An efficient virus- induced gene silencing (VIGS) system for gene functional studies in Miscanthus

Virus- induced gene silencing (VIGS) is a powerful tool for transient gene functional analysis in plants, especially for monocot species (e.g., grasses) that are recalcitrant to transformation. Despite various VIGS systems that have been developed in different plant species, none was previously available for the bioenergy crop Miscanthus. Here, we report the establishment of an efficient and robust VIGS system mediated by Tobacco Rattle Virus (TRV) in Miscanthus. We first investigated the impact of various factors that may affect gene silencing efficiency using the Miscanthus sinensis Phytoene Desaturase ( MsPDS ) gene as a visual indicator of photobleaching. Then, we optimized the TRV- elicited VIGS procedure using an orthogonal experimental design with four factors (sprout size, Agrobacterium concentration, vacuum infiltration time, and co- incubation time) each at three levels. The following led to the highest silencing efficiency ( ~ 76%):

and this led to compromised salt tolerance in the silenced Miscanthus plants. The TRV-based VIGS system established may, therefore, substantially facilitate functional genomic studies in Miscanthus.

| INTRODUCTION
Reverse genetics approach plays a crucial role in the elucidation of gene functions in plants. Generally, it relies heavily on stable genetic transformation to obtain the necessary loss-of-function mutants and overexpression lines (Gilchrist & Haughn, 2010). Notwithstanding stable transformation being well established in several model plant species (e.g., Arabidopsis thaliana and Oryza sativa), it is less developed in the majority of economically important agronomy and bioenergy crops, which are usually recalcitrant to transformation (Dommes et al., 2019). This technical barrier significantly hinders the improvement of desired traits for these crops.
Virus-induced gene silencing (VIGS) has emerged as an attractive alternative reverse genetic tool for transient gene knock-down in plants (Becker & Lange, 2010;Robertson, 2004;Senthil-Kumar & Mysore, 2011a). VIGS is a post-transcriptional gene silencing (PTGS) technique derived from the inherent mechanisms used by the plant immune system to cope with viral infections (Ruiz et al., 1998;Wang & Metzlaff, 2005). The VIGS technique circumvents the shortcomings of conventional transformation, which is usually laborious and time-consuming (Baulcombe, 1999;Burch-Smith et al., 2004;Huang et al., 2012;Ramegowda et al., 2014). In sharp contrast to conventional genetic transformation, VIGS is relatively quick and easily accessible. It attains a moderately high efficiency with a simple procedure requiring no special technical facilities (Burch-Smith et al., 2004). In addition, the VIGS-derived phenotypes usually occur at 1-2 weeks after inoculation, which substantially shortens the experimental period needed for gene functional analysis (Baulcombe, 1999). Moreover, VIGS is genotype-independent, making it widely applicable for a broad range of hosts (Burch-Smith et al., 2004;Wang & Waterhouse, 2002). For these reasons, it has gained ever-increasing popularity in gene functional studies especially for plant species recalcitrant to genetic transformation (Dommes et al., 2019;Purkayastha & Dasgupta, 2009;Ramegowda et al., 2014;Senthil-Kumar & Mysore, 2011a). Notably, it has also been successfully employed in large-scale forward and reverse functional genomics assays in tobacco and other Solanaceae species (Becker & Lange, 2010;Huang et al., 2012;Senthil-Kumar & Mysore, 2011a).
In the past two decades, approximately 40 VIGS vectors have been developed and successfully applied in a wide range of plant species (Shi et al., 2021). The vast majority of VIGS vectors are reportedly employed in dicot species. By contrast, only a few number of VIGS vectors have been successfully implemented in monocots, especially in grass species (Dommes et al., 2019;Ramanna et al., 2013;Scofield & Nelson, 2009). Among the various VIGS applications, the vector derived from the Tobacco Rattle Virus (TRV) is one of the most widely used vectors in previous studies (Dommes et al., 2019;Shi et al., 2021). TRV is a positive-strand RNA virus composed of bipartite RNA strands (i.e., RNA1 and RNA2). RNA1 encodes two replicase proteins (134 and 194 kDa), a movement protein (29 kDa), and a cysteine-rich protein (16 kDa), which are required for viral replication and movement (MacFarlane, 1999). RNA2 encodes a coating protein (22.4 kDa) associated with virion formation and two structural proteins (29.4 and 32.8 kDa) with no essential roles (MacFarlane, 1999). These two structural proteins in RNA2 are usually replaced by an exogenous gene of interest for the purpose of gene functional characterization in plants. The TRV vector holds several outstanding features that merit its broad applications in various plant species. Firstly, it has a wide range of hosts covering both dicot and monocot species (Shi et al., 2021). Secondly, it is highly efficient in knocking down endogenous gene expression via convenient Agrobacterium inoculation (Shi et al., 2021). Thirdly, it is capable of infecting the meristem, and systematically spreads among various tissues/organs (e.g., leaf, root, stem, flower, and silique/fruit) over time while conferring no or very mild viral symptoms Ratcliff et al., 2001). To date, TRVderived VIGS has been extensively applied in a wide range of dicot plant species such as tobacco (Nicotiana benthamiana) (Liu, Schiff, Marathe, & Dinesh-Kumar, 2002), Arabidopsis (Arabidopsis thaliana) (Burch-Smith et al., 2006;Wang et al., 2006), tomato (Solanum lycopersicum) (Fu et al., 2005;, cotton (Gossypium arboreum) (Qu et al., 2012), pepper (Capsicum annuum) (Chung et al., 2004), potato (Solanum tuberosum) (Brigneti et al., 2004), and petunia (Petunia hybrida) (Chen et al., 2005), as well as a few of monocot species including wheat (Triticum aestivum), and maize (Zea mays) (Zhang et al., 2017).
Miscanthus is a C 4 bioenergy crop with high photosynthesis efficiency, high biomass productivity, low requirement of water and nutrients, and excellent environmental adaptability, which make it very suitable for K E Y W O R D S miscanthus, phytoene desaturase (PDS), sprout infiltration, tobacco rattle virus (TRV), transient gene silencing, virus-induced gene silencing (VIGS) cultivation on marginal land (Wang, Kong, et al., 2021). Its biomass is rich in lignocellulosic materials that serve as an ideal feedstock for the next generation of bioethanol production (Lee & Kuan, 2015). The genome sequences of M. lutarioriparius, M. sinensis, and M. floridulus have been released (Miao et al., 2021;Mitros et al., 2020;Zhang et al., 2021), and abundant transcriptome datasets are publicly available (Hu, Xu, et al., 2017;Xing et al., 2018). These valuable resources laid the foundation for functional genomics studies in Miscanthus. However, Miscanthus is recalcitrant to genetic transformation, which significantly hinders the implementation of large-scale functional genomics studies. Although successful transformation and tissue culture regeneration of the βglucuronidase (GUS) reporter gene have been recently reported in M. sinensis, the transformation efficiency is relatively low, and the procedure is lengthy (at least 6 months) and cumbersome (Hwang et al., 2014;Wang et al., 2011;Wu et al., 2021). Therefore, a rapid and efficient transient VIGS system is an urgent need for functional gene analysis in Miscanthus. Unfortunately, to the best of our knowledge, no VIGS system has hitherto been developed for Miscanthus.
To bridge this knowledge gap in transient gene silencing studies and the lack of a VIGS toolkit, we established an efficient VIGS system for Miscanthus gene functional studies. An efficient TRV-based VIGS system was successfully developed in Miscanthus using the Phytoene Desaturase (PDS) gene as a visual indicator for gene silencing. We further optimized the parameters affecting the VIGS efficiency including seed germination stage, Agrobacterium density, vacuum infiltration time, coincubation time, and growth temperature. To verify the efficacy of the VIGS system, we successfully characterized the function of a MYB transcription factor MsMYB112 in salt stress tolerance in Miscanthus via the TRV-mediated VIGS approach. The VIGS system established herein is anticipated to greatly facilitate functional genomics studies in Miscanthus and other bioenergy crops.

