Abstract Nitrate (NO[3]^−) and ammonium (NH[4]^+) are the main inorganic nitrogen (N) sources absorbed by oilseed rape, a plant that exhibits genotypic differences in N efficiency. In our previous study, the biomass, N accumulation, and root architecture of two oilseed rape cultivars, Xiangyou 15 (high N efficiency, denoted “15”) and 814 (low N efficiency, denoted “814”), were inhibited under NH[4]^+ nutrition, though both cultivars grew normally under NO[3]^− nutrition. To gain insight into the underlying molecular mechanisms, transcriptomic changes were investigated in the roots of 15 and 814 plants subjected to nitrogen-free (control, CK), NO[3]^− (NT), and NH[4]^+ (AT) treatments at the seedling stage. A total of 14,355 differentially expressed genes (DEGs) were identified. Among the enriched Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway categories of these DEGs, carbohydrate metabolism, lipid metabolism, protein metabolism, and cell wall biogenesis were inhibited by AT treatment. Interestingly, DEGs such as N transporters, genes involved in N assimilation and CESA genes related to cellulose synthase were also mostly downregulated in the AT treatment group. This downregulation of genes related to crucial metabolic pathways resulted in inhibition of oilseed rape growth after AT treatment. Keywords: oilseed rape, nitrate, ammonium, physiology, transcriptome, gene expression 1. Introduction Nitrogen (N) is an indispensable macronutrient for plant growth and development that not only serves as a constituent of many crucial macromolecules, including proteins, enzymes, and nucleic acids, but also acts as a signal to regulate many biological processes from metabolism to resource allocation, growth and development. Nitrate (NO[3]^−) and ammonium (NH[4]^+) are the main N forms absorbed and utilized by plants, accounting for 70% of anion and cation absorption [[38]1]. The mechanisms used by plants to absorb and utilize NO[3]^− and NH[4]^+ differ, and NO[3]^− and NH[4]^+ can affect plant growth and physiological processes, such as dry matter accumulation, root morphology, photosynthesis, and N assimilation. Accordingly, researchers have carried out studies on the effects of different N forms on the growth and physiology of plants, but the conclusions of these studies vary due to differences among plant species and genotypes. For example, some plants such as rice, tea, and Vaccinium spp. prefer NH[4]^+ and grow better under NH[4]^+ nutrition [[39]2,[40]3,[41]4], whereas others such as wheat, watermelon, and Arabidopsis thaliana prefer NO[3]^− [[42]5,[43]6,[44]7]. Oilseed rape is a crucial source of plant oil worldwide, with a large demand for N fertilizer, and judicious application of N fertilizer can significantly improve the yield of oilseed rape [[45]8]. Furthermore, a certain amount of N can increase oilseed rape biomass and yield, though excessive N decreases the oil content of oilseed rape [[46]9]. Moreover, different oilseed rape genotypes exhibit distinct differences in their absorption and utilization of N fertilizers [[47]10]. In most dryland soils, NO[3]^− is the primary form of inorganic N; however, in natural environments, the N form may be highly variable, with NH[4]^+ often predominant in acidic and/or water-logged soils [[48]11,[49]12]. Dryland and acidic soils are both suitable for planting oilseed rape. To date, NO[3]^− accumulation and utilization have been a major emphasis of studies on N use in oilseed rape. In contrast, less is known about the growth performance and gene expression of these plants under NH[4]^+ nutrition and whether oilseed rape cultivars with different N efficiencies have the same or different reactions to NO[3]^− and NH[4]^+ nutrition. Hence, we evaluated the effects of NO[3]^− and NH[4]^+ on the growth and physiology of two oilseed rape cultivars with different N efficiencies. The results demonstrated that NH[4]^+ greatly inhibited the growth and root architecture of both cultivars but that they grew and developed normally under NO[3]^− nutrition [[50]13]. Although NO[3]^− and NH[4]^+ clearly have distinct effects on the growth of the two cultivars, studies of oilseed rape under NO[3]^− and NH[4]^+ nutrition at the molecular level are rare. Oilseed rape is closely related to Arabidopsis. Differences in gene expression in Arabidopsis under NO[3]^− and NH[4]^+ nutrition have been explored using microarray analysis, and the results suggest that the underlying bases of nitrate- and ammonium-specific patterns of gene expression differ [[51]14]. In general, comparative high-throughput RNA sequencing (RNA-Seq) analyses of different cultivars and treatments can provide a deep understanding of the actual metabolic pathways involved, and a rice transcriptomic encyclopedia has recently been used to explore a large number of ammonium-responsive genes [[52]2]. Moreover, Yang et al. [[53]15] explored the transcriptomic profiles of tea (Camellia sinensis) treated with different N forms using RNA-Seq. The development of these metabolic databases has greatly facilitated studies focusing on NO[3]^− and NH[4]^+ nutrition. Thus, RNA-Seq appears to be a promising technique for determining the genomic behavior of oilseed rape under NO[3]^− and NH[4]^+ nutrition. Here, the effects of NO[3]^− and NH[4]^+ application on the physiological responses of two oilseed rape cultivars, namely, Xiangyou 15 (with high N efficiency) and 814 (with low N efficiency), are explored. RNA-Seq of roots was applied to reveal molecular mechanistic differences between NO[3]^− and NH[4]^+ nutrition. Some candidate genes associated with N uptake, transport and assimilation, enriched Gene Ontology (GO) terms and enriched metabolic pathways were identified to provide a valuable resource for further understanding the major transcriptomic regulatory pathways responding to NO[3]^− and NH[4]^+ in the roots of two rape genotypes. 