Nano‐silicon fertiliser increases the yield and quality of cherry radish

Although silicon‐based nanomaterials (Si‐based NMs) can promote crop yield and alleviate biotic and abiotic stress, the underlying performance mechanisms are unknown. In the present study, the effect of the root application of Si‐based NMs on the physiological responses of cherry radish (Raphanus sativus L.) was evaluated in a life cycle experiment. Root exposure to 0.1% (w/w) Si‐based NMs significantly increased total fresh weight, total chlorophyll and carotenoids by 36.0%, 14.2% and 18.7%, respectively, relative to untreated controls. The nutritional content of the edible tissue was significantly enhanced, with an increase of 23.7% in reducing sugar, 24.8% in total sugar, and 232.7% in proteins; in addition, a number of nutritional elements (Cu, Mn, Fe, Zn, K, Ca, and P) were increased. Si‐based NMs exposure positively altered the phytohormone network and decreased abscisic acid content, both of which promoted radish fresh weight. LC‐MS‐based metabolomic analysis shows that Si‐based NMs increased the contents of most carbohydrates (e.g., α‐D‐glucose, acetylgalactosamine, lactose, fructose, etc.) and amino acids (e.g., asparagine, glutamic acid, glutamine, valine, arginine, etc.), subsequently improving overall nutritional values. Overall, nanoscale Si‐based agrochemicals have significant potential as a novel strategy for the biofortification of vegetable crops in sustainable nano‐enabled agriculture.


INTRODUCTION
The prevalence of food insecurity and malnutrition has increased due to greater population pressure, changing climates, economic shock and regional conflicts. 1,2pproximately 702-828 million people worldwide faced starvation in 2021 according to the latest data from the Food and Agriculture Organization of the United Nations (FAO). 3Nanotechnology has demonstrated potential as an effective strategy for promoting food security and may be more efficient and sustainable than other approaches, such as use of excessive fertiliser and other defensive agrochemicals. 4 Kah et al. conducted an analysis of the use efficiency and environmental impact of nanoscale fertilizers and conventional fertilizers, and the results showed that nanoscale approaches could increase crop yield by 20%-30% with minimal environmental impacts. 5ilbertson et al. evaluated the benefits of various engineered nanomaterials (ENMs) and found that foliar spray of selected ENMs could increase crop yield (up to 72% for ZnO NMs) with less energy cost (about 1 � 10 −4 MJ/ plant) compared with corresponding bulk materials or metal ion substitutes. 6Although the application of ENMs in agricultural production could bring great benefit, an understanding of the critical mechanisms of action, as well as the environmental and human health risks, need to be acquired prior to the wide scale implementation of these strategies.
Silicon (Si) is the second most abundant element in soils and is considered a beneficial element to plants, particularly for the growth and development of gramineous crops (rice, wheat, barley, etc.). 7The advantages of Si in promoting plant growth while alleviating biological/ abiotic stress have also been demonstrated. 8,9For example, the application of blast furnace slag-based silicate fertiliser in the northeastern regions of China significantly increased the average yield of wheat, lettuce, tomato and soybean crops by more than 10% while simultaneously exhibiting a significant economic advantage with a benefit/cost ratio of 42.9 for lettuce. 10Si can improve the response of plants to different abiotic stresses, such as salt, 11 drought, 12 temperature, 13 metal toxicity 14 and excessive nutrients. 15Importantly, nanoscale SiOx or Si-based fertilizers have demonstrated greater impacts at promoting crop yield and resisting biotic/abiotic stresses in comparison with traditional Si fertilisers. 16,17However, an understanding of the nanoscale specific mechanisms of action, as well as the optimisation of Si materials, needs to be further explored.
Investigations on the use of nano-silica to enhance crop nutrition and protection have been increasing over the past 20 years, and projections suggest a rapid publication growth rate of 83% over the next 6 years (2022-2028) (Figure S1 and S2).9][20][21][22][23] In addition, nano-silica has been used to alleviate heavy metal toxicity, as well as drought and salt stress in plants.Manzoor et al. reported that root exposure of 250 mg/kg-soil silica nanoparticles (SiO 2 NPs) significantly alleviated the adverse effects of chromium (Cr) stress on wheat (Triticum aestivum L.) by increasing the activities of antioxidant enzymes (CAT, APX, SOD and POD), reducing the level of reactive oxygen species, and decreasing the in planta Cr content. 24Valizadeh-rad et al. demonstrated that SiO 2 NPs significantly enhanced the resistance of both canola and wheat to drought-stress in comparison with silicon presented in a salt form (potassium silicate) at the same dose of 200 mg-Si/kg-soil. 25In general, the literature demonstrates that the application of nanosilica has a positive impact on crop yield.However, the physiological responses and biological metabolites directly impacted by nano-silica are unclear, and investigations on the environmental and human health risks related to this novel fertiliser are needed.
Cherry radish (Raphanus sativus L.) provides a rich source of nutrients such as carbohydrates, amino acids (AAs) and minerals (e.g., Ca, Fe, Mg and K), which contribute to human health and prevent chronic diseases. 26The global production of cherry radish is approximately 7 million tons per year, accounting for approximately 2% of the total global vegetable production. 27In the present study, Si-based NMs, made from mineral resources (mainly composed of quartz, kaolinite, and illite), were used as a novel fertiliser to promote the yield and quality of cherry radish.Physiological responses of cherry radish (Raphanus sativus L.) upon exposure to Si-based NMs were evaluated.The role of Si-based NMs in altering phytohormone networks and subsequent crop growth was evaluated.Additionally, the metabolomic profile of shoots and the edible tissue of cherry radish affected by Si-based NMs was analysed.Overall, our findings suggest a new strategy for enhancing crop growth and yield and subsequently strengthening crop resistance to environmental stress using sustainable nanoscale Si.