| Plant materials and growth conditions
Seeds of M. lutarioriparius (accession No. A0110) and M. sinensis (accession No. C0148) were harvested from the Miscanthus germplasm library in Changsha, Hunan province, China. The tassels of natural maturity were dried at 37°C overnight in an air-forced oven. The seeds were harvested by mechanical rubbing of tassels on a coarse surface, and stored at −20°C until use. After surface sterilization with 75% ethanol plus 0.01% Triton X-100, the wild type (WT) seeds or TRV-VIGS infiltrated seeds were sown onto water-saturated filter paper for germination, then transferred to the plug tray (28 × 54 × 6.5 cm, 72 plugs per tray) with soil composed of peat, coconut bran, perlite, humus, and vermiculite (12:3:2:2:1). Seedlings were kept in a controlled-environment growth chamber at 25 °C with a photoperiod of 16 h light/8 h dark and relative humidity kept at 60%-70%. The light intensity was kept at 110 mmol m −2 s −1 provided by tubular LED lights (T5, PHILIPS).
The MsMYB112 overexpression lines and WT (Col-0) seeds were surface sterilized with 75% ethanol containing 0.01% Triton X-100 for 5 min. After rinsing with sterile distilled water, seeds were placed onto 1/2 MS agar plates and stratified at 4°C for 3 days. The seeds were germinated in a growth chamber at 21°C under a long-day photoperiod (16 h light vs. 8 h dark). The 7-day-old seedlings were transferred to soil and grew in the same conditions as above.