2. Materials and Methods 2.1. Plant Materials and Treatments The two oilseed rape cultivars used in this study were previously characterized as having high (Xiangyou 15, denoted “15”) or low (814) N efficiency [[54]16]. Seeds of 15 and 814 were provided by the National Center of Oilseed Crops Improvement, Hunan Branch, Hunan, China. Mature seeds were sown in trays with growth medium containing a mixture of peat and vermiculite (1:1, v/v). After producing ~3–4 leaves (35 d after sowing), we transplanted uniform seedlings into different pots for physiological and transcriptomic experiments. To prevent NH[4]^+ from being converted into NO[3]^− by microorganisms, this study used pot culture (diameter: 30 cm; height: 30 cm) with quartz sand containing no nutrients. The nutrient solution used was based on the improved Hoagland nutrient solution [[55]17]. The physiological experiment was conducted in 2016. All pots with different N treatments contained the same levels of N (15 mmol/L) and dicyandiamide (DCD) (10 μmol/L) to prevent NH[4]^+ transformation. There were six treatments, i.e., 2 cultivars × 3 N treatments. Ten seedlings were used for each treatment, with only one seedling per pot, for a total of 60 seedlings. [56]Table 1 shows the macronutrient composition of the hydroponic solution of the three N treatments. The micronutrient composition was as follows: B, 1.0 mg/L; Mn, 0.5 mg/L; Zn, 0.05 mg/L; Cu, 0.02 mg/L; and Mo, 0.01 mg/L. The NT + AT (75% NO[3]^− + 25% NH[4]^+) treatment ([57]Table 1) received a conventional oilseed rape fertilizer and was set as a control in the physiological experiment. The hydroponic solution was refreshed every five days. To prevent reaction between Ca[2]^+ and CO[3]^2−/SO[4]^2−, CaCl[2] was applied separately. The total N input of each treatment was 2.8 g/pot, and all N was added to pots before 60 d after transplanting. Table 1. Macronutrient composition of the hydroponic solution. Treatment Nutrient Concentration (mmol/L) KNO[3] (NH[4])[2]SO[4] K[2]CO[3] CaCl[2] MgSO[4] KH[2]PO[4] Fe-EDTA NT 15 0 0 2 2 1 0.2 NT + AT 11.25 1.875 1.875 2 2 1 0.2 AT 0 7.5 7.5 2 2 1 0.2 [58]Open in a new tab NT, 100% NO[3]^− treatment; NT + AT, 75% NO[3]^− + 25% NH[4]^+ treatment; AT, 100% NH[4]^+ treatment. The transcriptome experiment was conducted in 2017. Most of the methods were similar to those used in the physiological experiment. The NT (100% NO[3]^−) and AT (100% NH[4]^+) treatments were the same as those in the physiological experiment. We used a nitrogen-free treatment (control, CK) instead of the NT + AT treatment to further distinguish the transcriptomic characteristics of oilseed rape roots in the NT and AT treatments. The 11.25 mmol/L KNO[3], 1.875 mmol/L (NH[4])[2]SO[4] and 1.875 mmol/L K[2]CO[3] ([59]Table 1, NT + AT) treatments were replaced by 7.5 mmol/L K[2]CO[3] (CK), and the other nutrients were the same. Twenty seedlings were included in each treatment, with only one seedling per pot, for a total of 120 seedlings. At 48 h after exposure to different N treatments, root tissues from five seedlings in each treatment group (three biological replicates) were collected, mixed, washed with double-distilled water, immediately frozen in liquid nitrogen, and stored at −80 °C for use in RNA extraction. 2.2. Measurement of Growth-Related Parameters For biomass and N concentration determination, five plants at the bud period (70 d after transplanting) from each treatment were individually harvested and separated into shoots and roots. All the plant samples were washed with double-distilled water, heated at 105 °C for 30 min and dried at 80 °C until their weight remained constant; the dry weight (DW) was recorded. The N concentrations of the shoots and roots were determined using a SmartChem Discrete Auto Analyzer 200 (AMS Systea, Roma, Italy) with H[2]SO[4]-H[2]O[2] digestion. Five roots from each treatment were first separately cleaned with double-distilled water to remove the growth medium. Each sample was placed in a glass rectangular sink (300 mm × 200 mm × 20 mm) in batches with a layer of water approximately 6–8 mm deep to allow the roots to extend fully and minimize overlap for capturing images; each sample was scanned at 600 dpi with an automatic scanning apparatus (Perfection V700, Epson, Nagano, Japan). The total root length (sum of path length of the main and lateral roots), root surface area (sum of areas of the main and lateral roots), average root diameter (average diameter of the main and lateral roots), and root volume (sum of volumes of the main and lateral roots) were acquired by analysis of root images using WinRHIZO Pro 2009 software (Regent Instruments, Québec, QC, Canada). 2.3. RNA Quantification and Qualification Total RNA was extracted manually from five mixed root tissues using a modified CTAB method [[60]18] (three biological replicates) for a total of 15 root tissues per treatment. RNA integrity was assessed at 2.2 ≥ OD260/280 ≥ 1.8 using an RNA Nano 6000 Assay Kit with a Bioanalyzer 2100 system (Agilent Technologies, Palo Alto, CA, USA). RNA purity was checked using a NanoDrop spectrophotometer (Thermo Fisher, Waltham, MA, USA). RNA concentrations were measured using a Qubit RNA Assay Kit and a Qubit 3.0 Fluorometer (Life Technologies, Carlsbad, CA, USA). Equal amounts of total RNA from three biological replicates per condition were pooled to construct the cDNA library; the remaining RNA was carefully preserved for later use (quantitative real-time polymerase chain reaction, qRT-PCR). Sequencing was performed using the Illumina HiSeq X Ten platform and 150-bp paired-end module. The sequence data were deposited in the NCBI database and are accessible via BioProject ID: PRJNA518759. 