In Silico Identification of Sugarcane Genome-Encoded MicroRNAs Targeting Sugarcane Mosaic Virus

: Sugarcane mosaic virus (SCMV) (genus, Potyvirus ; family, Potyviridae ) is widespread, deleterious, and the most damaging pathogen of sugarcane ( Saccharum officinarum L. and Saccharum spp.) that causes a substantial barrier to producing high sugarcane earnings. Sugarcane mosaic disease (SCMD) is caused by a single or compound infection of SCMV disseminated by several aphid vectors in a non-persistent manner. SCMV has flexuous filamentous particle of 700–750 nm long, which encapsidated in a positive-sense, single-stranded RNA molecule of 9575 nucleotides. RNA interference (RNAi)-mediated antiviral innate immunity is an evolutionarily conserved key biological process in eukaryotes and has evolved as an antiviral defense system to interfere with viral genomes for controlling infections in plants. The current study aims to analyze sugarcane ( Saccharum officinarum L. and Saccharum spp.) locus-derived microRNAs (sof-miRNAs/ssp-miRNAs) with predicted potential for targeting the SCMV +ssRNA-encoded mRNAs, using a predictive approach that involves five algorithms. The ultimate goal of this research is to mobilize the in silico-predicted endogenous sof-miRNAs/ssp-miRNAs to experimentally trigger the catalytic RNAi pathway and generate sugarcane cultivars to evaluate the potential antiviral resistance surveillance ability and capacity for SCMV. Experimentally validated mature sugarcane ( S. officinarum , 2 n = 8 X = 80) and ( S. spp., 2 n = 100–120) sof-miRNA/ssp-miRNA sequences ( n = 28) were downloaded from the miRBase database and aligned with the SCMV genome (KY548506). Among the 28 targeted mature locus-derived sof-miRNAs/ssp-miRNAs evaluated, one sugarcane miRNA homolog, sof-miR159c, was identified to have a predicted miRNA binding site, at nucleotide position 3847 of the SCMV genome targeting CI ORF. To verify the accuracy of the target prediction accuracy and to determine whether the sugarcane sof-miRNA/ssp-miRNA could bind the predicted SCMV mRNA target(s), we constructed an integrated Circos plot. A genome-wide in silico-predicted miRNA-mediated target gene regulatory network was implicated to validate interactions necessary to warrant in vivo analysis. The current work provides valuable computational evidence for the generation of SCMV-resistant sugarcane cultivars.

In plants, microRNAs (miRNA) are endogenously expressed small (19-25 nucleotides), evolutionarily conserved, non-coding (NC)-ss RNA molecules [20].In higher plants, the biogenesis and transcription of the miRNA gene (MIR) is controlled by RNA polymerase II, which is then transcribed into single-standard polycistronic primary transcripts (pri-miRNAs).They control a variety of biological processes in plants by regulating gene expression, cell growth, development, differentiation, and host-virus interactions [21,22].The miRNA-mediated RNAi is a post-transcriptional gene-silencing mechanism that provides antimicrobial innate immunity and regulates host-virus interactions to limit or inhibit viral infection [23].
An integrative multi-network approach based on SCMV infection assessment was used to identify target binding sites of sugarcane genome-encoded sof-miRNAs/ssp-miRNAs in the SCMV genome.The identification of multiple host-derived miRNA binding sites in the SCMV genome for the creation of transgenic sugarcane varieties resistant to SCMV is the main objective of this study.In this study, several miRNA prediction tools were evaluated and used to identify microRNA-mRNA binding sites in the SCMV genome for use in developing transgenic or non-transgenic modified sugarcane plants with resistance to SCMV and, potentially, closely related potyviruses.Potential targets of the most promising sugarcane miRNAs for breeding were also of interest to better understand potyvirussugarcane plant interactions during infection.Until now, there have been no reports on the use of an amiRNA-based strategy to develop SCMV tolerance in sugarcane plants, based on the prediction of homologous amiRNAs for silencing SCMV.The predicted locus-derived sof-miRNAs/ssp-miRNAs in the sugarcane genome were further evaluated to understand the complex interactions between sugarcane host planta and SCMV potyviruses and to identify novel antiviral targets.

