Effects of silver nanocolloids on plant complex type N-glycans in Oryza sativa roots

Silver nanomaterials have been mainly developed as antibacterial healthcare products worldwide, because of their antibacterial activity. However, there is little data regarding the potential risks and effects of large amounts of silver nanomaterials on plants. In contrast, N-glycans play important roles in various biological phenomena, and their structures and expressions are sensitive to ambient environmental changes. Therefore, to assesse the effects of silver nanomaterials, we focused on the correlation between N-glycans and the effects of silver nanomaterials in plants and analyzed N-glycan structures in Oryza sativa seedlings exposed to silver nanocolloids (SNCs). The phenotype analysis showed that the shoot was not affected by any SNC concentrations, whereas the high SNC exposed root was seriously damaged. Therefore, we performed comparative N-glycan analysis of roots. As a result, five of total N-glycans were significantly increased in SNC exposed roots, of which one was a free-N-glycan with one beta-N-acetylglucosamine residue at the reducing end. Our results suggest that the transition of plant complex type N-glycans, including free-N-glycans, was caused by abnormalities in O. sativa development, and free-N-glycan itself has an important role in plant development. This study originally adapted glycome transition analysis to environmental toxicology and proposed a new category called “Environmental glycobiology”.

Nanomaterial is a general term for small substances that are 1-100 nm in diameter. Nanomaterials have many unique electrical, chemical, and physical properties and are used in electronics, medicine, and healthcare fields. Silver nanomaterials have been developed and mainly used for their antibacterial activities in clothes, appliances, cosmetics, and plastics. However, there have been concerns that silver nanomaterials are likely released into the aquatic environment through factory and household wastewater on a large scale, and there are also concerns regarding the effects of silver nanomaterials on aquatic organisms and ecological systems. The toxicity of silver nanomaterials in the embryos of aquatic organisms such as medaka and zebrafish has been reported 1,2 . The toxicity affected the expression of morphogenesis-and cell proliferation-related genes and induced the increase of severe development abnormalities and mortality. The toxicity of silver nanomaterials was dependent on the particle size, shape, and capping materials 3,4 . Toxicity of silver nanomaterials has also been reported in plants, with the toxicity affecting germination, development, and photosynthetic efficiency because of the induction of oxidative stress, cytotoxicity, and genotoxicity 5,6 . For example, silver nanoparticle exposure significantly reduced root elongation, shoot and root fresh weights, and total chlorophyll and carotenoid contents 7 . Colman et al. showed that low silver nanoparticle concentrations caused a decrease in biomass 8 . Furthermore, the toxicity of silver nanoparticles affects the expression of several proteins that are mainly involved in primary metabolism and cell defense in wheat seedlings 9 . At a gene level, silver nanoparticles activate gene expression involved in plant cellular events, including cell proliferation, metabolism, and hormone signaling pathways 10 . The above-mentioned studies showed that silver nanomaterials have high toxicity in plants. Therefore, the risk assessment of silver nanomaterials in plants is important.
Asparagine (N)-linked glycans (N-glycans) are comprised several types of monosaccharides, forming complex compositions and linkage types. N-Glycan has a trimannosyl core structure [Man alpha1-6(Man alpha1-3) Man beta1-4GlcNAc beta1-4GlcNAc-Asn], which is a common feature in eukaryotes. Many N-glycan structures are linked to proteins or peptides and are closely involved in all life phenomena, such as development, signaling, SCiEnTifiC RepoRts | (2018) 8:1000 | DOI:10.1038/s41598-018-19474-z and cell-to-cell recognition. In plants, N-glycan structures are categorized into three main types: high-mannose, complex, and paucimannose types; except for hybrid types. A characteristic of plant-specific N-glycans is the addition of beta1,2-xylose and alpha1,3-fucose to the trimannosyl core structure. High-mannose type N-glycans are synthesized in endoplasmic reticulum (ER), and other type N-glycans are synthesized in the Golgi apparatus. Paucimannose type N-glycans are linked to vacuole proteins and complex type N-glycans are linked to secretory proteins. In addition, cell alterations are reflected in gene expressions through cell signaling, whereas N-glycan is synthesized as a result of the integral expression of glycosyltransferase genes, and N-glycan structure is sensitive to slight environmental changes 11 . Therefore, N-glycan structural analysis is valuable for the risk assessment of silver nanomaterial toxicity in plants. However, there is little data regarding the toxicity of silver nanomaterials in glycobiology. In this study, to assesse the effects of silver nanomaterials, we focused on the correlation of N-glycan structures and the effect of silver nanomaterials in Oryza sativa and analyzed the N-glycan structures in SNC exposed O. sativa seedlings.