2.4. Bioinformatic Analysis The original image data obtained by high-throughput sequencing were converted to sequence data by CASAVA base calling. Each sample generated more than 8 gigabytes of data. For further analysis, the reads were filtered from raw sequencing data by removing adaptors and trimming low-quality (30 Phred quality score (Q30) was >92% and the guanine-cytosine (GC) content >46% for the six samples ([83]Table S2). Approximately 83.44%–89.72% of the clean reads were successfully mapped to the B. napus genome ([84]http://brassicadb.org/brad/datasets/pub/Genomes/Brassica_napus/) using TopHat2 software, and 65.05%–70.34% of the clean reads matched unique genomic locations ([85]Table S2). In total, 57,749, 60,261, 56,295, 56,989, 58,473, and 57,842 genes were expressed in 15-CK, 15-NT, 15-AT, 814-CK, 814-NT, and 814-AT, respectively. These results suggested that the RNA-Seq data used in the present study were highly reliable. 3.3. Differentially Expressed Gene Analysis Using the six samples, we constructed six comparison groups, CKvsNT (15), CKvsAT (15), NTvsAT (15), CKvsNT (814), CKvsAT (814), and NTvsAT (814), and 14,355 unigenes ([86]Table S3) were identified as significant DEGs in the six comparison groups, which contained 6283, 1368, 6620, 5709, 1017, and 6515 DEGs ([87]Figure 3a), respectively. Two comparisons between the CK and NT treatments (15 and 814) revealed 4582 and 4203 upregulated and 1701 and 1506 downregulated DEGs in 15 and 814, respectively; 501 and 271 upregulated and 867 and 746 downregulated DEGs were found between the CK and AT treatments (15 and 814, respectively) ([88]Figure 3b). Furthermore, compared with NT, the AT treatment resulted in many more downregulated DEGs, with 5016 and 4944 in 15 and 814, respectively ([89]Figure 3b). Correlation of gene expression levels among samples is an important index for testing the reliability of experiments and the rationality of sample selection: the closer the correlation coefficient is to 1, the higher is the similarity between samples. [90]Table S3 shows Pearson correlations for gene expression levels between different comparison groups. The Pearson correlation of gene expression levels between 15-CK and 15-NT was only 0.261, whereas that between 15-CK and 15-AT was as high as 0.876 ([91]Table S4). As expected, the Pearson correlation between 814-CK and 814-NT was only 0.275, but that between 814-CK and 814-AT was as high as 0.946 ([92]Table S4). This finding indicated fundamental differences in gene expression levels between the NT and AT treatments and that gene expression levels in oilseed rape roots in the AT treatment group were similar to those in the CK treatment group. Figure 3. [93]Figure 3 [94]Open in a new tab (a) A Venn diagram illustrating overlaps among differentially expressed genes (DEGs) in different comparison groups, i.e., CKvsNT, CKvsAT, and NTvsAT, in 15 and 814; (b) a bar chart showing up- and downregulated DEGs in the six comparison groups. The red column shows upregulated DEGs, and the green column shows downregulated DEGs. CK: control. 3.4. Gene Ontology Enrichment Analysis of Differentially Expressed Genes GO enrichment analysis was performed to broadly classify all of the DEGs into corresponding biological process, molecular function, and cellular component categories. GO terms in each category were sorted from lowest to highest by corrected p-values. To understand the distinct response of the two genotypes to the NT and AT treatments, we analyzed the enriched GO terms of the comparison groups CKvsNT (15 and 814) ([95]Figure S1) and CKvsAT (15 and 814) ([96]Figure S2). There were more upregulated than downregulated DEGs in enriched GO terms for CKvsNT (15 and 814); more downregulated than upregulated DEGs were detected among enriched GO terms for CKvsAT (15 and 814). Moreover, the numbers of DEGs for CKvsNT (15 and 814) were much higher than those for CKvsAT (15 and 814). Furthermore, the number of downregulated DEGs among enriched GO terms was also greater than the number of upregulated DEGs for NTvsAT (15 and 814) ([97]Figure 4). In addition, highly enriched GO terms for NTvsAT DEGs according to the biological process category are involved in translation, RNA methylation, and macromolecule methylation; those according to the molecular function category included structural constituent of ribosome, structural molecule activity, and rRNA binding predominated, those and according to the cellular component category included ribosome, cytosolic ribosome and ribosomal subunit. These results indicate distinctive differences in the transcript levels of certain genes between the NT and AT treatments. Figure 4. [98]Figure 4 [99]Open in a new tab Enriched Gene Ontology (GO) terms for NTvsAT (15 and 814). The results are summarized under three top-level ontologies: biological process, molecular function, and cellular component. The x-axis represents the functional category of the DEGs. The y-axis indicates the number of annotated genes expressed in a given subcategory. The downregulated DEGs are represented by green, and the upregulated DEGs are represented by red. 3.5. Kyoto Encyclopedia of Genes and Genomes Metabolic Pathway Enrichment Analysis Pathway enrichment analysis of DEGs using the KEGG database ([100]http://www.genome.jp/kegg/) identified 122, 123, 118, 110, 124, and 124 KEGG pathways for the six comparison groups CKvsNT (15), CKvsNT (814), CKvsAT (15), CKvsAT (814), NTvsAT (15), and NTvsAT (814), respectively. The top 15 enriched pathways for the two genotypes contained more upregulated DEGs than downregulated DEGs for CKvsNT ([101]Figure S3) but more downregulated DEGs than upregulated DEGs for CKvsAT ([102]Figure S4). DEGs involved in ribosome, biosynthesis of secondary metabolites and phenylalanine metabolism were more enriched in CKvsNT, whereas DEGs involved in biosynthesis of secondary metabolites, phenylalanine metabolism and phenylpropanoid biosynthesis were more enriched in CKvsAT. In NTvsAT ([103]Figure 5), there were more downregulated than upregulated DEGs, and DEGs involved in ribosome, phenylalanine metabolism and biosynthesis of secondary metabolites were highly enriched. Figure 5. [104]Figure 5 [105]Open in a new tab Top fifteen enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways for NTvsAT (15 and 814). The x-axis represents the number of DEGs involved in each pathway; the y-axis depicts pathways arranged from bottom to top based on the degree of enrichment. 3.6. Genes Involved in Starch and Sucrose Metabolism, Fatty Acid Biosynthesis, Ribosome and Cell Wall Biogenesis Carbohydrates, lipids and proteins are the main organic materials that compose plant cells and provide energy. In the pathway enrichment analysis in this study, we found three metabolic pathways, namely, starch and sucrose metabolism (SSM) (ko00500), fatty acid biosynthesis (FAB) (ko00061) and ribosome (RIB) (ko03010), which are the primary pathways of carbohydrate metabolism, lipid metabolism and translation, respectively, to be enriched in CKvsNT and NTvsAT. DEGs involved in SSM, FAB, and RIB showed the same trend. The numbers of DEGs involved in SSM, FAB, and RIB in CKvsNT (15 and 814) and NTvsAT (15 and 814) were greater than those in CKvsAT (15 and 814) ([106]Table 2). Moreover, there were more upregulated DEGs in SSM, FAB, and RIB and fewer downregulated DEGs in the NT treatment group than in the CK treatment group, despite few differences in response to the AT treatment. There were fewer upregulated DEGs in SSM, FAB, and RIB and more downregulated DEGs in the AT treatment group than in the NT treatment group. Information on the DEGs involved in SSM, FAB, and RIB in the six comparison groups is provided in [107]Table S5. To further determine whether the molecular levels of carbohydrates, lipids and proteins were significantly altered in the different N treatments, we examined the numbers of up- and downregulated DEGs for each pathway in carbohydrate metabolism, lipid metabolism and translation in the six comparison groups ([108]Table S6). The results indicate that carbohydrate metabolism, lipid metabolism and protein metabolism were inhibited in the AT treatment group, which led to decreased growth. Table 2. Numbers of up- and downregulated genes in starch and sucrose metabolism (SSM), fatty acid biosynthesis (FAB), ribosome (RIB), and cell wall biogenesis (CWB) in the six comparison groups. Comparison Groups SSM FAB RIB CWB Up Down Up Down Up Down Up Down CKvsNT(15) 158 20 28 1 481 1 8 0 CKvsNT(814) 142 13 21 1 566 1 6 0 CKvsAT(15) 18 20 3 1 6 29 0 0 CKvsAT(814) 4 8 0 2 0 9 0 0 NTvsAT(15) 19 161 0 32 2 613 0 9 NTvsAT(814) 16 162 0 22 4 631 0 5 [109]Open in a new tab The cell wall is the major component of plant roots. In the GO term analysis performed in this study, we found that the numbers of DEGs involved in primary cell wall biogenesis (CWB) (GO: 0009833) differed between the comparison groups ([110]Table 2). There were 8 and 6 upregulated DEGs in CKvsNT (15 and 814), 9 and 5 downregulated DEGs in NTvsAT (15 and 814) and, interestingly, no DEGs in CKvsAT (15 and 814). Furthermore, DEGs associated with CWB are members of the cellulose synthase (CESA) gene superfamily, including CESA1, CESA2, CESA3, and CESA6 ([111]Table S5). Therefore, NT treatment enhanced the transcriptional levels of genes related to CWB, whereas AT treatment had no effect on primary CWB. 3.7. Genes Related to N Uptake, Transport, and Assimilation N metabolism in plants mainly involves three processes: N uptake, transport and assimilation. Many genes involved in NO[3]^−/NH[4]^+ uptake, transport and assimilation were differentially expressed in the six samples ([112]Tables S2 and S6). In the current study, 27 DEGs encoding NO[3]^− transporters (NRT1, NRT1.5, NRT 1.8, NRT2, NRT3, and CLC-B (chloride channel protein B)) were detected ([113]Figure 6, [114]Table S7). The abundances of the transcripts of these genes, except for 1 DEG (BnaC09g20910D), were increased in CKvsNT; transcript levels of 4 genes decreased in CKvsAT (15). In NTvsAT, the transcript levels of most DEGs were reduced in both 15 and 814, but 2 DEGs (BnaC09g20910D and BnaC02g36720D) were upregulated only in 15. A total of 5 genes encoding NH[4]^+ transporters were identified, of which BnaCnng01740D was responsive only in 15 among the six samples ([115]Table S7). Transcript levels of 3 DEGs were increased in CKvsNT (15). Two DEGs were upregulated in CKvsNT (814), though the other 2 DEGs were downregulated ([116]Figure 6). One DEG (BnaA01g23190D) was upregulated in NTvsAT (814). Two DEGs were downregulated in NTvsAT (15 and 814). Figure 6. [117]Figure 6 [118]Open in a new tab Heatmap showing the relative expression patterns of genes related to N metabolism between different comparison groups in 15 and 814. From left to right: 15-log[2](NT/CK), 814-log[2](NT/CK), 15-log[2](AT/CK), 814-log[2](AT/CK), 15-log[2](NT/AT), and 814-log[2](NT/AT). Colors