Overexpression of OsF3H modulates WBPH stress by alteration of phenylpropanoid pathway at a transcriptomic and metabolomic level in Oryza sativa

The whitebacked planthopper (WBPH), has become a devastating pest for rice crops, causes serious yield losses each year, and urgently needs biological control. Here, we developed a WBPH-resistant rice cultivar by overexpressing the OsF3H gene. A genetic functional analysis of the OsF3H gene confirmed its role in facilitating flavonoid contents and have indicated that the expression of the OsF3H gene is involved in regulation of the downstream genes (OsDFR and OsFLS) of the flavonoid pathway and genes (OsSLR1 and OsWRKY13) involved in other physiological pathways. OxF3H (OsF3H transgenic) plants accumulated significant amounts of the flavonols kaempferol (Kr) and quercetin (Qu) and the anthocyanins delphinidin and cyanidin, compared to the wild type, in response to the stress induced by WBPH. Similarly, OsF3H-related proteins were significantly expressed in OxF3H lines after WBPH infestation. The present study, indicated that the regulation of JA in OxF3H plants was suppressed due the overexpression of the OsF3H gene, which induced the expression of downstream genes related to anthocyanin. Similarly, the OsWRKY13 transcriptional factor was significantly suppressed in OxF3H plants during WBPH infestation. Exogenous application of Kr and Qu increased the survival rates of susceptible TN1 lines in response to WBPH, while decreased the survival rate of first instar WBPHs, indicating that both flavonols exhibit pesticide activity. Phenotypic demonstration also affirms that OxF3H plants show strong resistance to WBPH compared with wild type. Collectively, our result suggested that OsF3H overexpression led to the up-regulation of defense related genes and enhanced rice resistance to WBPH infestation.


Materials and methods
plasmid construction and cloning via the gateway method. The Cheongcheong cultivar was provided by the Plant Molecular Breeding lab, Kyungpook National University, and was used for RNA isolation and OsF 3 H gene amplification. Before sowing, seeds were soaked in an incubator for 3 days at 30 °C, and the water changed each day. Germinated seeds were transplanted to pots, kept in the dark for 3 days, and then transferred to greenhouses. Total RNA was isolated from young leaves of 14-day old rice seedlings, using RNeasy Plant Mini Kits from Qiagen. Standard cDNA was synthesized from total RNA using qPCRBIO cDNA Synthesis Kits, from PCRBIOSYSTEMS, according to the manufacturer's instructions. The full length ORF region of OsF 3 H (483 bp) was amplified by PCR, using gene-specific primers, with four additional nucleotides "CACC" attached to the 5′ end of the forward primer (Supplemental Table S1). The gateway cloning system was followed. Entry clones were generated by inserting template DNA into pENTR/D-TOPO cloning vectors, following the manufacturer's instructions [for details, see the user manual for the pENTR Directional TOPO cloning kit (Invitrogen)]. The TOPO cloning reaction was transformed into DH5α cells using heat shock, and spread on LB media containing kanamycin as a selection marker. Plasmids were isolated from selected colonies and grown overnight in liquid LB media containing kanamycin, using QIAprep Spin Miniprep Kits from Qiagen. Isolated plasmids were double digested with Not1 and Asc1 restriction enzymes, checked on gels (Supplemental Figure S1A), and confirmed by sequencing. For overexpression vector construction, entry clone fragments were inserted into BamH1 and Xho1 sites of the destination vector, pSB11, under the control of a 35S promoter, through LR reaction using Gateway LR Clonase enzyme mix kits (Invitrogen). The constructs were transformed into Agrobacterium cells LBA4404 (Takara) via heat shock and spread on spectinomycin-containing LB media. Plasmids were isolated and double digested with BamH1 and Xho1 restriction enzymes, and the ligation and transformation confirmed by running the digested reactions on gels (Supplemental Figure S1B).  