Silicon‐mediated multiple interactions: Simultaneous induction of rice defense and inhibition of larval performance and insecticide tolerance of Chilo suppressalis by sodium silicate

Abstract The rice striped stem borer (SSB, Chilo suppressalis) is one of the most destructive pests of rice plants. Si‐mediated rice defense against various pests has been widely reported, and sodium silicate (SS) has been used as an effective source of silicon for application to plants. However, there is quite limited information about the direct effects of Si application on herbivorous insects. SSB larval performance and their insecticide tolerance were examined after they had been reared either on rice plants cultivated in nutrient solution containing 0.5 and 2.0 mM SS or on artificial diets with 0.1% and 0.5% SS. SS amendment in either rice culture medium or artificial diets significantly suppressed the enzymatic activities of acetylcholinesterase, glutathione S‐transferases, and levels of cytochrome P450 protein in the midgut of C. suppressalis larvae. Larvae fed on diets containing SS showed lower insecticide tolerance. Additionally, RNA‐seq analysis showed that SS‐mediated larval insecticide tolerance was closely associated with fatty acid biosynthesis and pyruvate metabolism pathways. Our results suggest that Si not only enhances plant resistance against insect herbivore, but also impairs the insect's capacity to detoxify the insecticides. This should be considered as another important aspect in Si‐mediated plant–insect interaction and may provide a novel approach of pest management.

Silicon (Si) is the second most common element after oxygen in the earth's crust (Epstein, 1994). In addition, Si has been listed as a "beneficial substance" by the International Plant Nutrition Institute (IPNI, 2015), which has been widely reported to improve plant resistance against both abiotic and biotic stresses (Bhat, Nazir, Mahajan, Zargar, & Deshmukh, 2019;Liang, Nikolic, Bélanger, Haijun, & Alin, 2015;Liang, Sun, Zhu, & Christie, 2007;Ma, 2004;Meharg & Meharg, 2015;Wang et al., 2017). In the performance assay, the addition of Si to maize resulted in increased larval mortality of the true armyworm, Pseudeletia unipuncta (Haworth) compared with the maize plants without Si (Moise, McNeil, Hartley, & Henry, 2019). Therefore, application of Si is a potential management method to control a wide range of pests including leaf-chewing (Han, Lei, Wen, & Hou, 2015;Ye et al., 2013), sap-feeding (Dias et al., 2014;Goussain, Prado, & Moraes, 2005), and stem-boring insects (Hou & Han, 2010;Kvedaras & Keeping, 2007). However, the results of foliar-applied Si on plant resistance against biotic stress such as pests sometimes are considered controversial because current evidence suggests that Si needs to be absorbed by plant roots to trigger systemic resistance (Coskun et al., 2018). For enhanced resistance to pests by application of Si to plants, an alternative explanation is that insects could directly consume soluble Si which may have direct effects on insect physiology. However, little information has been drawn on the direct effect of Si on insects and its related mechanisms. Thus, the scenario beyond Si directly mediating plant-insect interactions deserved further investigation.
Sodium silicate (SS) has been used as an effective source of Si (Heckman, 2013). Application of SS to plants has been shown to influence insect performance. For example, Italian ryegrass (Lolium multiflorum) fertilized with SS demonstrated a reduction in colonization by stem-boring larvae of Oscinella frit compared to the control plants (Moore, 1984). Application of SS to wheat plants significantly reduced preference, longevity, and production of nymphs of Schizaphis graminum (Basagli et al., 2003;Moraes et al., 2004). Similarly, both foliar and soil application of another type of soluble Si (silicic acid) enhanced rice resistance against fall armyworm (Nascimento, Assis, Moraes, & Souza, 2018).
There is growing and compelling evidence that adaptation to toxic host plants has been a factor in the evolution of insecticide resistance in some herbivore species (Alyokhin & Chen, 2017;Ryan & Byrne, 1988). For example, the susceptibility of Spodoptera littoralis to pesticides differs with host plants by impacting detoxification enzyme levels (Abd El-Rahman, Salem, Yacoub, & Naguib, 2019). Because it is possible for SSB larvae to directly consume Si, we hypothesized that SS exposures may also directly impact the pest's ability to detoxify insecticides. Insect herbivores rely heavily on their detoxification enzymes typically including the glutathione S-transferases (GSTs), cytochrome P450 monooxygenases (P450s), and carboxylesterases to overcome the toxicity of allelochemicals in host plants and insecticides (Després, David, & Gallet, 2007;Terriere, 1984).
This work aimed to characterize the role of SS in SSB larval performance, resistance-related enzymes (AChE, GST, and CYP450), differential gene expression, and insecticide tolerance. Results of the present study may expand the current understanding of the beneficial aspects of Si to be used as an environment-friendly agent for pest management purpose.

