Nanopriming technology for enhancing germination and starch metabolism of aged rice seeds using phytosynthesized silver nanoparticles

Application of nanomaterials for agriculture is relatively new as compared to their use in biomedical and industrial sectors. In order to promote sustainable nanoagriculture, biocompatible silver nanoparticles (AgNPs) have been synthesized through green route using kaffir lime leaf extract for use as nanopriming agent for enhancing seed germination of rice aged seeds. Results of various characterization techniques showed the successful formation of AgNPs which were capped with phytochemicals present in the plant extract. Rice aged seeds primed with phytosynthesized AgNPs at 5 and 10 ppm significantly improved germination performance and seedling vigor compared to unprimed control, AgNO3 priming, and conventional hydropriming. Nanopriming could enhance α-amylase activity, resulting in higher soluble sugar content for supporting seedlings growth. Furthermore, nanopriming stimulated the up-regulation of aquaporin genes in germinating seeds. Meanwhile, more ROS production was observed in germinating seeds of nanopriming treatment compared to unprimed control and other priming treatments, suggesting that both ROS and aquaporins play important roles in enhancing seed germination. Different mechanisms underlying nanopriming-induced seed germination were proposed, including creation of nanopores for enhanced water uptake, rebooting ROS/antioxidant systems in seeds, generation of hydroxyl radicals for cell wall loosening, and nanocatalyst for fastening starch hydrolysis.

the key enzymes involved in degradation of starch during germination of cereal seeds and in subsequent seedling establishment 9 . This is the only enzyme which initiates hydrolysis of native starch granules and is de novo synthesized during the germination of cereal seeds and catalyzes the hydrolysis of α-1, 4 linked glucose polymers to release fragments that can be further broken down by other amylolytic enzymes 10 . Thus, the enhancement of α-amylase during seed germination is of interest in the field of carbohydrate research to promote economic plant growth.
All seeds stored under air dry conditions will have suffered a degree of deterioration 11 and seeds in long-term storage will eventually lose their viability due to spontaneous biochemical damage occurring at cellular level 12 , leading to natural seed aging and subsequently limiting crop productivity. Seed priming is a technique that partially hydrates seeds in natural or synthetic compounds under specific environment to a point where germination-related metabolic processes begin, but radicle emergence does not occur 13 . Seed priming has been found to be useful for enhancing seed quality, seedling establishment and crop yields as well as increasing tolerance to environmental stresses 8,13 . Seed priming can improve the germination of weak, damaged or aged seeds 11 or even under adverse environment 13 . A number of commonly used priming agents include polyethelyne glycol, inorganic salts, nutrients, and plain water 12,14,15 . However, different priming solutions have different properties, effectiveness, and optimization of priming agents is required for each crop species 14 . Therefore, there is a growing need to develop new priming agents to enhance seed germination of various crop plants.
In recent years, several metal-based NPs (e.g., AgNPs 16 , AuNPs 5 , CuNPs 17 , 18 , FeNPs 17 , FeS 2 NPs 19 , TiO 2 NPs 20 , ZnNPs 17 , 18 , ZnONPs 21 ) and carbon-based NPs (e.g., fullerene 22 and carbon nanotubes 23 ) have been applied as seed pre-treatment agents for promoting seed germination, seedling growth, and stress tolerance in some crop plants. Among these studies, only a few researchers have used seed priming strategy, in which seeds must be re-dried to their original moisture content before sowing. Thus, the mechanism behind seed nanopriming would be different from that of pre-sowing seed treatment without drying seeds. In addition, comprehensive studies on physiological and molecular mechanism of nanopriming effects on seed germination have not been elucidated, and thus there are many questions remained to be addressed, especially mechanism behind NPs-induced seed germination.
In this study, jasmine rice (Oryza sativa L. cv. KDML 105) was selected as a model plant since this cultivar is one of the world's best quality aromatic rice cultivars which is in high demand for world's markets, but it is sensitive to various environmental stresses, resulting in the reduction of yield 24 . Moreover, many seed lots of jasmine rice are naturally aged seeds due to improper storage conditions under ambient temperatures by most farmers. Therefore, the overarching aims of this study are to develop green method for the synthesis of AgNPs and present a new priming technique using phytosynthesized AgNPs to enhance the germination and starch metabolic process of rice aged seeds alternative to conventional priming. The mechanism behind AgNPs induced seed germination and starch mobilization was also proposed.