| Multiple sequence alignments and phylogenetic analysis
The full-length sequences of PDS genes of Miscanthus were obtained by tBLASTn search against the M. sinensis genome at Phytozome (https://phyto zome-next.jgi. doe.gov/info/Msine nsis_v7_1) with N. benthamiana PDS protein sequence as the query bait. The multiple sequence alignments of full-length PDS protein sequences of Arabidopsis, rice, tobacco and M. sinensis was carried out with Clustal X (v2.0). The aligned sequences were further edited with the BioEdit software (v7.1). The Maximum Likelihood (ML) phylogenetic tree was constructed with MEGA X using full-length PDS protein sequences, the bootstrap value was set as 1000. A threshold of credibility set at 50% was shown on each node.

| Vector construction
The TRV-derived VIGS vectors pTRV1 and pTRV2 used in the study were constructed as part of a previous study . To simultaneously knock-down the expression of PDS in the multiple gene copies present in Miscanthus, a 300 bp fragment targeted to the consensus region of four Miscanthus PDS coding sequences was amplified by Polymerase Chain Reaction (PCR) using cDNA template from M. sinensis leaves, with the primer pair (5′ to 3′) of Forward: ttctctagaaggcctccatg-gCACTC AAT TTC ATA AAT CCT GATGAGT, and Reverse: gagacgcgtgagctcggtaccGCTTG AAG ATA TCA ACT GGT GTTGC (homologous flanking sequence in lowercase and underlined). Subsequently, the antisense MsPDS fragment was ligated into the pTRV2 vector at the Kpn I and Nco I restriction sites via In-fusion homologous recombinant reactions using the ClonExpress Ultra One Step Cloning Kit (Vazyme Biotech, China) to generate the pTRV2-MsPDS construct. To silence the expression of MsMYB112 (Misin14G128900) in Miscanthus, a 300 bp gene-specific fragment was amplified by PCR using cDNA template from M. sinensis leaves with the primer pair (5′ to 3′) of Forward: gagacgcgtgagctcggtaccTGACG TCC AGC TGC ATGCG and Reverse: ttctctagaaggcctccatggTCGGT AAT GCT CTC GGA GAAGC (homologous flanking sequence in lowercase and underlined). The resulting MsMYB112 fragment was sub-cloned into the pTRV2 vector at the introduced Kpn I and Nco I restriction sites. The recombinant construct was designated pTRV2-MsMYB112.
To construct the MsMYB112 overexpression vector, the full coding sequence of MsMYB112 was amplified by PCR from cDNA template of M. sinensis leaves using the primers fused with homologous flanking sequences (in lowercase and underlined): Forward: 5′-gtcgacacgtggatccATGG CCGCAACACAG-3′, and Reverse: 5′-cggccgcgccggatccG TCTTTGGATGTGTA-3′. The amplicon was further subcloned into pCAMBIA1301 under the control of CaMV 35S promoter at the introduced BamH I restriction site via In-fusion homologous recombinant reactions. The recombinant construct was infiltrated into Arabidopsis Col-0 via Agrobacterium-mediated floral dip method. The positive transgenic lines were screened on 1/2 MS plates containing 25 mg L −1 hygromycin (Hyg).