Results and Discussion
Phenotype analysis of O. sativa seedling exposed to silver nanocolloids (SNCs). To observe the effect of SNCs on O. sativa seedlings, O. sativa seeds were grown with and without SNC exposure. Germination rate was 95% (control), 100% (SNCs 0.5 mg/L), 100% (SNCs 1.0 mg/L), 95% (SNCs 1.5 mg/L), 100% (SNCs 3.0 mg/L), 95% (SNCs 5.0 mg/L), 90% (SNCs 10 mg/L), 100% (SNCs 25 mg/L) after 48 h incubation. The result shows that there is no effect on germination rate at any concentration of SNC for 48 h exposure in O. sativa. Figure 1A shows the results of root and shoot elongation in O. sativa exposed to SNCs at 0 (control), 0.5, 1.0, 1.5, 3.0, 5.0, 10, and 25 mg/L for 96 h. Shoot and root length of control was 1.46 ± 0.08 cm and 0.98 ± 0.08 cm, respectively. The shoot elongation was not affected at any SNC concentration, whereas the root length increased from 0.5 to 10 mg/L based on Fig. 1A; however, in roots exposed to 25 mg/L SNCs, the lengths were two times lower than those of the control. Representative images of control and 25 mg/L SNC exposed O. sativa seedlings are shown in Fig. 1B. From the phenotype analysis, root length was seriously affected by 25 mg/L of SNC exposure.
The effect of SNCs was also observed in other plants. SNCs also have a significant effect on Arabidopsis thaliana and poplar development 12 . SNCs are mainly present in two forms: silver nanoparticles and free ions (Ag + ), which are derived from silver nanoparticles. Free Ag + is more poisonous than SNCs because of its oxidative potency. Wang et al. 12 demonstrated that free Ag + tends to accumulate in A. thaliana roots. Previous studies also showed that silver nanoparticles or free Ag + inhibited the growth of O. sativa roots 13 , and these materials affect cell metabolism-related proteins 14 . In addition, the effects of SNCs or free Ag + occur in ER-and vacuole-localized proteins of Eruca sativa 15 . Nair et al. reported that total sugar levels are decreased in SNC exposed O. sativa seedlings 7 . From these reports and our observation, SNCs and SNC derived molecules may affect N-glycans that are synthesized in the ER and Golgi apparatus of O. sativa roots. Therefore, we focused on the correlation of N-glycan structures and silver nanomaterials in O. sativa and analyzed the N-glycan structures in SNC exposed O. sativa roots. N-Glycan analysis of SNC exposed O. sativa roots. N-Glycans were prepared by hydrazinolysis, N-acetylation, and pyridylamination (PA). The resulting PA-N-glycans were separated according to their degree of saccharide polymerization by size-fractionation HPLC. Then, N-glycans of control and 25 mg/L of SNC exposed roots were compared (Fig. 2I, control and II, SNC treatments). Thirteen peaks were detected (indicated by bars). The areas of peaks B, E, and M increased in SNC exposed roots, and in particular, peak B was extremely increased in SNC exposed roots. Therefore, to identify each N-glycan structure in detail, reversed phase HPLC was performed, and branched N-glycan isomers were separated. Reversed phase HPLC analysis revealed three major peaks (peaks E1, E2, and E3) of peak E (Fig. 3). Comparing the HPLC elution times with known-position N-glycans, peaks E1, E2, and E3 coincided with those of GN M3FX, GN M3FX, and GN2M3X, respectively. Similarly, one major peak, M1, was detected in reversed phase HPLC analysis, and peak M1 coincided with the elution position of Gal2F2GN2M3FX. Table 1 shows the identified N-glycan structures and ratios derived from peaks E1, E2, E3, and M1, and each N-glycan is shown in terms of percentage proportion relative to GN2M3FX (peak G). The largest N-glycan Gal2F2GN2M3FX was increased three fold after SNC exposure. For other plant complex type N-glycans, GN M3FX, GN M3FX, and GN2M3X, SNC exposure caused up to five-or six-fold higher accumulations than the control. In general, complex type N-glycans are formed in from the cis-Golgi to medial Golgi apparatus, and higher complex modifications occur downstream in the synthetic