Potential Targets of Sugarcane MicroRNAs in the SCMV Genome
The prediction of effective microRNA-mRNA binding sites is a first step toward understanding microRNA-regulated gene regulatory networks.The accuracy of miRNA target site prediction can be affected by several factors, such as the specificity and sensitivity of the algorithm, the choice of reference sequence, and the length of the target sequence.Various in silico methods for effective silencing have been developed for the computational prediction of miRNA-mRNA target sites.A computational approach refers to the use of multiple computational methods, algorithms, or tools to analyze and interpret biological data.This approach combines different types of publicly available in silico algorithms, including miRanda [43,44], RNA22 [45,46], TAPIR [47], psRNATarget [48,49], and RNAhybrid [50] (Table 1).

miRanda
miRanda is one of the first miRNA target predictors, a highly versatile algorithm based on the seed-based interactions of miRNA target duplexes [43].It was implemented as a standard tool to detect potential miRNA binding sites.RNA-RNA duplex dimerization and sequence complementarity are features considered by the miRanda algorithm.It considers the cross-species conservation of target sites, which distinguishes it from other algorithms [44].The miRanda algorithm has been implemented in C, and its first version was published in 2003.The default parameters were selected for the analysis (Table 1).

RNA22
The RNA22 algorithm has a diverse, web-based application with implemented interactive exploration.It uses a pattern-recognition based approach to serve as a miRNA target discovery tool.It predicts statistically significant target patterns using maximum folding energy (MFE) [45,46].Site complementarity and non-seed-based interactions are important features.Its prediction is also based on highly sensitive and significant target patterns.The default parameters were chosen (Table 1).

TAPIR
The TAPIR algorithm is used to assess the seed-based interactions of plant miRNAs in the target sequence.It is a highly precise plant miRNA target prediction algorithm used to detect target binding sites in the target sequence.It is used to deliver precise miRNA target predictions, including target mimics, with FASTA and RNAhybrid search options [47].The default parameters were chosen (Table 1).

psRNATarget
The psRNATarget algorithm is a highly sensitive, newly designed web-based tool developed for plant miRNA prediction.The target binding sites of plant miRNAs are predicted based on complementary scoring schema.The algorithm predicts the inhibition pattern of cleavage [48,49].The default parameters were chosen (Table 1).

RNAhybrid
The RNAhybrid is a seed-based scanning algorithm based on intermolecular hybridization used to predict effective binding sites of miRNAs in the target sequence.It predicts target binding sites in a very easy and flexible manner [50].It is an online available tool.It is used for the rapid prediction of miRNA targets based on the MFE hybridization of mRNAs and miRNAs.The default parameters were chosen (Table 1).

RNAfold
The RNAfold algorithm is available on a web server implemented in the ViennaRNA package [51].

Statistical Analysis
The miRNA-mRNA target prediction biological data were further processed.Graphical representations of the miRNA data were prepared using the R language [52].

Prediction and Analyysis of Sugarcane MicroRNAs Targeting SCMV Genome
An integrative computational approach for identifying the possible interactions of high-confidence target sites of mature sugarcane miRNAs located in the SCMV positivesense single-stranded (+ssRNA) genome from among the 28 sugarcane miRNAs (sof-miRNAs/ssp-miRNAs) revealed sof-miRNA/ssp-miRNA-derived MIR genes at a high proportion of sugarcane miRNA gene loci [33,[53][54][55][56].The predicted SCMV +ssRNAencoded mRNA sequences were localized hypothetically by the sugarcane locus-derived sof-miRNA/ssp-miRNAs based on the miRanda algorithm predicted 19 miRNA-mRNA target pairs and RNA22: 15 sugarcane sof-miRNAs/ssp-miRNAs and 20 binding sites i.The TAPIR identified seven binding sites of mature sugarcane locus-derived sof-miRNA/ssp-miRNA target pairs.In total, 16 sugarcane miRNAs targeting 30 cleavable attachments sites in the SCMV genome were identified by the psRNATarget algorithm.RNAhybrid predicted 28 high-probability binding sites of sugarcane miRNAs in the SCMV genomic RNA sequence (Figures 1 and 2, File S1, Table S2).

