Internal ammonium excess induces ROS-mediated reaction and causes carbon scarcity in rice

Background: Overuse of nitrogen fertilizers is often applied to satisfy strong nitrogen demand of high-yielding rice, leading to persistent NH 4 + excess in the plant. However, the mechanisms constraining the effectiveness of elevated plant NH 4 + in plant growth and grain yield of rice are not suciently addressed. In the current study, we attempt to real the nature or mode-of-action of such internal NH 4 + excess in rice, and the ecient coordination measure with current high N fertilizer cropping systems is investigated. Results: By mimicking a rapid accumulation of plant NH 4 + and an RNA-Seq analysis, the present work reveals that internal NH 4 + excess in rice plant initiates a burst of reactive oxygen species (ROS) and triggers probably specically the activation of glutathione transferase (GST)-mediated glutathione cycling for ROS cleavage. Meanwhile, the suppression of the expression of genes involved in photon caption and the activity of primary CO 2 xation enzymes (Rubisco), provides implications of a reduction in photosynthetic carbon income. Along the progress of NH 4 + / ROS stresses, enhanced energy-producing processes (carbon breakdown) take place as indicated by strong induction of glycolysis related genes to meet the demand of energy consuming activation of ROS-cleavaging systems. The development of these defensive reactions causes a sugar scarcity in the plant that accumulatively leads to growth inhibition. To the opposite direction, a sucrose feeding treatment to the roots renders the accumulation of free NH 4 + and ROS, partly restores the activities of photosynthetic CO 2 xation and thus alleviates the scarcity in active carbon source. Conclusion: Our results demonstrate the necessity of ecient carbon coordination, aiming at improving the nitrogen performances under current N fertilizer overuse circumstances. Glutamine GSH: Glutathione; GST: Glutathione S-transferase; H 2 DCF-DA: 2′,7′-Dichlorouoresceindiacetate; LHCs: light-harvesting complexes; N, nitrogen; NH 4+ : ammonium; qRT-PCR: Quantitative Reverse Transcription Polymerase Chain Reaction; POD: peroxidase; ROS: Reactive Oxygen Species; RNA-Seq: RNA SOD: superoxiddismutase; Rubisco: Ribulose– 1, 5-bisphosphate carboxylase/oxygenase(cid:0) TCA acid cycle

excess in rice, and the e cient coordination measure with current high N fertilizer cropping systems is investigated.
Results: By mimicking a rapid accumulation of plant NH 4 + and an RNA-Seq analysis, the present work reveals that internal NH 4 + excess in rice plant initiates a burst of reactive oxygen species (ROS) and triggers probably speci cally the activation of glutathione transferase (GST)-mediated glutathione cycling for ROS cleavage. Meanwhile, the suppression of the expression of genes involved in photon caption and the activity of primary CO 2 xation enzymes (Rubisco), provides implications of a reduction in photosynthetic carbon income. Along the progress of NH 4 + / ROS stresses, enhanced energyproducing processes (carbon breakdown) take place as indicated by strong induction of glycolysis related genes to meet the demand of energy consuming activation of ROS-cleavaging systems. The development of these defensive reactions causes a sugar scarcity in the plant that accumulatively leads to growth inhibition. To the opposite direction, a sucrose feeding treatment to the roots renders the accumulation of free NH 4 + and ROS, partly restores the activities of photosynthetic CO 2 xation and thus alleviates the scarcity in active carbon source. Conclusion: Our results demonstrate the necessity of e cient carbon coordination, aiming at improving the nitrogen performances under current N fertilizer overuse circumstances. Background Nitrogen (N) limitation was a leading constraint on the grain yield of rice [1,2]. Leaf N accounted for the largest N sink of rice plant, ca. 80% of which was distributed in the chloroplast and stored as Ribulose-1,5-bisphosphate carboxylase⁄oxygenases (Rubisco), the primary carbon fixation enzymes of C 3 plants [3]. Photosynthesis was tightly correlated with leaf N content [4], and more than 80% of grain N was derived from leaves in rice [5]. Hence, insu cient leaf N storage would lead to reduction of photosynthetic carbon xation e ciency and was therefore considered as a major limitation to biomass and grain production of cereal ecosystems [6][7][8][9].
In China's rice farming, to satisfy the N demand of higher grain yield (6.4 t/hm 2 or above), the average N input was normally over 180 kg/hm 2 [10]. In the high-yielding rice farming areas, the N input could even reach to 300 kg/hm 2 and this was particularly the case for recent super-hybrid rice cultivars that achieved as high as >10 t/hm 2 of grain yields [11]. In soils, applied N fertilizers (e.g. urea form accounted for the majority of current N fertilizers) would rapidly convert to ammonium with the potent reactions of ureases.
