Physiological and Transcriptomic Responses of Antioxidant System and Nitrogen Metabolism in Tomato Roots Treated With Nitrogen Starvation and Re-Supply

Nitrogen (N) is one of the essential macronutrients that plays important roles in plant growth and development. To better understand the response of antioxidant system and N metabolism under N starvation and re-supply condition, physiological and transcriptomic analysis were performed in tomato roots. The malondialdehyde (MDA) and reactive oxygen species (ROS) contents increased signicantly in tomato seedlings after N starvation for 24 h. The activities of superoxide dismutase (SOD), catalase (CAT), ascorbate peroxidase (APX) and monodehydroascorbate reductase (MDHAR), the ratio of ASA/DHA and GSH/GSSG, the NO 3- contents, nitrate reductase (NR) activity were decreased after N starvation treatment and increased after N re-supply for 24 h. Compared with the control, 1766 genes were up-regulated and 2244 genes were down-regulated after N starvation in tomato. These differentially expressed genes (DEGs) are mainly enriched in functional items such as cellular process, metabolic process and catalytic activity. The KEGG pathways revealed that the DEGs were mainly involved in phenpropane biosynthesis, amino sugar and nucleotide sugar metabolism, and N metabolism. The expression patterns of tomato SlSOD, SlCAT, SlAPX, SlMDHAR, thioredoxin (SlTrxh), peroxiredoxin (SlPrx) and glutaredoxin (SlGrx) genes, and nitrate transporter SlNRT2.4, SlNR, glutamine synthetase (SlGS2), nitrite reductase (SlNiR) decreased after N starvation and increased after N re-supply, which were validated by qRT-PCR. Our results provide a basis for understanding the response of tomato to N deciency and re-supply and a theoretical reference for cultivation regulation.


Introduction
Nitrate as a nutrient, is absorbed by roots through low and high a nity nitrate transporters (NRT1 and NRT2), which is reduced to nitrite by nitrate reductase (NR), and to ammonium by nitrite reductase (NiR).
Ammonium is then incorporated into amino acids by glutamine synthetase (GS) and glutamate synthase (GOGAT) ( The response of nitrogen metabolism in plants to the change of nitrogen level is multifaceted, and the current research on nitrogen metabolism in plants is mostly concentrated in corn, cereals, wheat and other crops. For example, the activity of nitrate reductase in corn increased with the increase of nitrogen level, which indicated that proper nitrogen application could signi cantly improve the nitrogen use e ciency (Zhan and Lynch 2015). Studies have shown that the lack of nitrogen in the soil of cereal crops will lead to the change of root length and reduce the protein content of grains (York et al. 2015). In addition, extensive transcriptomic studies have investigated nitrogen metabolisms in plants at various levels, such as time points after N treatments (Krouk et al. 2010), N sources (Patterson et al. 2010), and rates (Wang et al. 2007), in cell (Picault et al. 2002), and tissue types (Wang et al. 2003). In pear roots resupplied by nitrate starvation, the KEGG pathways revealed that 15 unigenes were related to nitrogen metabolism and signi cantly differentially expressed in response to nitrate starvation and a nitrate resupply treatment ).
Under the stress of nitrogen de ciency, the dynamic balance between the production and removal of reactive oxygen species (ROS) in plants is broken, which leads to excessive production of ROS in plant cells. If it is not removed in time, it will cause peroxidation of the plasma membrane, thus disrupting the Under NaCl stress, the activity of SOD, POD and CAT in leaves increased by some extent. Under NaHCO 3 stress, the activity of SOD and POD signi cantly increased, while that of CAT decreased compared to that of control. The ascorbic acid-glutathione (AsA-GSH) cycle in mulberry seedling leaves was enhancement in both NaCl and NaHCO 3 stress (Huihui et al. 2020). Research on rice found that nitrogen stress induces changes in plant ROS content and antioxidant enzyme activity, helping plants resist the damage caused by oxidative stress (Kumagai et al. 2009). However, there is no report on the effect of nitrogen de ciency and nitrogen re-supply on the antioxidant system of tomato roots.
RNA-Seq, one of next-generation high-throughput sequencing technologies, has been widely used recently, due to low background noise, high sensitivity and reproducibility, great dynamic range of expression and base pair resolution for transcription pro ling (Marioni et al. 2008;Quan et al. 2016 Tomato is cultivated on a commercial scale across the world. According to National Bureau of Statistics of China, annual use of nitrogenous fertilizer exceeds 24 million tons (Xin et al. 2017), a fair share of this nitrogenous fertilizer is utilized for tomato production. However, little research is available on tomato seedlings in response to N de ciency and N re-supply. In this study, an RNA-sequencing approach was applied to gain a comprehensive picture of transcriptional mechanisms underlying the response to N deprivation and N re-supply of tomato roots. In addition, to better characterize the overall effects of these treatments on plant physiological status, the antioxidant enzyme system and N metabolism were also assessed in the experiment. Our results provide a basis for understanding the molecular mechanisms of tomato's response to N recovery after N de ciency stress and laid a theoretical foundation for cultivation.