39 , with minor modifications. For callus induction, good quality mature Nagdong rice seeds were selected, gently dehulled, and sterilized with 70% ethanol for 5 min with continuous shaking, and then washed three times with double distilled water. The seeds were then sterilized with 3% sodium hypochlorite for 10 min with shaking, rinsed with sterilized water three times, and dried for 1 h on a clean bench. Dried seeds were inoculated into callus induction medium, with 10-15 seeds per plate, and placed in dark conditions for 12 days. All medium types and compositions are presented in Supplemental  Table S2. After growing for 12 days, calluses were pre-cultured into small pieces of calli and inoculated for 3 days in callus induction media under dark conditions. At the same time, Agrobacterium strain LBA4404 harboring the pSB11 plasmid containing the full length OsF 3 H gene was cultured by selecting single colonies from the transformed plates, inoculating them in 5 ml LB medium containing 50 mg/l spectinomycin, and incubation for 16-18 h in a shaking incubator at 28 °C. Further cultures were prepared in autoclaved baffled flasks containing 100 ml medium, under the same conditions, and the cells harvested when O.D. reached 600. The pelleted cells were resuspended in MS medium fortified with acetosyringone, and the calli were immersed in the suspension for 30 min with continuous shaking. Excess Agrobacterium cells were removed by drying for 30 min on sterilized filter paper and then inoculated into co-cultivation medium for 3 days under dark conditions. Excessive Agrobacterium growth was controlled by washing three times with 500 mg/l carbenicillin, drying for 30 min, and inoculating into first selection medium containing 50 mg/l spectinomycin under light conditions (16/8 photoperiod) for 12 days. The calli were transferred to second selection medium for 10 days, followed by third selection medium for 5 days in light conditions (16/8 photoperiod 43 . Equal amounts of protein were boiled for 5 min and separated on 10% SDS-PAGE at 100v for 150 min, and then transferred to a NC membrane (Whatman Japan) by a semi-dry method running for 90 min at 19v, using a Trans-Blot DS semidry transfer cell (Bio Rad). The membrane was blotted in TBST (0.1% Tween 20 in TBS) and 5% non-fat dry milk (w/v) for 2 h at room temperature. Proteins were further blotted with primary rabbit anti-F3H antibodies in 5% non-fat dry milk (w/v) and TBST overnight at 4 °C, and rinsed three times for 10 min in TBST solution. The membranes were then incubated in Gt anti-Ms IgG (H + L) secondary antibodies (Invitrogen USA), at a dilution of 1:1,000, for 2 h at room temperature, and rinsed three times for 10 min in TBST solution. The blot was developed with Amersham ECL (GE Healthcare UK), and protein bands were exposed on X-ray film. Western blotting and quantification analyses were performed in at least two biological replications (Supplemental Figure S2).  Figure S3) were grown on 3 ppm Norflurazon (Sigma-Aldrich) subjected filter paper in petri plates, as norflurazon is an herbicide that eliminates autofluorescence chlorophyll and carotenoid. After growing for 10 days, approximately 10 WBPH (male and female) were inoculated into each plate (5 plants) and samples were collected for DPBA staining when symptoms appeared 44  Samples were collected at 2-, 12-, and 24 h, and approximately eight to ten plant leaves were randomly collected in triplicate, following the protocol described by Bilal et al. 46 with slight modifications. Approximately 1 g freezedried leaves of wild-type and OxF 3 H plants was ground into a fine powder in liquid nitrogen, with a chilled mortar and pestle. About 0.3 g of the ground sample was mixed in an extraction mixture of acetone and 50 mM citric acid (70:30, v/v), and kept overnight at a low temperature to evaporate highly volatile solvents. The remaining crude extract was filtered through Whatman filter paper and further extracted with 10 ml diethyl ether for 5 repetitions. The extract was then loaded on the solid phase extraction cartridge (500 mg sorbent, aminopropyl), and the cartridge washed with 8.0 ml trichloromethane and 2-propanol (3:1, v/v). The extracted JA and standard were diluted with 10 ml diethyl ether and acetic acid (97:3, v/v). Samples were evaporated, the residue esterified with diazomethane, and the volume adjusted to 50 ml with dichloromethane. The purified extract was then subjected to GC-MS (6890N network GC system and the 5,973 network mass selective detector; Agilent Technologies, Palo Alto, CA, United States). The ion portion was checked at m/z D 83amu, analogous to the JA base peaks. The quantification of endogenous JA was measured from the peak areas compared with the relevant standards.  www.nature.com/scientificreports/ ance to WBPH and WBPH rearing") were collected in three replicates, after 10 days of WBPH infestation, at three time points within one week intervals. Approximately 0.5 g lyophilized tissues were ground into fine powder in liquid nitrogen and homogenized with 80% ethanol (2 ml) at 80 °C for 20 min, following previously described methods 47 . The homogenates were pelleted by centrifugation at 10,000 rpm for 15 min. Supernatants were carefully removed, and the pellets were resuspended in 6 ml distilled water and filtered through 0.2 mm filter paper. Sugar contents were analyzed with HPLC separated with Bio-Rad Aminex 87 C columns (300 × 7.8 mm). Water was used as the eluent at a 0.6 ml/min flow rate.