Sugarcane miRNAs Targeting 6K1
The 6K1 ORF (3269-3469 nucleotides) encode a 6K1 protein that functions in viral genome replication.It mediates cell-to-cell movement, controlling defense mechanism and gene regulation.It is a key component of the 6K2-induced viral replication complex (VRC) and regulation [60,61].The 6K1 had the least number of predicted sugarcane sof-miRNAs/ssp-miRNAs.The ssp-miR444c-3p was predicted to optimally target 6K1 at nucleotide position 3441, according to the psRNATarget algorithm (Figure 2D, File S1, Tables S2 and S3).

Identification of the miRNA-mRNA Regulatory Network
A Circos plot represents the predicted host-virus interactions of sugarcane miRNAs and SCMV target genes.A Circos plot was generated to visualize a comprehensive master miRNA regulatory network with novel antiviral targets (Figure 4).The generation of the miRNA-mRNA regulatory network was conducted using 'Circos' software [84].

Identification of the miRNA-mRNA Regulatory Network
A Circos plot represents the predicted host-virus interactions of sugarcane miRNAs and SCMV target genes.A Circos plot was generated to visualize a comprehensive master miRNA regulatory network with novel antiviral targets (Figure 4).The generation of the miRNA-mRNA regulatory network was conducted using 'Circos' software [84].

RNA Secondary Structures
The computationally predicted locus-derived mature miRNAs in the sugarcane genome were analyzed by generating their secondary structures using the original precursor sequences.Pre-miRNA hairpin sequences were used for manual curation.The main parameters of the predicted stable secondary structures were evaluated (Table 4).The stable secondary structures of the potential consensus sugarcane precursor sequences were predicted by the RNAfold algorithm [51].

RNA Secondary Structures
The computationally predicted locus-derived mature miRNAs in the sugarcane genome were analyzed by generating their secondary structures using the original precursor sequences.Pre-miRNA hairpin sequences were used for manual curation.The main parameters of the predicted stable secondary structures were evaluated (Table 4).The stable secondary structures of the potential consensus sugarcane precursor sequences were predicted by the RNAfold algorithm [51].