Si-based NM preparation and characterisation
Raw materials (mainly composed of quartz, kaolinite, and illite) for nanomaterial synthesis were collected in the Songshan Mountains and surrounding areas in Henan Province, China (34°72 0 12″N; 112°30 0 08″E) and were washed repeatedly with distilled water to remove surface impurities, followed by drying for 3 h in a drying oven at 70°C.The materials were then crushed into uniformly sized particles (5-10 mm) using a jaw crusher (WFY-60, Wheatstone), and were further finely ground using a planetary ball mill (QM-3SP4, Nanda Instrument) with water as the medium.The operating conditions were as follows: milling time: 10 h; rotation speed: 500 rpm; ball-to-powder ratio: 8:1.The NM suspension was centrifuged at 10,000 rpm for 5 min followed by repeated washing with deionised water and centrifugation; the process was repeated three times.The Sibased NMs were obtained by drying at 70°C for 3 h.
Field emission transmission electron microscopy (TEM, FEI Talos F200S) and field emission scanning electron microscopy (SEM, FEI NOVA NANO450) were used to determine the morphology and structural characteristics of the Si-based NMs.The thickness of the Si-based NMs was measured using atomic force microscopy (AFM, Bruker Dimension ICON), and AFM 3D-image and corresponding height curve were constructed by NanoScope Analysis 1.7 software.The hydrodynamic particle size and surface charge of MODERN AGRICULTURE Si-based NMs were determined by dynamic light scattering (DLS, Malvern Zetasizer Nano ZS90).Fourier transform infrared spectroscopy (FTIR, Nexus 670) was used to evaluate the molecular structure and functional groups present within the Si-based NMs.The particle elemental composition was analysed using SEM with energy dispersive X-ray spectroscopy (EDS).The composition of Si-based NMs was further investigated by wide-angle x-ray diffraction (XRD), and the diffraction pattern was compared with standard substances using MDI Jade 6 software.

Plant growth experiment
Cherry radish (Raphanus sativus L., Dongsheng Seed Industry Co., Ltd, Beijing, China) seeds were planted in plastic boxes filled with 500 g of agricultural soil.Details of soil characterisation are provided in Table S1.Subsequently, the soil was homogenously mixed with different doses of Si-based NMs (mass ratio: 0, 0.1%, 0.5%, 1.0%, 2.0% and 5%); six biological replicates were established for each treatment.All plants were maintained under greenhouse conditions (temperature: 25/ 20°C, day/night; light period: 16/8 h, day/night; humidity: 75%).The plants were watered with 40 mL tap water daily for 35 days.Additionally, 30 mL half-strength Hoagland's solution was applied to each pot once per week to support growth.Phenotypic images of cherry radish were taken at harvest, and the total fresh weight, shoot fresh weight, and edible tissue fresh weight were recorded.The shoot and root samples were ground into a fine powder using a tissue grinder (SCIENTZ-48) in liquid nitrogen and stored at −80°C for subsequent analysis.

