind the ToLCNDV mediated leaf abnormality.
Here, we report that ToLCNDV agroinfection can significantly deregulate the host miRNA expression and the corresponding targets as well. Moreover, this ToLCNDV induced differential shift in miRNA levels was specific to leaf tissues since we did not observe the similar changes of either miRNA or its target in tissues other than the leaves. This tissue specific deregulation of miRNA levels is perhaps required to establish favorable conditions for virus survival and perhaps leads to leaf deformation. The expression levels of miR159/319 and miR172 were observed to be associated with disease progression, thereby making these as potential biomarkers for ToLCNDV infection.
Discussion
The involvement of miRNAs in diverse abiotic responses (salt, temperature, drought and nutritional starvation) has been demonstrated by various research groups [
23,
24,
34]. Recent studies have also confirmed their roles in conferring immunity against bacterial responses [
35]. Although siRNA-mediated viral defense responses are well studied in plants, the roles of host miRNAs in plant viral immunity/sensitivity have not been well investigated. The miRNA profiling is a good indicator of many diseases, especially cancers, where strategies to cure rely on the early disease detection [
36]. There is growing evidence that certain cancerous tissues exhibit deregulated levels of miRNAs, thus supporting the notion that these molecules are promising therapeutic agents or drug targets [
37,
38]. However, the plant miRNAs as biomarkers of disease are at the stage of exploration. Here, we examined the possibility that miRNAs can be used as diagnostic markers in response to a leaf curl disease caused by ToLCNDV agroinfection in tomato plants. These miRNA biomarkers can eventually be manipulated developing antiviral strategies.
In our present study, we have identified a set of miRNAs, the levels of which were differentially altered under ToLCNDV (2A+2B) infection in tomato leaves. We speculate that both virus and host utilize miRNAs as efficient weapons to fight against each other. A major proportion of the genes targeted by these miRNAs are reported to play crucial role in various defense responses highlighting their role in viral defense mechanism. For instance, the genes targeted by miR398, miR399, miR162 and miR168 are, either directly or indirectly, responsible for responding to adverse stress stimuli. Interestingly, we found upregulation of miR168 and miR162 in our microarray data. These miRNAs target DCL1 and AGO1 proteins respectively, that are primarily involved in regulating global miRNA flux and their functionality can profoundly affect the overall miRNA levels [
39,
40]. Although DCL1 and AGO1 are demonstrated to participate predominantly in miRNA biogenesis, recent studies have extended their roles in the generation of certain siRNAs. For instance, DCL1 is shown to be required during ta-siRNAs and lsiRNA/nat-siRNAs production and these are necessary in both abiotic and biotic stress responses [
41,
42]. The miR398 targets Copper superoxide dismutases (CSD1 and CSD2) that are known to control abiotic stress response [
43] and we also observed increase in the levels of miR398 following ToLCNDV infection. Thus a cross-talk between the pathways of abiotic and biotic stress-responses might be a reality. Similarly, most of the miRNAs (miR395, miR396 and miR399) known to be involved in abiotic stress conditions showed enhanced expression under ToLCNDV infection and the levels of the corresponding targets were reduced. These target genes are required to overcome oxidative and nutrient stress responses and their repression could be beneficial for establishment of viral infection and disease expression. Till date, only few miRNAs have been identified to regulate leaf development
viz., miR165/166, miR159/319, miR164 and miR160. We also observe modest changes in the levels of either precursor or mature miRNAs. In addition, our data reflect that many other miRNAs also exhibit altered expression in response to viral infection, suggesting their probable role in basal defense activity and leaf morphogenesis.
Microarray and northern hybridization results show that most of the deregulated miRNAs were induced and only few (miR160, miR164, miR169, miR171 and miR391) were down-regulated following ToLCNDV infection. AC4 of
African Cassava mosaic virus (ACMV) has been demonstrated to destabilize single stranded miRNAs by interacting directly with them and thus AC4 over-expressing transgenic plants showed reduced accumulation of miRNAs [
27]. It is possible that, similar to ACMV AC4, ToLCNDV AC4 might act to destabilize miRNAs which explains the reduction in the levels of certain miRNAs. However, as ToLCNDV infected leaves predominantly show induction in the expression of miRNAs, it likely appears that there exist some other mechanisms to achieve this deregulation both at the miRNA and pre-miRNA stages. ToLCNDV encoded suppressors,
viz., AC2 and the pre-coat protein (AV2), might also increase the level of miRs and all of these three VSRs (AC2, AC4 and AV2) might behave differently [
44].
