Enhancing effect of Na₂SeO₃ on the growth and physiological parameters of Vitis vinifera × labrusca 'Shuijing' under nitrogen deficiency and underlying transcriptomic mechanisms

The experiment was conducted in the Agricultural Practice Park of Kunming College (E 102°42′, N 25°02′, Kunming, China) from 2022 to 2023. Vitis vinifera × labrusca "Shuijing' grape cuttings were obtained from Dongfeng Farm Management Bureau, Mile City, Yunnan, China. Pest- and disease-free cuttings with healthy growth were selected and transplanted into black nutrient pots (internal diameter = 22 cm and depth = 25 cm). Each pot was filled with 7 L of substrate (vermiculite: perlite: peat = 1: 1: 3) and contained one cutting. The seedlings were cultivated for 10 days in a glass greenhouse at 20–30°C and relative humidity of 80%–90%.

Seven treatment groups were set, repeating five plants per treatment: Control, 0.1Se + 15N (0.1 mmol · L⁻¹ Na₂SeO₃ + 15 mmol · L⁻¹ NO₃^–), 0.2Se + 15N (0.2 mmol · L⁻¹ Na₂SeO₃ + 15 mmol · L⁻¹ NO₃^–), 0.4Se + 15N (0.4 mmol · L⁻¹ Na₂SeO₃ + 15 mmol · L⁻¹ NO₃^–), 0.1Se (0.1 mmol · L⁻¹ Na₂SeO₃), 0.2Se (0.2 mmol · L⁻¹ Na₂SeO₃), and 0.4Se (0.4 mmol · L⁻¹ Na₂SeO₃). To the substrate in each group, 500 mL of treatment solution was poured every 3 days. The specific formulation of the treatment solution is given in Table 1. The experiment was conducted for 70 days. After the treatment, the morphological characteristics of the plants were measured and photographed. Further, healthy leaves in the middle of each plant were picked, washed with DI water thrice, dried, frozen in liquid nitrogen, and stored at –80°C for the determination of growth and physiological indicators and transcriptome sequencing. The experiment included four biological replicates.

Plant height and stem diameter were determined using a tape measure and vernier calliper, respectively; net increase of plant height = plant height at planting – plant height at the end of treatment; net increase of stem diameter = stem diameter at planting – stem diameter at the end of treatment; node spacing was assessed using a ruler; root volume was determined using the drainage method; leaf thickness was determined by picking 10 leaves and considering leaf thickness as 1/10 of the measured value; biomass was determined by the whole plant weighing; root/shoot ratio was calculated as the fresh weight of the belowground part/fresh weight of the aboveground part ×100%.

The leaves dried at 80°C and ground into fine powder. Soluble sugar content was determined according to the method by Yemm and Willis (1954). Then, 0.1 g of this powder and 6 mL of 80% (v/v) ethanol were taken in a centrifuge tube, incubated at 80°C for 30 min, and centrifuged. The supernatants were made up to 25 mL with distilled water. The precipitate was used to determine starch content according to the method by Clegg (1956). The precipitate was mixed with 2 mL deionised water and placed in a boiling water bath for 15 min. To this, 2 mL of 9.2 mol · L⁻¹ perchloric acid was added and stirred for 15 min. The mixture was centrifuged, the supernatant was collected, and the volume was made up to 50 mL with water. This solution was used to determine starch content using anthrone reagent.

The fresh leaves were cleaned, dried at 85°C, and ground into fine powder. Further, 0.1 g of the powder was digested with K₂Cr₂O₇–H₂SO₄. Total nitrogen content was determined by the Kjeldahl method using a Kjeldahl nitrogen analyser (NKD6280, Shanghai Wanghai Environmental Science and Technology Co., Ltd., CHN). To 0.2 g of the powder, 10 mL of 50 mmol · L⁻¹ sodium phosphate (pH = 7.8) containing 2 mmol · L⁻¹ EDTA and 80 mmol · L⁻¹ L-ascorbic acid was added. The mixture was centrifuged. The supernatant was collected and used to determine soluble protein content using Bradford G-250 reagent (Bradford, 1976).