Results
Synthesis and characterization of phytosynthesized AgNPs. In the present study, the color of kaffir lime leaf extract changed from pale yellow to dark brown after exposure with AgNO 3 solution for about 1 h at room temperature, indicating reduction of Ag + to Ag 0 (Inset of Fig. S1 in Supporting information). The surface plasmon resonance (SPR) of AgNPs showed a peak centered near 443 nm at UV-vis spectra (Fig. S1), confirming the reduction of silver ions to colloidal silver.
TEM image (Fig. 1a) of the AgNPs sample shows nanoparticles exhibiting a typical spherical and an ellipsoidal morphology. The approximate particle diameter was found to be in the range of 6-36 nm. EDX analysis also confirmed the presence of elemental silver signal of the synthesized AgNPs (Fig. S2, see Supporting information). The selected area electron diffraction (SAED) pattern in Fig. 1b showed the bright circular rings, suggesting the crystalline nature of AgNPs. The diffraction rings of the AgNPs exhibited Debye-Scherrer rings assigned (111), (200), (220), and (311) lattice planes of the face centered cubic (fcc) silver, respectively, which is consistent with XRD result (Table S1 in Supporting information).
X-ray diffraction (XRD) was also used to confirm the crystalline nature of synthesized AgNPs. Figure 1c displays four prominent diffraction peaks at 2θ value of 38.2°, 44.3°, 64.6° and 77.4° indexed as (111), (200), (220) and (311) miller indices, respectively (Table S1), which are characteristic of fcc crystalline structure of metallic silver (JCPDS file No.04-0783) 25 . The unassigned peaks marked with asterisks were also observed, suggesting that the crystallization of bioorganic phase occurred on the surface of the nanoparticles. The average nanocrystalline size of synthesized AgNPs is calculated by Debye-Scherrer's equation 25 was found to be around 11.2 nm with respect to high intense peak (111) as shown in Table S1.
FTIR analysis was conducted to identify the surface structure of the AgNPs and identify the possible functional groups (phytomolecules) of the plant extracts involved in the synthesis and stabilization of the AgNPs. In FTIR of kaffir lime leaf extracts (Fig. 2b), a number of peaks reflecting a complex nature of plant metabolites were recorded: for example, 3323 cm −1 (stretching vibrations of the hydroxyl (-OH) groups of alcohol and phenolic compounds) [26][27][28] , 2922 cm −1 (stretching vibrations of secondary amines in the N-H bond) 26,29,30 , 1614 cm −1 (amide I band arising due to carbonyl (-C=O) stretch of the proteins) 29, 31 , 1241 cm −1 (amide III band of protein) 29,30 , and 1020 (C-N stretching vibration of aliphatic amines or to alcohols or phenols) 29 . By comparing the FTIR spectra of the plant extract and synthesized AgNPs, the band shift and intensity decrease of the hydroxyl, carbonyl and amine groups was observed (Fig. 2a), indicating that the major phytochemicals from the extract were involved in reduction and stabilization processes and coated on the AgNPs surface.
Ag + ion release in priming solution. To determine bioavailability and Ag + ion release in AgNPs and AgNO 3 priming solutions, we measured Ag + concentration in the priming solutions at 1 h and 24 h after dissolution using ICP-OES technique. At 1 h, the release of Ag + ions in both concentrations of AgNO 3 priming solutions (10 and 20 mgL −1 ) was significantly higher than those of AgNPs priming solutions (Table S2), suggesting the rapid release of Ag + ions in AgNO 3 . At 24 h of dissolution, Ag + ion release in both AgNO 3 10 and AgNO 3 20 drastically increased with concentration change of about 1.12-6.7 times compared to their respective solutions at 1 h of dissolution. In contrast, Ag + ion release in both AgNPs10 and AgNPs20 solution slowly increased with concentration change of about 1.2-1.5 times compared to their respective solution at 1 h of dissolution. At 24 h, the Ag + concentrations in AgNO 3 10 and AgNO 3 20 solutions were 12.7-21.6 and 25.5-43.5 times higher than AgNPs10 and AgNPs20 priming solution, respectively. No Ag + concentration was observed from deionized water used as hydropriming solution.
AgNPs could mediate hydroxyl radicals as determined by Electron Spin Resonance (ESR) analysis. Hydroxyl radical (•OH) generation in the priming solution was analyzed using ESR, which is the most accurate and sensitive technique for the detection of free radicals. The spin trap DMPO, which reacts covalently with •OH, gives rise to stable secondary radical adduct that can be detected at room temperature using ESR. A four line spectrum with a 1:2:2: intensity ratio is of indicative of the formation of DMPO-OH spin adduct. ESR analysis indicated that AgNPs were able to generate •OH compared with standard Fenton reaction (Fig. 3). In contrast, deionized water and AgNO 3 did not produce ESR signal under the same conditions, indicating that Ag + did not produce •OH. Acceleration of seed germination and seedling growth by nanopriming. As shown in Fig. 4a, priming the rice seeds with nanopriming solutions (AgNPs10 and AgNPs20) could accelerate the early seed germination and germination percentage compared to those of the unprimed and other primed groups (p < 0.05). For example, while the control seeds germinated on the fourth day after incubation, 16.6% of both nanoprimed seeds were able to germinate during the second day after incubation (p < 0.05). Moreover, both nanoprimed seeds had reached 93% of germination on the fourth day after incubation. A significant increase of seed germination was observed for nanoprimed seeds on the third day. At this time, the germination percentage for seeds that were primed with AgNPs10 and AgNPs20 averaged 73.3% and 93.3%, respectively, while seed germination percentage of the hydropriming, AgNO 3 10 and AgNO 3 20 treatments averaged 46.6, 30.0 and 33.3%, respectively (p < 0.05). The germination percentage of nanoprimed seeds was significantly higher than that of both groups of AgNO 3 -primed seeds, but was not significantly different from hydropriming group (Table 1).
In addition, the germination index (GI) and germination rate (GR) of the tested seeds were also calculated, and it was found that nanopriming treatment greatly increased GI and GR values compared to those of unprimed and other primed seeds ( Table 1). The phenotypic characteristics of the rice seedlings developed from all nanoprimed seeds are normal without the sign of toxicity (Fig. 4b). After 6 days of germination, nanopriming significantly increased vigor index compared to unprimed control and other priming treatment (Table 1). Also, priming with nanosolution significantly increased the lengths of roots and shoots and seedling biomass (mg  weight) compared to the unprimed and other primed groups (p < 0.05) (Fig. S3a-d). Overall, priming the aged seeds of jasmine rice with phytosynthesized AgNPs at optimum doses (10 and 20 mgL −1 ) increased seed germination, vigor strength and seedling growth.

Seed water uptake.
To test whether different priming agents can affect seed imbibition at the early germination process, seed water uptake at 4 and 24 h of imbibition was determined. It was found that all primed seeds adsorbed water faster than those unprimed at the initial (4 h) and later (24 h) periods of imbibitions (p < 0.05) ( Table 1). Nanoprimed seeds (both AgNPs10 and AgNPs20) imbibed water faster than other primed groups at the very beginning (4 h) of imbibition (p < 0.05). At 24 h, seed water uptake of nanopriming treatment was still higher than hydropriming and AgNO 3 priming treatments.
Internalization of Ag in rice seeds. Internalization and accumulation of silver (Ag) of primed seeds were analyzed by ICP-OES. As shown in Table S3, no Ag was detected in unprimed and hydroprimed seeds (detection limit of Ag analysis by ICP-OES is 0.001 ppm), indicating that the Ag element is not present in rice seeds. On the contrary, both groups of seeds treated with AgNO 3 or AgNPs priming solutions showed the accumulation of Ag, albeit very low concentration. This implied that Ag + ions released from AgNO 3 or AgNPs solutions were able to penetrate the seed coat into endosperm and embryonic tissues. The concentration of Ag increased as the external  concentration of AgNPs or AgNO 3 increased (Table S3). Ag concentrations of AgNPs10-primed seeds were significantly lower than both groups of AgNO 3 treatments, while those of AgNO 3 20-primed seeds were the highest. Although Ag concentrations of AgNPs20 were significantly lower than AgNO 3 20, they were not statistically different from AgNO 3 10. The higher concentrations of Ag found in AgNO 3 20-primed seeds could be due to higher solubility and bioavailability of AgNO 3 solution compared to those of AgNPs solution. This result corresponded well with the data on Ag + ion release (Table S2), demonstrating that AgNO 3 priming solution had higher Ag + concentration than AgNPs priming solution.
Nanopriming can enhance starch metabolism for rice seedling growth. Starch metabolism in rice seedlings was assessed in terms of α -amylase activity and total soluble sugars. After 6 days of seed germination, nanopriming treatments significantly enhanced α-amylase activity in rice seedlings compared with control and other primed seedlings (Fig. 5a). The α-amylase activity of seedlings increased 2.6 and 2.5 folds in AgNPs10 and AgNPs20 priming treatments, respectively, compared to control seedings. The α-amylase activities of seedlings of AgNPs10 and AgNPs20 treatments were almost 1.3 and 1.2 times greater than those in the hydropriming group, while the α-amylase activities in the AgNO 3 10 and AgNO 3 20-treated seedlings were not significantly different from the control. Likewise, the total soluble sugar contents of seedlings were induced from 2.0 to 2.2 folds by nanopriming treatments (AgNPs10 and AgNPs20) compared to the control, while 1.8-fold increase was observed in the hydroprimed seedlings (Fig. 5b). In the AgNO 3 priming treatments, the total soluble sugars remained almost unchanged compared to control. We also confirmed above results by measuring amylase production and secretion from embryoless half-seeds using a starch agar plate α-amylase assay method. Clear zones appeared after IKI staining when isolated endosperms of half seeds were incubated in starch agar containing gibberellins (GA), indicating expression and secretion of α-amylases that hydrolyzed starch 32 . After 3 days of incubation, numbers of clear zone recorded from nanoprimed seeds were higher than that from other primed seeds and unprimed control (Fig. S4), corresponding with rapid germination of nanoprimed seeds compared to other treatments. Besides, clear zone areas observed in nanoprimed seeds were larger than in other priming treatments, indicating the higher levels of α-amylase expression and production from nanoprimed seeds. Control plate without GA showed no amylase production from embryoless half-seeds.
Starch granule morphology. We also observed the morphological change of isolated starch in both unprimed and primed seeds after 24 h of imbibition. It can be observed from SEM results that in general individual starch granules (ISGs) of rice seeds of unprimed and primed groups were polyhedral in appearance with irregular shapes (Fig. 6), corresponding with previous reports 33,34 . Some compound starch granules (CSGs) were observed in control, hydroprimed and SN-primed seeds, suggesting that starchy endosperm of these treatments had low starch decomposition rate at initial phase of germination. On the contrary, starch granules of AgNPs-primed seeds did not form the CSGs and contained some smaller ISGs, indicating higher degrees of starch decomposition. Moreover, some starch granules of nanoprimed seeds had small pits on their surface, indicating a higher amylase activity compared to control and other primed seeds.
Nanopriming enhanced dehydrogenase activity in rice seedlings. Dehydrogenase activity is an indicator of root vitality and used as comprehensive assessment index reflecting the metabolic activity level and root's ability to absorb nutrients and water 35 . After 6 days of seed germination, pronounced increase (∼2.3 to 2.5-fold) in root dehydrogenase activity over unprimed seedlings was found in nanopriming treatments (Fig. 5c). Root dehydrogenase activity of AgNPs10 and AgNPs20 priming treatments was also significantly higher (∼1.3and 1.2-fold) than that of the hydropriming group, while in the silver nitrate treatments it remained almost unchanged compared to the control.
Nanopriming enhanced aquaporin gene expression in aged rice seeds. In order to elucidate the molecular basis for changes in seed metabolism in response to priming treatments, the expression levels of two aquaporin genes, PIP1;1 and PIP2;1, which are responsible for water uptake were monitored in germinating seeds at 24 h of imbibition. The expression levels of the two aquaporin genes in AgNO 3 -primed seeds and unprimed control were not obviously different. On the other hand, priming with AgNPs markedly enhanced the expression of aquaporin genes, especially PIP2;1 as compared to the control and other priming treatments (Fig.S5), coinciding with the higher percentage of seed water uptake at 24 h ( Table 1). The transcript levels of both PIP1;1 and PIP2;11 in hydroprimed seeds also increased as compared with AgNO 3 -primed seeds and control, but their expression levels were lower level than nanopriming treatments.
Nanopriming modulated antioxidant enzyme activity in seeds after priming. Antioxidant enzymes (SOD and CAT) activities in unprimed and primed seeds varied significantly ( p < 0.05) in response to different seed priming treatments (Fig. 7a,b). In dry seeds (0 h of imbibition), SOD activities of both unprimed control and primed seeds were on the same level (Fig. 7a). However, when seeds were imbibed in water for 24 h, activity levels of SOD in both nanoprimed seeds and hydroprimed seeds were higher than those in the unprimed control and AgNO 3 -primed seeds, although the values of AgNPs10-primed seeds were not significantly different from hydroprimed seeds. Interestingly, considerable increase in CAT activity in nanoprimed seeds was observed in dry seeds (0 h of imbibition) after priming with AgNPs compared to control and other primed seeds (Fig. 7b). After 24 h of imbibition, activity levels of CAT in AgNPs10-and AgNPs20-primed seeds increased by 71% and 61%, respectively, compared with the control, which coincides with the rapid radicle protrusion and confirms the connection between the raise in CAT activity and radicle protrusion. Hydroprimed seeds also had higher CAT activity compared to control and AgNO 3 -primed seeds, which had lowest CAT activities, corresponding with their delay in germination (Fig. 4a).