indicate the relative differential gene expression: green, downregulated; red, upregulated; and blue, not differentially expressed. Some of the DEGs encode key enzymes in NO[3]^−/NH[4]^+ assimilation, including 4 DEGs encoding nitrate reductase (NR), 10 DEGs encoding cytochrome B5 (CB5) of NR, 2 DEGs encoding nitrite reductase (NIR), 5 DEGs encoding glutamine synthetase (GS), and 5 DEGs encoding glutamate dehydrogenase (GDH) ([119]Figure 6, [120]Table S7). In CKvsNT, there were more upregulated DEGs than downregulated DEGs encoding NR, CB5, NIR, and GS in the two cultivars. In CKvsAT, there were more downregulated DEGs than upregulated DEGs encoding NR, CB5 and GS in the two cultivars. In NTvsAT, 1 DEG encoding NR, 5 DEGs encoding CB5, and 1 DEG encoding NIR were downregulated; 1 upregulated DEG and 1 downregulated DEG encode GS1.3 and GS2, respectively ([121]Table S7). Interestingly, DEGs encoding GDH were always downregulated in CKvsNT and upregulated in NTvsAT. 3.8. Validation of Genes Using Quantitative Real-Time PCR To confirm the accuracy and reproducibility of the Illumina RNA-Seq results, 15 representative genes were chosen based on the above analysis, and their expression levels in 15 and 814 in the different N treatment groups were validated by quantitative real-time PCR (qRT-PCR) ([122]Figure 7a). The 15 representative genes encoded low- and high-affinity NO[3]^−/NH[4]^+ transporters, NR, NIR, GS, GDH, and catalytic subunits of cellulose synthase, which are key proteins or enzymes in N metabolism and root development. Among these DEGs, the qRT-PCR profiles for the 15 representative genes mostly agreed with those obtained from RNA-Seq, with minimal differences in expression ([123]Figure 7a). Pearson correlation (coefficient 0.9576) revealed strong agreement between the qRT-PCR data for these genes and the RNA-Seq results ([124]Figure 7b). Therefore, the RNA-Seq data presented herein are reliable. Figure 7. [125]Figure 7 [126]Open in a new tab Quantitative real-time PCR (qRT-PCR) validation of 15 differentially expressed genes (DEGs). (a) Transcript levels of 15 DEGs. Vertical bars indicate SDs (n = 3). The columns represent relative expression obtained by qRT-PCR. (b) Comparison of relative expression obtained from RNA-Seq data and qRT-PCR. RNA-Seq relative expression (x-axis) is plotted against qRT-PCR relative expression (y-axis). The correlation coefficient (R^2) is indicated in the figure. FPKM: fragments per kilobase of exon per million fragments mapped. 4. Discussion NO[3]^− and NH[4]^+ are the main N sources for oilseed rape, but studies investigating the effects of NO[3]^− and NH[4]^+ on the physiology and transcriptome levels of oilseed rape are scarce. Our primary goal in this work was to differentiate the physiological and molecular effects of the two most common sources of N utilized by oilseed rape: NO[3]^− and NH[4]^+. There were clear, distinct differences in physiology and transcriptome levels between NO[3]^− and NH[4]^+ nutrition. 4.1. Oilseed Rape Seedling Growth Was Inhibited by AT treatment In the physiological experiment, compared with a high concentration of NO[3]^− as the sole N source, a high concentration of NH[4]^+ significantly reduced oilseed rape biomass, root growth, and N accumulation. These results indicate that a high concentration of NH[4]^+ as the sole N source is a source of stress for oilseed rape but that NO[3]^− or NO[3]^− plus a small amount of NH[4]^+ is a suitable N nutrition environment for growth. Some studies have found that NO[3]^− plus a small amount of NH[4]^+ is beneficial for the growth of crops that prefer NO[3]^− [[127]6] and that NH[4]^+ plus a small amount of NO[3]^− is beneficial for the growth of crops that prefer NH[4]^+ [[128]26]. Additionally, some studies have confirmed that a single application of NH[4]^+ or overapplication of NH[4]^+ is conducive to the growth of most dryland crops [[129]5,[130]6]. As NO[3]^− taken up by plant cells can be stored in large quantities in vacuoles, this form of N is non-toxic for plants. NO[3]^− concentrations in plant cell vacuoles and the cytoplasm range 30–50 mol/m^3 and 3–5 mol/m^3 [[131]27,[132]28], respectively, whereas NH[4]^+ is rarely found in vacuoles. Indeed, NH[4]^+ taken up by plant cells must be combined with organic acids as soon as possible, leading to the synthesis of amino acids or amides and the acidification of cells. Recent studies have suggested that NH[4]^+ can restrain uptake of K^+, Ca^2+, Mg^2+, and other cations by crops [[133]6]. A high concentration of NH[4]^+ can also suppress the activities of key enzymes in crop N metabolism, such as NR, GS, and GOGAT (Glutamate synthase) [[134]29]. Therefore, a high concentration of NH[4]^+ is likely to result in NH[4]^+ toxicity in dryland crops. In summary, oilseed rape appears to be a nitrate-preferring crop. The N efficiency of a crop has two meanings. On the one hand, it refers to the amount of N absorbed by crops at the same N supply level; on the other hand, it refers to high use efficiency of the N absorbed and the amount of dry matter produced per unit of N absorbed [[135]30]. In this experiment, compared with the N-inefficient cultivar 814, the N-efficient cultivar 15 displayed better root architecture and higher N use efficiency under the NT treatment. The root system is the main organ for absorbing water and nutrients and is important for organic matter assimilation, transformation, and synthesis [[136]31]. The 15 cultivar showed a better root length, root surface, and root volume than did 814, which led to more N uptake and accumulation in 15 than in 814 under the