The contents of flavonoids were measured using a method reported previously (Liu and Zhu, 2007). In brief, 0.5 g of fresh leaves were cut into filaments and immersed in a mixture of 10 mL of 70% ethanol and 0.1 g of CaCO₃ for 24 h. Further, 0.5 mL of the extracted solution was taken in a test tube, to which 1.5 mL of 70% ethanol and 0.3 mL of 5% NaNO₂ were added and incubated for 6 min. Then, 0.3 mL of 10% Al(NO₃)₃ solution was added and incubated for 6 min. Finally, 2 mL of 4% NaOH was added and incubated for 10 min. The absorbance was determined at 510 nm using a spectrophotometer. The flavonoid content was calculated based on the standard curve.

After 70 days of treatment as indicated in Section 2.1, through the comprehensive analysis of morphological and physiological indicators, grape leaves from the Control, 0.2Se + 15N, and 2Se groups were finally selected as the test materials, with three replicates for each treatment. Total RNA was extracted from the grape leaves using the TRIzol method. The concentration and purity of the total RNA were examined using Nanodrop 2000 (Thermofisher Scientific, US). The integrity of RNA and DNA or protein contamination was examined using agarose gel electrophoresis. The RIN was determined using Agilent 5300 (Agilent Technologies, US). After passing the test, the library construction and non-parametric transcriptome sequencing (Illumina Novaseq 6000, Illumina Corporation, US) were commissioned to Shanghai Major Biomedical Technology Co., Ltd., CHN. The raw reads obtained from sequencing were subjected to quality control by removing reads with adapters, N ratio >10% (indicating that base information could not be determined), and all A bases and low-quality reads (bases with a quality value of Q_phred ≤ 20 accounting for more than 50% of the entire reads) to obtain high-quality clean reads for subsequent analysis (Liu et al., 2019).

Using Hisat2 software to perform sequence alignment between clean reads and the grape reference genome (http://plants.ensembl.org/Vitis_vinifera/Info/Index), obtaining positional information on the reference genome and unique sequence feature information of the sequencing sample. The reads per fragments per kilobase of transcript per million mapped reads (FPKM values) were calculated. Then, the gene and transcriptional expression levels were detected using the RSEM tool, and the p-adjust and fold change (FC) values were calculated using the Nbinom Test function. p-adjust ≤0.05 and |Log₂ FC|>1 were used as the screening criterion to obtain differentially expressed genes (DEGs) under different treatments. Finally, differentially expressed genes based on the GO Database (http://geneontology.org/) and KEGG Database (http://www.genome.jp/kegg/) perform gene function annotation. Perform GO enrichment analysis on DEGs using Goatools software (https://github.com/tanghaibao/GOatools) and use Fisher's exact test; using KOBAS software (https://bioinformaticshome.com/tools/rna-seq/descriptions/KOBAS.html) for KEGG enrichment analysis of DEGs, the calculation principle was the same as GO functional analysis. To control the false positive rate, the Bonferroni method was used for multiple tests, with a p-adjust value ≤0.05 as the threshold. GO and KEGG pathways that meet this condition were defined as pathways significantly enriched in differentially expressed genes.

Experimental data were organised using Microsoft Excel 2019 software. SPSS 27 software (International Business Machines Corporation, US) was used for oneway ANOVA and Duncan’s test to compare growth and physiological indicators. The results were expressed as the mean ± standard deviation of four replicates. p < 0.05 was considered significant. Graphs were plotted using GraphPad Prism 9 software (GraphPad software, LLC., US) and Origin 2021 software (OriginLab Corporation, US).