Photosynthetic pigment and nutritional composition
Photosynthetic pigment contents: The content of photosynthetic pigments, including chlorophyll (chla, chlb), total chlorophyll and carotenoids, were measured as described by Lichtenthaler. 28Briefly, 100 mg of fresh leaves were ground in liquid nitrogen and then transferred into a centrifuge tube containing 10 mL of 95% (v/v) ethanol.The pigments were extracted in the dark for 48 h.The absorbance of chlorophyll and carotenoids was measured using a UV-visible spectrophotometer (GENESYS, Thermo Scientific) at 665, 649 and 470 nm.The contents of leaf chlorophyll and carotenoids were calculated as follows: Chl a ¼ 13:95A 665 -6:8A 649 ; Chl b ¼ 24:96A 649 -7:32A 665 ; Nutritional composition: The nutritional composition of radish edible tissue, including total sugar, reducing sugar, protein, and AAs, was measured.The total sugar and reducing sugar were measured using the dinitrosalicylic acid (DNS) method, 29 and the AA and soluble protein content were determined using the spectrophotometric ninhydrin method and Bradford method, respectively. 30

Elemental content measurement
The shoot and edible tissues of cherry radish were lyophilised for 48 h in a freeze dryer (Gold SIM-FDFD, USA) and digested as described by Ma et al. 31 Briefly, 100 mg of shoot or edible tissue samples were weighed into a digestion tube, and 3 mL of concentrated HNO 3 was added for a 40 min digestion at 115°C.After cooling to ambient temperature, 0.5 mL of H 2 O 2 was added for another 20-min digestion at 115°C.Then, pre-cooled digests were diluted to 50 mL with deionised water.Elemental Si content and other essential elements, including micronutrients (e.g., Cu, Mn, Zn, Fe) and macronutrients (K, Ca, P, S), were measured in shoot and fruit tissues by inductively coupled plasma optical emission spectroscopy (ICP-OES, Thermo Fisher iCAP 700).A reference sample with known elemental concentrations was measured every 30 samples to ensure the accuracy of the data.

Phytohormone content
The content of abscisic acid (ABA), jasmonic acid methyl ester (JA-ME), brassinosteroids (BR), gibberellins (GA 3 and GA 4 ), dihydrozeatin riboside (DHZR), zeatin (ZR), and auxins (indole-3-acetic acid, IAA, and indolepropionic acid, IPA) in shoot and edible tissues were measured using enzyme-linked immunosorbent assay (ELISA) as described by Hao et al. 32 The enzyme-linked immunosorbent assay kits were provided by Professor BM Wang from China Agricultural University, Beijing, China. 33,34Briefly, the tissues were ground in liquid nitrogen to fine powder, and then 0.5 g of sample was weighed into a tube containing 2 mL of extraction solution (a mixture of 80% methanol (v/v) and 1mmoL BHT).The samples were incubated at 4.0°C for 4 h and centrifuged at 400 rpm for 15 min.The supernatant was then passed through a C18 Sep-Pak filter cartridge (Waters Corp., Millford, MA, USA), concentrated under nitrogen, and diluted 500 times with phosphate-buffered saline containing 0.1% (v/v) Tween-20 and 0.1% (w/v) gelatine.The diluted samples were incubated in 3.0% (v/v) skimmed dry milk at 37°C for 1 h and washed repeatedly with phosphate buffered solution (PBST).The standard substances, test samples, and antibodies were sequentially added to an ELISA plate, incubated at 37°C for 30 min, and then washed with PBST to eliminate non-specific binding.The o-phenylenediamine (OPD)/H 2 O 2 reaction substrate system was added, and samples were incubated for 30 min.The absorbance of all samples was recorded at 492 nm using a multifunctional ELISA reader (BioTek Synergy HTX, USA).