| VIGS assay
The pTRV1, pTRV2, pTRV2-MsPDS, and pTRV2-MsMYB112 constructs were individually transformed into Agrobacterium tumefaciens strain GV3101 via electroporation. After confirming the presence of the correct plasmid in the colony with PCR, the positive transformants were cultured in 5 mL of YEB medium containing 25 μg mL −1 Rifampicin (Rif) and 50 μg mL −1 Kanamycin (Kan). The culture was scaled up to 100 mL YEB containing the same antibiotics with shaking at 220 rpm. To determine the optimal Agrobacterium density, the Agrobacterium cells were adjusted to different optical density at 600 nm (OD 600 ) values of 0.4, 0.7 and 1.0. A mixture of pTRV1 with pTRV2, or pTRV2-MsPDS, or pTRV2-MsMYB112 at a ratio of 1:1 were prepared for subsequent VIGS inoculation. These combinations were designated pTRV:00, pTRV:MsPDS, and pTRV:MsMYB112, respectively. The Agrobacterium cells were collected by centrifugation at 5000 rpm, and the pellets were re-suspended in the infiltration medium consisting of 100 μM acetosyringone (AS), 3.3 mM cysteine (Cys), and 5 mg L −1 Tween 20 as described previously (Zhang et al., 2017). The Agrobacterium cultures were stabilized for 3 h at room temperature before the inoculation.
The M. lutarioriparius plants at the four-leaf stage were subject to leaf-rubbing VIGS inoculation. The newly expanded leaves at the four-leaf stage were mechanically rubbed with fine sand paper, then covered with cotton saturated with the mixture of Agrobacterium harboring pTRV:00 or pTRV:MsPDS. The inoculated plants were incubated at 20°C in dark for 24 h, then transferred to grow at 25°C with a 16 h light vs. 8 h dark photoperiod. The newly developed leaves were observed for photobleaching phenotypes after 14 days of inoculation.
To investigate the effect of seed germination stages on VIGS efficiency, Miscanthus seeds were first surfaced sterilized with 75% ethanol plus 0.01% Triton-X100 for 5 min, then placed onto filter paper saturated with sterile water for germination. The seeds were germinated at 30°C for 48 h and 60 h, and developed into sprouts of 0.5-1.0 mm and 1.0-2.0 mm in length, respectively. The ungerminated seeds, and seedlings with sprouts of different lengths, were subjected to the VIGS assay.
The germinating sprouts with different sizes (0, 0.5-1.0, and 1.0-2.0 mm) were inoculated with Agrobacterium harboring pTRV:00 or pTRV:MsPDS at various concentrations (OD 600 = 0.1, 0.4, 0.7, 1.0, and 1.5). The inoculators were subject to vacuum infiltration at a − 95 kPa pressure for different duration times (1, 30, 60, 90, and 120 min). After vacuum infiltration, the seedlings were co-incubated with Agrobacterium for different duration of time (0, 5, 15, and 25 h) at different ambient temperatures (18, 22, and 25°C). Afterwards, the inoculated sprouts were rinsed with distilled water and transferred to grow in soil in a controlled growth chamber with the same conditions as described above. The photobleaching frequency of the seedlings was counted and recorded after 14-days growth.
To determine the optimum parameters for the inoculation procedure of the Miscanthus VIGS assay, an orthogonal experiment consisting of four factors at three levels was designed (Table 1). The setup of four factors at three levels are Agrobacterium OD (0.4, 0.7, and 1.0), vacuum infiltration duration (30, 60, and 120 min), sprouts in different sizes (0, 0.5-1.0, and 1.0-2.0 mm), and co-incubation time (0, 5, and 14 h). The inoculation and infiltration procedure was kept the same as described above.
For each treatment, the experiment was carried out with at least three biological replicates and no less than 20 seeds/sprouts were used for each repeat. The gene silencing efficiency was calculated by the ratio of photobleaching occurrence out of the total inoculated seeds/sprouts at 2 weeks after infection.

| Gene expression analysis
Total RNA was isolated from Miscanthus leaves of pTRV:00-, pTRV:MsPDS-or pTRV:MsMYB112-inoculated plants with Trizol (TransGen, China) in accordance with the manufacturer's instruction. The quality and concentration of total RNA was analyzed with the NanoDrop 2000 (Thermo Fisher Scientific). The first-strand cDNA synthesis was carried out with 2.0 μg RNA and Oligo (dT) primer using the TransScript® One-Step gRemoval and cDNA Synthesis SuperMix kit (TransGen, China).

| Chlorophyll content measurement
Miscanthus leaf samples (20 mg) were grounded into fine powder in liquid nitrogen. The chlorophyll was extracted with 0.4 mL of 100% methanol at 4°C overnight. The contents of chlorophyll a (Chla) and b (Chlb) were calculated by determining the absorbance at 653 and 666 nm following a previously described method (Lichtenthaler & Wellburn, 1983).

| Salt stress tolerance assay
The two MsMYB112 over-expression lines and WT (Col-0) seeds were germinated on 1/2 MS agar plates supplemented with various concentrations of NaCl (0, 125, and 150 mM). The percentage of green cotyledons was calculated and representative photos were taken at 10 days after sowing.
Miscanthus seedlings inoculated with pTRV:00 or pTRV:MsMYB112 were grown in soil for 2 weeks. Then, the seedlings were transferred to hydroponic culture with 1/2 Hoagland's solution supplemented with 100 mM NaCl for salt stress treatment. After 5 days of stress treatment, the survival rates of the seedlings were recorded, and representative photos were captured. For each experiment, three biological replicates were performed with at least five seedlings.

| Statistical data analysis
The orthogonal experiment with four factors at three levels was analyzed using the Orthogonal Experimental Design Assistant II software (Sharetop Software Studio, v3.1). The significance of differences among treatments was determined using two-tailed Student's t-test or oneway analysis of variance (ANOVA) with Tukey's multiple comparison test. All statistical analyses were performed with the SPSS software (v18.0).