exogenous application of kaempferol and quercetin into WBpH inoculated plants.
chlorophyll content measurements. Chlorophyll contents of both the wild-type and OxF 3 H plants infested with WBPH (discussed in "Test of OxF3H plants' resistance to WBPH and WBPH rearing") were determined using a SPAD-502 chlorophyll meter (Minolta Camera Co., Osaka, Japan), at three time points within one week intervals, following the method described by Lu, et al. 48 . The third leaf of each selected plant was measured after one week of WBPH inoculation. The average of readings from five replicates were taken to measure the mean SPAD measurement for each plot.

Statistical analysis.
All experiments for each section were performed in triplicate, and the data from each replicate were pooled together. Data were analyzed using two-way ANOVA, followed by the Bonferroni post hoc test (significant difference: p ˂ 0.05). A completely randomized design was used to compare the mean values of different treatments. Data were graphically presented and the statistical analyses were performed using Graph-Pad Prism software (version 5.01; GraphPad, San Diego, CA, USA).

OsF 3 H genetic transformation and generation of the OxF 3 H line. The ORF region of the OsF 3 H
gene was cloned into a pSB11 expression vector using the gateway system, pENTR Directional TOPO Cloning Kit. Ligation to the pENTR/D-TOPO vector was confirmed by double digestion (Supplemental Figure S1A), while ligation to the expression vectors through LR reaction were confirmed by double digestion and sequencing with company (Supplemental Figure S1B). To prepare the insert, the gene (Os04g0662600, https ://www.grame ne.org/) was amplified with the forward and reverse primers. The inserted sequence was further evaluated by blasting through the Gramene database (https ://www.grame ne.org/), and our alignment was 100% identical to the database genome (Supplemental Figure S1C). The gateway system is graphically presented in Supplemental Figure S3, with further details in the Material and Methods section. The OsF 3 H gene was transformed to white calluses using LBA4404 agrobacterium (Supplemental Figure S4A). After two weeks of transformation, green spots appeared on the calluses, which were considered to be the initiation of embryos (Supplemental Figure S4B). The non-gentamycin-resistant calluses turned brown and black, finally dying. After four weeks, embryos emerged from the living callus with gentamycin resistant cells (Supplemental Figure S4C). The embryos carrying the OsF 3 H gene were regenerated into buds, and finally differentiated into leaves in plates (Supplemental Figure S4D). The 18 OxF 3 H plants that were regenerated from the plates were transferred to appropriate medium in test tubes to enhance root growth (Supplemental Figure S4E). After three weeks, the plants were transplanted to soil in a green house and kept until seeds developed (Supplemental Figure S4F). Furthermore the transgenic lines were confirmed by genotyping (Supplemental Figure S5).  Figure S6). The phenotypic variations showed that overexpression of OsF3H also enhances agronomic traits. Such as, shoot length, panicle length, grain weight and fertility rate were higher in transgenic line as compared to wild type. This investigation validated that OsF3H is not only involved in defense mechanism but can also enhance agronomic traits. Previously unrecorded quality data were gathered during phenotypic evaluations under WBPH stress. WBPH are sap-sucking pests that mostly attack stems and leaf midribs (Fig. 1A,B). After WBPH inoculation, the number of infected plants were counted and the infection rate was significantly higher (p ˂ 0.05) in wild-type plants. After 10 days of inoculation, 16.6% of the wild-type plants were infected, which increased to 50% and 100%, after 50 and 70 days, respectively. The number of infected OxF 3 H plants was 0%, 5%, and 16.6%, after 10, 50, and 70 days, respectively (Fig. 1C). The lesion length of WBPH infection was also higher in the wild-type plants, visualized at a later stage of development in the stem (Fig. 1D). Considerable color variation was observed for infected wild-type and OxF 3 H plants.