NT treatment. Our previous work [[137]16] showed that compared with the N-efficient cultivar 15, the N-inefficient cultivar 814 stores more NO[3]^− absorbed by the root system in root vacuoles, resulting in a decrease in NO[3]^− transported from the root system to the aboveground part. However, to meet the growth and developmental needs of the aboveground part, the root system of the 814 cultivar needs to absorb more NO[3]^−. In other words, the 814 cultivar must absorb more NO[3]^− to produce the same biomass as 15, which leads to lower N use efficiency in the former. GO ([138]Figure S6) and KEGG ([139]Figure S7) analyses of DEGs between 15 and 814 at 48 h after NT treatment (15vs814(NT)) revealed some differences in gene expression between the two cultivars, especially DEGs related to the cell wall ([140]Table S8) and N metabolism ([141]Table S9). Overall, 1877 DEGs (734 upregulated and 1143 downregulated) were identified ([142]Figure S5). GO analysis showed high enrichment of DEGs involved in the organonitrogen compound metabolic process, oxidoreductase activity, catalytic activity and the cell wall, which was consistent with a recent report using 15 and 814 as test materials treated with low NO[3]^− (0.30 mm/L) for 0 h, 3 h, and 72 h [[143]32]. The expression levels of both BnaA02g13640D (NRT1) and BnaA01g26960D (NRT1), which are responsible for efficient NO[3]^− uptake, were higher in the roots of 15 than in 814 in our study. In contrast, mRNA levels of BnaA09g47380D (NRT1.1) and BnaC08g43380D (NRT2.1) were much higher in the roots of 15 than in 814 in the recent report [[144]32], particularly under low NO[3]^−. Because the N solution concentration, N treatment time and seedling period were different between our study and the previous report, we conclude that the discrepancy may be attributed to gene expression being affected by the growth period and N treatment. In previous hydroponic and sand culture experiments, we found no significant differences in phenotypes (growth and physiology indexes) between 15 and 814 under NT treatment during the early seedling period (0–50 d after seedling emergence). Therefore, to analyze the transcriptomic results between 15 and 814 under NT treatment at different growth periods, it is helpful to comprehensively understand the molecular mechanism of N efficiency differences in oilseed rape. Although a high concentration of NH[4]^+ as the sole N source is detrimental to oilseed rape growth at the physiological level, the related transcriptomic changes remain largely unknown. In our study, we applied RNA-Seq to evaluate gene expression under different N treatments in the roots of seedlings of the two genotypes and examined many of the genes that were differentially expressed among the CK, NT and AT treatments. These genes are mainly involved in N metabolism, CWB, and carbohydrate, lipid and protein metabolism. 4.2. N Metabolism of Oilseed Rape Was Inhibited by AT treatment Compared with CK, most genes involved in N metabolism were markedly elevated by the NT treatment, though these genes were not induced by the AT treatment. The latest transcriptomics results for N limitation (0.3 mmol/L NO[3]^−) in oilseed rape show that genes related to N uptake and allocation, such as BnaNRT1.1 s (B. napus nitrate transporter 1.1 s), BnaNRT1.4 s, BnaNRT1.5 s, BnaNRT1.9 s, BnaNRT2.1 s, BnaNRT2.4 s, and BnaNRT3.1 s, are induced in roots by N limitation, and N metabolism genes such as BnaGS1s and BnaGS2s are also elevated in roots under N limitation; in contrast, genes such as BnaNRT1.8 s, BnaNIA1s, and BnaNIA2s are downregulated in roots [[145]33]. These results are not consistent with our findings. The majority of genes related to N metabolism were induced in the roots of oilseed rape exposed to N limitation, whereas these genes were mostly strongly repressed in roots by the N-free (CK) treatment. This might be caused by the different adaptive mechanisms of oilseed rape in response to N limitation or N-free conditions. Overall, increased N uptake and transport and enhanced assimilation of inorganic N into amino acids might be helpful for the adaptability of oilseed rape to N limitation [[146]33]. NT treatment enhanced expression of BnNRTs but AT treatment decreased that of BnAMTs in oilseed rape. Four types of transporters, namely, low- and high-affinity NO[3]^− transporters (NRTs)/NH[4]^+ transporters (AMTs), take up and reallocate NO[3]^− and NH[4]^+ in angiosperms, respectively [[147]34]. NO[3]^− induction of GmNRT2 mRNA expression was accompanied by a fourfold increase in net NO[3]^− uptake by soybean roots under 100 μM external NO[3]^− [[148]35]. In addition, correlation coefficients between ^13NO[3]^− influx from 100 μM and 5 mM [NO[3]^−] suggest that in A. thaliana, high- and low-affinity transport systems are mediated primarily by AtNRT2.1 and AtNRT1.1, respectively [[149]36]. In our physiology experiment, N accumulation in the NT treatment group was significantly higher than that in the AT treatment group. Therefore, upregulated expression of most BnNRTs under NT treatment greatly promoted NO[3]^− uptake by oilseed rape. Recent research has demonstrated that downregulation of AtAMT1 genes, which mediate the majority of NH[4]^+ uptake in Arabidopsis roots, is associated with decreased NH[4]^+ accumulation in the AtNRT1.1 mutants chl1-1 and chl1-5 under conditions of high NH[4]^+ [[150]29]. The adjustment of oilseed rape to excessive NH[4]^+ conditions via downregulation of the transcript levels of BnAMTs might be an adaptive mechanism ([151]Figure 8) for reducing NH[4]^+ uptake. Other research also supports this mechanism. For