Nontargeted metabolomics analysis
The metabolomic profile of shoots and edible tissue in the control and 0.1% Si-based NM treatments were analysed.There were four biological replicates in each treatment.Sixty mg of fresh tissues were weighed into a pre-cooled 1.5 mL centrifuge tube containing 600 μL of methanol-water (v:v = 7:3, containing 4 μg/mL L-2chlorophenylalanine).All samples were ultrasonicextracted in an ice water bath for 30 min and then left to stand overnight at −40°C.The extracts were centrifuged at 12,000 rpm, 4°C for 10 min and then 150 μL of the supernatant was filtered using a 0.22 μm organic phase pinhole filter, and transferred into a LC injection vial for analysis by liquid chromatographytandem high-resolution mass spectrometry system (LC-MS) (Dionex U300 UHPLC and QE plus, Thermo Scientific, USA).The chromatographic conditions were as follows: Column: ACQUITY UPLC HSS T3 (100 mm � 2.1 mm, 1.8 μm); Column temperature: 45°C ; Mobile phase: A-water (containing 0.1% formic acid), B-acetonitrile; Flow rate: 0.35 mL/min; Injection volume: 5.0 μL.The mass spectrometry conditions were as follows: Positive and negative ion scanning modes were used for sample mass spectrometry signal acquisition.
The LC-MS data were baseline-filtered, peakidentified, integrated, retention time-corrected, peakaligned, and normalised using the software Progenesis QI V2.3 (nonlinear, kinetic, Newcastle, UK).Qualitative analysis was performed using the human metabolome database, Lipidmaps (V2.3),Metlin, EMDB, and PMDB databases.The variable importance in projection (VIP) values obtained from the orthogonal partial least squares discrimination analysis model were used to rank the overall contribution to group discrimination for each variable.A two-tailed Student's t-test was used to assess whether the differences in metabolites between groups were significant.Metabolites with VIP values > 1 and p-values <0.05 were identified as significantly different metabolites.Partial least-squares discriminant analysis (PLS-DA) was performed using the mixOmics package in R 4.2.1 software.The clustering diagram was generated using the pheatmap package in R 4.2.1 software. 35Bubble plots and volcano plots were generated using the ggplot2 package (version 3.3.2) in R 4.2.1 software. 36ircular diagrams were generated using the ggraph package in R 4.2.1 software.

Data analysis
The mean and standard deviation across all treatments were calculated from 4 to 6 biological replicates.A oneway ANOVA was used to determine statistical significance (p ≤ 0.05) for different Si-based NM treatments.A Student's t-test was used to compare the differences between the control and the Si-based NM treatments.Data analysis was performed using IBM SPSS Statistics 25.

Characterisation of Si-based NMs
The prepared Si-based NMs exhibit a "flake-like" morphology and are approximately 100-1000 nm in length (Figure 1a,b).AFM imaging shows that the thickness of the Si-based NMs is 10-20 nm (Figure 1c, Figure S3A,B).The layered nanostructures of Si-based NMs have the potential to serve as nano-delivery systems.For example, Hao et al. demonstrated the enhanced pesticide loading capacity of boron nitride nanoplatelets due to their layered nanostructure. 37The average hydrodynamic diameter of the Si-based NMs is 420 nm and the single peak of the hydrated particle size curve indicates that the material is relatively uniform (Figure 1d).The Si-based NMs had a zeta potential of −24.3 mV, indicating reasonable colloidal stability.The FTIR spectrum shows absorption bands at 3343 and 1040 cm −1 that are attributed to hydroxyl and -O-Si, respectively (Figure 1e).The EDS spectrum shows that the main elements are O (40.82%), C (33.25%), Al (9.80%), Si (9.07%) and Fe (3.91%) (Figure S4).The XRD spectrum indicates that the material compositionally includes carbonyl iron carbide, kyanite, and SiO 2 , according to the inorganic crystal structure database (Figure 1f).The comparison results with the standard spectrum are shown in Table S2.