| Insects
The C. suppressalis population was initially collected in 2016 from rice paddy fields on the campus of Fujian Agriculture and Forestry University (Fuzhou, China) and maintained under laboratory conditions. Larvae were reared on artificial diets at 25 ± 2°C and 70 ± 5% relative humidity with a photoperiod of 16:8 hr (L:D), and artificial diets were prepared according to the protocol of Liu, Li, Han, Peng, and Hou (2012).

| Plants
Rice (O. sativa L. cv. Nipponbare) seeds were surface-sterilized with 2% sodium hypochlorite (v/v) for 20 min and then were rinsed with distilled water twice. The seeds were soaked in Milli-Q water for 2 days and then were placed on moist filter paper for a week in Petri dishes at 25℃. After germination, rice seedlings with uniform size were transferred to the polyethylene pots and hydroponically cultured in a nutrient solution that contains both macronutrients and micronutrients for the growth of rice plants (Yoshida, Forno, Cock, & Gomez, 1976). Plants were grown in a greenhouse with a day:night temperature regime of 32℃ (12 hr): 26℃ (12 hr), 75% relative humidity, and nature daylight.
Nutrient solutions were replenished every 3 days, and rice plants were used for all the experiments 20 days after transplanting.

| Effect of Si application to rice plants on SSB performance and insecticide tolerance
To examine whether Si supplement induced rice defense against SSB larvae, rice plants were exposed to a nutrient solution containing 2.0 mM sodium silicate (SS, Na 2 SiO 3 •9H 2 O). For Si negative treatments, 2.0 mM sodium chloride was added to balance sodium levels. A plastic straw (diameter: 3 cm, length: 6 cm) was used to fix larvae on the rice stem with cotton plugging the both ends to prevent insects from crawling away. Newly molted 3rd-instar SSB larvae (20 ± 5 mg) were selected and allowed to feed on the stem.
Boring behavior was also observed within 90 min by recording the time of penetration and start of boring (Hou & Han, 2010). Each treatment had 48 replicates. Percentage of penetration was calculated as the number of larvae successfully penetrating divided by the total number of larvae inoculated. After 3 days, the larvae were removed and weighted. The Si concentration in the rice stem was measured using molybdenum-blue spectrophotometry method described by Jia, Yang, Zhang, Fang, and Chen (2016). Larvae with similar mass (25 ± 5 mg) were selected and then treated with 2 μl of chlorpyrifos with 98% purity purchased from J&K Scientific Ltd. dissolved in acetone on the larval pronotum. The concentration of chlorpyrifos was 250 μg/ml, which was determined based on the lethal dose (LD 50 ). After 48 hr, the larval mortality was recorded.
Each treatment had three replicates, and each replicate had 20 individuals.