Quantification of reactive oxygen species (ROS) in unprimed and primed seeds. The contents
of two different ROS species, i.e. hydrogen peroxide (H 2 O 2 ) and superoxide anions (O 2 •-) were quantified spectrophotometrically in the embryos of seeds at 0 and 24 h. At 0 h (dry seeds), the O 2 •contents in the embryos of all tested seeds were not significantly different (Fig. 7c). In contrast, when seeds were imbibed in water for 24 h, O 2 •contents of both AgNPs10-and AgNPs20-primed seeds significantly increased by ∼42% and 33%, respectively compared to control. The contents of O 2 •in hydroprimed seeds also slightly increased (∼12%) compared to control, while there were no significant changes of O 2 •between AgNO 3 treatments and control. Interestingly, priming treatments with AgNPs can reduce H 2 O 2 contents as both nanoprimed seeds showed lower levels than control and AgNO 3 -primed seeds at 0 h of imbibition ( Fig. 7d), whereas AgNPs10 showed no significant difference from hydroprimed seeds. At 24 h of imbibition, H 2 O 2 contents of all tested seeds leveled up. Interestingly, H 2 O 2 contents of AgNPs10-and AgNPs20-primed seeds increased by ∼33% and 32%, respectively, compared to control. Also, levels of H 2 O 2 in nanoprimed seeds were higher than hydroprimed and AgNO3-primed seeds. Hydroprimed seeds showed a slight increase in H 2 O 2 contents (∼15%) compared to control, whereas no significant differences between unprimed control and AgNO 3 -treated seeds were observed.
in vivo fluorescence imaging and histochemical localization of ROS production. The distribution and accumulation of H 2 O 2 in germinating seeds at 24 h were visualized with a specific fluorescent probe, DCF-DA. After 24 h of imbibition, coleorhiza of both hydroprimed and nanoprimed seeds began to protrude, while AgNPs-primed seeds and unprimed control did not. The fluorescence signal intensities for nanoprimed and hydroprimed treatments were stronger than those of the unprimed and AgNO 3 -primed treatments, whereas the H 2 O 2 -excited fluorescence intensity was stronger in nanoprimed seeds than hydroprimed seeds (Fig. 8).
Fluorescence imaging also showed that H 2 O 2 was located mainly in coleorhizal zone and partly in the aleurone layer in germinating seeds. After 36 h of imbibitions, we localized H 2 O 2 and O 2 •in germinating seeds using DAB and NBT staining, respectively. Results showed that in nanoprimed and hydroprimed seeds embryo and aleurone layer were somewhat stained by DAB, whereas embryo showed stronger staining than aleurone layer. Coleorhiza and radicle of nanoprimed seeds, which were more developed than unprimed and other primed seeds, showed higher accumulation of brown spots indicating the higher level of H 2 O 2 . Besides, the production and accumulation of H 2 O 2 were greater in coleorhiza and radicle than in coleoptiles of germinating seeds (Fig. 9). Also, NBT staining indicated a marked increase of O 2 •in germinating nanoprimed seed than unprimed control and other primed seeds (Fig. 9). Although NBT stained strongly as dark purple color product in coleorhiza and radicle of nanoprimed seeds, it also stained coleoptiles with pale purple, indicating O 2 •could participate in cell elongation in these structures during germination.