Miscanthus
To easily discern the gene silencing effect, various visual indicators are commonly employed in VIGS applications. The Phytoene Desaturase (PDS) gene is one of the most used visual indicators. PDS acts as a key enzyme required for the synthesis of carotenoid in plants (Chamovitz et al., 1993). Silencing or knocking-down of PDS expression results in a visually discernable photobleaching phenotype in leaves (Kumagai et al., 1995). Since no information of PDS orthologs was available for Miscanthus, we first mined Miscanthus PDS orthologs by tBLASTn searching against the genome sequences of M. sinensis using tobacco PDS gene as the query. To this end, four Miscanthus PDS genes: Misin01G458800, Misin02G453200, Misin11G228100, and Misin12G241800 were obtained. Multiple sequence alignments revealed that PDS genes in Arabidopsis, rice and tobacco shared high levels of sequence identity with its four orthologs in Miscanthus (Figure 1a). In addition, the protein sequence identities between the four MsPDS family members ranged from 53.5% (Misin01G458800 vs. Misin02G453200) to 98.1% (Misin11G228100 vs. Misin12G241800) (Figure 1b). Phylogenetic analysis showed that the four MsPDS formed a separate clade, with Misin11G228100 and Misin12G241800 sharing the highest percentage identity (Figure 1c). Moreover, the four MsPDS genes also shared a relative high sequence similarity of nucleotides in the coding region. To simultaneously knock down the expression of four MsPDS genes in Miscanthus, we constructed the pTRV2-MsPDS VIGS vector targeted to a highly conserved region across the MsPDS genes ( Figure S1).

| Factors affecting VIGS gene silencing efficiency in M. sinensis
To establish an efficient VIGS system for Miscanthus, we examined the effects of various factors on gene silencing efficiency including agro-inoculation methods, seed germination stages, Agrobacterium density, vacuum infiltration duration, co-incubation time, and growth temperature. The pTRV:MsPDS was agro-inoculated into Miscanthus leaves or germinating sprouts, and the effect of various factors on PDS gene silencing efficiency was evaluated by the occurrence of photobleaching phenotypes.
We firstly tried the leaf rubbing method for agroinoculation with Agrobacterium harboring pTRV:MsPDS at an OD 600 of 0.7. The inoculated plants grew normally with healthy green leaves without distinguishable differences from the wild type (WT). After 14 days of inoculation, none of the fully expanded leaves exhibited photobleaching phenotypes ( Figure S2). This suggested that leaf rubbing method may be not suitable for agroinoculation of TRV virus in Miscanthus.
We next sought to explore better methods for agroinoculation using germinating seeds at different stages. A simplified flow chart depicting the VIGS procedure is shown in Figure 2a. The un-germinated seeds and germinating sprouts (0.5-1.0 mm, and 1.0-2.0 mm) were inoculated with Agrobacterium harboring pTRV:MsPDS at OD 600 of 1.0. After 10 days of inoculation, the Miscanthus seedlings derived from different inoculators displayed various extents of photobleaching phenotypes. The agroinoculated germinating sprouts (0.5-1.0 mm) displayed the highest PDS gene-silencing efficiency (~40%), while the un-germinated seeds had the lowest gene-silencing efficiency (~3.5%) (Figure 2b). This result indicated that sprouted seeds are more suitable for TRV-mediated VIGS for Miscanthus.
Subsequently, we investigated the co-incubation time (0, 5, 15, and 25 h) and growth temperature (18, 22, and 25°C) on the effect of gene silencing. The seeds and sprouts without co-incubation had a gene silencing efficiency lower than 12.5%. Co-incubation for 5 h significantly improved the gene silencing efficiency to approximately 45%. However, longer incubation times of 15 or 25 h had no additional improvement on gene silencing efficiency (Figure 2c). The growth temperature range explored had no significant influence on gene silencing efficiency. No significant difference in gene silencing efficiency was discernable among treatments of different temperatures (Figure 2d).
To examine the effect of initial Agrobacterium culture titer on the gene-silencing efficiency, we inoculated the germinating sprouts (0.5-1.0 mm) with Agrobacterium suspensions of different OD 600 (0.1, 0.4, 0.7, 1.0, and 1.5). The highest gene-silencing efficiency (~65%) was achieved with Agrobacterium at OD 600 of 0.7, whereas the agrobacterium at OD 600 of 0.1 gave the lowest gene silencing efficiency (~8.5%). Inoculation with Agrobacterium with OD 600 higher than 1.0 significantly decreased the gene-silencing efficiency (Figure 2e). This suggested that the Agrobacterium of OD 600 of 0.7 is the optimum density amongst the five different OD 600 conditions for MsPDS gene silencing.
To test whether the period under vacuum during the infiltration has an effect on gene silencing efficiency, we applied different vacuum infiltration duration times (1, 30, 60, 90, and 120 min) for 0.5-1.0 mm sprouts inoculated with Agrobacterium (OD 600 of 0.7). The gene silencing efficiency was lower than 15% when vacuum infiltration was applied for 1 min and 30 min. However, the gene-silencing efficiency was significantly increased when the vacuum infiltration time was prolonged, and the highest gene-silencing efficiency (~50%) was attained by vacuum infiltration of 90 min. When the vacuum infiltration was extended to 120 min, the gene-silencing efficiency substantially declined to ~35% (Figure 2f). This implicated that vacuum infiltration for 90 min gives most efficient TRV infection amongst the five vacuum infiltration durations explored for Miscanthus.