Infected wild-type plants were a lighter green than OxF 3 H plants, possibly due to nutrient deficiencies as WBPH are cell sap suckers that can reduced nutrient availability. The feeding ratios of WBPH was much higher in wildtype plants, which was evaluated visually, as quantitative calculations was difficult due to the high population numbers of WBPH. Our results demonstrate that wild-type plants were ideal for WBPH feeding and development, with a large number of first instar larvae found in these plants. Only adult WBPH were found on OxF 3 H plants (Fig. 1E,F, respectively). Only male WBPH were found in OxF 3 H plants, with more female WBPH present on wild-type plants. Our study confirmed that WBPH significantly affects plant length, illustrating that significant WBPH infections with continued feeding causes dwarfism and pre-maturation in wild rice (Fig. 1G).

Expression of flavonoid-related genes during WBPH stress.
To evaluate the molecular mechanisms of WBPH-specific induction of flavonol and anthocyanin-related genes, a genetic screening was executed for wild-type and OxF 3 H plants under WBPH stress. Ten leaves from wild-type and OxF3H infected plants were randomly selected in triplicate for RNA isolation, to relatively quantify related genes. We investigated the Scientific RepoRtS | (2020) 10:14685 | https://doi.org/10.1038/s41598-020-71661-z www.nature.com/scientificreports/ OsF 3 H, OsFLS, OsDFR, and OsWRKY13 genes, which are induced by WBPH infestation, using q-PCR analysis ( Fig. 2A-D). Results indicated that the patterns of OsF 3 H, OsFLS, and OsDFR transcription were similar for wild-type and OxF 3 H plants ( Fig. 2A-C) (Fig. 3B). Phenotypic evaluations also indicated that WBPH adversely damaged control plants, while both www.nature.com/scientificreports/ Qu and Kr application protected plants against WBPH damage. However, Kr had better results than the control, but lower than Qu (Fig. 3B). Evaluation of WBPH survival rates showed that most WBPH died after the first week of Qu application, followed by Kr, as compared to control plants. The means from 5 weeks of survival rate evaluations showed that WBPH survival rates on control plants was 85%, kaempferol was 37%, and quercetin was 21% (Supplemental Figure S7A,B). Additionally, OxF 3 H plant resistance was checked with susceptible TN1 lines, demonstrating that all TN1 plants died one week after infection. However, OxF 3 H plants exhibited resistance post infection (Fig. 3B). It was also observed that more female than male WBPH died, confirming the increased susceptibility of female WBPH. Large numbers of instars were found in control plants, with fewer in kaempferol-treated plants, and even less in quercetin-treated plants.

Detection and localization of kaempferol and quercetin in rice seedlings under WBPH stress. OsF 3 H functions as part of a regulatory complex of flavonoids that are inhibited by WBPH induced
stress. Kr and Qu inhibits stress in vivo, therefore endogenous Kr and Qu accumulation and localization should increase in rice seedlings under WBPH stress. Therefore, we used a flavonoid-specific florescence dye, diphenylboric acid 2-aminoethyl ester (DPBA), for endogenous Kr and Qu differential localization in WBPH stressed rice seedlings. Optical sectioning by confocal laser scanning microscopy (CLSM) exhibited the area of Qu and  Figure (A) was more cropped because same sample was run in two replicates in separate gels with other samples due to limited time and resources and then transferred to one X-ray film and again replicated on another X-ray film to make it more clear ( Figure S8). The blots were cut after the protein transfer to incubate with different antibodies to save time, reagents and materials. (B-left picture) KS is kaempferol sprayed, KT is kaempferol treated, QT is quercetin treated and QS is quercetin sprayed. Both the compounds were applied in 3 ppm quantity. The best result shown by quercetin treated and sprayed however, the result of kaempferol treated and sprayed was also better than control plants. ( (Fig. 4). Green indicates Kr presence, orange indicates Qu, and red indicates naringenin. We evaluated Kr and Qu presence in the leaves, stems, and roots of wild-type and OxF 3 H plants. Wild-type seedlings exhibited only naringenin accumulation as a red autofluorescence in the leaf and stem area, with an undifferentiated region of Kr and Qu accumulation indicated by arrows in the stomata zone in leaves and the epidermal zone of stems (Fig. 4A,B). This indicates that wild-type plants activate the flavonoid biosynthesis pathway under WBPH stress, but cannot efficiently convert naringenin due to a lack of OsF 3 H activity. However, CLSM analysis of OxF 3 H plants showed that a significant quantity of Kr and Qu accumulated in leaves and stems under stress conditions, with a lower accumulation in roots. This increased leaf and stem accumulation indicates a rapid stress condition response in the form of naringenin conversion, due to OsF 3 H overexpression. Kr mostly accumulated in the epidermal regions and stomata cells, while Qu accumulated in the xylem and phloem (Fig. 4C). Accumulation was also higher in OxF 3 H plant leaves, with most of the Kr localized to stomata cells and some detected in mesophyll cells. A small quantity of naringenin was observed in leaf mesophyll tissue (Fig. 4D). Due to high Kr accumulation in stomata cells, it was predicted that rice plants close the stomata under WBPH induced stress (Fig. 4E). Kr and Qu were also detected in roots, with higher Kr levels compared to Qu, indicated by the arrow. Kr localized to the epidermal regions, while Qu was present in interior cells (Fig. 4E).