Significant differences were observed in the growth status of grape plants in various groups after 70 days of treatment (Figure 1A). The 0.2Se + 15N group exhibited the most vigorous growth, and the Control exhibited the weakest growth of above- and belowground parts. Compared with the Control, the 0.1Se + 15N, 0.2Se + 15N, 0.4Se + 15N, 0.1Se, 0.2Se, and 0.4Se groups exhibited an increase in the net growth of grape plants by 114.32%, 145.02%, 73.00%, 78.88%, 93.15%, and 26.19%, respectively; net growth of stem diameter by 21.54%, 38.31%, –23.87%, –34.37%, 6.93%, and –18.96%, respectively; internodal spacing by 27.55%, 30.41%, 11.27%, 36.88%, 44.97%, and 6.46%, respectively; root volume by 82.46%, 144.52%, 72.44%, 46.62%, 82.59%, and 56.63%, respectively; leaf thickness by –5.90%, 5.62%, 1.76%, 9.84%, 13.39%, and 5.39%, respectively; biomass by 84.83%, 112.51%, 52.03%, 55.76%, 76.89%, and 55.94%, respectively; and root/shoot ratio by 25.14%, 36.67%, 26.90%, 30.58%, 48.62%, and 22.10%, respectively (Figure 1B). Additionally, they exhibited an increase in flavonoid content by 12.36%, 21.98%, 15.16%, –5.27%, –9.56%, and –14.78%, respectively (Figure 1C); starch content by 14.58%, 22.12%, 8.60%, 41.18%, 61.86%, and 51.52%, respectively; soluble sugar content by 24.74%, 59.93%, 37.83%, 27.40%, 11.55%, and 7.89%, respectively; total nitrogen content by 79.98%, 73.17%, 69.97%, 33.61%, 49.06%, and 42.04%, respectively; and soluble protein content by 38.86%, 89.22%, 10.78%, 5.10%, 51.85%, and 6.09%, respectively (Figure 1D), compared with the Control. Overall, Na₂SeO₃ treatment considerably increased root volume, biomass accumulation, and root/shoot ratio and promoted plant height and nitrogen uptake under various nitrogen conditions.

Figure 1.

Effects of Na₂SeO₃ treatment on the growth and physiological characteristics of grape under various nitrogen conditions. (A) The growth status of grape plants; (B) the net growth of plant height, the net growth of stem thickness, root volume and internode length of grape; (C) leaf thickness, biomass, root shoot ratio and flavonoids content of grape; (D) starch content, soluble sugar content, nitrogen content and soluble protein content of grape. The results were shown as mean ± standard deviation (n = 4), and one-way ANOVA was used to compare the significant differences among the treatments. Different lowercase letters on the bar graph indicate significant differences at the p < 0.05 level. Control group is Control; 0.1 mmol · L⁻¹ + 15 mmol · L⁻¹ N is 0.1Se + 15 N; 0.2 mmol · L⁻¹ + 15 mmol · L⁻¹ N is 0.2Se + 15 N; 0.4 mmol · L⁻¹ + 15 mmol · L⁻¹ N is 0.4Se + 15 N; 0.1 mmol · L⁻¹ Na₂SeO₃ is 0.1Se; 0.2 mmol · L⁻¹ Na₂SeO₃ is 0.2Se; 0.4 mmol · L⁻¹ Na₂SeO₃ is 0.4Se.

The Control, 0.2Se + 15N, and 0.2Se groups were selected for transcriptome analysis based on the results described in Section 3.1. Transcriptome sequencing was performed on 9 samples using the Illumina platform, and 54.99 Gb clean bases and 0.38 Gb raw reads were obtained. The clean reads, error rate, GC content for each sample, and percentage of Q20 and Q30 bases were 0.37 Gb, <0.0258%, 45.69%–46.82%, >97.73%, and >93.32%, respectively (Table 2). This indicated that the accuracy of the sequencing data was high, which met the requirements of quality control and facilitated data analysis at the later stage. Using Trinity to assemble high-quality sequences from scratch, 147924 transcripts were obtained, with an N50 length of 2268 bp, and the C value was 89.7%, with an S value of 64.2% and a D value of 25.5%. A total of 89421 unigenes were obtained, with an N50 length of 1954 bp, and the C value was 73.5%, with an S value of 70.2% and a D value of 3.3% (Table 3). The clean reads of each sample were compared with the reference sequences obtained from the Trinity assembly to obtain the mapping results of each sample, which was also the basis for the quantification of the genes and transcripts of each sample. The mapped ratio of each sample was >87.96% (Table 4). Overall, the quality of the samples was good and suitable for subsequent bioinformatic analysis.