Plant physiological response
Soil application of Si-based NMs significantly increased the growth and yield of cherry radish compared to untreated controls; additionally, in the 5% Si-based NM treatment, no phenotypic phytotoxicity was evident, demonstrating the overall biocompatibility of the material (Figure 2a and Figure S5).The addition of 0.1% Si-based NMs significantly increased the total fresh weight by 36.0%and edible tissue fresh weight by 71.5% as compared to the control (Figure 2b,c).However, there was no significant difference in shoot fresh weight as a function of treatment (Figure S6).There is evidence in the literature demonstrating that Si-based NMs can promote plant growth and health.For example, Sun et al. reported that root exposure to 1000 mg/L mesoporous Si-based NMs significantly increased the fresh weight of shoots in wheat (Triticum aestivum) and lupin (Lupinus angustifolius) by 25.8% and 23.6%, respectively, as compared with untreated controls. 38Similarly, root exposure to Si-based NMs at 400 mg/L significantly increased the root weight by 52.9% in tuberose (Polianthes tuberosa L.); importantly, this increase was significantly greater than that of bulk-scale silicon materials. 39Here, exposure to 0.1%-5.0%Si-based NMs increased the total chlorophyll and the carotenoid content by 14.2%-75.7%and 18.7%-86.4%,respectively, relative to the control (Figure S7A-D).The beneficial effects of Si-based NMs on the photosynthetic machinery have been reported, including enhancing photosynthetic pigment content or improving electron transport MODERN AGRICULTURE systems.For example, foliar exposure to 100 mg/L Si nanoparticles significantly increased the total chlorophyll content by 12.1% and the net photosynthesis rate by 20.8% in Mentha piperita L. as compared to the control. 40Aligning with our findings, root exposure to 250 mg SiO 2 NMs/kg-soil significantly increased the total chlorophyll by 13.3% and the carotenoid content by 14.1% in barley (Hordeum vulgare). 413][44] However, the ratios of total chlorophyll/carotenoid and Chla/Chlb were unchanged across all the treatments (Figure S7E,F), indicating that Sibased NMs might not adversely impact photosynthesis.
Carbohydrates are an important source of energy, while proteins and AAs are essential nutrients that participate in structural composition and photosynthesis in plant cells.The root application of 0.1% Sibased NMs increased the content of reducing sugars and total sugars in the edible tissue by 23.7% and 24.8%, respectively, relative to controls (Figure 2d).Exposure to 0.1% Si-based NMs significantly increased the protein content of cherry radish by 2.3-fold; however, 0.1% Si-based NMs significantly reduced the AAs content by 21.6% as compared to controls (Figure 2e).Karimian et al. reported that exposure to 400 mg/L Sibased NMs via both root and foliar pathways not only promoted the growth of tuberose (Polianthes tuberosa L.) but also increased the soluble sugar content by 156.0% and 176.9%, and the total soluble protein content by 61.9% and 59.7%, respectively, when compared to the control. 39Si can promote the synthesis of proline, whereas carbohydrate biosynthesis could be a function of the upregulation of osNAC protein expression. 45

Nutrient content
Root exposure to 0.1% Si-based NMs significantly increased the Si content in shoots and edible tissue by 47.5% and 79.2%, respectively, as compared to controls (Figure 2f).The Cu content in Si-based NM-treated shoots and edible tissue was increased by 100.3% and 21.3% relative to the control (Figure 2g); similarly, 32.1% and 151.6% increases in the Mn contents in shoots and edible tissue were observed upon exposure to Si-based NMs (Figure 2h).Additionally, the root application of Si-based NMs also elevated the content of other essential nutrients in radish tissues (Figure S8).Previous studies have also shown that Si increased the accumulation of mineral nutrients (N, P, K, Ca and Mg) in the leaves of tomatoes and potatoes, and ultimately promoted plant growth. 42,46Kheyri et al. reported that soil exposure to Si-based NMs (392 kg Si ha −1 ) increased the Si and Zn content by 44.6% and 20.6%, respectively, in the straw of rice (Oryza sativa L.) as compared to the control. 47Foliar application of 20-60 mg/L SiO 2 NPs significantly increased the content of Si and micronutrients (e.g., Fe, Cu, and Mn) in cucumber (Cucumis sativus L.) roots. 48Si-based NMs may elevate the content of mineral nutrients by modulating the activities of enzymes involved in nutrient metabolism; additionally, Si-based NMs can improve soil structure and nutrient availability, both of which could further contribute to increasing tissue nutrient content. 49,50Since the addition of 0.1% Si-based NMs has shown the best performance in terms of radish yield and nutritional contents, in the following sections, we mainly focused on the impacts of 0.1% Si-based NMs on phytohormone networks and metabolic alteration in radish shoots and edible portions.