M. sinensis
In order to develop an optimized procedure of TRVmediated VIGS for Miscanthus, we set up an orthogonal experiment taking into consideration of all the above factors. A total of nine treatments, consisting of four factors at three levels, were employed: Agrobacterium OD 600 (0.4, 0.7, and 1.0), seed germination stages (un-germinated, 0.5-1.0 mm sprouts, and 1.0-2.0 mm sprouts), vacuum infiltration time (30, 60, and 120 min), and co-incubation time (0, 5, and 14 h) ( Table 1).
Analysis of the orthogonal results revealed that the size of germinating sprouts exerted the most influential effect on the VIGS efficiency, followed by co-incubation time and Agrobacterium OD 600 . By contrast, the vacuum infiltration time had the least effect on the VIGS efficiency. The highest gene silencing efficiency (~76%) was attained with the following combination of factors and levels: sprouts of 1.0-2.0 mm, Agrobacterium OD 600 of 0.4, vacuum infiltration of 90 min, and co-incubation time of 5 h ( Table 2).
The variance analysis of the orthogonal experiment indicated that the effects on the VIGS efficiency of all the factors examined except vacuum infiltration time reached statistically significant levels (p < 0.05, and p < 0.01). It was noteworthy that the effect of geminating sprout size reached a higher significance threshold (p < 0.01). In contrast, the effect of vacuum infiltration time (30, 60, and 90 min) showed no significant contribution to the VIGS efficiency (Table 3).
Among the inoculated seedlings showing MsPDS genesilencing phenotypes, almost 90% of seedlings developed leaves and shoots with complete photobleaching symptoms, while only a minor percent (~10%) of seedlings displayed partial photobleaching with patchy or streaked albino leaves (Figure 3a-c). The chlorophyll content was decreased by ~85% (ranging from 81% to 87%) in the photobleaching seedlings compared to the vector control plants (Figure 3d). Accordingly, RT-qPCR analysis showed that the expression of MsPDS was reduced by ~65% (ranging from 61% to 67%) in seedlings with photobleaching phenotypes relative to those of the vector control plants (Figure 3e).  (1, 30, 60, 90, and 120 min). Data are means ± standard deviation (SD) of three independent biological replicates. Different letters denote significant differences among treatments of each factor as determined by Tukey's multiple range test (p < 0.05).

M. lutarioriparius
To examine if the established TRV-based VIGS technique also works well in other Miscanthus species, we tested its efficacy in M. lutarioriparius using the above optimized procedure. A comparable gene-silencing efficiency of MsPDS was attained (Figure 4) between M. sinensis and M. lutarioriparius. Likewise, more than 90% of the infected seedlings exhibited systematic photobleaching phenotypes, while only a minority (~10%) exhibited patchy or streaked photobleaching leaves (Figure 4a-c). Quantification of chlorophyll content and MlPDS expression in seedlings with photobleaching phenotypes suggested that the leaf chlorophyll content was substantially reduced and the endogenous MlPDS expression was successfully knocked down by VIGS (Figure 4d,e). This result implicated the potential application of TRV-mediated VIGS in various Miscanthus species.