Discussion
Characterization of phytosynthesized AgNPs. Characterization of photosynthesized AgNPs revealed some useful information. For example, reduction of Ag + ions into AgNPs during exposure to the plant extracts could be visualized by color change, with kaffir lime leaf extracts turning from pale yellow to dark brown color (Fig. S1), indicating the reduction of Ag + ions into AgNPs. Change in color was due to the excitation of the surface plasmon resonance (SPR) in AgNPs 36 . UV-vis absorption spectrophotometer is used to investigate the SPR phenomenon. In this study, a single, strong, and broad SPR peak at 441 nm was observed in the UV-vis spectrum, confirming the successful synthesis of AgNPs. Previous studies suggested that a SPR peak located between 410 and 450 nm exists for AgNPs and could be attributed to spherical shaped NPs 36 . The SPR peak is reported to be very sensitive to the size and shape of the NPs, amount of plant extract, AgNO 3 concentration, and the type of phytochemicals or biomolecules present in the leaf extract 37 . Broadening of UV-visible peak observed in this study also suggested that phytosynthesized AgNPs were polydisperse in nature 38 .
TEM micrographs of the phytosynthesized AgNPs indicated that the particles were spherical and ellipsoidal, having an average size of 6-36 nm (Fig. 1a). SAED pattern with bright circular rings indicates the crystalline nature of AgNPs, which are similar to previous report 26 . Our SAED data were also corroborated with the XRD pattern of the same materials (Fig. 1c). XRD pattern of phytosynthesized AgNPs matches with the characteristics of fcc crystalline structure of metallic silver (JCPDS file No.04-0783) 25 . Also, the intensity and broadening of peaks lend further support that the obtained products are crystalline in nature and small crystallite size, respectively. The unassigned peaks marked with asterisks could be due to the crystallization of bio-organic phase, e.g. polysaccharides and proteins present in the leaf extracts, occurring on the surface of AgNPs 31 .
Possible mechanism of AgNPs formation. Our FTIR analysis showed various functional groups of phytochemicals found in kaffir lime leaf extract coated onto AgNPs surface, especially hydroxyl, carbonyl and amine groups. These functional groups could be derived from flavonoids, phenolic compounds, carbohydrates, sugars and proteins, which could act as reducing and stabilizing agents 39 . Kaffir lime leaves have been reported to have high levels of natural antioxidant polyphenols, including flavonoids and tannins, terpenoids, and volatile compounds, e.g. citronellal monoterpenoid (linalool), acyclic terpene alcohol and viridiflorol 40 . Based on the phytochemicals abundant in the plant leaves, the plausible mechanism of AgNPs formation using phytochemicals is shown in Fig. 10. Upon the addition of plant extracts to the Ag + aqueous solution, the Ag + -phytomolecule complex intermediate is formed, and the phenolic groups could donate electron or H + to Ag + ions resulting in the reduced Ag 0 . After the phenol ring undergoes oxidation, it converts to its quinone form. The electrochemical potential difference between Ag + ions and phytomolecules drives the reaction, and the formed AgNPs are stabilized by the lone pair of electrons and π-electrons of quinone structures 28 . Carbohydrates and sugars present in the plant leaf extracts can also act as mild reducing agents 41 . Moreover, the carboxylate group present in the proteins which has greater affinity to act as surfactant by forming a protein layer on NPs could stabilize protein through negatively charged carboxylate group 26 . Further studies using advanced analytical tools, such as gas chromatography (GC) are required to determine the exact phytochemicals present in kaffir lime leaves, which are responsible for the bioreduction of silver.
AgNPs generated hydroxyl radicals, but Ag + did not. In the cell free system or priming solution, only AgNPs solution generated •OH (Fig. 3), which correspond with previous reports using a cell free system 42 and cell culture system 43 . In contrast, Ag + derived from AgNO 3 did not produce •OH in our test system, which is consistent with a previous study of Li et al. 42 who reported that the addition of Ag + inhibited the ESR signal or did Since the valence of Ag changes from zero to one, the AgNPs may be viewed as "Fenton nanoparticles", and the concentration dependence should be predicted by reaction 43 . That is, the production of •OH could be accelerated by increasing AgNPs or H 2 O 2 concentration, or decreasing the pH of the buffer solution. Conversely, •OH production could be inhibited by addition of Ag + or increasing the pH of the buffer 43 . Therefore, the presence of Ag + as observed in our study could suppress the generation of •OH. The role of AgNPs mediated •OH generation for enhancing seed germination is discussed below.
Nanopriming can stimulate seed germination of rice aged seeds by enhancing water uptake and starch metabolism. Seed priming with phytosynthesized AgNPs solutions at 10 and 20 mg L −1 induced more rapid water uptake at 4 and 24 h of imbibtion (Table 1). This data corresponded well with the early protrusion of radicles and higher germination percentage in nanoprimed seeds compared to other priming treatments and control. Previous literature highlighted that priming treatment can improve seed water uptake, as primed seeds exhibited a faster imbibition in comparison with non-primed ones, although pre-treated seeds were dried after priming to reach the same water content as non-primed ones 44 . Besides, seed priming with AuNPs also increased water uptake in maize plant 5 . Water uptake in seed germination process is essential because mature seeds are quite dry and need sufficient amount of water to start cellular metabolism and growth 45 . As the activities of dehydrogenase, a key enzyme in cellular respiration, in seedling roots of nanopriming treatments were higher than those of unprimed control and other treatments, this could be attributed to the consequence of rapid water Figure 10. The plausible mechanism of the formation of AgNPs mediated by phytochemicals from kaffir lime leaf extract. uptake occurred in nanoprimed seeds. Thus, higher water uptake can enhance seed germination and seedling development through complex networks.
We also observed that two aquaporin genes, PIP1;1 and PIP2;11, involved in water uptake were up-regulated in germinating seeds at 24 h of imbibitions (Fig. S5). Qian et al. 46 also found that, in Arabidopsis seedlings, the transcript level of the aquaporins, including PIP 1;2, PIP 2;1, PIP 2;2, SIP 1;1, and TIP 1;1 increased approximately 2-fold over the control levels after treatment with 0.2 or 0.5 mgL −1 AgNPs for one week. Aquaporins (water channels) are transmembrane proteins which not only facilitate the diffusion of water across biological membranes and regulate water homeostasis via a complex mechanism 46,47 , but also facilitate gases such as CO 2 , nutrients, e.g., boron (B), silicon (Si), and reactive oxygen species (ROS) 47 . Previous studies indicated that certain concentrations of carbon nanotubes (CNTs) (40 μg mL −1 ) and AgNPs (0.2 and 0.5 mg L −1 ) can significantly activate the expression of aquaporin genes in tomato roots 48 and Arabidopsis seedlings 46 , respectively. Besides, CNTs (multiwalled CNTs) can stimulate the expression of aquaporin genes and could play a role in the enhancement of seed water uptake during germination of some vegetable crops 49 . Thus, higher expression of aquaporin genes observed in nanoprimed seeds could be one of the mechanisms which enhanced seed germination.
We noted that at the time of water uptake or aquaporin gene expression increased, more H 2 O 2 accumulation was also observed in nanoprimed seeds at 24 h of imbibition compared to unprimed and other primed seeds. This suggests that there is connection between ROS and aquaporin genes in controlling seed germination. Past studies suggested that a complex interplay may exist between aquaporins and biological relevant ROS 50 . Specific aquaporin isoforms can facilitate the passive diffusion of H 2 O 2 across biological membranes and regulate the membrane permeability of H 2 O 2 , thereby acting in redox signaling process 50,51 . Thus, the upregulation of aquaporin genes observed in this study could facilitate the diffusion of H 2 O 2 in germinating rice seeds, which further lends support to our argument that both aquaporins and ROS could participate in enhancing seed germination (discussed later).
As demonstrated by colorimetric and starch-agar plate assays, we also found that nanopriming was the most effective treatment in improving starch metabolism of rice seeds and seedlings. Our results confirmed the existing literature that priming can enhance starch metabolism during germination 15,52 . α-A mylase is hydrolytic enzyme responsible for the degradation of reserve carbohydrate to soluble sugars to maintain active respiratory metabolism for seed germination and subsequent seedling growth until photosynthesis is sufficient to support growth 9 . An increase in germination rate and seedling vigor of nanoprimed seeds observed in this study could partly be explained as a consequence of increased activity of α-amylase allowing faster rate of starch hydrolysis in germinating nanoprimed seeds. This resulted in more availability of soluble sugars necessary for generating the energy required for seedling growth, thereby leading to increased speed of germination.
We also found that nanopriming can enhance starch degradation during seed germination compared to control and other primed seeds (Fig. 6). Starch granules of nanoprimed seeds did not form the CSGs and contained some smaller ISGs, indicating higher degrees of starch decomposition. Moreover, some starch granules of nanoprimed seeds had small pits on their surface. A study by Man et al. 53 found that α-amylase pitted the starch granule surface first, then penetrated into the interior and hydrolyzed the granule from the inside out, implying that α-amylase activity was induced higher in nanoprimed seeds during the first 24 h of imbibition. The increase in amylase activity and production in nanoprimed seeds resulted in higher starch degradation and rapid seed germination.
Earlier, Srivastava et al. 19 suggested that pyrite NPs (FeS 2 NPs) could produce H 2 O 2 in situ on their surface in the presence of water, and are also considered as pyrite-only Fenton-like (PF) reagent, which could generate highly reactive •OH. These researchers also highlighted that FeS 2 NPs could mimic the enzymatic activity of amylase enzyme to enhance the breakdown of starch in spinach seeds. Our study also revealed that AgNPs could act as Fenton like reagent and mediate the •OH generation as discussed above. Thus, AgNPs might behave in a similar manner to pyrite NPs in enhancing amylase activity in rice seed.

