Background
Virus-induced gene silencing is an antiviral defense mechanism in plants, in which the vital elements involved are virus-derived small interfering RNAs (vsiRNAs) which are mainly 20–24 nucleotides (nt) in length [
1,
2]. The vsiRNAs are produced by Dicer-like proteins (DCLs) from viral double-stranded RNA (dsRNA) replication intermediates during the viral replication process, or from highly structured single-stranded RNA molecules present in virus-infected host plants [
1‐
4]. These vsiRNAs are templates generated in host cells that are loaded into Argonaute proteins (AGOs)-containing RNA-induced silencing complexes (RISCs), which then guide the target genomic and subgenomic viral mRNA in a sequence-specific manner to interfere with virus replication, translation and movement, and, in some cases, eliminate the viral infection [
3,
5‐
11].
It is well known that DCLs, AGOs, and the RNA-dependent RNA polymerases (RDRs) participate in the antiviral silencing pathways as key silencing factors, and their RNA silencing activities are varied under different temperatures [
4,
12‐
15]. In the model plant
Arabidopsis, among the four relatively well-studied DCLs, DCL4 and DCL2 function in RNA silencing against RNA viruses by producing 21- and 22-nt vsiRNAs, respectively. It has been demonstrated that the activity of DCL2 in producing specific 22-nt vsiRNAs derived from
Turnip crinkle virus (TCV) in
Arabidopsis is enhanced by higher temperatures [
16,
17]. DCLs-generated vsiRNAs are associated with specific AGO complexes, a process partially dependent on the 5'-terminal nucleotides. For instances, vsiRNAs with a 5'-terminal uridine or adenosine are recruited preferentially by AGO1 and AGO2 [
18]. Recent studies have also shown that AGO2 plays an antiviral role in the temperature-dependent survival of TCV- and
Potato virus X (PVX)-infected
Arabidopsis plants [
17,
19,
20]. In addition, RDRs like RDR1, RDR2, or RDR6 are involved in the biogenesis of secondary vsiRNAs to further enhance the antiviral RNA silencing efficiency [
21‐
23]. In
N. benthamiana, silencing of
RDR6 increases viral RNA accumulation and facilitates viruses to invade the meristem tissue [
21]. High temperatures intensify the RDR6 activity in the antiviral RNA-silencing defense response [
15]. It has been documented that RDR6 plays a tissue-specific role in the inhibition of
Chinese wheat mosaic virus (CWMV) accumulation and vsiRNA biogenesis at higher temperatures [
7,
15,
24,
25].
Apple stem grooving virus (ASGV), a member of the genus
Capillovirus in the family
Betaflexiviridae [
26]. The ASGV genome is a positive-sense ssRNA with 6.5 kb in length that contains two overlapping open reading frames (ORFs). The larger ORF1 encodes a polyprotein of 240 kDa, in which the N-terminal region contains replicase domains including methyltransferase (Met), papain-like protease (P-pro), NTP-binding helicase (Hel), and the RNA dependent RNA polymerase (RdRp), and the C-terminal region is the coat protein (CP) of 27 kDa [
27]. ORF2 is embedded within the ORF1 and encodes a movement protein (MP) of approximately 36 kDa [
27]. MP and CP may be produced through the 3′-coterminal subgenomic RNAs (sgRNAs), and CP expression from sgRNA is essential for ASGV systemic infection in the host [
28,
29]. Phylogenetic analysis of 16 ASGV full-length genomic sequences clusters them into two groups with no correlations to host and geographical origins [
30,
31]. ASGV infection is symptomless on most commercial cultivars of apple and pear, but does induce the typical symptoms of stem pitting and grooving on some cultivars of citrus, lily, kiwifruit, and pear [
32‐
35]. In asymptomatic apple plantlets, ASGV infection induces global gene expression changes, suggesting that extensive host genome-wide gene expression changes do not necessarily lead to disease symptoms [
36]. In pear, ASGV infection often deteriorates fruit quality [
37]. In the past several years, an increasing incidence of ASGV infection was observed in the pear-growing areas of China, leading to substantial economic losses [
37,
38]. High temperature in combination with shoot meristem tip culture is an effective way to obtain virus-free germplasm to control viral diseases of fruit trees [
39,
40]. The absence of viruses in the shoot meristem tip tissues is of practical importance because virus-free clones can be generated from infected shoots by culturing excised meristem tips. The effect of temperature on the RNA-silencing activities in plants has been investigated. Accumulated evidence suggests that low temperature inhibits RNA silencing-mediated defense by limiting the generation of small interfering RNA (siRNA) molecules, and high temperature promotes this innate immunity via increasing siRNA accumulation levels [
15,
41‐
43]. In virus-infected plants, viral symptoms disappear in new leaves at high temperatures, resulting from the higher temperature-mediated acceleration of the host antiviral gene silencing system in the meristem tip [
9,
39,
44]. In a recent study, we found that thermotherapy elimination of ASGV from Asian pear is associated with the high temperature-induced mixed action of a number of miRNA-mediated target genes related to disease defense and hormone signal transduction pathways in the apical meristem of pear shoots [
45]. These data suggest that elevated temperatures may enhance vsiRNA-mediated antiviral gene silencing activity, which in turn reduces the accumulation of viral RNA in the infected meristem tip cells of in vitro-cultured pear shoots.