The correlation coefficients of the transcriptome sequencing samples between different replicates of the Control, 0.2Se, and 0.2Se + 15N groups were >0.94, 0.99, and 0.97, respectively (Figure 2A), indicating good reproducibility within the groups. These groups contained 18077, 17922 and 17555 unigenes, respectively, with 16861 shared unigenes (Figure 2B), that the expression estimated for transcripts assembled from all samples simultaneously. Based on FDR < 0.05 and |log₂FC| ≥ 1, the DEGs in various groups were identified. Visualisation of results using a volcano plot indicated that 0.2Se + 15N versus Control, 0.2Se versus Control, and 0.2Se versus 0.2Se + 15N had 1196 (395 up- and 801 downregulated), 2238 (520 up- and 1718 downregulated), and 1980 (678 up- and 1302 downregulated) DEGs, respectively (Figures 2D–2F), and 0.2Se + 15N versus Control, 0.2Se versus Control, and 0.2Se versus 0.2Se + 15N had 294, 753, and 294 DEGs, respectively. These comparison groups had 179 co-expressed unigenes (Figure 2C). Overall, Na₂SeO₃ treatment considerably promoted the production of DEGs and grape growth under various nitrogen conditions.

Figure 2.

Transcriptome analysis of grape leaves after Na₂SeO₃ treatment under various nitrogen conditions. (A) Correlation between samples; (B) unigene distribution and group overlap; (C) Venn diagram displaying the number of DEGs in each comparison group and group overlap; (D–F) The DEGs in 0.2Se + 15N versus Control, 0.2Se versus Control, and 0.2Se + 15N vs 0.2Se are plotted in a volcano plot. Control group is Control; 0.1 mmol · L⁻¹ + 15 mmol · L⁻¹ N is 0.1Se + 15 N; 0.2 mmol · L⁻¹ + 15 mmol · L⁻¹ N is 0.2Se + 15 N; 0.4 mmol · L⁻¹ + 15 mmol · L⁻¹ N is 0.4Se + 15 N; 0.1 mmol · L⁻¹ Na₂SeO₃ is 0.1Se; 0.2 mmol · L⁻¹ Na₂SeO₃ is 0.2Se; 0.4 mmol · L⁻¹ Na₂SeO₃ is 0.4Se.

To clarify the biological functions associated with the DEGs after Na₂SeO₃ treatment under various nitrogen conditions, GO pathway enrichment analysis was conducted on the top 40 DEGs with FDR < 0.05 in 0.2Se + 15N versus Control, 0.2Se versus Control, and 0.2Se + 15N versus 0.2Se (Figure 3). DEGs present in all three comparisons were involved in catalytic activity (GO:0003824), oxidoreductase activity (GO:0016491), transferase activity, transferring acyl groups (GO:0016746), transferase activity, transferring acyl groups other than amino-acyl groups (GO:0016747), tetrapyrrole binding (GO:0046906), heme binding (GO:0020037), extracellular region (GO:0005576), cell wall (GO:0005618), external encapsulating structure (GO:0030312), oxidation-reduction process (GO:0055114), response to abiotic stimulus (GO:0009628), and cell wall organisation or biogenesis (GO:0071554). DEGs only present in 0.2Se + 15N versus Control were involved in DNA binding (GO:0003677), DNA-binding transcription factor activity (GO:0003700), lyase activity (GO:0016829), plasmodesma (GO:0009506), polymeric cytoskeletal fibre (GO:0099513), supramolecular polymer (GO:0099081), response to stimulus (GO:0050896), transmembrane transport (GO:0055085), lipid biosynthetic process (GO:0008610), small molecule biosynthetic process (GO:0044283), microtubule-based process (GO:0007017), response to oxygen-containing compound (GO:1901700), response to external stimulus (GO:0009605), ion transmembrane transport (GO:0034220), mitotic cell cycle process (GO:1903047), organic acid biosynthetic process (GO:0016053), carboxylic acid biosynthetic process (GO:0046394), regulation of cell cycle (GO:0051726), inorganic ion transmembrane transport (GO:0098660). DEGs only present in 0.2Se versus Control were involved in monooxygenase activity (GO:0004497), Golgi apparatus (GO:0005794), protein phosphorylation (GO:0006468), external encapsulating structure organisation (GO:0045229), and cell wall organisation (GO:0071555). DEGs only present in 0.2Se versus 0.2Se + 15N were involved in apoplast (GO:0048046), organic substance catabolic process (GO:1901575), on transport (GO:0006811), regulation of biological quality (GO:0065008), homeostatic process (GO:0042592), cation transport (GO:0006812), metal ion transport (GO:0030001), response to inorganic substance (GO:0010035), and inorganic ion homeostasis (GO:0098771). Overall, the high-frequency occurrence of GO-enriched DEGs after Na₂SeO₃ treatment may be a key factor in promoting grape growth under various nitrogen conditions.