| Direct effects of SS application in artificial diets on SSB growth and tolerance against insecticides
Newly molted 3rd-instar SSB larvae (20 ± 5 mg) were fed on artificial diets containing sodium silicon (SS) at concentrations of 0.1% and 0.5% (w/w) for 48 hr. For the control treatment, larvae were fed on artificial diets without SS addition. Artificial diets were made based on soybean (Glycine max) powder and fresh water bamboo (Zizania caduciflor) (Liu et al., 2012). The pH of diets containing sodium silicate was amended with hydrochloric acid to get as similar pH (8.5 ± 0.5) as the control diet. Twenty newly molted 3rd-instar larvae were used for each treatment, and three independent replicates were conducted for all the treatments.
Larval mass was measured at both the beginning and the end of the bioassay for calculating mass gain. After feeding on artificial diets with SS for 48 hr, larvae with similar mass were selected and used for measuring the mortality against chlorpyrifos as described above.

| Toxicity of mixture of sodium silicate and insecticide to SSB larvae
The mixture of three different doses of chlorpyrifos ranging from 0.015% to 0.025% (w/w) and two different doses of SS at 0.1% and 0.5% (w/w) were applied into artificial diets synchronously. The newly molted 3rd-instar larvae were reared on diets supplemented with the mixture of SS and chlorpyrifos for 48 hr, and then, larval mortalities were recorded.

| Enzyme extraction and assays
The midguts of larvae fed on both rice plants and artificial diets with or without SS amendment were dissected, and midguts from three larvae were pooled as one replicate for the enzyme assay. To measure the activities of the detoxification-related enzymes, the dissected midguts were homogenized in 0.1 M PBS (pH 7.6). The homogenate was centrifuged at 4℃, 14,000 g for 20 min. The supernatant of each sample was used immediately for the measurement of enzyme activities. Detection kits from Nanjing Jiancheng Bioengineering were used to measure the activities of AChE and GST . The protein concentration was determined using Bradford assay (Bradford, 1976). For CYP450 enzyme measurement, ELISA Kit (Shanghai Enzyme-linked Biotechnology Co., Ltd) was used to determine CYP450 levels in insect midguts by The primers used for qRT-PCR using purified insect CYP450 antibody according to manufacturer's instructions, and the concentration of CYP450 in the samples was then determined by comparing the OD value of the samples to the standard curve.

| Total RNA isolation
Total RNA was extracted from the midguts of SSB larvae fed on artificial diets with or without SS amendment by use of Total RNA Extraction Kit (Promega Corporation) according to the manufacturer's protocol. Quantity of RNA was confirmed with Nanodrop (Bio-Rad), and quality of RNA was monitored by electrophoresis gel analysis. Small aliquots of the isolated RNA were stored in −80°C for quantitative real-time PCR (qRT-PCR), and the remaining of RNA from the three replicates was used for RNA sequencing.

| Library preparation and sequencing
Sequencing libraries were generated using NEBNext® Ultra™ RNA Bioinformatics Institute, and paired-end reads were generated.

| Transcriptome sequencing analysis
Raw data in FASTQ format were processed through in-house Perl scripts to remove reads containing adapters, reads containing ploy-N, and low-quality reads, and the sequencing quality was assessed by measuring Q20, Q30, and GC-content of the cleaned data. A de novo transcriptome was assembled using the short-read assembly program Trinity v.2.1.1 (Grabherr et al., 2011). Reads were mapped to unigenes by the Bowtie software (Langmead, 2010), and then, the number of reads mapping to each unigene was converted as FPKM (Fragments Per Kilobase per Million mapped reads) by using RSEM software (Li & Dewey, 2011). To annotate the obtained unigenes, the NCBI nr (nonredundant) and the Swiss-Prot databases with an E-value cutoff of 10 −5 was searched. Functional annotation by KEGG with an E-value cutoff of 10 -10 was conducted by searching against the KEGG databases. The read counts were normalized using the edgeR Bioconductor (Robinson, McCarthy, & Smyth, 2009) with the TMM method (Strorey, 2003), and the DESeq R package provided statistical routines for determining differential expression between the control and SS-treated samples using a model based on the negative binomial distribution. We used "fold changes ≥ 2 and q < 0.005" as the threshold to assess DEGs between the