example, ^13NH[4]^+ influx and transcript levels of OsAMT1.3 in roots decreased within 48 h when plants acclimated to 10 µm of external NH[4]^+ for 3 weeks were transferred to 10 mm NH[4]^+ [[152]37]. Figure 8. [153]Figure 8 [154]Open in a new tab A proposed model for N metabolism and other primary metabolism in oilseed rape roots in NT and AT treatment groups. The colored rectangles indicate the following: red, N transporters, enzymes or physiological processes that increased; and green, N transporters, enzymes or physiological processes that decreased. Red or green solid arrows represent the strengthening or weakening of physiological processes, and black solid arrows represent physiological processes that did not significantly change. Compared with AT treatment, NT treatment led to upregulation of the NRT/AMT genes in roots and facilitated NO[3]^− uptake. Simultaneously, NR and NIR activities increased, and BnNRT1.5 gene expression was upregulated, which facilitated more NO[3]^− transport into shoots from roots for assimilation under NT treatment. This enhanced transport was beneficial for oilseed rape growth because NO[3]^− assimilation efficiency is known to be higher in shoots than in roots due to the utilization of solar energy by oilseed rape shoots. Furthermore, carbohydrate metabolism (CAR), lipid metabolism (LIP), protein metabolism (PRO), and CWB were promoted by NT treatment, which facilitated oilseed rape root growth and development. Thus, NT treatment was more beneficial to oilseed rape growth than was AT treatment. NT treatment increased expression of BnNRT1.5 while decreasing that of BnCLC-B. Long-distance transport and distribution of NO[3]^− between roots and shoots are regulated by two genes encoding components involved in long-transport mechanisms. NRT1.5 is responsible for NO[3]^− xylem loading, and NRT1.8 is responsible for NO[3]^− xylem unloading [[155]38,[156]39]. BnNRT1.5 expression in the roots of oilseed rape increased significantly with NO[3]^− treatment, driving greater NO[3]^− long-distance transport from roots to shoots [[157]16], which was beneficial for oilseed rape growth because NO[3]^− assimilation efficiency is known to be higher in shoots than in roots due to solar energy utilization by shoots [[158]40]. The NT treatment resulted in overexpression of all three NRT1.5 genes in both genotypes ([159]Tables S3 and S7), indicating that more NO[3]^− was transported to the shoot from the root for assimilation under NT. Thus, the NO[3]^− assimilation efficiency of oilseed rape was promoted by the NT treatment. AtCLC-B, one of seven AtCLC transport proteins in A. thaliana that are responsible for vacuolar NO[3]^− short-distance transport and are the main channels for NO[3]^− movement between vacuoles and the cytosol, conducts strong outwardly rectifying anionic currents that are greatest in the presence of NO[3]^− [[160]16,[161]41,[162]42]. Highly expressed BnaC02g36720D (BnCLC-B) most likely led to increased transport of NO[3]^− from vacuoles into the cytosol to satisfy the demand for NO[3]^− in the AT treatment group ([163]Figure 6 and [164]Figure 8). This result further verified that oilseed rape is a nitrate-preferring crop. NT treatment led to upregulation of more genes encoding NR, CB5, and NIR than did AT treatment. NO[3]^− reduction involves concerted reactions in which NO[3]^− is reduced to nitrite (NO[2]^−) by NR and NO[2]^− is reduced to NH[4]^+ by NIR [[165]43]. Each NR subunit contains 3 prosthetic groups, namely, FAD, heme (CB5) and a molybdenum cofactor (MoCo) [[166]44]. As previously reported for many higher plant species, transcriptional and nitrate-inducible regulation is one of the major mechanisms of NR activity in rapeseed [[167]45]. Previous studies have also demonstrated that nitrate-supplied Arabidopsis roots exhibit large increases in the abundance of NR (At1g77760), NR (At2g15620) and NADH-GOGAT (At5g53460) transcripts and a >10-fold increase in NR enzyme activity [[168]14]. Hence, our results are consistent with the above results. Furthermore, BnGDH2 was specifically induced by the AT treatment ([169]Figure 6 and [170]Figure 8). GDH may play a major role in maintaining a subtoxic NH[4]^+ level in the cytoplasm [[171]46] and might operate via amination. GDH responds positively to abiotic stress and is important for the detoxification of NH[4]^+ under stress [[172]47,[173]48,[174]49]. In addition, transcriptomic analysis of Arabidopsis roots supplied with NO[3]^− or NH[4]^+ demonstrated that AtGDH2 is specifically induced by NH[4]^+ [[175]14], which is consistent with our results. Hence, we can conclude that oilseed rape roots endured the excessive but subtoxic NH[4]^+ level in the cytoplasm under AT treatment by upregulating the transcript levels of BnGDH2 ([176]Figure 8), which further suggests that a high concentration of NH[4]^+ tends to result in NH[4]^+ toxicity in oilseed rape roots at the transcriptome level. 