To explore the possible roles of vsiRNA in the interaction of ASGV and pear plants at higher temperatures, we determined the complete genomic sequence of the ASGV-Js2 isolate and profiled the vsiRNAs in the meristem tip of in vitro-cultured pear shoots at 24 and 37 °C. We analyzed the distribution of vsiRNAs along the viral genome. The corresponding accumulation levels in relation to viral titer and the expression profiles of vsiRNA were also characterized in the shoot tip tissues by RT-qPCR. Furthermore, we determined the PpDCL2,4, PpAGOs, and PpRDR1 mRNA sequences and assessed their relative accumulation levels in the ASGV-infected pear shoot tips at 24 and 37 °C. Collectively, these results assist in a better understanding of the roles of vsiRNAs in ASGV infection in vitro-grown pear meristem tips in response to high temperature.
Discussion
Viral infection triggers the host gene silencing response, leading to vsiRNA accumulation [
54‐
60]. Temperature plays a key role in the RNA silencing-mediated antiviral defense in plants due to its effect on the control of siRNA generation [
15,
17,
42,
43]. Although vsiRNAs associated with virulent virus infections under various environmental conditions have been relatively well studied in model plant species, their involvement in latent virus infection in woody plants, especially fruit trees remains poorly characterized [
61,
62]. Our study is the first report on the characterization of a population of vsiRNAs in pear plants infected by a latent virus, ASGV-Js2, in response to high temperature treatment.
In this study, small RNA sequencing data revealed the presence of vsiRNAs in the tip of the ASGV-infected
P. pyrifolia shoots (Figs.
2,
3 and
4). We found that approximately 0.05 % (7,495/13,741,468 reads) and 0.06 % (7,949/14,071,933 reads) reads from the ASGV-infected samples matched the ASGV-Js2 genome from the 24 and 37 °C libraries, respectively. The results revealed that vsiRNAs accounted for a relatively small proportion of small RNAs in the ASGV-infected pear shoots when compared to the levels of vsiRNAs in other virus-host pathosystems in which virus-infected leaves are often used as materials for siRNA profiling studies [
11,
57]. Therefore, the low levels of ASGV-derived vsiRNAs may be attributed to the meristem tips of in vitro-cultured pear shoots used in this study. It is well known that there are endogenous restrictions preventing viral genomes from moving into plant meristems. In this study, we also found a continuous and uneven distribution of plus- and minus-sense vsiRNAs throughout the ASGV-Js2 genome in pear at either 24 or 37 °C (Figs.
2b and
5). This is different from findings from several previous reports that vsiRNAs are increasingly distributed toward the 3' end of the viral genome [
61‐
63]. It is not clear if the distribution pattern of vsiRNAs found in this study is also related to the particular tissues, e.g., pear shoots used in this study.
Real-time PCR analysis using primers specific for the MP coding region (Fig.
7) or RdRp sequence (data not shown) demonstrated a reduction in ASGV accumulation in the pear meristem tissues in response to high temperature treatment, similar to what was found in our previous study [
45]. Also the accumulation of vsiRNA4379 (+) and vsiRNA5839 (−) derived from the RdRp and CP regions of the ASGV genome increased accordingly (Figs.
5 and
6). These data suggest that vsiRNA abundance is negatively correlated with the levels of the ASGV viral RNA in response to high temperature (37 °C) treatment. This is consistent with the results obtained using other virus-host pathosystems [
15]. An exemption is the case of a DNA virus,
Cucurbit leaf crumble virus (CLCV). The relative abundance of CLCV-derived siRNA is apparently positively correlated with viral titers in pumpkin [
55]. A possible explanation for this discrepancy is that each virus-host combination might reflect unique characteristics; specifically, a dynamic equilibrium established during viral infection may affect sRNA levels in different virus-host systems [
54,
55,
57,
64]. It has been demonstrated that DCLs, RDRs, AGOs and other factors involved in RNA silencing also participate in antiviral defense in model plant species such as
Arabidopsis, tobacco, and rice [
2,
13,
15,
20,
23,
25,
65‐
67]. To explore the association of the corresponding homolog proteins with ASGV infection in pear in response to high temperature treatment, we cloned and obtained the partial sequences of
PpDCL2, PpDCL4,
PpAGO1, PpAGO2, PpAGO4, and
PpRDR1 and determined their relative expression levels (Figs.
8 and
9). Overall these genes were up-regulated in the ASGV-infected pear shoots at 37 °C, which was accompanied with the reduced level of viral RNA (Fig.
7) and the elevated levels of vsiRNAs (Fig.
6). These data support that high temperature treatment may enhance the RNA silencing capacity in the pear meristem tissue via up-regulating the expression of key components of the antiviral pathway to cope with ASGV infection. Future study is directed to elucidate how temperature regulates gene silencing in the ASGV-infected pear shoots and if this is tissue- or ASGV-specific. Such work would help better understand and improve thermotherapy for the effective control of virus diseases in fruit trees.
Conclusions
This study represents the first report on the characterization of vsiRNA in the in vitro-grown ASGV-infected pear shoots in response to high temperature treatment. The profiles of vsiRNAs showed an uneven distribution along the ASGV-Js2 genome, and that 21- and 22-nt vsiRNAs preferentially accumulated when cultured at higher temperature. ASGV-specific siRNAs from all libraries had a similar distribution of 5'-terminal nucleotides. U was the most frequent among the 5’ terminal nucleotides, and its frequency was slightly higher at 37 °C. The expression levels of the viral mp gene and vsiRNAs were characterized by RT-qPCR. We also cloned PpDCL2,4, PpAGO1,2,4 and PpRDR1 partial sequences and examined their expression patterns, and found their expression levels were up-regulated in the ASGV-infected pear shoots at 37 °C. This up-regulation was accompanied with the reduced level of viral RNA and the elevated levels of vsiRNAs. Taken together these data suggest that high temperature may induce and enhance the RNA silencing capacity in the pear meristem tissue.