Figure 3.

GO pathway enrichment analysis of DEGs in grape after Na₂SeO₃ treatment under various nitrogen conditions. The numbers listed on the horizontal axis represent the top 40 GO-enriched entries in different comparison groups (P_adjust < 0.05). Control group is Control; 0.2 mmol · L⁻¹+ 15 mmol · L⁻¹ N is 0.2 Se + 15 N; 0.2 mmol · L⁻¹ Na₂SeO₃ is 0.2 Se. 1. Catalytic activity; 2. oxidoreductase activity; 3. DNA binding; 4. transferase activity, transferring acyl groups; 5. transferase activity, transferring acyl groups other than amino-acyl groups; 6. DNA-binding transcription factor activity; 7. tetrapyrrole binding; 8. heme binding; 9. lyase activity; 10. cytoskeletal protein binding; 11. extracellular region; 12. cell wall; 13. external encapsulating structure; 14. supramolecular complex; 15. plasmodesma; 16. anchoring junction; 17. cell-cell junction; 18. cell junction; 19. polymeric cytoskeletal fibre; 20. supramolecular polymer; 21. response to stimulus; 22. oxidation-reduction process; 23. response to stress; 24. transmembrane transport; 25. response to chemical; 26. cell cycle process; 27. response to abiotic stimulus; 28. lipid biosynthetic process; 29. small molecule biosynthetic process; 30. microtubule-based process; 31. response to oxygen-containing compound; 32. response to external stimulus; 33. ion transmembrane transport; 34. response to oxidative stress; 35. mitotic cell cycle process; 36. organic acid biosynthetic process; 37. carboxylic acid biosynthetic process; 38. regulation of cell cycle; 39. inorganic ion transmembrane transport; 40. cell wall organisation or biogenesis; 41. transferase activity, transferring glycosyl groups; 42. transferase activity, transferring hexosyl groups; 43. hydrolase activity, acting on glycosyl bonds; 44. UDP-glycosyltransferase activity; 45. oxidoreductase activity, acting on paired donors, with incorporation or reduction of molecular oxygen; 46. hydrolase activity, hydrolysing O-glycosyl compounds; 47. monooxygenase activity; 48. iron ion binding; 49. intrinsic component of membrane; 50. integral component of membrane; 51. membrane; 52. plasma membrane; 53. Golgi apparatus; 54. carbohydrate metabolic process; 55. catabolic process; 56. protein phosphorylation; 57. polysaccharide metabolic process; 58. cellular carbohydrate metabolic process; 59. cellular polysaccharide metabolic process; 60. external encapsulating structure organisation; 61. cell wall organisation; 62. apoplast; 63. organic substance catabolic process; 64. ion transport; 65. regulation of biological quality; 66. homeostatic process; 67. cation transport; 68. metal ion transport; 69. response to inorganic substance; 70. inorganic ion homeostasis.

To further reveal the growth- and physiology-related functions of DEGs affected by Na₂SeO₃ treatment under various nitrogen conditions, based on the results of GO pathway enrichment analysis, the entries with P_adjust < 0.05 and a higher number of enriched genes were further selected for KEGG pathway enrichment analysis (Figure 4). DEGs present in all three comparisons were involved in flavonoid biosynthesis (map00941), phenylpropanoid biosynthesis (map00940), starch and sucrose metabolism (map00500), plant hormone signal transduction (map04075), MAPK signalling pathway— plant (map04016), and plant-pathogen interaction (map04626). DEGs only present in 0.2Se + 15N versus Control were involved in stilbenoid, diarylheptanoid, and gingerol biosynthesis (map00945) and glycine, serine, and threonine metabolism (map00260). DEGs only present in 0.2Se versus Control were involved in tyrosine metabolism (map00350), phagosome (map04145), and endocytosis (map04144). DEGs only present in 0.2Se versus 0.2Se + 15N were involved in protein processing in the endoplasmic reticulum (map04141), galactose metabolism (map00052), and oxidative phosphorylation (map00190). These results indicate that flavonoid biosynthesis (map00941), phenylpropanoid biosynthesis (map00940), starch and sucrose metabolism (map00500), and plant hormone signal transduction (map04075), which are relatively abundant and present in all three comparison groups, play important roles in promoting grape plant growth.