And in well-ventilated dryland soils, ammonium could convert to another root available nitrogen formnitrate through the actions of nitri cation microbes. However, in rice paddy soils, ca. 70-80% of the growth period was water ooded, causing an anaerobic environment that largely prevented the process of nitri cation. To this end, retention of higher concentration of ammonium around the roots is a generalized fact in rice paddies. Recent results showed that over-dosed N fertilizers strengthened excessive N accumulation in rice plant that however did not accordingly result in increased grain yields [12]. Therefore, low N use e ciency and its agronomic bene ts were major problems of high N input rice farming.
Plant responses to toxic concentrations of NH 4 + supplied exogenously had been extensively demonstrated centering the (molecular) mechanisms involved in the con guration adjustment of root architectures. In Arabidopsis, root tip contact was essential for triggering the inhibitory growth of primary roots subjected to high NH 4 + treatments [13]. Whereas leaf contact and absorption of toxic NH 4 + obstructed primary polar transportation of IAA to the roots through the pathway of AUX1 thereby inhibited the emergence of lateral roots [14]. In rice, high concentrations of NH 4 + supplied to roots strongly inhibited seminal root elongation and led to reduction of plant growth [15][16][17][18][19]. Mechanisms of plant responses to external NH 4 + stresses had been summarized as the accumulative consequences of ion imbalance, intracellular pH disturbance, carbon limitation, charge/hormone imbalance and oxidative stresses [20][21][22][23][24]. The work on Arabidopsis hsn/vtc mutants provided further molecular evidence that GDP-mannose pyrophosphorylases were involved in protein N-glycosylation modulated inhibition of root elongation under NH 4 + stresses [25][26][27]. Morever, phytohormone signals were reported to interact with NH 4 + supply thus to regulate plant metabolism, growth and development [13,14,24,25,27,28]. And these responsive components are closely related to high external NH 4 + supply but seemingly independent of internal NH 4 + accumulation. A previous study implied that the adjustment of carbohydrate metabolisms could be a notable feature in responding NH 4 + status in rice in a short time period [31]. Environmental stress stimuli such as salinity, drought etc. could induce overproduction of reactive oxygenspecies (ROS) and promptly trigger oxidative defense responses [32][33]. Meanwhile, the reduced photosynthetic CO 2 xation e ciency and alteration of carbohydrate metabolism were speculatively major causes leading to compromised carbon gain and growth retardance [32,34].
Based on the above descriptions, the present work aimed at revealing the speci c nature and mode-of-  (Fig. 1a). The inhibition was more profound in roots showing a biomass reduction of up to 67% (Fig. 1a)and the root/shoot ratio was signi cantly lowered from approximately 0.5 down to 0.2 (Fig. 1b). Meanwhile, 7 and 5 folds higher concentrations of free NH 4 + were measured in roots and shoots, respectively (Fig. 1c).The inhibitory effect by high NH 4 + subjected to plant roots was however, a well known phenomenon as described by numerous reports elsewhere (see Introduction). Efforts have be extensively made on the elucidation of molecular mechanisms involved in root architecture adjustments in response to the accumulation of relatively longterm (several days or longer) stress effects impended by high NH 4 + treatments. Here to reveal early responsive reactions that could be the trigger of the accumulative responses (growth modi cations), a prompt status of internal NH 4 + excess is necessarily to be established without causing visible changes in plant growth (especially roots). Therefore, L-methionine-D,L-sulfoximine (MSX), a potent inhibitor of the primary NH 4 + assimilation pathway mediated by the activity of glutamine synthetases [35] was applied (1 mM) for 4 h in the presence of high NH 4 + (20 mM). This treatment resulted in similar increases in free NH 4 + and rapidly achieved 'saturable' NH 4 + excess in both roots and shoots (Fig. 1d), providing a facilitated approach for subsequent identi cation of genes or processes involved in this response. In line with the accumulation of free NH 4 + , bursts of reactive oxygen species (ROS) were observed (Fig. 1e), implying possible occurrence of oxidative injuries or ROS-induced reactions triggered by internal NH 4 + excess.
RNA-Seq analysis for preliminary identi cation of genes modulated by NH 4 + excess According to above description, rice seedlings were treated with high NH 4 + in the presence of 1mM MSX for 4 h to establish an internal environment of NH 4 + excess. Then RNA-Seq analyses were carried out to seek for molecular responses related to this circumstance. Respectively 1077 and 1040 differentially expressed genes (DEGs) were obtained from roots and shoots, with > 2 fold changes in their transcriptional levels (Additional le1). Based on the GO classification, these genes mainly belonged to "metabolic process", "molecular function", "binding" and "biological process" (Additional le 2). Further KEGG pathway analysis revealed possible involvements of the responsive genes (DEGs) in stress response, photosynthetic adjustment, carbohydrate and amino acid metabolisms, preparation of hormone signaling pathways and re-adjustment of NH 4 + transport (Additional le 3). The signi cantly regulated genes were further summarized below within the framework of major processes they participate.