Comparison of flavonols and anthocyanin in response to WBPH.
We also determined the flavonol (kaempferol, quercetin) and anthocyanin (delphinidin, cyanidin) concentrations in relation to the flavonoid biosynthesis pathway (Fig. 2F). WBPH treated plants exhibited remarkable variation in flavonol and anthocyanin accumulation, compared to non-treated plants. We found that Qu and delphinidin accumulation was increased  OsF 3 H overexpression suppresses OsSLR1 to mitigate dwarfism. SLR1 is one of the key genes that encodes the DELLA protein and functions as a GA response suppressor in the GA signaling pathway. Our phenotypic study found that high WBPH numbers caused devastating damage to rice seedlings and plant death (Fig. 3B). However, a large population of WBPH in the developing stage caused severe damage, dwarfism, and pre-maturation (Fig. 1G). To determine the mechanism of how OsSLR1 causes dwarfism and pre-maturation at a transcriptional level under WBPH induction, we identified the OsSLR1 expression levels at different time points in wild-type and OxF 3 H plants. OsSLR1 was significantly (p < 0.05) upregulated in wild-type plants, with irregular expression in OxF 3 H plants (Fig. 2E). In OxF 3 H plants, OsSLR1 was upregulated 2 h after infestation, downregulated at the second timepoint, and upregulated again at the third time point. Thus, OsSLR1 expression seems to influence dwarfism and pre-maturation of wild-type plants.

Quantification of sugar and chlorophyll contents under WBPH infestation. WBPH infestations
significantly reduced sugar content (Fig. 6B-D). Analyses showed decreased amounts of sucrose, glucose, and fructose in wild-type plants at each time point, while OxF 3 H plants showed significant increases (p < 0.05). In wild-type plants, sucrose, glucose, and fructose were decreased by 42%, 43%, and 22%, respectively, between the first and last time points, and increased by 20%, 61%, and 14%, respectively, in OxF 3 H plants. Chlorophyll content was also reduced by 29% in wild-type plants and increased by 17% in OxF 3 H plants, between the first and last time points (Fig. 6D).