pathway. Recent reports showed that protein-linked complex type N-glycans are related to proper targeting and functioning of linking proteins 16,17 and also showed that complex type N-glycans play an important role in resistance to external stresses such as salt stress 11,18,19 . These reports showed that the Golgi-localized N-glycan synthetic enzymes are related to plant growth and development, and their defect inhibited growth and caused abnormalities. Therefore, the transition of complex type N-glycans may be related to the disorder of Golgi-localized N-glycan synthetic enzymes and genes in SNC exposed roots. The relatively small N-glycans GN M3FX, GN M3FX, and GN2M3X were affected by SNCs more significantly. These results imply that the upstream part of the N-glycan complex pathway was affected by SNCs; therefore, the downstream part of the synthetic pathway was less affected for more complicated N-glycans such as Gal2F2GN2M3FX. These results may provide evidence that the intermediate complex type N-glycans play an important role in plants under excessive stress conditions. Free-N-glycan analysis of SNC exposed O. sativa roots. Reversed phase HPLC analysis revealed that peak B1 were eluted at around 3 min; therefore, to purify peak B1, twice reversed phase HPLC was performed (Fig. 4A). Peak B1 was further analyzed using MALDI-TOF mass spectrometry and sequential enzyme digestion. Mass spectrometry analysis of peak B1 showed that the m/z ratio was 1143.90 (Na + ), which corresponded to (Hex) 3 (HexNAc) 2 (Pent) 1 -PA. The m/z ratio and elution position on reversed phase HPLC revealed that the N-glycan structure of peak B1 was predicted to be a free-GNM3X structure [GlcNAc 1 Man 3 Xyl 1 GlcNAc 1 -PA] with one GlcNAc residue at the reducing end. To ascertain the N-glycan structure of peak B1, peak B1 was enzymatically digested with two exoglycosidases. Peak B1 [GlcNAc 1 Man 3 Xyl 1 GlcNAc 1 -PA] was converted to Man 3 Xyl 1 GlcNAc 1 -PA, releasing one GlcNAc residue by beta-N-acetylhexosaminidase (Fig. 4B-II), and Man 3 Xyl 1 GlcNAc 1 -PA was further converted to Man 1 Xyl 1 GlcNAc 1 -PA, releasing two mannose residues by alpha-mannosidase (Fig. 4B-III). Thus, peak B1 was assigned as free-GNM3X structure with one GlcNAc residue at the reducing end ( Fig. 4B-I).
Peak G is a major component of O. sativa roots (Fig. 2). As a result of reversed phase HPLC, two major peaks were detected; as a result of N-glycan two-dimensional mapping, the retention time of peak G2 corresponded to that of known position N-glycan, GN2M3FX 20 , and the other one was eluted at around 3 min (Fig. 5A, peak  G1). In a similar way to peak B1, peak G1 was analyzed using MALDI-TOF mass spectrometry and exoglycosidase digestion. Mass spectrometry analysis showed that the m/z ratio was 1456.67 (Na + ), which corresponded to (Hex) 7 (HexNAc) 1 -PA. The mass value and elution position of reversed phase HPLC revealed that peak G1 was predicted to be a free-M7 structure [Man 7 GlcNAc 1 -PA]. To confirm the N-glycan structure, peak G1 was enzymatically digested with alpha-mannosidase. As a result, peak G1 [Man 7 GlcNAc 1 -PA] was converted to Man 4 GlcNAc 1 -PA, Man 3 GlcNAc 1 -PA, Man 2 GlcNAc 1 -PA, and Man 1 GlcNAc 1 -PA, releasing from three to six mannose residues by alpha-mannosidase (Fig. 5B-II); therefore, peak G1 was assigned as free-M7 structure.
As shown in Table 2, the proportion of free-GNM3X (peak B1) increased six fold after SNC exposure. In contrast to free-GNM3X, free-M7 increased three fold after SNC exposure (Table 2). To date, high-mannose type free-N-glycans have been detected in plant during development 21,22 , and other complex type free-N-glycans have been detected in the culture broth of rice cultured cells 23 and Egeria densa 24 , and most of them have Lewis a structure [Gal beta1-3(Fuc alpha1-4)GlcNAc beta1-] that is characteristic of the N-glycan of   24 . In animal cells, the accumulation of sialyl free-N-glycans is caused by a decline in free-N-glycan metabolism by basal autophagy 25 .