Activation of GSH cycle for ROS scavenging
Following the acute plant NH 4 + excess and the bursts of ROS (Fig. 1c, d, e), a most remarkable response was the strong induction of glutathione S-transferases (GST) genes ( Fig. 2). Eleven GST genes were typically upregulated for >7 or even some tens to hundreds fold both in roots and shoots ( Fig. 2a, b, genes#1-11). Among those GSTs, a OsGSTU4 (Os10g0528300, Fig. 2a, gene#11) was the most severely induced by >300 and >600 fold in roots and shoots respectively, followed by 2 putative GST genes (Os10g0481300 and Os10g0527800) that were upregulated by 50-100 fold in both parts. Whereas Os10g0525500 (77 fold) and Os03g0785900 (90 fold) showed strong induction in roots and shoots respectively ( Fig. 2a, b). Since GSTs catalyze the transfer of superoxide free radicals to reductive glutathione (GSH) that leads to the detoxi cation of the oxidants, these changes in GST gene expression provide indications for the critical involvement of the GSH cycle in scavenging the NH 4 + excess induced ROS.
The enhanced GST activity accelerates the consumption and conversion of GSH to its oxidized form (GSSG). In line with strengthened demand of reducing power, a putative glutathione reductase gene (Os10g0415300) responsible for the recruitment of GSH was moderately upregulated (~ 8 fold) in roots and vigorously enhanced by 70 fold in shoots (Fig. 2a). Meanwhile, a NADH dehydrogenase gene (Os07g0564500) was stimulated by 127 folds in shoots, partly re ecting the coupling of energization and reducing power with the operation of the GSH cycle (Fig. 2a).
In addition to profound changes related to the GSH cycle, 7 peroxidase genes were suppressed in roots whereas a putative 1-Cys peroxiredoxin B gene (Os07g0638400) was signi cantly induced in both roots (19 fold) and shoots (179 fold) (Fig. 2a), corresponding to the contradictory roles of peroxidases in the cleavage / homeostasis maintenance of ROS [36].
Suppression of photosynthesis components and contrasting regulation of energy producing carbohydrate metabolism The chlorophyll a/b binding proteins of light-harvesting complexes (LHCs), also known as antenna proteins, are involved in gathering light energy (photons) of the primary reaction of photosynthesis [37]. Then trapped photons and electrons are transported to reaction center for further photochemical reactions. Disruption of these processes by photodamage, herbicides, or accumulation of highly active radicals will obviously hinder the progress of photosynthesis. Upon a prompt (4 h) NH 4 + excess treatment, 6 genes coding for the LHC antenna proteins (4 LHC II and 2 LHC I, respectively), a PS I and a PS II reaction center genes were almost evenly suppressed by approximately 5 fold (Fig. 3), indicating the onset of the reduction of e ciencies of photon gathering and transfer. It would be easily supposed that apparent suppression of photosynthesis would accumulate along the progress of NH 4 + excess stress and growth inhibition would consequently occur. Meanwhile, Os12G0292400 coding for the small chain of Rubisco, the key enzyme catalyzes the xation / assimilation of CO 2 , was downregulated by ~ 5 fold (Fig.   3), providing further indication of compromised photosynthetic carbon production. Therefore, plant NH 4 + excess initiates and probably also develops the disruption of photosynthesis by interfering in the primary reaction and the Calvin Cycle.
As the AMT1 (OsAMT1;1(Os04g0509600/LOC_Os04g43070), OsAMT1;2(Os02G0620500/LOC_Os02g40710) and OsAMT1;3(Os02G0620600/LOC_Os02g40730)) expression level that previously suppressed by high NH 4 + in roots (respectively by 3, 67 and 6 fold) subsequently restored to close to their initial levels (at 1 mM NH 4 + ) with the supplement of sucrose to the high NH 4 + hydroponics (Fig. 5c). This implied a release of ammonium transporting activity facilitated by the supply of sucrose, thus to contribute to enhanced NH 4 + accumulation in roots under high NH 4 + plus sucrose condition, whereas the reduced NH 4 + content under the same condition in shoots indicated probably the e cient utilization of NH 4 + upon the addition of sucrose (Fig. 5b). Meanwhile the GS (Fig.   5d) and GOGAT (Fig. 5e) activities were respectively enhanced by 17 % (GS) and 29% (GOGAT) in roots following the sucrose feeding treatments, indicating a restoration of NH 4 + assimilation activities from initial suppression by NH 4 + excess.