MsMYB112 in salt stress tolerance
To test the efficacy of the established TRV-mediated VIGS system for Miscanthus gene functional analysis, we attempted to knock down the expression of MsMYB112 via TRV-mediated VIGS to characterize its potential role in abiotic stress tolerance.
As no functional information was available for MsMYB112, we first characterized its potential role in salt stress response in transgenic Arabidopsis (Figure 5a). Two Arabidopsis homozygous overexpression lines (MsMYB112-OX-6 and MsMYB112-OX-7) with high MsMYB112 mRNA levels were used in the salt tolerance assay ( Figure S3). The seedlings of MsMYB112-OX-6 and MsMYB112-OX-7 exhibited comparable growth properties to the WT under normal growth conditions. However, when exposed to 100 and 125 mM NaCl, the percentage of seedlings with green cotyledons of the MsMYB112 T A B L E 2 Optimization of VIGS procedure with the orthogonal test.  overexpression lines was significantly lower compared to WT (Figure 5a,b). This result indicated that MsMYB112 overexpression exacerbates the response to salt stress. We subsequently used the VIGS approach to identify any role for MsMYB112 in the salt stress response in Miscanthus. To this end, the 300 bp gene-specific fragment of MsMYB112 was targeted for silencing by TRVmediated VIGS. A total of 17 infected lines were obtained. The expression of MsMYB112 was decreased by ~70% in the pTRV:MsMYB112 infected lines relative to the vector control plants, indicating the successful silencing of MsMYB112 expression in Miscanthus (Figure 5c). The pTRV:MsMYB112 lines exhibited growth properties comparable to the vector control plants under normal growth conditions (Figure 5d). When subjected to 100 mM NaCl treatment, the majority of vector control plants exhibited serious leaf wilting phenotypes and eventually died after the 5-day of salt treatment. By contrast, more than 80% of pTRV:MsMYB112 lines kept healthy green leaves with only minor necrosis on leaf tips and eventually survived the salt stress (Figure 5d). Accordingly, the survival ratio of pTRV:MsMYB112 lines was significantly larger than that of the vector control plants after the salt treatment (Figure 5e). This indicated that silencing of MsMYB112 enhanced salt stress tolerance in Miscanthus. MsPDS expression in vector control plant was set as 1. Data are mean and SD from three biological replicates. Asterisks indicate significant differences based on Student's t-test (*p < 0.05, **p < 0.01). Ramanna et al., 2013;Ramegowda et al., 2014;Scofield & Nelson, 2009). To facilitate functional genomics studies for the bioenergy crop Miscanthus, we established a robust and highly efficient TRV-mediated VIGS system using germinating sprouts via optimization of various influencing factors. To the best of our knowledge, this is the first TRV-mediated VIGS system reported for Miscanthus species.