Discussion
The SCMV is a monopartite potyvirus suspected as an etiological agent that has spread to Pakistan and China due to its high transmissibility and has become an increasingly potential long-lasting threat to sugarcane and maize production in the last two decades [13,17,85].In our previous studies, we have investigated experimentally validated mature locus-derived microRNAs in the sugarcane genome, which were predicted to be targets of SCBGAV, SCYLV, and SCBV based on in silico criteria [37][38][39].Several studies have identified complex host-virus interactions and have investigated miRNAs targeting plant viruses using an in silico approach [86][87][88][89][90][91][92].miRNAs have emerged as novel endogenous targets for multiple levels of miRNA gene-level regulation [53,93,94].Several studies have shown that the efficacy of amiRNA-based RNA interference leads to specific gene silencing in transgenic crops to reduce host plant virus infection [27,28,[95][96][97].In this computational research, mature sugarcane sof-miRNAs/ssp-miRNAs were aligned with the genomic sequence of the SCMV target to identify miRNA-mRNA binding sites hypothesized to understand complex host-virus specific interactions with the P1, HC-Pro, P3, 6K1, CI, 6K2, NIa-VPg, NIa-Pro, NIb, and CP of SCMV.The P1 is the least-conserved hypervariable that modulates host responses and is essential for the replication of the viral +ssRNA genome [98,99].Host adaptation is a key process for virus genome evolution [100,101].P1 is also related to virus-host adaptation [102].The HC-Pro is a multifunctional, nonstructural dimeric helper component proteinase.It has been reported as a viral suppressor.HC-Pro is required to enhance expression via the fusion of P1, symptom development, and viral replication [103][104][105][106][107][108].Until now, the potential for exploiting the regulation of sugarcane genome-encoded miRNA to abate infection by SCMV has not been investigated as a strategy for developing tolerant or resistant sugarcane cultivars.The results of this study provide the first computationally based evaluation of mature locus-derived miRNAs in the sugarcane plant genome to enable the prediction of effective miRNA binding sites and provide new tools for better understanding the molecular and omic interactions between sugarcane plant host cells and SCMV-encoded mRNAs/protein.
Based on our findings, the SCMV genome (HC-Pro, CI, NIa-VPg, and CP) is susceptible to nine consensus sugarcane miRNAs.We found that nine miRNAs could theoretically originate from the sugarcane genome (Tables 2 and 3).In silico tools, RNA22, TAPIR, and psRNATarget, identified a genomic consensus base pair complementarity in sof-miR159c at nucleotide position 3847 (Figure 2 and Table 2).The ssp-miR444c-3p was predicted by all five algorithms, making it the only unique sugarcane miRNA identified in this study (Figure 1 and Table S2).We identified the maximum folding energy of the consensus functional miRNA-mRNA target pair, which is −18.00Kcal/mol, using RNA22.RNA22 is a highly sensitive algorithm that uses a pattern-based approach to target miRNAs.Using the psRNATarget algorithm, we estimated an expectation score of 5.50 for a consensus target pair (Table 2) [109].The RNA22 and psRNATarget algorithms predicted target sites using a non-seed-based approach.Experimentally determining miRNA-mRNA interactions can be expensive and time-consuming; making the accurate computational prediction of miRNA targets a high priority.The limitations and bottlenecks of existing algorithms and approaches are interpreted using the union and intersection level of the predictions in this study.The miRNA targeting relies on the base pairing of miRNA-mRNA targets [110].These results suggest that the predicted consensus miRNA-mRNA duplex represents a 'true target'.Our results indicate that sugarcane miRNAs likely play a role in the interaction between host and virus.Our results highlight the interaction of SCMV ss-RNA with the sugarcane miRNA target interaction network.
Potyvirus cylindrical inclusion helicase (CI) is required for the initiation of the viral replication mechanism, cell-to-cell movement, and plant-host, protein-virus interactions [62,63,111].Computational predictions and analyses revealed that the sugarcane consensus sof-miR159c is a high-confidence target site potentially targeting the CI ORF (Table 3).The conserved precursor MIR159 is considered to be controlled by plant growth and fertility [112].In our previous study, the consensus sof-miR159e (Accession ID: MI-MAT0001661), predicted to have an effective target binding site at nucleotide position 5535 in the SCBV genome, was identified as the most effective miRNA by the miRanda, RNA22, and RNAhybrid algorithms.
While miRNA-mRNA target pair interactions between locus-derived miRNAs in the sugarcane genome and SCMV have been determined, the development of amiRNAbased constructs and further transformations in sugarcane to control SCMV are not fully understood.We have performed a comprehensive analysis of SCMD-associated Potyvirus for the first time, which is a first step toward the development of miRNA-based antiviral therapy.An amiRNA construct relies on the high-level specificity of a nucleotide base pairing to control deleterious off-target effects.The small size of amiRNA is a unique feature for the development of a single gene expression vector to control multiple potyviruses in transgenic sugarcane.This approach offers specificity and sensitivity and complements existing molecular approaches for analyzing targets for SCMV disease abatement.A number of environmental concerns have been raised regarding the large-scale use of virusresistant transgenic plants [113][114][115][116][117][118].As amiRNAs have high specificity to the designed target gene, detrimental off-target effects can be minimized, permitting their silencing expression to be stably transmitted to future generations [119][120][121][122][123]. The results indicate that the use of in silico tools provides better results than a single algorithm when developing amiRNA-based mdm-miRNA therapeutics to target SCMV and other plant viruses as well.Despite the frequent use of RNAi in biology and agriculture, there are several drawbacks and challenges in designing efficient silencing constructs.Furthermore, the small size of amiRNA permits for the insertion of multiple and distinct amiRNAs within a single gene expression cassette, which can then be transformed to develop transgenic plant resistant to multiple viruses simultaneously [27,95,124].The in silico analysis was designed for experimental validation to show whether these predicted miRNAs could make the plants resistant to SCMV.Future work will be focused on transiently expressing these miRNAs or injecting RNA hairpins in N. benthamiana to show its efficacy against SCMV.

Conclusions and Future Directions
The SCMV, which infects sugarcane crops worldwide, is the most damaging potyvirus pathogen associated with an ongoing SCMD epidemic that reduces yield in all sugarcane cultivars cultivated in China.This study involved in silico tools and approaches to characterize the target binding sites of mature sugarcane locus-derived miRNAs in the SCMV genome.Among the 28 sugarcane miRNAs from the miRBase database, only one, sof-miRNA (sof-miR159c), was identified as the most effective, naturally occurring sof-miRNA biomolecule for targeting the SCMV genome (nucleotide 3847 onward), based on the consensus of multiple algorithms used herein.This approach offers specificity and sensitivity and complements existing molecular approaches for analyzing targets for SCMV disease abatement.The current focus of attention is the development of SCMV-resistant sugarcane plants that abate the effects of SCMD.