| Quantitative real-time PCR analysis
To validate the DEG analysis results, qRT-PCR experiments were performed in a 10 µl reaction volume consisting of 5 µl of 2× SYBR GoTaq qPCR Master Mix (Promega Corporation), 0.4 µl of each gene-specific primer (10 µM), 1 µl cDNA equivalent to 50 ng total RNA and sterilized water to reach the final volume. PCR conditions were set as one cycle of 95℃ for 10 min; 40 cycles of 95℃ for 15 s, 55℃ for 30 s, and 72℃ for 30 s. The reference gene elongation factor 1 alpha (EF-1α) was used as the internal control.
A dissociation curve analysis program was performed to check the homogeneity of the PCR product. The relative mRNA levels were normalized against EF-1α using the 2 −ΔΔCt method (Livak & Schmittgen, 2001). All the primers used for qRT-PCR were listed on Table 1.

| Statistical analysis
Gut enzyme assays and Gene expression in the guts of SSB larvae were compared using Student's t test. Larval mass gain and larval mortality to insecticides were determined with one-way ANOVA followed by Fisher's least significant difference (LSD) test.
Minitab 17 (Minitab Inc., State College) was used for all analyses.
A p-value ≤ .05 for difference between means was considered as significant, and all data were presented as the mean ± SEM.

| Larval performance and insecticide sensitivity of SSB fed on rice plants with sodium silicate amendment
To verify whether SS amendment enhances rice resistance to

| Larval performance and insecticide sensitivity of SSB fed on artificial diet with sodium silicate amendment
The mass of larvae fed on artificial diet supplemented with 0.096% sodium chloride, 0.1% sodium silicate, and 0.126% potassium silicate showed no difference to the controls, while the mass of larvae fed on artificial diet containing 0.5% sodium silicate significantly decreased compared to that of other four treatments (Figure 3a, F(4, 95) = 22.4, p < .05). Additionally, 0.096% sodium chloride had the same concentration of sodium as 0.1% sodium silicate, and 0.126% potassium silicate had the same concentration of silicic acid as 0.1% sodium silicate in the artificial diets. After treatment with chlorpyrifos, higher mortality of larvae fed on artificial diets containing 0.1% and 0.5% sodium silicate, as well as 0.126% potassium silicate, was observed compared to the control larvae fed on artificial diet containing 0.096% sodium chloride (Figure 3b, F(4, 10) = 13.29, p = .0005).
These results indicated that it was silicon rather than sodium in the artificial diets that increase larval mortality of SSB to chlorpyrifos.

| Insecticide sensitivity of SSB larvae fed on artificial diet containing the mixture of chlorpyrifos and sodium silicate
In order to investigate whether Si has synergistic effects on insecticide mortality, the mixture of chlorpyrifos and sodium silicate was applied to artificial diets to examine mortality of SSB larvae (Figure 4).
The addition of 0.1% and 0.5% SS showed synergistic effects with chlorpyrifos by increasing the mortality of SSB larvae when the concentrations of chlorpyrifos reached 150 µg/g (F(2, 12) = 14.73, p = .0006) and 200 µg/g (F(2, 12) = 44.11, p < .001). Furthermore, the synergistic effect of 0.5% SS was significantly more effective than that of 0.1% SS, and when concentration of chlorpyrifos reached 250 µg/g, the synergistic effect of 0.5% SS was still very significant (F(2, 12) = 23.7, p < .001). The results showed that Si can enhance the toxicity of chlorpyrifos to SSB larvae.

F I G U R E 4 Silicon enhanced the toxicity of chlorpyrifos to
Chilo suppressalis larvae. Newly molted 3rd-instar larvae were fed on artificial diets containing the mixture of both sodium silicate and chlorpyrifos for 2 days, and then, the larval mortality was measured. Values are mean ± SEM (n = 5). Different letters indicate significant differences (ANOVA, p < .05, Fisher's test)

| Effects of sodium silicate amendment on detoxification enzymes in SSB larvae fed on rice plants and artificial diets
The activities of insect detoxification-related enzymes were assayed in the midguts of SSB larvae fed on rice plants for 3 days and on artificial diets containing sodium silicate for 2 days ( TA B L E 3 Downregulated genes in SSB larvae fed on artificial diet with sodium silicate diets suppressed the activities of detoxification-related enzymes in the midguts of SSB larvae.