Mkhikian et al. demonstrated that alternative N-glycan structures were generated under unusual growth conditions 26 . These observations suggest that the occurrence of specific N-glycan structures is involved in cell conditions under excessive stress conditions. The elongation of O. sativa shoots was unaffected by SNC exposure (Fig. 1A), and N-glycan structures were also unaffected. We suggested that a root defense mechanism serves to protect shoot development from SNC toxicity, and O. sativa roots may have a defense mechanism against soil environmental changes. Although the generation of high-mannose type free-N-glycans is generally caused by the deglycosylation from misfolded glycoproteins in ER-associated degradation system 27 , there is almost no information about the biological significance of complex type free-N-glycan. Our results showed that the increase of complex type free-N-glycans was caused by the effect of SNCs, suggesting that complex type free-N-glycan itself in O. sativa roots has an important role for resistance mechanism against excessive environmental changes. N-Glycan is one of the post-translational  modifications, and the structure and quantity are sensitive to ambient environment. In addition, N-glycans are generated by the result of the integrated multiple gene expression, and the biosynthesis is strictly controlled by many glycosyltransferases and glycosidases in each specific organelle. Though, it is difficult to identify which genes are specifically affected under stress condition, N-glycan analysis can derive the related genes and synthetic regions. Therefore, N-glycan analysis can predict to the initial response to environmental changes at comprehensively, and the transition is valuable for studying the relationship between environmental changes and biological response. As the case for monitoring the signaling pathway of plant under environmental changes, novel plant nanobionics approach has been reported 28 . Our results revealed a correlation between free-N-glycans and plant development under excessive stress conditions and also demonstrated that free-N-glycan transitions are valuable as stress markers for assessing trace environmental changes. This is the first report of the relationship between "environmental changes and glycome transition", and the present study originally adapted glycome transition to environmental toxicology and proposed a new category called "Environmental glycobiology".

Plant materials and sample preparation. Seeds of O. sativa (Koshihikari) were supplied by Itakura
Agricultural Cooperative Society, Japan. Before each experiment, seeds were washed five times with water and then washed twice with deionized water. The washed twelve seeds were then placed in a plate and immersed in 7 mm depth of water (control) or each concentration of SNCs suspension (0.5, 1.0, 1.5, 3.0, 5.0, 10.0, and 25.0 mg/L of SNCs). The plates were placed in a controlled environmental chamber, and kept at 37 °C for 96 h in the dark. The resulting seedlings were washed and then dried.
Preparation of pyridylaminated N-glycans from O. sativa seedlings. N-Glycan preparation was performed according to the method of Natsuka et al. 30 . Dried shoots and roots were ground in a mortar at room temperature, and a ten milligram sample was used. N-Glycans were prepared by hydrazinolysis, N-acetylation. The reducing ends of the liberated N-glycans were then tagged with a fluorophore, 2-aminopyridyne (pyridylaminated N-glycans; PA-N-glycans), as described in previous paper 20 . These preparations were performed following details in Hase et al. 31 with minor modifications. Separation of PA-N-glycans. Size-fractionation HPLC was performed in a Cosmosil 5NH 2 -MS column (4.6 ID × 150 mm) at a flow rate of 0.8 mL/min at 40 °C. PA-N-glycans were detected with a fluorescence spectrophotometer at 310 nm excitation and 380 nm emission. Reversed phase HPLC was performed on a Cosmosil 5C 18 -P column (4.6 ID × 150 mm) at a flow rate of 1.5 mL/min at 40 °C. The detection of PA-N-glycans performed by use of fluorescence spectrophotometer at 315 nm excitation and 400 nm emission. Each HPLC conditions were described in previous paper 20 . Mass spectrometry analysis of PA-glycans. MALDI-TOF mass spectrometry analysis was then performed using an AXIMA resonance instrument (Shimadzu) in reflector mode. Sample preparation was described in previous paper 20 .
Glycosidase digestion of PA-N-glycans. A two picomoles of PA-N-glycans was prepared in 1 microL of accessory reaction buffer (5 mM CaCl 2 , 10 mM ammonium acetate buffer, pH 4.5) and 2 microL of D. D. W, 1 microL of beta-N-acetylhexosaminidase (0.05 units/microL) was added, and the mixture was incubated at 37 °C for 4 h, and then 1 microL of jack bean alpha-mannosidase (19 units/mg) and 5 microL of 10 mM ammonium acetate buffer (pH 4.5) were added, and the mixture was incubated at 37 °C for 1 h. alpha-Mannosidase digestion was described in previous paper 20 . To stop all reactions, the mixtures were boiled for 5 min at 95 °C, and the mixture was analyzed by size-fractionation HPLC.