Discussion
We selected an appropriate gene, OsF 3 H, by using quantitative trait locus (QTL) data, after WBPH rice plant infestation. The WBPH response in the OxF 3 H F1 generation was evaluated significant differences were noted between transgenic and wild plants (Fig. 1C-G). Our results indicated that WBPH prefer wild-type plants, confirmed by the presence of larger feeding populations and the increased number of infected wild-type plants (Fig. 1C). The preference of WBPH for wild-type rice could be due to the high flavonoid biosynthesis rates in OxF 3 H plants. Our results also identified that wild-type lesion lengths were higher than in OxF 3 H (Fig. 1D), due to high lignin biosynthesis levels that increases the cell wall mechanical support, possibly reducing tissue damage. Previous reports have shown that oxidative stress activates the phenylpropanoid metabolic pathway related genes involved in the synthesis of lignin and flavonoids 50 . It is possible that OsF 3 H overexpression could produce more lignin due to its prominent position in the flavonoid biosynthesis pathway, potentially providing a mechanical defense mechanism against WBPH infestation. www.nature.com/scientificreports/ OsF 3 H is a prominent gene in the flavonoid biosynthesis pathway, supporting the association of flavonoids with WBPH resistance. To confirm whether OsF 3 H induces WBPH resistance and the regulation of downstream genes, we carried out a genetic functional analysis through overexpression. Our results indicated that WBPH regulates the OsF 3 H gene in wild-type and OxF 3 H rice, but that transcriptional and translational expression was significantly higher in OxF 3 H plants (Figs. 2A, 3D). This shows that OsF 3 H actively participated in WBPH resistance, and positively regulated WBPH response characteristics and flavonoid accumulation. Furthermore, in addition to OsF 3 H, our study extends to the regulation of the downstream signaling genes, OsFLS and OsDFR, during WBPH infestation. These genes are involved in flavonol and anthocyanin synthesis. OsF 3 H, OsFLS, and OsDFR were also significantly upregulated in OxF 3 H plants during WBPH infestation. A previous report showed that the activity of downstream genes responsible for the reduction of flavonol and anthocyanin accumulation were reduced in rice tissues, due to a lack of F 3 H gene activity 51 . Researchers evaluated the function of OsF 3 H in flavonoid biosynthesis and BPH resistance, through the development of OsF 3 H overexpression and the use of RNAi in plants. Results showed that overexpressing plants were more resistant, and plants treated with RNAi were more susceptible, compared with wild-type plants. Additionally, flavonoid contents were increased in overexpressing-, and decreased in RNAi plants 52 . OsF 3 H, OsFLS, and OsDFR genes were co-regulated as one regulatory unit during WBPH induced stress, indicating that flavonoid, flavonols, and anthocyanins are synthesized as a single unit during stress conditions. Our results predict that OsFLS, which is not a structural gene, is co-expressed OsF 3 H. The OsF 3 H gene converts naringenin into dihydrokaempferol or dihydroquercetin (flavanonol), which is used by OsFLS as a substrate to synthesize Kr and Qu (flavonol) (Fig. 7) 52 . The TF families of WRKY and MYB are broadly involved in regulating various metabolic pathways in plants under stress conditions 57 . To demonstrate whether the OsWRKY13 TF significantly regulated genes involved in other signal transduction pathways responsible for disease resistance, we analyzed its quantitative expression during WBPH stress in rice plants. We found that OsWRKY13 expression was initially higher, but was downregulated at later stages of WBPH infestation. However, expression was still higher in the wild-type than in the OxF 3 H plants. OsWRKY13 TF downregulation could be due to OsF 3 H overexpression, which can enhance the production of downstream genes responsible www.nature.com/scientificreports/ for the biosynthesis of flavonols and anthocyanins, which are responsible for stress resistance. We expected a downregulation of OsWRKY13 in wild-type plants after 12 and 24 h of WBPH infestation, due to the infection severity caused by the high densities and extended feeding times of WBPH. However, OsWRKY13 expression was higher in the initial stages. Reports show that the WRKY TF binding site (a W-box) occurs in the F 3 H and DFR promoter regions, regulating various processes related to defense against pathogenic and abiotic stressors 57 . Previous investigation additionally reported that OsWRKY13 induces the F 3 H gene 31 . This observation provides strong evidence that WRKY TF is involved in plant resistance against WBPH induced stress, via the induction of gene expression in the flavonoid biosynthesis pathway. We further extended our experiments to include the exogenous application of Kr and Qu against WBPH (Fig. 3B). We found that Kr and Qu have significant pesticidal