Characterization of small interfering RNAs derived from Sugarcane mosaic virus in infected maize plants by deep sequencing

Abstract

RNA silencing is a conserved surveillance mechanism against viruses in plants. It is mediated by Dicer-like (DCL) proteins producing small interfering RNAs (siRNAs), which guide specific Argonaute (AGO)-containing complexes to inactivate viral genomes and may promote the silencing of host mRNAs. In this study, we obtained the profile of virus-derived siRNAs (vsiRNAs) from Sugarcane mosaic virus (SCMV) in infected maize (Zea mays L.) plants by deep sequencing. Our data showed that vsiRNAs which derived almost equally from sense and antisense SCMV RNA strands accumulated preferentially as 21- and 22-nucleotide (nt) species and had an adenosine bias at the 5'-terminus. The single-nucleotide resolution maps revealed that vsiRNAs were almost continuously but heterogeneously distributed throughout the SCMV genome and the hotspots of sense and antisense strands were mainly distributed in the HC-Pro coding region. Moreover, dozens of host transcripts targeted by vsiRNAs were predicted, several of which encode putative proteins involved in ribosome biogenesis and in biotic and abiotic stresses. We also found that ZmDCL2 mRNAs were up-regulated in SCMV-infected maize plants, which may be the cause of abundant 22-nt vsiRNAs production. However, ZmDCL4 mRNAs were down-regulated slightly regardless of the most abundant 21-nt vsiRNAs. Our results also showed that SCMV infection induced the accumulation of AGO2 mRNAs, which may indicate a role for AGO2 in antiviral defense. To our knowledge, this is the first report on vsiRNAs in maize plants.

Figure 1. 21- and 22-nt vsiRNAs accumulated at high levels in SCMV-inoculated maize.

A: Diagram showing the stepwise computational extraction of vsiRNA reads from small RNA libraries recovered from mock-inoculated and SCMV-inoculated systemic leaves. B: Histogram representation of total vsiRNA reads in each size class. C: Size distribution of total small RNAs in libraries prepared from either mock-inoculated or SCMV-inoculated maize plants. D: Size distribution of total small RNAs in the library from SCMV-inoculated maize plants.

Figure 2. Relative frequency of 5′-terminal nucleotide of vsiRNAs and accumulation of sense and antisense vsiRNAs.

A: Relative frequency of distinct 5′-terminal nucleotides in 21- and 22-nt vsiRNAs of SCMV-inoculated library. B: Accumulation of sense and antisense vsiRNAs. Percentage for each class of vsiRNAs from the SCMV-inoculated library is shown within the pie graph.

Figure 3. Profile of vsiRNAs derived from SCMV-inoculated library.

A: Schematic diagram of SCMV genome. B: Maps of 21- and 22-nt vsiRNAs from SCMV-inoculated maize plants at single-nucleotide resolution. The graphs plot the number of 21- and 22-nt vsiRNA reads at each nucleotide position of the 9595-nt SCMV genome; Bars above the axis represent sense reads starting at each respective position; those below represent antisense reads ending at the respective position. C: Northern blotting of vsiRNAs from distinct regions. “+” indicates vsiRNAs derived from sense strand of SCMV genome; “-” indicates vsiRNAs derived from antisense strand of SCMV genome.

Figure 4. The expression level of the predicted target mRNAs of vsiRNAs in mock- (blue) and SCMV-inoculated (red) maize plants.

For each target, the asterisk(s) indicates significant differences (*P<0.05; **P<0.01) of SCMV-inoculated versus mock-inoculated maize plants. The information and primer sequences of the predicted targets were listed in Table S5.

Figure 5. Functional classification of some predicted vsiRNAs target transcripts according to BLAST2GO.

The GO classification includes biological process, molecular function and cellular component. The detailed GO annotation information of each target transcript was listed in Table S4.

Figure 6. qRT-PCR analysis of the expression of maize DCLs and AGO2 mRNAs in mock- (blue) and SCMV-inoculated (red) maize plants.

For each gene, asterisk indicates significant differences (*P<0.05) of SCMV-inoculated versus mock-inoculated maize plants. The information and primer sequences used for amplification of ZmDCLs and ZmAGO2 were listed in Table S6.