Plant Material and Growth Conditions
Tomato seeds were soaked in warm water at 55°C for 1 ~ 2 h, then placed in a petri dish covered with two layers of wet lter paper and put in a thermostat at 28°C for germination. After 2 days, seeds with similar bud potential are sown in perlite with nutrient solution. When tomato seedling grows to a true leaf, they are moved to 4 L square basins for Hoagland's nutrient solution with twelve seedlings each basin (Siddiqi et al. 2002). At three-leaves stage, seedlings were exposed to nitrogen-free nutrients solution (N starvation) for 2 days, after which the tomato seedlings were transferred into Hoagland's nutrient solution

Determination of nitrate contents and nitrate reductase activities
The nitrate ion concentration was determined according to Cataldo's methods (CATALDO and Analysis 1975). The activity of nitrate reductase was measured by Living body method (Hageman and Reed 1971). Tomato roots after different nitrogen treatments were rinsed with distilled water and dried with absorbent paper, cut into small pieces with scissors, weighed 0.5 g with a small balance, and put into a triangular ask. First add 1 mL of 30 % trichloroacetic acid solution to the control triangular ask, then add 4 mL of 0.1 mol L − 1 phosphate buffer (pH:7.5) and 5 mL 0.2 mol L − 1 KNO 3 to each triangular ask after mixing, put it in a desiccator immediately and evacuate it for 30 minutes. During this period, let in air several times, and then evacuate to make the blade sink completely into the bottom of the bottle, and then put it at 25°C for 30 minutes, add 1 mL of 30 % trichloroacetic acid solution to the triangular ask to stop the reaction. After shaking each triangular ask and let stand for two minutes, then, 2 mL of the supernatant was treated with 4 mL 1 % sulfonamide and 0.2 % α-naphthylamine before incubating for 15 min in a 35°C water bath with agitation. The nitrate content and nitrate reductase activity were then measured at 410 nm and 540 nm, respectively, using a spectrophotometer.

RNA extraction and detection
RNA was extracted using TransZolUp (TRANS, Company) with approximately 0.1 g sample according to the TRNAS kit instructions. After the RNA extraction is completed, RNA is subjected to 1.0 % gel electrophoresis to ensure the purity and integrity.
Transcriptome sequencing sample preparation and library establishment After extracting total RNA from the sample, enrich the mRNA with magnetic beads with Oligo (dT), adding fragmentation buffer to the obtained mRNA to make fragments therefore into short fragments, then the mRNA after the fragment is taken as a template. The rst strand of cDNA was synthesized with random hexamers. Add buffer, dNTPs, RNase H and DNA polymerase I to synthesize cDNA second chain. The cDNA was puri ed by QiaQuick PCR test kit and eluted with EB buffer solution, and subjected to end repair. Add base A, add sequencing linker, and then recover the target size fragment by agarose gel electrophoresis, and carry out PCR ampli cation to complete the whole library preparation work. The Transcriptome data assembly First, we lter the Raw data, remove low quality and connectors to obtain Clean data, and use TopHat to compare and remove rRNA-containing reads respectively. Then,the reads of the ltered rRNA are compared to a reference genome. Finally, transcript reconstruction is carried out by cu inks to obtain all transcripts.