properties against WBPH, with Qu having a greater effect than Kr (Fig. 3B). Visual observations showed that the WBPH death rate was higher with exogenously applied Kr and Qu, compared to untreated plants. This indicates that Kr and Qu are strong pesticides which adversely inhibit insect digestion, acting as deterrents of WBPH. A previous investigation reported that rice flavone glucosides are strong inhibitors of Nilaparvatalugens digestion, acting as a deterrent 58 . Qu glucoside also increases mortality and inhibits Lymantria dispar and Spodopteralitura development 59 . Kr and Qu also have repellent properties against nematodes, which feed on plants like Radopholussimilis and Meloidogyne incognita 60 . Our observations indicated that the development rates of newly hatched nymphs was negligible, supporting the assertion that Kr and Qu inhibits hatching. Kr restricts hatching, and some other flavonoids prevent insects from laying eggs 61,62 . Kr and Qu are strong ROS scavengers that reduce oxidative stress in cells 63,64 . Confocal microscopy provided strong evidence for Kr and Qu accumulation in various seedling tissues under stress conditions, when stained with DPBA (Fig. 4). Our results showed that Kr and Qu accumulated in rice seedlings, parallel with OsF 3 H and OsFLS expression under WBPH induced stress, as previously reported 66 . Naringenin accumulated at high levels in leaves and stems of non-infected plants, indicating that it is not converted into Kr and Qu under normal conditions (Fig. 4A,B). However, large amounts of Kr and Qu accumulated in the epidermal region of stems and the stomata of leaves in infected plants (Fig. 4C,D). A smaller amount of naringenin was detected in leaves (Fig. 4D). Large amounts of observed Kr in the epidermal region of stems suggests that it initially www.nature.com/scientificreports/ accumulates in the infectious region to protect the plant from damage. Kr and Qu are released in vascular tissues for transportation to different parts of the plant to boost defensive mechanisms against the pathogen throughout the entire plant. This observation indicates that naringenin is an unstable compound and is rapidly converted into Kr and Qu when exposed to stressors. A previous report indicated that the flavonoid biosynthetic pathway is under feedback control, with naringenin inducing the transcription of genes encoding its biosynthetic enzymes 66 . This result supports the hypothesis that naringenin is rapidly converted due to WBPH induced gene expression. The presence of naringenin in the stem vascular regions indicates that conversion is enhanced only in epidermal tissues due to direct contact with WBPH, as naringenin was only detected in vascular regions (Fig. 4C). Kr and Qu were also highly localized to the stomata, possibly keeping the stomata closed during WBPH infestation. Stomata closure is a defensive mechanism against stressors. However, low accumulation in the roots of infected plants indicates a reduced response to WBPH induced stress. Flavonols and anthocyanins have received attention as indicators of pest and pathogen resistance, and excessive light, drought, and cold stress 67 . Although the role of flavonols and anthocyanin against abiotic stressors in plants is well documented, their role in plant-herbivore interactions remains unclear 68 . Here, we observed that Kr, Qu, delphinidin, and cyanidin levels in the OxF 3 H treated line were significantly (p < 0.001) higher than in OxF 3 H non-treated and wild-type treated plants (Fig. 6A). The levels of related compounds in the OxF 3 H nontreated line was also higher than in the wild-type control and treated plants. High flavonol and anthocyanin levels provides strong evidence of OxF 3 H line resistance against WBPH induced stress. Our results showed that Kr and cyanidin did not accumulate in wild-type plants, indicating notably low production not detected by LCMS-MS, or the possibility that these are not regulated during WBPH induced stress. Similar results were observed in confocal laser microscopy analysis where Kr was not detected in wild-type plants, while Qu was detected at low magnification. (Fig. 4A,B). Previous reports suggested that Kr constitutes lower antioxidant activity compared to Qu, which is a strong antioxidant, and that Qu biosynthesis was higher than Kr during the stressed condition 70 , which favors our result of lower Kr production compared to Qu. It has been reported that MYB75 overexpression, which is involved in the flavonoid biosynthesis pathway, significantly regulates anthocyanins, Kr, and Qu biosynthesis during caterpillar or aphid stress, with increased levels in over-expressors compared to wild-type plants. Kr and Qu have been shown to accumulate in cotton crops during insect feeding, while anthocyanins were effective in bacterial blight resistance in cotton leaves 71 . Results from LCMS-MS and the peaks detected for different compounds are presented in Supplemental Figure S8.