Phytohormone crosstalk
Plant hormones, also known as phytohormones or plant growth regulators, play critical roles in regulating developmental processes in plants, including growth, differentiation and responses to environmental stress. 51In the shoots, 0.1% Si-based NMs significantly increased the IAA content by 10.8% relative to the control, while the ABA, GA 3 and IPA contents were significantly reduced by 19.0%, 19.2% and 11.2%, respectively, as compared to controls (Figure 3a-I).In the edible tissue, 0.1% Sibased NMs significantly reduced the content of multiple phytohormones, including ABA, JA-ME, BR, GA 3 , GA 4 , ZR, IAA and IPA, by 9.2%-46.3%(Figure 3a-I).To determine the important factors that promote crop growth as a function of phytohormone crosstalk, we constructed a structural equation model to explore the relationships between phytohormones and radish fresh weight.The results show that ABA was negatively correlated with the fresh weight of shoots, while GA 3 , IAA and IPA were positively correlated with the fresh weight of the shoots.Specifically, the decrease in ABA content and the increase in GA 3 content were two important factors leading to the increased fresh shoot weight (Figure 3j).For the radish edible tissue, ABA was also negatively correlated with the fresh weight of the edible tissues, while GA, ZR and IPA were positively correlated with the fresh weight of the edible tissue.The decrease in the ABA content was the main factor leading to increased radish fresh weight (Figure 3k).ABA plays a vital role in regulating various physiological processes in plants, and is negatively correlated with plant growth. 52Recent studies have shown that Si-based NMs modulate the effect of ABA on plants.For example, soil exposure to 150 mg/kg SiO 2 NPs significantly reduced the ABA content by 44.1% in wheat (Triticum aestivum L.) as compared to the control. 53Similar findings were reported with 100 mg/L SiO 2 NP-treated cotton. 33In addition, GA 3 can positively regulate plant growth.For example, exogenous application of GA 3 promoted plant growth, increased stem elongation, increased seed germination rate and enhanced the yield of lettuce (Lactuca sativa L.) by 41.0% and rocket (Eruca sativa L.) by 25.5%. 54,55Zhang et al. observed that soil exposure of Glycyrrhiza uralensis Fisch to 0.6 g SiO 2 /kg-soil significantly increased the GA 3 content by 12.0% and reduced the ABA content by 35.8%, subsequently enhancing the dry weight by 24.0%. 56Similarly, 825 kg/ha Si-based NM treatment of potato (Solanum tuberosum L.) significantly increased the GA 3 content by 45.9% and decreased the ABA content by 45.8%. 57The significant decreases in the content of IAA, GA 3 and IPA may be due to the ripening process of the edible tissue, where the balance of phytohormones shifts to a decrease in auxins and an increase in ethylene, which promotes edible tissue ripening and overall senescence. 58

Metabolomics of radish shoots
In total, 5600 metabolites were identified across all samples, including fatty acyls (12.7%), organooxygen compounds (10.5%), carboxylic acids and derivatives (8.0%), prenol lipids (6.5%), benzene and substituted derivatives (4.4%), polyketides (4.4%), flavonoids (4.1%) and glycerophospholipids (2.7%) (Figure S9A).A PLS-DA model shows a clear separation between the nano-silica treatment and the control, indicating that Sibased NMs significantly shifted the shoot metabolic profile (p < 0.05, R 2 = 0.81) (Figure 4a).Similarly, Li et al. reported that exposure to 20 mg/L SiO 2 NMs resulted in significant separation of the metabolic profile of cucumber roots from controls. 48A heatmap of the metabolic profile also demonstrates that Si-based NMs and the control were present in different sub-clusters (Figure 4b).A co-occurrence analysis shows that Sibased NM treatment did not significantly impact the positive/negative correlation among identified metabolites, but increased the number of clusters, indicating that the Si-based NM application increased metabolome complexity (Figure S9C,E).The volcano plot shows that the Si-based NM treatment upregulated 241 metabolites and downregulated 256 metabolites (Figure 4c).The circular packing plot of differential metabolites indicates that Si-based NMs primarily affected metabolites of organic acids and derivatives, phenylpropanoids and polyketides and lipids and lipidlike molecules (Figure 4g).A total of 120 differential metabolites were enriched in 56 KEGG pathways, including ABC transporters, alanine, aspartate and glutamate metabolism and aminoacyl-tRNA biosynthesis metabolic pathways (Figure 4d).
In order to further understand the impact of Si-based NMs on nutritional component changes in the shoots, we focused on differential metabolites in AAs and carbohydrates.Root application of Si-based NMs significantly altered the shoot AA metabolites.Nine AAs, including proline, glutamic acid, valine, arginine, aspartic acid, isoleucine, aspartic acid, (3S,5S)-3,5-diaminohexanoate and homomethionine, were down-regulated with Si-based NM treatment and one AA, N-Acetyl-L-aspartic acid, was up-regulated (Figure 4e).AAs can serve as nutrients for plants and participate in critical physiological processes such as photosynthesis, respiration and nitrogen metabolism. 59,60In addition, Si-based NMs also significantly upregulated nine carbohydrates, including sucrose, α-lactose, lactose, deoxyribose 5-phosphate, gluconic acid, fructose 1-phosphate, xylobiose, coniferin and glucoraphanin, and decreased six carbohydrates, including allose, α-D-glucose, glucose, diketogulonic acid and β-D-ribosylnicotinate in shoots (Figure 4f).Carbohydrates serve as a source of energy for plant growth and development and as a material for storing energy, as well as participating in the construction of plant cell walls. 61Zhu et al. reported similar findings that root exposure to 0.3 mM Si significantly increased the contents of sucrose, glucose, fructose, and starch in cucumber leaves. 62