| Analysis of DEGs in the midgut of SSB larvae fed on Si-amended artificial diets
To dissect the molecular mechanism underlying the SSB responses to sodium silicate treatment, we utilized RNA-seq to explore differences in gene expression in midgut tissues of SSB larvae maintained on artificial diets with or without SS for 48 hr. Then, the midgut tissues were dissected and the DEGs were analyzed by RNA-Seq. In total, 33 downregulated genes with transformation (log 2 fold ≤ −2) were observed in the midgut of SSB larvae fed on artificial diets containing SS (Table 3). Most DEGs in SS-treated caterpillars compared to control larvae were enriched in biological processes (oxidation-reduction and cellular processes) and molecular functions (structural constituent of cuticle and oxidoreductase activity) ( Figure 5). These results showed that numerous putative metabolic enzymes especially those having oxidoreductase activity were suppressed in SSB subjected to SS treatment, which may help to understand why the detoxification-related enzymes were mostly inhibited.

| Verification of RNA-seq results by qRT-PCR
To validate the gene expression data obtained from RNA-seq, we selected four genes which found to be downregulated in samples from sodium silicate treated larvae for qRT-PCR analysis ( Figure 6). FASN (fatty acid synthase) is a gene in fatty acid biosynthesis pathway, and ALDH (aldehyde dehydrogenase gene), HOGH (hydroxyacylglutathione hydrolase), and DLDH (dihydrolipoamide dehydrogenase) were genes in pyruvate metabolism pathway. In addition, we also found two significantly downregulated genes called CYP6AE60 and EST (Esterase) in SS-treated larvae. All of the downregulated genes from RNA-seq data were also found downregulated in the qRT-PCR analysis (FASN, 7.5-fold; ALDH, 15.7-fold; HOGH, 3.8-fold; DLDH, 12-fold; CYP6AE60, 7.6-fold; and EST, 31.3-fold, respectively). These results indicated that the RNA-seq approach provided reliable DEG data for this assay and demonstrated that SS treatment has significantly adverse effects on these resistance-related genes in the midguts of SSB larvae.