| DISCUSSION
One of the most important factors that determine the efficiency of VIGS lies in the successful delivery of viral vectors into plant cells (Ramanna et al., 2013;Shi et al., 2021). During the last two decades, different inoculation methods have been developed for transferring VIGS vectors into plant cells in various plant species. Leaf infiltration is one of the most frequently used methods especially in dicot plant species (e.g., tobacco and tomato) Liu, Schiff, Marathe, & Dinesh-Kumar, 2002). Various leaf infiltration methods including infiltration with needle-less syringe, vacuum infiltration, and high-pressure spraying are exploited for different combinations of plants and VIGS vectors. However, the efficiency of leaf infiltration is relatively lower for some dicots (e.g., soybean) (Nagamatsu et al., 2007) and even more so for monocots (e.g., barley, maize, and wheat) (Holzberg et al., 2002;Mei et al., 2016;Scofield et al., 2005). Agrodrench provides an alternative inoculation method for efficient VIGS in roots in some Solanaceous species (e.g., tobacco, and tomato) (Ryu et al., 2004). In addition, direct injection of Agrobecterium has been developed to successfully induce VIGS in flowers and fruits in several plant species such as tomato, eggplant, and walnut (Fu et al., 2005;Wang & Fu, 2018;Wang, Huang, et al., 2021).
In addition to the above infiltration methods, agroinoculation via co-incubation with germinating spouts has proven to be an efficient and convenient method for VIGS assays (Yan et al., 2012;Zhang et al., 2017). Infiltration with sprouts has several advantages over leaf infiltration and agrodrench methods (Yan et al., 2012). Firstly, sprout infiltration is easily applicable and time saving. The germinated sprouts are infiltrated under vacuum, and then directly sowed in soil. Moreover, the gene silencing phenotype is visible for 1-2 weeks after the sprout infiltration Data are mean ± SD in triplicate. At least five seedlings were used for each replicate. Asterisks denote significant differences between the transgenic lines and control plants as determined by Student's t-test (*p < 0.05, **p < 0.01). (Figures 3 and 4). Compared with leaf infiltration, sprout infiltration significantly reduces the experimental period from 4-5 weeks to 1-2 weeks (Yan et al., 2012;Zhang et al., 2017). Additionally, there is no excess requirement for plant growth and maintenance prior to sprout infiltration. Secondly, sprout infiltration is highly efficient and induces systematic gene silencing across various tissues/ organs. By contrast, VIGS via leaf infiltration usually induces gene silencing in leaves (Burch-Smith et al., 2006;Liu, Schiff, Marathe, & Dinesh-Kumar, 2002;Wang et al., 2006). Therefore, sprout infiltration is extremely useful to study the functional roles of genes that are associated with earlier seedling development, or the control and development of reproductive organs such as flowers and fruits (Yan et al., 2012). Thirdly, VIGS mediated by sprout infiltration is particularly amenable for large scale reverse or forward genomics studies owing to the above advantages over the leaf infiltration method (Yan et al., 2012).
It is generally accepted that the monocot species, especially the grasses, are by far the most recalcitrant to VIGS infection (Ramanna et al., 2013;Scofield & Nelson, 2009). Therefore, it was not unexpected that we failed to infect Miscanthus via leaf infiltration ( Figure S2). However, we discovered that sprout infiltration is an attractive alternative for VIGS in Miscanthus (Figures 3 and 4). Apart from the infiltration method, many other factors including virus type, developmental stage, Agrobacterium density, vacuum infiltration, co-incubation time, and temperature have substantial effects on VIGS efficiency (Shi et al., 2021). After optimization of various factors affecting VIGS efficiency through a series of single factor and multifactor orthogonal experiments ( Figure 2 and Table 2), the size of germinated sprouts (1.0-2.0 mm) represented the most critical factor determining the VIGS efficiency in Miscanthus, followed by suitable Agrobacterium OD 600 , moderate co-incubation time, and appropriate vacuum infiltration time (Table 2). Temperature has been deemed as one of the most important factors affecting gene silencing efficiency in VIGS. Lower temperature has been implicated in conferring reduced anti-viral defense of plants, thus facilitating viral spreading in plant cells. Several previous studies showed that the plant growth temperature significantly affects the VIGS efficiency (Chung et al., 2004;Wang et al., 2006;Zhang et al., 2017). However, our results showed that although the VIGS efficiency was slightly higher at 18°C compared to those of 22°C and 25°C, the difference was not significant among different temperatures examined (Figure 2d). The effect of growth temperature on VIGS efficiency in Miscanthus warrants further investigation.
Although VIGS usually induces transient gene silencing in various plants, there is some evidence that the gene silencing effect can be transmitted to progeny (Marton et al., 2010;Senthil-Kumar & Mysore, 2011b). The duration of the gene silencing effect generally lasts for 1-3 months, depending on the combination of virus type and plant species (Ratcliff et al., 2001;Ryu et al., 2004;Wang et al., 2006). The infected Miscanthus plants developed systemic photobleaching phenotypes, which seriously affected the vitality of the seedlings and caused a high percentage of mortality (Figures 3 and 4). Since the majority of pTRV:MsPDS infected Miscanthus plants were not survived, we currently could not accurately estimate the duration time of the photobleaching effect. The lasting of the gene silencing effect of TRV-mediated VIGS in Miscanthus will be further investigated in future studies. It can be expected that if the gene silencing effect persists sufficiently long in Miscanthus, germplasm generated by the VIGS technique could be genetically maintained indefinitely via clonal propagation. Similar results have been reported in vegetatively propagated plants, including potato (Solanum spp.) and ginger (Zingiber officinale) (Faivre-Rampant et al., 2004;Renner et al., 2009). Therefore, the VIGS technique may hold a potential application in the genetic improvement of asexually propagated species such as Miscanthus.
MYB transcription factors play important roles in various abiotic stress responses in plants (Wang, Niu, & Zheng, 2021). We used the salt-inducible MsMYB112 gene to test the feasibility of the VIGS system described here in Miscanthus gene function analysis. The expression of MYB112 was successfully knocked down using TRV-mediated VIGS, and the pTRV:MsMYB112 infected Miscanthus plants displayed increased tolerance to salt stress ( Figure 5). This illustrates the successful application of the TRV-mediated VIGS system in the analysis of Miscanthus gene function. The VIGS system established here can now be implemented for high-throughput functional genomic studies in Miscanthus.

| CONCLUSIONS
In this study, we developed a robust and efficient VIGS system using germinating sprouts and PDS gene as a visual indicator in two Miscanthus species. We optimized the various factors affecting gene silencing efficiency and achieved a maximum gene silencing efficiency of ~76% for MsPDS. Moreover, the efficacy of the VIGS system was verified by knocking-down the MYB transcription factor MsMYB112 to achieve enhanced salt stress tolerance. In summary, we established a convenient and efficient TRV-mediated VIGS system for transient gene functional analysis in Miscanthus. The VIGS system developed in the current study will substantially facilitate future functional genomic studies in Miscanthus.