Differential toxicity of Ag + (AgNO 3 ) and AgNPs. Previous studies comparing the biological effects of
AgNPs and AgNO 3 have led to mixed conclusions; some studies revealed higher toxicity and/or bio-uptake for AgNPs than AgNO 3 54,55 ; other studies report the opposite conclusion [56][57][58] . In the present study, Ag accumulation in seeds primed with AgNPs was lower than those of AgNO 3 (see supporting information, Table S3). Similar results were also observed in Allium 57 , castor 59 , and maize plants 58 , although treatment and culture methods of these studies were different from our approach. In addition, we also observed that seedlings developed from AgNO 3 priming treatments showed stunted and slow growth compared to AgNPs priming treatments, indicating that AgNO 3 showed phytotoxic effect than AgNPs. Our study confirmed the findings about stronger phytotoxic effect of AgNO 3 compared to AgNPs previously reported by other authors [57][58][59] . Taken together, these data indicated that higher Ag accumulation could inhibit seed germination and seedling growth.
Although AgNO 3 at appropriate concentrations, especially ultra-high dilution can be beneficial to plant growth 60 , silver ions at high concentration can cause phytotoxic effects in several plants [57][58][59] . It is known that Ag + ions can interact metabolically with Cu and Se and replace H 2 from the sulfhydryl groups of the photosynthetic enzymes, changing their structure and inactivating them. In addition, Ag can form complexes with amino acids, nitrogenous bases, and nucleotides, as well as with their corresponding macromolecular forms, indicating its potential to be either highly toxic or easily inactivated by the plant 61 . Besides, respiratory enzymes can be interfered by Ag + ions due to their high affinity with thiol group of the enzymes 62 . Thus, higher concentration of Ag internalized into AgNO 3 -primed seeds can induce oxidative stress in seeds, leading to lower germination indices and seedling vigor index compared with AgNP-primed seeds.
It is known that the major factors that influence AgNPs toxicity and uptake in plants are attributed to the physico-chemical properties of particles, such as size, shape, surface coating, and experimental conditions including concentration, exposure time, method of exposure, and cell type or plant species 57, 63, 64 . Although the mechanism of toxic effects of AgNPs has not yet been clearly elucidated, it is generally believed that the toxicity of AgNPs on cells and organisms is partly driven by a release of Ag + ions 57 . A previous study suggested that the oxidative stress effects of AgNPs were delayed as compared to AgNO 3 due to the slower release of Ag + ions from the AgNPs 65 . There was also evidence that Ag + (released from AgNO 3 ) exerted stronger toxicity on poplar and Arabidopsis growth than AgNPs 66 . In the present study, the release of Ag + ions from AgNPs priming solution was relatively low (Table S2), corresponding with lower amount of Ag internalized into seeds (Table S3). Based on our data and past literature, it can be surmised that during the 24 h priming exposure, rice seed uptake of Ag + from AgNPs was less than the AgNO 3 due to slow release of AgNPs.
However, the effects of AgNPs are not simply due to the release of Ag + ions into the surrounding environment as Ag + and AgNPs commonly coexist 57 . It was suggested that AgNPs themselves can induce cellular oxidative stress and genotoxicity through overproduction of ROS 42 . Additionally, Li et al. 42 demonstrated using ESR analysis that AgNPs can produce •OH directly, while AgNO 3 did not. These data indicated that the mechanism of toxicity of AgNPs and Ag + ions are different, and the nanoparticles, but not the released ions, mainly contribute to the AgNPs genotoxicity. ESR analysis also revealed that only AgNPs generated •OH. Thus, differential toxicity of AgNPs and Ag + (AgNO 3 ) could be due to the fact that AgNPs generate ROS slowly via their surface, while Ag + produces ROS via perturbation of the cell respiratory chain 42 .
As mentioned above, toxicity of AgNPs also depends on surface chemistry and coating. It has been reported that AgNPs synthesized using phytochemicals are coated by a thin layer of some capping organic material from the plant extracts and thus, this organic coating reduces their phytotoxicity than those synthesized using chemical methods 64 . Additionally, cellular uptake and variation of toxicity effects of NPs could be related to experimental conditions used, especially concentration and exposure time, and this would change (positive effect to negative or vice versa) if condition varies 63,67 .
The concentrations of NPs used in past toxicological tests were extremely high and not conducted using a realistic environmental dose, which can lead to biased or misleading conclusion on the toxicity of these NPs, especially in low concentration ranges and limiting the potential beneficial applications of NPs in agriculture and other related technology 68 . Our study showed beneficial effects of AgNPs in term of nanopriming, in which condition used only low concentration of AgNPs and short exposure time, i.e. 24 h. This method not only reduces the toxicity of AgNPs, but also enhances the germination and starch metabolism of rice aged seeds. In addition, our ICP-OES analysis confirmed that no Ag uptake was observed in plant shoots and roots derived from seeds primed with nanopriming (Table S3). This implied that translocation of Ag element from seeds into plant vegetative organs did not occur or the Ag content in vegetative parts was below detection limit of the ICP-OES. This might be because the concentrations of AgNPs used in priming treatments were relatively low and seeds were exposed to the solution only for a short period. Thus, nanopriming would be alternative method for promoting seed germination for agricultural practices.
Nanopriming modulated ROS-antioxidant enzyme system of aged seeds. Generally, seeds stored for a longer period are prone to oxidative damage due to the cumulative increase of ROS coupled with decline of the antioxidant potential of the cells, leading to seed aging and a loss of germination ability 69,70 . In this study, we found that aged seeds without priming (unprimed seeds) showed lower activities of antioxidant enzymes, but higher levels of ROS, especially H 2 O 2 at 0 h of imbibition (dry seed) (Fig. 7c,d). On the contrary, priming seeds with AgNPs can alter the activities of the antioxidant enzymes, especially CAT to higher levels, while reduce ROS production and accumulation after re-drying seeds (0 h of imbibition) compared to unprimed control and other treatments (Fig. 7b). As determined by spectrophotometric, in vivo fluorescence imaging, and histochemical localization assays, nanoprimed seeds produced higher H 2 O 2 and O 2 •after 24 and 36 h of imbibition. Coincidently, antioxidant enzymes, i.e. SOD and CAT also increased significantly in nanoprimed seeds compared to unprimed and other primed seeds. Based on these evidences, therefore, alteration of ROS production and antioxidant enzyme activities in rice aged seeds after priming can be linked with higher germination in nanoprimed seeds compared with the unprimed ones.
In a recent concept of plant seed physiology, it has been proposed that the germination is completed only when the ROS content is within an oxidative window that allows ROS signaling 71,72 . Above or below the 'oxidative window for germination' , low or high amounts of ROS would not permit progress towards germination. Additionally, H 2 O 2 is one of key ROS molecules acting as signalling molecule for the regulation of seed germination, and the precise regulation of H 2 O 2 content by antioxidant machinery is essential to achieve a balance between oxidative signalling that stimulates germination and oxidative damage that prevents or delays germination 73 . Thus, higher level of H 2 O 2 observed in nanoprimed seeds can act as signal molecule and was consistent with oxidative window concept, which resulted in better germination and faster seedling growth compared with that in unprimed and other primed seeds.
Antioxidant systems also play an important role in seed physiology 71 . In order to act as the positive regulator of germination, ROS accumulation must be tightly controlled by antioxidant systems. In the present study, activities of two antioxidant enzymes, i.e. SOD and CAT are relatively high in nanoprimed seeds at 24 h of imbibition, coinciding with high levels of ROS. Previous reports suggested that primed seeds exhibited more robust antioxidant system than unprimed ones during germination or early seedling establishment 8 , and increased antioxidant properties in primed seeds can enhance fitness of seedlings and plants subjected to stress conditions 52 . Similar to our finding, Kibinza et al. 74 demonstrated that osmopriming sunflower seeds can increase the expression levels of a gene encoding catalase, and suggested that CAT is the key enzyme responsible for the recovery of vigor in aged seeds. SOD is the primary antioxidant enzyme that scavenges O 2 •to H 2 O 2 , while CAT catalyzes the conversion of H 2 O 2 to H 2 O 75 . Considering the relationship between ROS and antioxidant enzymes, the high amount of antioxidant enzymes observed in nanoprimed seeds can be due to induction by AgNPs-mediated ROS, which is a tightly controlled mechanism to balance ROS in oxidative window range for stimulating seed germination as discuss above.