Differential expression gene (DEG) analysis and gene function annotation
The original data is standardized by using the DESeq's own standardized method. In the process of difference analysis, the negative binomial distribution method is used to estimate the distribution of Read count, after evaluating and calculating P value, multiple hypothesis tests are performed on P value to reduce false positives. Differentiated genes were screened according to edgeR's general ltering criteria (log 2 |Fold Change| > 1 & FDR < 0.05), and the screened differentially expressed genes were enriched with gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) to determine the major metabolic pathways.
Quantitative real-time PCR (qRT-PCR) analysis Total RNA was extracted and cDNA was synthesized by Prime Script TM RT regent kit with gDNA Eraser (Perfect Real Time), and uorescence quantitative assay was carried out by abm® Eva Green qPCR Master Mix-no dye kit. The primers were designed with Software Premier 5.0 based on cDNA Fragments (Table S1). The Pre-denaturation of reaction program was 95°C, 5 min. The number of reaction cycles is 40, the cycle process is 95°C, 10 s, 60°C, 30 s; 72°C, 20 s. The speci city was evaluated by dissolution curve and size estimation of ampli cation products. The expression level of DEGs was calculated with 2 −△△ct . Actin gene is used as the standardization of the determination gene. Value of each stage is taken as the average of three technical repetitions for each biological repetition.

Statistical analysis
Three replicates of each sample were used for statistical analysis. Data were analyzed with Student's ttest indicated as follows: *, P < 0.1; **, P < 0.05. The Graphpad Prism 7.0 software (GraphPad Software, La Jolla California USA, www.graphpad.com) was used for making gures.

Effects of N de ciency and re-supply treatments on lipid peroxidation and ROS accumulation in tomato roots
Membrane lipid peroxidation often occurs when plant organs are aged or damaged under adversity. MDA is the nal decomposition product of membrane lipid peroxidation, and its content can re ect the degree of damage to plants under adversity. In this experiment, the MDA content of samples was measured at four treatment time points: CK, CK-N, T1 and T2. After two days of N de ciency treatment, the MDA content increased by 89.90 % compared with that of normal tomato seedlings (Fig. 1a). Compared with the control group, MDA content still increased by 38.57 % after re-supplying N for 6 h, and basically returned to normal level after re-supplying N for 24 h.
The ROS content in tomato roots was labeled with uorescent probe H 2 DCF-DA (Fig. 1b). The uorescence degree of ROS in roots treated with N de ciency was increased dramatically, which indicated that the oxidative damage caused by N de ciency was aggravated. The uorescence intensity decreased after re-supplying N, which indicated that the oxidative damage was relieved to some extent after resupplying N.
Effects of N de ciency and re-supply treatments on antioxidant enzyme activities and ratio of AsA/DHA and GSH/GSSG in tomato roots As shown in Fig. 2, the activities of SOD, CAT, APX, and MDHAR in tomato roots were all decreased after N de ciency stress compared with the control. SOD activity decreased by 38.02 %, 22.01 % and 19.71 % after 2 days of N de ciency, N re-supply for 6 h and 24 h (Fig. 2a). CAT activity decreased signi cantly by 48.50 % after N de ciency stress, and continued to decrease to 59.72 % after 6 h of N re-supply (Fig. 2b).
APX is an antioxidant enzyme that mainly catalyzes AsA to capture H 2 O 2 in chloroplasts and cytoplasm of plant cells. The APX activity in the roots of tomato seedlings decreased by 32.78 % after N de ciency, and continued to decrease to 43.45 % after 6 h of N re-supply and then increased by 25.02 % after 24 h of N re-supply (Fig. 2c). MDHAR is an important enzyme to regenerate AsA in AsA-GSH cycle. MDHAR activities decreased by 60.13 % after N de ciency and then increased after N re-supply (Fig. 2d).
The ratio of ASA/DHA and GSH/GSSG in tomato seedling roots decreased after N de ciency and then increased after N re-supply (Fig. 2e, f). Compared with the control group, the ratio of GSH/GSSG and ASA/DHA decreased by 29.15 % and 36.22 %, respectively, after 2 days of N de ciency. The ratio of ASA/DHA increased by 16.15 %, 21.59 % and the ratio of GSH/GSSG increased by 8.93 %, and 17.18 %, respectively, after 6 h and 24 h of N re-supply.
Effects of N de ciency and N re-supply treatments on NO 3 − contents, nitrate reductase activity in tomato roots After N de ciency for 2 days, the NO 3 − content decreased by 39.61 %, compared with the control (Fig. 3a).
The NO 3 − content in tomato roots increased by 17.56 % after re-supplying N for 6 h. The N re-supply for 24 h treatments basically restored NO 3 − accumulation to normal levels.
Nitrate reductase (NR) is a key enzyme in the plant nitrate-assimilation process. Therefore, effects of N de ciency and the N re-supplying treatment on the NR activity in tomato roots were investigated. As shown in Fig. 3b, N de ciency treatments signi cantly reduced the NR activity by 54.78 % in tomato roots. However, the inhibitory effect under conditions of N de ciency was relieved by the N re-supply treatment.