Additionally, high-density WBPH infestation caused rice seedling death, but led to pre-maturation and dwarfism, and eventually decreased yield in the tillering stage. Previous researchers reported that WBPH infestation significantly reduced shoot length and plant vigor, enhanced leaf discoloration, and inhibited tillering emergence and grain setting [72][73][74] . These results also support our finding regarding the decreased chlorophyll contents of wild-type plants, and its regulation in OxF 3 H plants under stress conditions (Fig. 6D). To investigate the cause of dwarfism, we extended our study to OsSLR1, which produces the DELLA protein, enhances immunity, and regulates the JA hormone during WBPH induced stress. We found that OsSRL1 significantly enhanced defenses against WBPH in wild-type plants due to lack of OsF 3 H activity. However, due to OsF 3 H expression in OxF 3 H plants, OsSLR1 expression was reduced, although non-significantly. This suggests that OsSLR1 enhances the basal defensive mechanism under stressed conditions. Figure 1E shows that high-density infestation inhibited growth in wild-type plants, although the plants still survived possibly due to the activation of plant defensive mechanisms, which usually take place at the expense of growth. Previous reports investigated whether this conflict between growth and defense is supported by the principle that plant-resources are restricted and can only be utilized for either growth or defense, depending on exterior and interior conditions 75 . In OxF 3 H plants, stress was reduced due to OsF 3 H overexpression, resulting in the inhibition of OsSLR1 expression. Researchers found that the SLR1 gene positively regulates the defense mechanism by regulating the SA and JA signaling pathways 76 . Although we did not evaluate GA regulation, previous investigations have reported antagonism of JA and GA. Our results identified positive regulation of JA in wild-type plants, compared to OxF 3 H plants, providing strong evidence for GA restriction under stressed conditions, resulting in growth inhibition 75,77 . However, decreased JA levels in OxF 3 H plants avoids stress mitigation induced by WBPH, through OsF 3 H expression. The mutual antagonism between JA-GA is an important strategy in maintaining the balance between defense and growth, through physical interactions between DELLAs and JAZs 75,77 . These results indicate that along with OsF 3 H, DELLA proteins also contribute to defense against WBPH stress, whether in the form of JA upregulation-which is the ultimate source of anthocyanin biosynthesis, or in the form of hijacking of the GA crosstalk mechanism.
Along with the induction of the OsSLR1 gene and JA-GA antagonistic mutualism, sugar content reduction is another inhibitor of plant length. Sugar is crucial for plant growth and development as it is a major source of energy. It has been reported that exogenous application of sugar enhanced plant growth. Sugar also upregulates genes associated with the defense mechanism and biosynthesis of secondary metabolites. Sugar content inhibition in wild-type plants is also associated with slow photosynthesis rates due to a decrease in chlorophyll contents and the high density of WBPH feeding which extract sugar and other nutrients from the phloem. The second hypothesis for the decrease in sugar content in wild-type plants is the deviation of total energy toward defense mechanisms, which utilizes large quantities of sugar as a carbon source. The use of large quantities of sugar as a defensive tool critically affects plant growth and development, ultimately resulting in stunted growth.

conclusion
WBPH is a significant biotic stressor that can seriously impact rice yield in several countries. Previously, conventional breeding was the main tool for selecting the most effective, easily adaptable, and resistant crop varieties. However, molecular breeding techniques are currently used to protect agricultural crops by developing new, resistant variants. Using molecular breeding techniques, we developed a highly WBPH-resistant OxF 3 H rice Scientific RepoRtS | (2020) 10:14685 | https://doi.org/10.1038/s41598-020-71661-z www.nature.com/scientificreports/ cultivar by selecting the gene of interest through QTL analysis. Plants respond to pest attacks through a complex network of transcriptional, proteomic, metabolomic, and phytohormonal reprogramming. In this study, we evaluated all possible mechanisms of regulation of complex responses, like regulation of the OsF 3 H gene at RNA, protein, metabolite, and hormone levels. OsF 3 H overexpression in rice, which resulted in elevated anthocyanins and flavonol production, was used to study the potential roles of anthocyanins and flavonols in plant-WBPH interactions. Overexpression of OsF3H has been reported to upregulate structural genes of the flavonoid biosynthesis pathway and its related proteins in OxF3H plants, while directly or indirectly downregulating JA. The results also demonstrated that the expression of structural anthocyanins and flavonols biosynthesis pathway genes in OxF3H plants were higher than in wild-type plants, demonstrating that anthocyanins and flavonols are crucial to the WBPH response. This is novel information identifying that anthocyanins and flavonols, especially Qu, plays a key role in WBPH resistance. Similarly, our study confirmed Qu as a significant deterrent of WBPH when applied exogenously, which predicts that Qu could be used as strong pesticide. Phenotypic evaluations of OxF3H plants, compared to wild-type plants, provided significant evidence that anthocyanins and flavonols play a critical part in producing WBPH resistance.