Metabolomics of radish edible tissue
Only 11.0% of the common differential metabolites were found in both the shoot and edible tissue of cherry radish (Figure S9B).5a).A heatmap of the metabolic profile also demonstrates that the metabolites of the edible tissue in the Si-based NM treatment and the control were clustered separately (Figure 5b).The positive/negative correlation among identified metabolites did not change significantly, although the metabolite network did become more complex (number of clusters: Control = 2, nano-Si = 3) upon exposure to Sibased NMs (Figure S9D,F).4][65] The volcano plots show differential metabolites in the Sibased NM treatment and the control, including the upregulation of 130 metabolites and down-regulation of 157 metabolites (Figure 5c).The differential metabolites in the edible tissue are mainly influenced by Sibased NMs at the super class levels of lipids and lipid-like molecules, organooxygen compounds, and phenylpropanoids and polyketides (Figure 5g).The KEGG pathway enrichment showed that 110 differential metabolites were mapped to 51 KEGG pathways, including citrate cycle (TCA cycle), arginine biosynthesis and alanine, aspartate and glutamate metabolism (Figure 5d).

The citric acid cycle
The TCA cycle in plants is an important metabolic pathway involved in energy production, biomolecule synthesis, and the regulation of growth and development.The metabolism of AAs and carbohydrates is closely linked to the TCA cycle. 67Root application of Sibased NMs resulted in the downregulation of 7 different metabolites and the upregulation of one metabolite associated with the shoot TCA cycle (Figure 6).Citric acid, isocitrate, and malic acid are important TCA cycle intermediates and have important physiological functions, including energy production, acid-base balance, and antioxidant activity.
In the radish edible tissue, Si-based NMs upregulated 8 metabolites (glucose, (3S,5S)-3,5-diaminohexanoate, asparagine, malic acid, histidinol, glutamate, N-acetyl-Lglutamate 5-semialdehyde, L-arginine) and downregulated three metabolites (citric acid, Isocitrate, 2oxoglutarate) associated with the TCA cycle (Figure 7).Among them, glucose is the most common monosaccharide in plants and is the primary product of photosynthesis.The increase in glucose can enhance edible tissue growth and yield, as well as improve sweetness and taste.In addition, (3S,5S)-3,5-diaminohexanoate, asparagine, histidinol, glutamate, Nacetyl-L-glutamate 5-semialdehyde, and L-arginine were all increased; these molecules are involved in the protein synthesis and nitrogen metabolism and are important sources of nitrogen for plants.The increased contents of these metabolites suggest that Si-based NMs may have a positive impact on plant protein synthesis and nitrogen metabolism.9][70] Zhao et al. reported that maize root exposure to 100 mg/kg-soil Sibased NMs altered nitrogen and carbon metabolism, including a significant increase in the nitrogen-containing compounds 4-aminobutyric acid and putrescine. 71