The RT-PCR analysis of pre-miRNAs, in most of the cases, reveals that the expression of pre-miRs was modulated analogous to their respective mature miRNAs. This suggests that ToLCNDV infection might lead to altered transcription of certain miRNA genes. Since AC2 and CP are known to localize into host nucleus, it is plausible to assume that these VSRs might interact with host transcription factors (TFs) and can modulate transcription of several downstream genes including those of the miRNAs [
29,
45]. We speculate that ToLCNDV infection leads to transcriptional activation/repression of certain miRNA genes and at the same time certain ToLCNDV encoded VSRs can bind to and stabilize/destabilize both, pre- and/or mature miRNAs.
ToLCNDV spreads systemically, but the virus is abundant in leaves and rarely present in flowers (Figure
1b). Similarly, the virus induced changes in the transcriptome (including miRNA genes) between the leaves and the flower tissues were observed to be dramatically different. Expression studies of pre-miRNA and miRNA targets in flowers, reveals that most of these transcripts did not show levels similar to that observed in leaf tissues. Nonetheless, the flowers of ToLCNDV agroinfected plants exhibit the disease symptoms (Figure
1a). Further, since leaf tissues are the primary site for viral entry, they suffer drastic alteration in the transcriptome following ToLCNDV infection. Therefore, it is not surprising that the deregulation of ToLCNDV induced transcript remain mostly localized in the leaf tissues.
The expression analysis of miR159 and miR172 at different dpi of ToLCNDV (2A+2B) infected Pusa Ruby leaves clearly shows induction of these miRNA during disease progression (Figure
8a). These results suggest that miR159 and miR172 could be used as potential indicator of ToLCNDV infection. Since the miR159 targets (MYB TFs) are well established factors determining leaf structure, the observed leaf deformation in ToLCNDV infected tomato plants could be due to the altered levels of miR159. Another viral induced miRNA
viz., miR319 targets TCP TF family member, including TCP4. TCP4 is a well known suppressor of growth in Arabidopsis [
46]. The accumulation of miR319 in ToLCNDV infected leaves will bring down the expression of tomato TCP4 homologs. This consequently would lead to uncontrolled cell growth and that might reflect in the form of leaf deformation. Thus, the altered expression of miR159/319 could serve as a prospective indicator of leaf curl disease.
Our microarray data revealed that the levels of miR396 were induced (~2 folds) in ToLCNDV infected leaves. MiR396 targets Growth-Regulating Factor (GRFs) transcription factors that are involved in cell division. Moreover, either miR396 over-expression or
GRF mutation leads to reduced leaf size in Arabidopsis [
47,
48] and it is noteworthy that the leaves of ToLCNDV agroinfected plants were also of small size compared to those of the uninfected plants. Further, studies by Rodriguez et al. [2010; [
49]] demonstrated that miR396 levels were increased in Arabidopsis
soj8 mutants (miR319-resistant TCP4 lines). Thus, it appears that leaf curl phenotype observed during ToLCNDV agroinfection could be a manifestation of high levels of miR319 and miR396, where both might act synergistically leading to a pronounced leaf curling. Significantly, we also compared the results of agroinfection with the same of field-infection. The induction of various miRNAs and the levels of disease progression were almost similar in both cases. Thus, the conclusions of the additional field level data reinforced the notions derived from the agroinoculation studies.
All these data together indicate that the defense responses are mediated by induction and repression of large array of genes that includes the miRNA transcripts. Although some of these altered transcripts could help host to defend against diverse pathogens and abiotic stresses, viruses can also utilize this disturbed transcriptome for their own benefit i.e., to make suitable environment for their survival.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
ARN carried out all the experiments and wrote the manuscript. QMRH helped in drafting the manuscript. SKM conceived of the study, and participated in its design and wrote the manuscript in its final form. All authors read and approved the final manuscript.