Figure 4.

KEGG pathway enrichment analysis of DEGs in grape after Na₂SeO₃ treatment under various nitrogen conditions. The numbers listed on the horizontal axis represent the top 15 KEGG enrichment items in different treatment comparisons (P_adjust < 0.05). Control group is Control; 0.2 mmol · L⁻¹ + 15 mmol · L⁻¹ N is 0.2 Se + 15 N; 0.2 mmol · L⁻¹ Na₂SeO₃ is 0.2 Se. 1. Flavonoid biosynthesis; 2. phenylpropanoid biosynthesis; 3. starch and sucrose metabolism; 4. stilbenoid, diarylheptanoid, and gingerol biosynthesis; 5. glycine, serine, and threonine metabolism; 6. glycerolipid metabolism; 7. cysteine and methionine metabolism; 8. glutathione metabolism; 9. pentose and glucuronate interconversions; 10. monoterpenoid biosynthesis; 11. glycolysis/gluconeogenesis; 12. plant hormone signal transduction; 13. MAPK signalling pathway–plant; 14. circadian rhythm–plant; 15. plant-pathogen interaction; 16. glycerophospholipid metabolism; 17. amino sugar and nucleotide sugar metabolism; 18. tyrosine metabolism; 19. phagosome; 20. endocytosis; 21. protein processing in endoplasmic reticulum; 22. galactose metabolism; 23. oxidative phosphorylation.

The growth, morphology, and distribution of roots are affected by the nutrients in soil (Lynch, 2011). In this study, grape root growth was significantly better in the Se and Se + N groups than in the Control groups, and the combination of Se and N was more effective than Se alone. Se and N were closely related to the growth and development of grape. The application of 0.2 mmol · L⁻¹ Na₂SeO₃+ 15 mmol · L⁻¹ NO₃^– significantly increased the root volume, plant height, and biomass of grape compared with other treatments. Grape plant growth was inhibited when Na₂SeO₃ concentration increased to 0.4 mmol · L⁻¹. These were consistent with previous studies (El-Hendawy et al., 2017; Wang et al., 2019).

The Se application can significantly increase the contents of soluble solids in plants (Fan et al., 2024). In this study, Se + N application significantly increased soluble sugar content in grape leaves, with 0.1 mmol · L⁻¹ Na₂SeO₃ treatment exhibiting the most significant increase. Se + N or Se application increased nitrogen and soluble protein contents in grape leaves; however, the increase was more after Se + N application than after Se application. This may be because Se fertiliser could regulate the electron transfer in plant photosynthesis and respiration, thus improving the photosynthetic capacity of crops and increasing N accumulation. Additionally, when an appropriate concentration of N fertiliser was supplied, flavonoid content in grape leaves increased. This was consistent with previous studies (Cao et al., 2012; Aly et al., 2015) and may be attributed to the fact that N increases the activities of some enzymes involved in flavonoid synthesis.

Flavonoids are low-molecular-weight secondary metabolites synthesised by plants and play an important role in their growth and development (Shi and Xie, 2014). They can increase the uptake of elements such as N and P by roots (Pei et al., 2020) and regulate seed germination, root growth, and photosynthetic pigment synthesis (Tan et al., 2019). Additionally, when plants are subjected to adverse conditions, large amounts of flavonoids are accumulated in plants to remove reactive oxygen species, activate defense-related signalling pathways, and improve the resistance of plants to adverse conditions (Landi et al., 2015). In this study, we performed transcriptome sequencing of grape leaves using a high-throughput sequencing platform (Illumina), and several DEGs related to flavonoid biosynthesis were enriched in the three comparison groups. In the flavonoid metabolism, stilbene synthase (CHS/ST) is involved in the first step of the catalytic reaction, and the expression level of its gene directly affects the quantity of flavonoid metabolites generated (Yeou et al., 2021). In 0.2Se + 15N versus Control, most DEGs related to CHS/ST were significantly upregulated, resulting in the accumulation of flavonoids in grape leaves. In 0.2Se versus Control, DEGs related to CHS/ST were not significantly upregulated or downregulated; however, the absolute number of downregulated unigenes was more. This led to lower flavonoid content in grape leaves in the 0.2Se group than in the Control.