Mechanism of enhancing seed germination and starch metabolism by nanopriming.
To date, the mechanism underlying nanopriming-induced seed germination has not yet been reported in the literature. In the present study, the mechanism of beneficial roles of AgNPs in inducing seed germination was hypothetically proposed, including (i) creation of the seed coat nanopores, (ii) mild stress-induced agent or ROS-generating agent, and (iii) nanocatalyst for enhancing starch-degrading enzyme activity.
Although the precise mechanism behind nanoparticles action on physiological enhancement of seed germination is not clearly understood, one of the possible mechanisms to explain this scenario is based on Khodakovsakaya et al. 's hypothesis 76 . These authors hypothesized that carbon nanotubes were able to penetrate seed coat by creating small pores, resulting in increased water uptake and up-regulation of the expression of aquaporin genes involved water uptake. Our study also revealed that priming with AgNPs could enhance rapid water uptake and expression of aquaporin genes in germinating seeds. Meanwhile, a higher accumulation of ROS (H 2 O 2 and O 2 •-) was observed in germinating seeds of nanopriming treatments than that of unprimed control and other priming treatments, indicating that more ROS are required for stimulating seed germination. Based on these evidences, both aquaporins and ROS could participate in activating seed germination. Previous studies suggested that aquaporins not only help water uptake, but also facilitate H 2 O 2 or ROS diffusion across biological membranes 51 . Also, the interplay of aquaporins with ROS has been reported for regulating plant development 47,50 . Thus, we further hypothesized that upon the formation of nanopores, the influx of the water into seeds becomes rapid, and also trigger the expression of aquaporin genes. Once aquaporin genes are up-regulated, they can facilitate the diffusion of ROS, especially H 2 O 2 through biological membrane. The elevated ROS must be tightly controlled by seed antioxidant systems to produce ROS in optimum range of oxidative window as signalling molecules for triggering essential metabolic activity of seed for promoting seed germination and seedling development (Fig. 11). Figure 11. Proposed mechanism of AgNPs induced seed germination of rice. (a) Seed without AgNPs priming treatment has low metabolic activity due to slow water uptake; in this condition, the hydrolysis of starch occur gradually and thus available sugars is also low in the initial phase of imbibition, resulting in slow seed germination and seedling growth. (b) Seed priming with AgNPs can enhance seed germination with at least three possible scenarios. Firstly, AgNPs could penetrate seed coat by interacting with cell walls of seed coat to create small pores and thus enhance water uptake into seeds. Rapid water uptake can induce higher metabolic activity and starch hydrolysis of seed during initial phase of imbibition. Secondly, penetrated AgNPs could interact with α-amylase as NPs-amylase complex to fasten the starch hydrolysis and thus seed can generate higher available sugars to support embryo growth. Thirdly, AgNPs could mediate generation of ROS being in oxidative window range for acting as signalling molecules and participating in cell wall loosening and endosperm weakening. See texts for further explanation.
SCIEnTIFIC REPORTS | 7: 8263 | DOI:10.1038/s41598-017-08669-5 Furthermore, we observed that amylase activity increased in response to nanopriming treatment resulting in increased soluble sugars. Increasing sugar concentrations in the cells can lead to lowering the osmotic potential, and hence water potential. This increases the difference (gradient) between the water potential outside and inside the tissues facilitating water movement into the seeds by osmosis. As soluble sugar contents increased in parallel to increased amylase activity in nanoprimed seeds, the enhancement of water uptake in their seeds could also be attributed to the change of internal osmotic potential due to increased soluble sugars (solutes).
Although plant cell growth is driven by water uptake, the growth is restricted by cell wall. As plant cell walls determine shape and the rate and direction of growth of individual cells as well as the mechanical resistance of whole tissues, the loosening of cell walls is an important process in all stages of plant development, requiring elongation growth or tissue weakening 77 . It has been approved that ROS, including •OH can act as a key component in cell wall loosening, testa and endosperm weakening, which is essential for radicle protrusion in seed germination [77][78][79] . As observed by ESR technique, AgNPs could mediate the generation of •OH, and thus act as ROS-generating agent. Hypothetically, it is likely that during the short duration (i.e. 24 h) of soaking seeds in nanopriming solutions, •OH generated by AgNPs could help loosening seed coat cell walls as well as endosperm weakening for enhancing seed germination. Besides, although ROS at high levels could induce oxidative stress, their optimum levels can be beneficial for seed germination. A slightly enhanced level of oxidative stress has been observed to stimulate seed germination 80 . Thus, priming seeds with AgNPs solutions could also induce mild oxidative stress for stimulating seed germination in addition to loosening cell wall.
We also propose another hypothesis that AgNPs penetrating the seeds might be important for their role as nanocatalyst in the hydrolysis of starch catalyzed by α-amylase and thereby increasing the reaction rate. It has been recently reported that AgNPs have the potential to act as nanocatalyst, and the rate of enzymatic digestion of soluble starch increased rapidly in the presence of nanosilver 81 . However, the exact mechanism behind the enhancement of starch hydrolysis by AgNPs is still unclear. A plausible explanation for one of the possible scenarios could be that, in the presence of AgNPs inside the seed, the α-amylase could possibly interact with functional groups of biomolecular ligands on the surface of phytochemicals-coated AgNPs through the thiol linkages, and thus the enzyme molecule is immobilized on the surface of AgNPs 81,82 . Consequently, the α-amylase-bound nanoparticles complex (α-amylase-NPs complex) is able to enhance the rate of starch hydrolysis (Fig. 11). Structurally, α-amylase consists of two cysteine thiol groups which can expose to the medium or substrate and away from the active site 83 . It was suggested that AgNPs could be stabilized by the protein molecule through the thiol linkages and thus the enzyme molecule could be immobilized 81 . Therefore, it is possible that the α-amylase induced inside seeds could form the thiol linkages with functional groups of biomolecular ligands on the surface of phytochemicals-coated AgNPs with the active site being available for catalysis. The faster rate of starch hydrolysis of NP-amylase complex can be due to the enzyme conformational change, but the enzyme still retains its active site for catalysis 82,83 . It has been well documented that upon a nanocarrier support, the efficiency of the enzyme increased compared to its free form 81,84 . Therefore, the enhancement of enzyme activity and stability resulting in higher starch degradation may be attributed to the conformational change in the enzyme after interacting with molecules coated onto the surface of Ag nanocarriers. However, the exact mechanism warrants further investigation.