Quality analysis of sequencing results
To investigate the molecular response of N starvation and re-supply to tomato seedling, RNA-seq technology was used. The sample of tomato roots of the control and the treatment group were sequenced by Illumina HiSeqTM. In order to ensure the quality of the data, the original data should be quality-controlled before information analysis. Removed reads with adapter, reads with N ratio greater than 10 %, and low-quality reads, CK, CK-N, PCA results showed that there was a signi cant difference in gene expression between the CK group and the CK-N group. The gene expression of the T1 group showed a recovery phenomenon, and the gene expression of the CK group samples was similar to that of the T2 group samples (Fig. S1). The results showed that the gene expression level basically recovered to that of the CK group of the hydroponic tomato roots resumed N culture for 24 h after 2 days of N de ciency.

Analysis of differences between samples
The R-based software package edgeR was used to process the RNA-seq data for pairwise samples or between groups with signi cant differences. FDR and log 2 FC were used to screen for differential genes, and the screening conditions were FDR < 0.05 and | log 2 FC | > 1. As shown in Fig. 4a To identify common and unique DEGs in response to N starvation and re-supply treatments, venn graphs were plotted. We analyzed DEGs that were transcriptionally regulated at different treatment, and a total of 430 common DEGs were identi ed among the four libraries (Fig. 4b).

GO function annotation analysis of different expressed genes
The number of genes with signi cant differences in GO function enrichment between each 2 groups is shown in Table S2. These differentially expressed genes are annotated into three directions: biological processes, cell components and molecular functions, involving a total of 45 functional entries. The GO classi cation charts of Ck-N-vs-T1, Ck-N-vs-T2 and T1-vs-T2 are shown in Fig S2, S3, S4. The GO classi cation map of CK and CK-N are shown in Fig. 5. In biological processes, differentially expressed genes are most abundant in cellular processes, metabolic processes and single-organism processes. In cell components, differentially expressed genes are mainly concentrated in cell, cell part and membrance. In terms of molecular function, differentially expressed genes are mainly enriched in catalytic activity and binding function items.

Differential gene KEGG Pathway analysis
In order to further understand the biological function of genes and determine the most important biochemical metabolic pathways and signal transduction pathways involved in DEGs, we performed pathway signi cant enrichment analysis. 7135 differentially expressed genes were annotated in the KEGG database. In CK and CK-N, CK-N and T1, CK-N and T2, T1 and T2, they were annotated to 124, 126, 125, and 113 pathways, respectively. The top 20 pathways with abundant genes are shown in Fig. 6 and Fig. S5, S6, S7. Among them, the differentially expressed genes are mainly involved in phenylpropanoid biosynthesis, amino sugar and nucleotide sugar metabolism, cysteine and methionine metabolism, starch and sucrose metabolism, N metabolism and other pathways.