Phenylpropane metabolism is a key pathway of secondary metabolism in plants (Douglas, 1996). PAL and 4CL are the two key enzymes in plants (Qiao et al., 2013) because they can promote plant cell differentiation and lignification and catalyse the formation of metabolites that are chemical barriers to pathogenic organisms and environmental stresses in plants (Heldt and Piechulla, 2011). In this study, in 0.2Se + 15N versus Control, seven DEGs related to PAL and one DEG related to 4CL were significantly upregulated, resulting in enhanced activities of PAL and 4CL and promoting the synthesis and accumulation of phenylpropanes. This is consistent with previous studies (Song et al., 2023). In 0.2Se versus Control, most unigenes were significantly downregulated, suggesting that the supply status of Se and N is closely related to the production and accumulation of phenylpropanes in grape.

Plant hormones not only participate in regulating growth and development, signal transduction, and adversity resistance but also regulate nitrogen metabolism in plants. ABA and IAA are closely related to signal transduction of nitrogen and directly affect the ability of plants to adapt to low-nitrogen stress (Chen et al., 2023). Pretreatment with IAA and GA can significantly increase nitrogen metabolism in Morella rubra Lour. under salt stress (Chakrabarti and Mukherji, 2003). In this study, in 0.2Se + 15N versus Control, only one auxin-related DEG encoding SAUR was significantly upregulated, whereas three auxin-related DEGs encoding SAUR and AUX1 were significantly downregulated. In 0.2Se + 15N versus Control, eight auxin-related DEGs encoding SAUR, AUX1, CH3, and IAA were significantly upregulated, whereas seven DEGs were significantly downregulated. Two ABA-related DEGs encoding PYL were significantly downregulated; one GA-related DEG encoding GID1 was significantly upregulated (VIT_07s0104g00930); one JAZ-related DEG encoding JAZ was significantly downregulated (VIT_01s0146g00480). Hence, Na₂SeO₃ treatment had negligible effects on hormone metabolism of grape under appropriate nitrogen supply, which ensured healthy and rapid plant growth. However, under nitrogen deficiency, Na₂SeO₃ treatment could greatly activate the expression of genes related to plant hormone metabolism, accelerate nitrogen metabolism by accumulating more auxin and GA and reducing ABA accumulation, and improve the adaptation of plants to nitrogen deficiency stress. This was consistent with previous studies (Chakrabarti and Mukherji, 2003; Chen et al., 2023). Hence, Se could regulate the expressions of genes related to hormone metabolism under nitrogen deficiency to promote the adaptation of grape plants to low-nitrogen environment.

Carbon assimilated by photosynthesis in leaves is used for the formation of starch in chloroplasts or transported to the cytoplasm for the synthesis of sucrose. Most photosynthetic products required for plant growth and development are supplied and transported in the form of sucrose. Sucrose phosphate synthase (SPS) is an important control point of the sucrose synthesis pathway and a key enzyme required for the entry of sucrose into various metabolic pathways. Its activity can reflect the status of sucrose biosynthesis pathways (Harbron et al., 1981; Liu et al., 2005). Some studies reported that SPS is negatively correlated with starch accumulation and positively correlated with sucrose synthesis (Huber, 1983; Wang et al., 2000). In this study, in 0.2Se versus Control, three DEGs encoding SPS were significantly downregulated, suggesting that sucrose accumulation was proportional to the activity of SPS. Accumulation and conversion of starch and sucrose were closely related to growth and development of plants and activities of enzymes in plants (Vizzolo et al., 1996; Suthumchai et al., 2007). In 0.2Se versus Control, three DEGs encoding β-glucosidas, one DEG encoding alpha-trehalose-phosphate synthase, one DEG encoding endo-1,4-β-D-glucanohydrolase, two DEGs encoding fructokinase, three DEGs encoding SPS, one DEG encoding β-fructofuranosidase, and one DEG encoding hexokinase were significantly downregulated. This led to decreased activity of the relevant enzymes, resulting in decreased soluble sugar content. Additionally, most DEGs in 15N + 0.2Se versus Control were upregulated. Combined with Figure 1, it could be inferred that this phenomenon was due to the lower starch content and higher soluble sugar content in the Se + N groups than in the Se groups.