Conclusion
In the present work, eco-friendly synthesis of AgNPs was successfully demonstrated through the use of kaffir lime leaves extracts as natural reducing and stabilizing agents. Various characterization techniques (UV-vis, TEM, EDX, SAED, XRD and FTIR) confirmed the formation of phytochemicals-capped AgNPs. We demonstrated that AgNPs can be applied as nanopriming agent for enhancing seed germination and starch metabolism of rice aged seeds. It was evident that AgNPs can internalize seed coat and support water uptake inside seeds, leading to promote seed germination and starch metabolism. Besides, our study indicated that priming with AgNPs priming could up-regulate the expression of aquaporin genes, thus facilitating water and H 2 O 2 diffusion. Meanwhile, elevated ROS, especially H 2 O 2 observed in nanoprimed seeds during germination could act as signal molecules for stimulating germination process. These evidences suggested that ROS and aquaporins play essential roles in enhancing seed germination. Furthermore, as evidenced by ESR technique, AgNPs could mediate the •OH generation, while AgNO3 did not. Thus, we hypothesized that when soaking seeds with nanopriming solution, •OH could help loosening seed coat cell wall and endosperms. Additionally, AgNPs penetrated into seeds, as determined by ICP-OES, could possibly interact with α-amylase enzyme or act as nanocatalyst; thereby enhancing seed starch degradation for seed germination and seedling growth. Since our nanopriming technique works with low concentration of AgNPs and does not apply NPs into soils, it prevents the dispersal of large amounts of NPs into the ecosystems. Nanopriming could be further exploited by seed companies as a new technique for commercial seed priming. Our study can serve as an initiative of nanopriming application for sustainable agricultural practices and agri-seed industry in the future.

Methods
Chemicals and materials used for AgNPs synthesis. AgNO 3 (99.98%) was purchased from Merck (Germany). All aqueous solutions were prepared using double distilled deionized (DI) water. All reagents were of analytical grade. All glassware was washed with aqua regia solution and rinsed three times with DI water. Plant materials. Kaffir lime (Citrus hystrix D.C.), a small tree belonging to family Rutaceae, has been widely used as flavoring in Southeast Asian cuisines due to its highly aromatic and unique citrus flavors and strong fragrances. Kaffir lime leaves possess some important bioactive compounds (e.g. polyphenols), and given its good antioxidant properties 40 we selected kaffir lime leaves (Inset of Figure S1 in Supporting information) as natural material source for AgNPs biosynthesis.
Preparation of plant extract. Preparation of plant extract as reducing and stabilizing agents for AgNPs synthesis was done according to the method of Karthik et al. 37 with some modifications. Briefly, the fresh leaf samples were initially rinsed with running tap water and then washed with DI water to remove the adhering dust particles and other contaminants. After blot drying, leaves (20 g) were cut into small pieces, blended in a kitchen blender with 100 ml of DI water. The blended extract was heated at 80 °C for 5 min and cooled at room temperature. Then, the extract was filtered through Whatman No.1 filter paper before centrifuging at 5000 rpm for 15 min to remove the heavy batteries. The filtered extract was kept in refrigerator (4 °C) up to 1 week for it to be used as reducing and stabilizing agents.

Synthesis of silver nanoparticles (AgNPs).
In a typical reaction procedure, 3 mL of kaffir lime leaf extract was added to 12 mL of 1 mM AgNO 3 aqueous solution and incubated in the dark for 1 h with gentle stirring. The solution turned to yellowish brown color after 1 h incubation and the reaction completed within 24 h. The AgNPs obtained from the solution were purified by repeated centrifugation at 10,000 g for 15 min followed by dispersion of the pellet in sterile DI water 4 times. The water-suspended nanoparticles were lyophilized for 48 h followed by characterization for the structure and composition.
Characterization of phytosynthesized AgNPs. The bioreduction of the silver ions in the solution was determined under UV-visible spectroscopy using Perkin-Elmer Lambda 2 UV198 between 300 to 800 nm. Morphological details of purified AgNPs were measured with transmission electron microscopy (TEM) at 200 kV using FEI TECNAI G 2 20 equipped with selected area electron diffraction pattern (SAED). Simultaneously, the elemental compositions of the samples were also recorded using Energy Dispersive X-ray analysis. X-ray powder diffraction (XRD) using Cu Kα radiation with λ = 0.15418 nm (Bruker D2 phaser) was used to investigate the crystalline nature and particles size of the synthesized AgNPs. Attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectra of the freeze-dried samples were recorded using a Bruker Tensor 27 equipment to determine the surface functional groupsof the AgNPs.
Preparation of priming solutions. The phytosynthesized AgNPs were used for seed priming test. AgNPs concentrations at 10 and 20 mg L −1 , designated as AgNPs10 and AgNPs20 nanopriming agents, respectively, were freshly prepared by dispersing the particles in deionized water using ultrasonic vibration (100 w, 40 kHz) for 30 min. Similarly, for silver nitrate (AgNO 3 ) priming solutions, two concentrations of AgNO 3 at 10 and 20 mg L −1 designated as AgNO 3 10 and AgNO 3 20, respectively, were dissolved in deionized water and kept in dark bottle. Deionized water was used as hydropriming.
Seed priming method. Seeds of jasmine rice (Oryza sativa L. cv. KDML105) were obtained from Khon Kaen Rice Research Centre, Thailand. These were naturally aged seeds due to long period of storage under ambient temperature (25-30 °C) for 3 years. The initial seed moisture content (MC) was 8.9% (on dry weight basis). Healthy seeds were selected from the same seed lot and used for all the experiments. For seed priming method, seeds were surface-sterilized in 3% H 2 O 2 and then rinsed with deionized water for several times. Seeds were then soaked in AgNO 3 or AgNPs priming solutions for 24 h with continuous aeration. The ratio of seed weight to solution volume was 1:4 (gmL −1 ). Seeds soaked in deionized water were defined as hydropriming (HP) and were washed three times with deionized water (3 min) and surface-dried on paper towel. Seeds were dried back to their original moisture content in shed at room temperature (25 ± 2 °C), sealed in polythene bags and stored at 4 °C until further use. The untreated seeds were used as the unprimed control (UP).
Germination assay and growth measurement. Germination tests on the primed and unprimed seeds were performed in triplicates. Briefly, a piece of sterile filter paper (Whatman no.1) was placed on a sterile plastic Petri dish (90 × 15 mm) and moistened with 5.0 ml deionized water. Ten sterilized seeds were allowed to equilibrate under room temperature for 24 hrs, then placed in each plate, covered with lid and sealed with Parafilm M ® to avoid moisture loss. All Petri plates were kept in an incubator under dark at 25 °C.
After initiation of germination assay, seed germination was monitored daily for 6 days and seeds were considered germinated when the radicle was extended to more than 5 mm. Besides, at the end of germination experiment, length and weight of seedlings (roots and shoots) were measured. The data on germination percentage (GP), germination index (GI), germination rate (GR), and seedling vigour index (SVI) were calculated according to the equations provided elsewhere 85, 86 . Seed water uptake and dehydrogenase activity. The water uptake by seeds during imbibition at 4 and 24 h was determined following the method described elsewhere 87 . Dehydrogenase activity of the root tips was determined according to the method given by Singh et al. 88 Starch metabolism and starch agar plate α-amylase assay. Starch metabolism in germinating seed and seedlings was assessed in terms of α-amylase activity and total soluble sugar content 15 . The α-amylase was determined following the method of Bernfeld 89 . Germinating seeds at 24 h of imbibition and seedling samples including shoot and root (1 g) were ground, mixed with 100 ml distilled water, and left standing for 24 h at 4 °C. The clear supernatant was collected and stored at −80 °C until analysis. The enzyme activity was determined from the supernatant by 3,5-dinitrosalicyclic acid (DNS) method 90 . In order to determine total soluble sugar contents, ground seedling sample (1.0 g) was mixed with 10 ml distilled water, left for 24 h at 25 °C 15 . The mixture was filtered with Whatman No. 42 and the final volume was made to 10 ml with distilled water. Total soluble sugar contents were determined by the phenol sulfuric method 91 .
The production of α-amylase using starch-agar plate assay was performed as described previously 32 . The embryoless half-seeds were surface sterilized with 1% sodium hypochlorite for 15 min and washed 5 times with sterilized water. The sterilized half-seeds were placed perpendicularly on autoclaved starch-agar plate [2% agar plate containing 0.2% soluble potato starch, 10 mM sodium acetate and 2 mM CaCl 2 (pH 5.3)] and incubated at SCIEnTIFIC REPORTS | 7: 8263 | DOI:10.1038/s41598-017-08669-5 25 °C in the dark for 3 days. GA plates were made by adding 1 μM gibberellic acid (GA 3 ) to the cooled agar, while no GA 3 was added to negative control plate. They were then removed and amylase production was examined by staining agar plate with IKI solution (0.1% I 2 and 1% KI). Clear zones (transparent halos) appear if α-amylases are synthesized in endosperms and secreted into the starch agar, resulting in starch hydrolysis. Four endosperms of each treatment were placed in the same plate. All measurements were performed in triplicates.