CONCLUSIONS
Nano-enabled agrochemicals are attracting increasing interest as a feasible solution to achieve agricultural sustainability and alleviate food insecurity.In the present study, root exposure to Si-based NMs significantly promoted the growth and yield of cherry radish, while enhancing the content of photosynthetic pigments (chlorophyll and carotenoids) and nutritional components (such as sugars and proteins) of the edible tissue.Additionally, the accumulation of Si and other essential micro-/macro-nutrients in radish tissues were increased.The crosstalk of phytohormones in Si-based NM-treated radish demonstrates that the increased fresh weight of radish shoots was positively correlated with the decreased ABA and the increased GA 3 content, while the increase in edible tissue mass was closely correlated with decreased ABA.The metabolomic profile also demonstrates that Si-based NMs positively altered edible tissue metabolites and increased the levels of various AAs and carbohydrates, indicating that Si-based NMs could significantly improve the nutritional composition and quality of the edible tissue.Overall, our findings demonstrate that nano-enabled agrochemicals could provide a novel strategy for not only enhancing crop yield but also for biofortifying crops to achieve enhanced nutritional content to address food insecurity and hidden hunger.

F I G U R E 1
Morphological characterisation and imaging of silicon-based nanomaterials.(a) TEM image; (b) SEM image; (c) AFM 3Dimage; (d) The hydrodynamic size distribution; (e) FTIR spectrum; and (f) XRD spectrum.F I G U R E 2 Effects of Si-based NMs on crop yield of cherry radish.(a) Photographs of plant treated with different proportions Si-based NMs for 35 days; (b) the total fresh weight; (c) the fresh weight of edible tissues; (d) the content of reducing sugar and total sugar in edible tissues; (e) the content of protein and amino acid in edible tissues; (f) Si content in shoots and edible tissues; (g) Cu content in shoots and edible tissues; (h) Mn content in shoots and edible tissues.A single asterisk (*) in each panel indicates significantly different between the control and the Si-based NM treatment at p < 0.05.

F I G U R E 3
Effects of Si-based NMs on cherry radish hormones.(a) ABA content; (b) JA-ME content; (c) BR content; (d) GA 3 content; (e) GA 4 content; (f) DHZR content; (g) ZR content; (h) IAA content; (i) IPA content in shoots and edible tissues; (j) Structural equation model of Si-based NMs-altered phytohormones elevating the shoot growth of radish; (k) the Structural equation model of Si-based NMs-altered phytohormones elevating the fresh weight of edible tissues of radish.The width of the arrows indicates the strength of the standardised path coefficient.The red lines indicate positive path coefficients, blue lines indicate negative path coefficients and grey lines indicate indirect paths.

F I G U R E 4
The metabolome profiles of cherry radish in the shoot affected by Si-based NMs.(a) PLS-DA score plot between control and Sibased NMs; (b) The cluster heatmap of metabolome profiles in the shoot; (c) The volcanic map of the differential metabolites; (d) the top 20 KEGG pathway enrichment of differentiated metabolites in the shoot of cherry radish; (e) The AAs in differentiated metabolites in the shoot of cherry radish; (f) The carbohydrates in differentiated metabolites in the shoot of cherry radish; (g) the circular packing of differential metabolites in the shoot cherry radish.The size of the circle represents the abundance of differential metabolites.
The PLS-DA score plot shows that the Si-based NM treatment and the control were clearly divided (p < 0.05, R 2 = 0.62), indicating that the addition of Si-based NMs significantly altered edible issue F I G U R E 5 The metabolome profiles of radish edible tissues as affected by Si-based NMs.(a) PLS-DA score plot between control and Sibased NMs; (b) The cluster heatmap of metabolome profiles in the edible tissue; (c) The volcanic map of the differential metabolites; (d) The top 20 KEGG pathway enrichment of differentiated metabolites in the edible tissue; (e) The AAs in differentiated metabolites in the edible tissue; (f) The carbohydrates in differentiated metabolites in the edible tissue; (g) The circular packing of differential metabolites in the edible tissue.The size of the circle represents the abundance of differential metabolites.

F I G U R E 6
Effects of Si-based NMs on the TCA cycle in radish shoots.The squares and the colour palette represent the relative content of different metabolites (Significance level: **p < 0.01; ***p < 0.001).
In summary, Si-based NMs can increase the AA and sugar F I G U R E 7 Effects of Si-based NMs on the TCA cycle in radish edible tissues.The squares and the colour palette represent the relative content of different metabolites (Significance level: **p < 0.01; ***p < 0.001).
plant edible tissues, thereby improving the fruit quality and nutritional value.