Figure 1 .
Figure 1.Five-set Venn diagram representing mutually common binding sites of mature sugarcane miRNAs predicted to potentially target the SCMV genome.The in silico prediction was established using computational tools (miRanda, RNA22, TAPIR, psRNATarget, and RNAhybrid) to identify potential targets of sugarcane-encoded miRNAs.The areas of overlap among computational tools show miRNA binding sites.The high-order intersection of five algorithms revealed the most potent mature sugarcane miRNA-ssp-miR1444c-3p.

Figure 1 .
Figure 1.Five-set Venn diagram representing mutually common binding sites of mature sugarcane miRNAs predicted to potentially target the SCMV genome.The in silico prediction was established using computational tools (miRanda, RNA22, TAPIR, psRNATarget, and RNAhybrid) to identify potential targets of sugarcane-encoded miRNAs.The areas of overlap among computational tools show miRNA binding sites.The high-order intersection of five algorithms revealed the most potent mature sugarcane miRNA-ssp-miR1444c-3p.

Figure 2 .
Figure 2. Individual sugarcane sof-miRNAs/ssp-miRNAs and their predicted high-confidence binding sites in the SCMV genome were predicted based on the 'five algorithms' approach.(A) miRNA sites were detected by miRanda.(B) Several miRNA target sites were detected by RNA22.(C) TAPIR identified sugarcane miRNA binding sites.(D) psRNATarget predicted several binding sites of sugarcane miRNAs.(E) The prediction of miRNA sites by RNAhybrid.(F) Union plot representing all predicted binding sites detected by all the algorithms used.Multiple copies of miRNA target binding sites are represented by colored dots.Targeted genes of SCMV are indicated by different colors.

Figure 2 .
Figure 2. Individual sugarcane sof-miRNAs/ssp-miRNAs and their predicted high-confidence binding sites in the SCMV genome were predicted based on the 'five algorithms' approach.(A) miRNA sites were detected by miRanda.(B) Several miRNA target sites were detected by RNA22.(C) TAPIR identified sugarcane miRNA binding sites.(D) psRNATarget predicted several binding sites of sugarcane miRNAs.(E) The prediction of miRNA sites by RNAhybrid.(F) Union plot representing all predicted binding sites detected by all the algorithms used.Multiple copies of miRNA target binding sites are represented by colored dots.Targeted genes of SCMV are indicated by different colors.

Figure 3 .
Figure 3. Intersection plot shows the consensus high-confidence binding sites of mature suga miRNAs predicted by at least two computational tools.The colored dots represent suga miRNA binding sites targeting different genes of SCMV.

Figure 3 .
Figure 3. Intersection plot shows the consensus high-confidence binding sites of mature sugarcane miRNAs predicted by at least two computational tools.The colored dots represent sugarcane miRNA binding sites targeting different genes of SCMV.

Figure 4 .
Figure 4. Integrated Circos plot shows multiple targets of sugarcane-encoded miRNAs.The colored connection lines are targeted genes (ORFs) in the SCMV genome.Construction, exploration, target predictions, and interactions between the sugarcane miRNAs and SCMV genes are mapped.

Figure 4 .
Figure 4. Integrated Circos plot shows multiple targets of sugarcane-encoded miRNAs.The colored connection lines are targeted genes (ORFs) in the SCMV genome.Construction, exploration, target predictions, and interactions between the sugarcane miRNAs and SCMV genes are mapped.

Table 1 .
Different features and parameters of algorithms applied for miRNA target predictions.

Table 2 .
Predicted high-confidence binding sites of consensus sugarcane miRNAs targeting the SCMV genome detected by different computational algorithms.

Table 3 .
Predicted consensus sugarcane-encoded miRNA target sites localized in the different target genes of SCMV-SO.

Table 2 .
Predicted high-confidence binding sites of consensus sugarcane miRNAs targetin SCMV genome detected by different computational algorithms.
* MFE: minimum free energy measured in Kcal/ml.** MFE: maximum folding energy for he duplex measured in Kcal/mol.

Table 3 .
Predicted consensus sugarcane-encoded miRNA target sites localized in the differen get genes of SCMV-SO.

Table 4 .
Features of the predicted precursors of sugarcane were determined.