| D ISCUSS I ON
The most significant finding of this study is that larval feeding on ei- Many studies have tried to elucidate the mechanisms for Sienhanced plant resistance to insect herbivores, including the direct defense through increased physical barrier (Han, Li, et al., 2016;Massey, Ennos, & Hartley, 2006)  The different genes are assigned to two main categories: biological process (BP) and molecular function (MF). The y-axis shows the number of genes matched in a category amendment improves rice Si content and impedes stalk penetration and prolongs penetration duration by C. suppressalis larvae, and larvae fed on the control plants without silicon treatment are bigger than those on Si-treated rice plants (Hou & Han, 2010).
Consistent with previous findings, here we found that the growth and penetration rate of SSB larvae on SS-treated rice plants were significantly inhibited compared to control treatments ( Figure 1a,b).
However, insoluble silicon may not influence the insect growth.
The growth rates of the southern armyworm (Spodoptera eridania) at the highest level of insoluble silica treatment in artificial diets showed no difference to the control (Peterson & Coors, 1988).
Similar study has found that addition of wollastonite at rates of up to 3.3% Si into artificial diet had no significant effect on larval weight of Helicoverpa spp. (Stanley, Baqir, & Mclaren, 2014).
Thus, we suspected that soluble Si may be a key biochemical and physiological mediator. Therefore, in this study SSB larvae were reared on artificial diets containing a range of soluble Si levels, and Si significantly inhibited larval growth and tolerance to chlorpyrifos ( Figure 3). It should be highlighted that the insect may be encountering many more different forms of silicon in the plant that in artificial diet, because sodium silicate absorbed by the rice plant occurred the conversion from silicic acid to silica. However, the major form of silicon in the xylem has been identified as monomeric silicic acid in rice, and the high concentration of silicic acid is transiently present in the xylem sap (Mitani, Ma, & Iwashita, 2005;Ma & Yamaji, 2006). Soluble silicic acid is formed by neutralization of SS with acid or hydrolyzed in solution (Danilovtseva, Aseyev, Karesoja, & Annenkov, 2009;Owusu, 1982), so SS could also be converted to silicic acid in artificial diet. Therefore, it may not be SS that is directly affecting the performance of SSB larvae, but the converted compound silicic acid. A previous study showed that exposure to silicic acid enhanced cytokinin biosynthesis and delayed senescence in both Arabidopsis and Sorghum (Markovich et al., 2017). In addition, silicic acid was able to inhibit the feeding of the brown planthopper on rice plants (Yoshihara et al., 1979 pyruvate metabolism pathways, and several genes related to these two pathways were suppressed in SSB larvae (Table 4). Chlorpyrifos is one of the most widely used organophosphorus insecticides (Lee, Strand, & Doty, 2012) and is widely used in paddy fields to control pests such as SSB and planthoppers (Kumar, Praveenkumar, Jeon, & Thajuddin, 2014). In addition to genes related to fatty acid biosynthesis and pyruvate metabolism pathways, many other genes, such as P450s, CarEs, and glutathione transferases, may also mediate Sitriggered responses in SSB. Chlorpyrifos kills insects by competing with acetylcholine and inhibiting the activity of AChE, and organophosphorus resistance has been related to metabolic changes in target insects (Cheng et al., 2010;Fournier & Mutero, 1994;Jiang et al., 2009;Qu et al., 2003). According to our observation, another two significantly downregulated genes EST (Esterase) and CYP6AE60 were found in SS-treated larvae (Figure 6), and these two genes have been found to play a role in insecticide detoxification and xenobiotic metabolism (Ping, 2014;Wang et al., 2014). In insects, many pheromones and other chemical molecules are types of esters that could be hydrolyzed by esterases (Montella, Schama, & Valle, 2012).
Therefore, further functional studies using targeted RNAi or gene editing approaches are needed to evaluate the impacts of silicate enzymes and genes in larval midguts.
Insecticides are often used in combination with organosilicon adjuvants to facilitate the wetting and the spread of droplets on leaves, which result in a more uniform distribution of active ingredients (Srinivasan, Hoy, Singh, & Rogers, 2008;Stevens, Kimberley, Murphy, & Policello, 1993), and organosilicon adjuvants may also show insecticidal activity to several pests (Cocco & Hoy, 2014).
Organosilicon added to chlorpyrifos showed synergistic activity to increase the mortality of C. suppressalis larvae in laboratory by spraying method (Zhang, Chen, Lu, Bao, & Deng, 2009). However, no information is available about whether inorganic silicon such as sodium silicate could be used as an adjuvant to insecticides.
In current study, larvae fed on artificial diets containing Si and chlorpyrifos showed higher mortality compared to larvae fed on diets containing only chlorpyrifos ( Figure 4). Sodium silicate was observed to have synergistic activity to chlorpyrifos, suggesting that application of sodium silicate to either plants or insects has a great potential to effectively control pests using a lower amount of chemical insecticides.
Si has been considered a plant resistance elicitor to control pests (Debona, Rodrigues, & Datnoff, 2017;Keeping & Kvedaras, 2008;Liang et al., 2015;Reynolds, Keeping, & Meyer, 2009;Sakr, 2017;Ye et al., 2013). Current study indicates that sodium silicate has direct effects on insect performance by interfering larval growth and inhibiting activities of detoxification related with enzymes. Additionally, sodium silicate showed a synergistic activity to chlorpyrifos and accelerates mortality of SSB larvae. In conclusion, sodium silicate not only can be used as a resistance inducer for plants, but also can be used as a potential and effective addition to integrated pest management tactics. This expands an interesting and unique dimension to our current understanding of how Si may mediate plant-insect interactions.