Starch granules morphology.
After 24 h of imbibition, dehusked seeds from unprimed and primed groups (HP, SNP10 and SN10) were steeped in 0.2% NaOH for 2 days and homogenized in a blender. The homogenate was squeezed through five layers of cotton cloth and then filtered with 100-, 200-, and 400-mesh sieves, successively. Starch granules were washed and further purified according to a method described by Wei et al. 33 The isolated starch granules were mounted on the specimen stub and sputter-coated with gold before viewing with scanning electron microscope (SEM, LEO 1450VP) at an accelerating voltage of 15 kV.
Measurement of reactive oxygen species and antioxidant enzyme activities. Hydrogen peroxide (H 2 O 2 ) content in the rice seed embryos imbibed for 24 h was measured according to the method described by Patterson et al. 92 , while O 2 •content was determined according to the method of Elstner and Heupel 93 . Antioxidant enzymes were extracted from aged rice embryos imbibed for 24 h. Seed embryos (~0.1 g) were homogenized in 2 mL of 50 mM potassium phosphate buffer (pH 7.8) containing 2 mM dithiothreitol (DDT), 0.1 mM ethylenediaminetetraacetic acid (EDTA), and 1.25 mM polyethylene glycol 4000 (PEG 4000) under cold conditions. Homogenate was centrifuged at 10,000 g for 15 min at 4 °C and the supernatant was used for the determination of superoxide dismutase (SOD, EC 1.15.1.1) and catalase (CAT, EC 1.11.1.6) as described elsewhere 70 . in vivo localization of hydrogen peroxide (H 2 O 2 ) in germinating seeds. To in vivo localize and image H 2 O 2 in germinating seeds, 2′,7′-dichlorofluorescein diacetate (DCF-DA) fluorescence probe was used to stain dehusked seeds after 24 h of imbibition according to the method of Bailly and Kranner 94 with minor modifications. Briefly, imbibed seeds were stained with 25 µM DCF-DA in 20 mM potassium phosphate buffer (pH 6.0). Seeds were incubated at 25 °C in the dark for 30 min before rinsing three times with 20 mM potassium phosphate buffer (pH 6.0) for 5 min each. As negative control, seeds were preincubated with 5 mM ascorbate in darkness for 1 h at 25 o C before adding the probe. The stained seeds were viewed at 4X on a Nikon ECLIPSE 80i fluorescence microscope with NIS-Element AR 3.2 sofware (Melville, NY, USA). The excitation and emission wavelengths for fluorescence probe are 488 and 525 nm, respectively.

Histochemical localization of H 2 O 2 and O 2
•in germinating rice seeds. To localize H 2 O 2 and O 2 •-, DAB and NBT were used to stain seeds, respectively, as described by Chen et al. 95 After 36 h of imbibition, five whole seeds were incubated with 1 mg mL −1 DAB solution (pH 3.8) or 1 mM NBT in 10 mM Tris-HCl (pH 7.0) at room temperature for 30 min. The stained seeds were washed with double-distilled water, and photographed through a bright field mode of Nikon ECLIPSE 80i fluorescence microscope.
Determination of silver contents in seeds, seedlings and priming solutions. Unprimed and primed seeds (10 seeds from each replication) were dehusked and oven-dried at 80 °C for 48 h. Seeding shoots and roots (6 days old) were also oven-dried and used for Ag analysis. A representative sample of up to 0.5 g (seeds) or 0.1 g (shoots or roots) was digested in 9 mL of concentrated nitric acid and 3 mL of hydrofluoric acid for 15 min using the microwave heating system, diluted with MilliQ water and analyzed using inductively coupled plasma atomic emission spectroscopy (Optima 2100 DV ICP-OES, Perkin Elmer Instruments, USA) as described elsewhere 96 . Similarly, Ag + ions in priming solutions, i.e. AgNPs10, AgNPs20, AgNO 3 10, AgNO 3 20, and DI water were also analyzed using ICP-OES.
Electron Spin Resonance (ESR) analysis of hydroxyl radicals. The ESR analysis of hydroxyl radicals (•OH) was carried out according to the method of Li et al. 42 , except that after addition of AgNPs or AgNO 3 to spin trap, the particle materials were removed by centrifugation and filtration according to the recommendation of Jeong et al. 97 All ESR measurements were performed at room temperature using a Bruker Elexys500 spectrometer. The ESR parameters were as follows: 1 G field modulation, 100 G scan range and 20 mW microwave power for detection of spin adducts using 5-dimethyl-pyrroline N-oxide (DMPO) spin trap (Santa Cruz Biotechnology).

RNA extraction and gene expression analysis.
Total RNA was extracted from fresh rice embryos imbibed for 24 h using the PureLink ® RNA Mini Kit (Life Technologies, USA) according to the manufacturer's instructions. Isolated RNA was further purified using the PureLink TM DNase Set (Invitrogen, USA) following the manufacturer's protocol. The purity and concentration of RNA were checked using the NANOdrop spectrophotometer. First strand cDNA synthesis was performed using SuperScript ® III First-Strand Synthesis System for RT-PCR (Invitrogen, USA) according to the manufacturer's instructions. Expression analysis of the two aquaporin genes using semi-quantitative reverse transcriptase-polymerase chain reaction (sqRT-PCR) was essentially as described previously 98 . Primer sequences for the studied aquaporin and actin genes were designed according to Sakurai et al. 98 and Nounjan et al. 99 , respectively.

Statistical analysis.
A one-way analysis of variance (ANOVA) was performed between treatment samples in a completely randomized block design using three replications and data were analyzed using IMB SPSS statistics version 22. The significant levels of difference for all measured traits were calculated and means were compared by Duncan's Multiple Range Test at 5% probability level. The p value < 0.05 was considered as statistically significant.