4.3. Expression Levels of Genes Involved in Cell Wall Biogenesis Decreased with AT treatment In this work, the root length, root surface area and root volume in the NT treatment group were significantly greater than those in the AT treatment group, which demonstrated that AT inhibited root development. All plant cells are surrounded by an extensible primary cell wall, which contains cellulose, pectin and noncellulosic polysaccharides [[177]50]. Additions and rearrangements of cell wall components are required throughout growth and development, such as for cell elongation and cell proliferation in developing roots [[178]51]. The AtCESA gene superfamily, which encodes the catalytic subunits of cellulose synthase, has been identified in hundreds of seed plant species and well characterized in Arabidopsis [[179]52]. Plants have evolved complex regulatory mechanisms, including expression of CESA genes and modification of CESA proteins, to control cellulose biosynthesis and assembly in cell walls [[180]53]. In this work, genes related to primary CWB (GO: 0009833) ([181]Figure 8) were differentially expressed under the different N treatments, including BnCESA1, BnCESA2, BnCESA3, and BnCESA6, which encode cellulose synthase family proteins ([182]Table S4). In A. thaliana, AtCESA1 (A. thaliana cellulose synthase 1), AtCESA3, and AtCESA6 are preferentially expressed in expanding tissues [[183]54], but AtCESA2 is poorly understood. The NT treatment greatly increased the transcript levels of BnCESA1, BnCESA2, BnCESA3, and BnCESA6, whereas the AT treatment had no distinct effects ([184]Table S5, [185]Figure 8). This difference might constitute one line of evidence explaining why AT treatment inhibited root development in oilseed rape in this study. In summary, there were significant differences between NO[3]^− and NH[4]^+ nutrition with regard to biomass, N accumulation and root architecture in oilseed rape cultivars. The potential molecular mechanisms may be explained by most of the downregulated genes being related to N metabolism, cell wall biosynthesis, and carbohydrate, lipid, and protein metabolism in oilseed rape exposed to a high concentration of NH[4]^+. To the best of our knowledge, this is the first comprehensive transcriptomic analysis of oilseed rape grown under NO[3]^− and NH[4]^+ conditions, and the findings provide valuable resources for better understanding the responses of oilseed rape to NO[3]^− and NH[4]^+ and subsequent improvement of N utilization efficiency. 5. Conclusions In the present study, the biomass, N accumulation and root architecture of two oilseed rape cultivars with different N efficiencies were significantly and analogously influenced by NO[3]^− and NH[4]^+ nutrition. To reveal the molecular mechanisms responsible for the differences in the two cultivars in response to CK, NO[3]^−, and NH[4]^+ conditions, RNA-Seq of root samples collected at 48 h after exposure to different N treatments was performed to analyze gene expression patterns related to N metabolism and other crucial metabolic pathways in plants. NH[4]^+ directly decreased the biomass, N accumulation, and root architecture of the two oilseed rape cultivars and affected the overall metabolism of shoots and roots; this result was related to downregulated expression of NO[3]^−/NH[4]^+ transporters or channels and crucial metabolic pathways (including carbohydrate, lipid, and protein metabolism) in the two oilseed rapes. The root expression levels of CESA genes encoding cellulose synthase decreased under NH[4]^+ nutrition, inhibiting root growth. These results provide insight into the regulatory mechanisms of some candidate genes, which will reveal the molecular mechanisms of decreased oilseed rape growth in response to a high concentration of NH[4]^+ as the sole N source. Supplementary Materials The following are available online at [186]https://www.mdpi.com/2073-4425/10/5/391/s1, Figure S1: Enriched Gene Ontology terms in comparison group CKvsNT (15 and 814). Figure S2: Enriched Gene Ontology terms in comparison group CKvsAT (15 and 814). Figure S3: Top fifteen enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways in CKvsNT (15 and 814). Figure S4: Top fifteen enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways in CKvsAT (15 and 814). Figure S5: Numbers of up- and downregulated DEG and fold change were analyzed in 15vs814 (NT). Figure S6: Enriched Gene Ontology (GO) terms in 15vs814 (NT). Figure S7: Top fifteen enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways in 15vs814 (NT). Table S1: List of primer sequences used for qRT-PCR analysis. Table S2: Summary of transcriptomic data. Table S3: Information for the 14,355 significantly differentially expressed genes (DEGs) detected in the six cDNA libraries. Table S4: Pearson correlation of gene expression levels between different comparison groups. Table S5: Information on the differentially expressed genes (DEGs) involved in starch and sucrose metabolism, fatty acid biosynthesis, ribosome and plant-type primary cell wall biogenesis in the six comparison groups. Table S6: Numbers of up- and downregulated genes in each pathway involved in carbohydrate metabolism, lipid metabolism and translation in the six comparison groups. Table S7: Genes encoding NO[3]^−/NH[4]^+ protein transporters and assimilation enzymes differing in expression between genotypes in response to different nitrogen treatments. Table S8: Numbers of up- and downregulated DEGs related to cell wall in 15vs814(NT). Table S9: Important DEGs related to N uptake, transport and assimilation in 15vs814(NT). [187]Click here for additional data file.^ (3.6MB, zip) Author Contributions W.T., L.L., and C.G. conceived and designed the experiments. W.T., F.W., and C.S. prepared the experimental materials. The bioinformatics analysis was carried out by W.T. and X.H. W.T. wrote the manuscript. L.Q., X.H., and Z.Z. provided intellectual input and revised the manuscript. All authors read and approved the final manuscript. Funding This work was supported by the Earmarked Fund for China Agriculture Research System (CARS-13), the National Science and Technology Support Project of China (2014BAD11B03) and the Oil Crop Innovation Team in Yunnan Province, China (2017HC021). Conflicts of Interest The authors declare no conflict of interest. References