Differential expression of N metabolism genes in tomato
The metabolic process of N in plants includes complex mechanisms such as absorption and transport, assimilation and reuse. As shown in Fig. 7, differentially expressed genes were found in this N metabolism pathway. Compared with the control group, 7 DEGs expressions were down-regulated and 1 DEGs expression was up-regulated in hydroponic tomato roots after N de ciency. The genes that are down-regulated or up-regulated due to N de ciency stress gradually returned to normal expression levels after 1 day of N restoration.
The expression of antioxidant enzyme genes in qRT-PCR and RNA-seq analysis qRT-PCR analysis was performed on the transcription level of SlSOD, SlCAT, SlAPX and SlMDHAR (Fig. 8). After 2 days of N de ciency in tomato seedlings, the relative expression of SlSOD, SlCAT, SlAPX and SlMDHAR genes in tomato roots decreased by 97.37 %, 94.86 %, 73.83 % and 83.45 %, respectively, compared with the control group. The relative expression of SlSOD and SlCAT genes increased by 6.55 and 3.76 times respectively after N re-supply for 6 h. The relative expression of SlAPX and SlMDHAR genes decreased by 74.63 % and 51.24 % respectively after N re-supply for 6 h. The relative expression of SlSOD, SlCAT, and SlMDHAR genes increased by 7.54, 8.53 and 1.91 times, and the relative expression of SlAPX genes decreased by 43.89 %, after N re-supply for 24 h.
The mRNA expression of SlTrxh, SlPrx and SlGrx was also analyzed by qRT-PCR in the roots of tomato seedling after the N de ciency and re-supply. SlTrxh, SlPrx and SlGrx expression decreased by 56.73 %, 93.27 % and 87.31 %, respectively, compared with the control group after 2 days of N de ciency. SlTrxh expression decreased by 53.74 % compared with the control group after 6 h of N re-supply. SlPrx and SlGrx expression increased by 4.77 and 1.03 times of the control group after 6 h of N re-supply. The mRNA expression of SlPrx and SlGrx increased signi cantly after 24 h of N re-supply.
We analyzed the transcriptome, and the results are basically consistent with the real-time PCR data (Table  S3), indicating that the transcriptome data is reliable.
The expression of N metabolism genes in qRT-PCR and RNA-seq analysis Compared with the control group, the transcription levels of SlNRT2.4, SlNR, SlNiR and SlGS in hydroponic tomato seedlings were down-regulated by 65.43%, 13.22%, 73.11% and 54.72%, respectively, after N starvation (Fig. 9). Compared with the control group, the transcription level was up-regulated by 1.17, 2.24, 2.15 and 0.59 times after N re-supplying for 24 h. Compared with the control group, the transcription level of SlGOGAT gene after N starvation treatment was increased by 2.19 times, and after 24 h of N resupplying, the transcription level was only 1.07 times that of the control group, which basically recovered to the level of the control group.  2). The APX in tomato gradually lost its activity after N de ciency treatment, which may be due to excessive ROS attacking biological functional molecules in defense system under N de ciency stress (Hasanuzzaman et al. 2019). In the cycle of AsA-GSH in plants, AsA is oxidized to MDHA, and MDHA regenerates AsA under the action of MDHAR. MDHAR activity decreased after N de ciency, N re-supply for 6 h and 24 h, compared with the control group (Fig. 2d). Previous results showed that AsA synthesis in tomato roots was damaged after N de ciency (Hasanuzzaman et al. 2019). In our study, the ratio of AsA/DHA decreased after N starvation. Transcripts involved in antioxidant responses were strongly upregulated when T. suecica was cultured under N starvation (Lauritano et al. 2019). However, in our study, the expression of antioxidant enzymes of SlSOD, SlCAT, SlAPX, SlMDHAR were all decreased after N starvation in tomato roots, indicating the decreased ROS detoxi cation capability.

Discussion
Most of the studies have implicated N transport genes, N assimilation genes, and GS/GOGAT cycle genes involved in NUE. The amount of nitrate ion in tomato root decreased after N de ciency treatment, which is consistent with the previous studies that the N starvation condition led to a marked reduction in total N content in roots (Curci et al. 2017  In our study, comparing the two groups of tomato seedling roots treated with N de ciency and N resupply, the DEGs were involved in metabolic process, nutrient reservoir activity and catalytic activity responding to N de ciency and N re-supply treatment. KEGG pathway analysis can help us to further understand the biological functions of genes and how these genes interact (Kanehisa et al. 2004). In our study, the differentially expressed genes are mainly involved in the pathways of phenylpropanoid biosynthesis, amino sugar and nucleotide sugar metabolism, cysteine and methionine metabolism, starch and sucrose metabolism, N metabolism and other pathways. In previous report, cotton under N starvation and re-supply treatment, enriched to a pathway similar to the results of our experiment (Iqbal et al. 2020). The KEGG pathways revealed that 15 unigenes, including one NRT gene, two NR genes, one NiR gene, two GDH genes, six GS genes and three GOGAT genes, were related to nitrogen metabolism and signi cantly differentially expressed in response to nitrate starvation and a nitrate re-supply treatment . From these pathways explained the possible roles of N de ciency and re-supply on the N metabolism of tomato.
In sum, antioxidant enzyme system and N metabolism were analyzed based on environmental conditions that restored N supply after two days of N de ciency in tomato seedlings by physiological and RNA-seq analysis. This study provided a valuable resource for better understanding of tomato seedling roots in responses to N starvation and re-supply and for understanding the genes and pathways involved.