Upon the compensation of sucrose source, the total ROS contents in both roots and shoots were lowered down by 20-30%, close to the levels determined at control (1 mM NH 4 + ) conditions (Fig. 6a). Accordingly, the GSH content and GST activity were signi cantly reduced to the initial levels (at 1 mM NH 4 + ), no longer showing strong induction by NH 4 + excess (Fig. 6b, c). Unexpectedly, no signi cant changes were observed with the activities of classical defense enzymes CAT, POD and SOD under either treatment (Fig.  6d, e, f). Together with the gene expression analyses (Fig. 2), our results demonstrated that the activation of GSH reducing pathway is probably a featured response of rice in dealing with NH 4 + excess and ROS accumulation. Finally, in consistent with the decreased level of ROS, Rubisco activity was elevated by 24% (compared with high NH 4 + ) in shoots with the presence of sucrose feeding (Fig. 6g), suggesting enhanced e ciency of primary CO 2 xation activity.
Taken together, this set of experiments indicated that sucrose feeding could effectively alleviate rice plant from carbon scarcities exerted by internal NH 4 + excess and ROS stresses.

Discussion
Due to particular water-ooding and anaerobic growth environment, NH 4 + is the major form of N nutrient available to paddy rice. Current farming practice of the overuse of N fertilizers to satisfy the N demand of high-yielding rice creates a persistent circumstance of NH 4 + excess around the roots and inside the plant for rice to cope with [10-12, 38, 39]. Therefore, a study focuses on such farming-intervened special circumstances would be helpful in discovering 'bottlenecked' constraints and adaptation strategies related to the (molecular) physiological and agronomic respects of N performances in rice.

NH 4 + stresses and toxicities have been considered as a major human-intervened environmental distress
exerted on plants and attracted extensive research interests. Researches on these topics have focused on the identi cation of mechanisms or pathways that primarily modulate the biological modi cations of root architectures [20][21][22][23][24]. Solid evidences have shown the re-con guration of plant root morphology in response to NH 4 + stresses is tightly controlled through the interactions with plant hormone signaling pathways [13,14,24,25,27]. Whereas NH 4 + toxicities could be attributed to ion imbalances [20], intracellular pH disturbance [40], energy consumption due to invalid NH 4 + cycles in roots [13,41].
Assessments of NH 4 + stress responses in plant roots and its biological toxicities, to a great extent, rely on the establishment of measurable growth phenotypes that requires effects or reactions to accumulate for a desired time course. These analyses are obviously important in addressing the mode-of-action of physiological effects or processes developed along the progresses of the treatments. To the other hand, since plants grow and develop during the experimental periods, these accumulative observations might be not satisfactory for capturing the initial reactions or the nature of NH 4 + excess stresses. Therefore, in the present work, we established prompt methods for identi cation of the early responses especially to internal NH 4 + excess without causing visible growth changes by blocking the primary NH 4 + assimilation with MSX.
Our results demonstrate that internal NH 4 + excess (accumulation of free NH 4 + ) correlates to a burst of ROS radicals (Fig. 1c, d, e) and activates (may be speci cally) the GST-mediated GSH cycling rather than classical antioxidant defense enzymes (such as CAT, POD and SOD) for ROS scavenging (Fig. 2, 6).
Therefore, our results support the notion that the initial nature of NH 4 + excess in rice plant is probably the induction of ROS bursts that triggers the downstream reactions.
Following the ROS radical bursts, genes involved in photosynthetic photon caption and the most important enzymes for primary CO 2 assimilation-Rubisco (gene expression and enzyme activity as well) were suppressed (Fig. 3, 6g), indicating the carbon income is reduced from its origin of production. Next, to support the sharp energy demand of ROS cleavage through the GSH cycling, we observe strong inductions especially in shoots, of genes involved in ATP-generating reactions (glycolysis and glycogen breakdown, Fig. 4) and enhanced gene expressions related to processes of gluconeogenesis and structural sugar breakdown to ll up the consumption of 'active' carbons (Fig. 4). To this end, the carbon consumption and thereby the carbon scarcity associated with NH 4 + excess in rice plant resemble the metabolism alterations caused by well known abiotic stresses, such as drought or salinity stresses described in other plant species [32,34].
The oxidative inhibition of photosynthesis and the scarcity in active carbon source initiated by internal NH 4 + excess are further supported by the experiments of sucrose feeding (Fig. 5, 6).  (Fig. 5c) and reduction of GS and GOGAT activities (Fig. 5d, e) provide further evidence of such feedback regulation in rice. Here, upon the carbon compensation by sucrose feeding, the NH 4 + assimilation activities restore to normal rates (normal NH 4 + , 1mM) and the AMT expression levels are accordingly enhanced (Fig. 5c, d, e), supportedly suggesting that carbon scarcity may be the major cause that leads to feedback inhibition of NH 4 + uptake.
This speculation could nd further supports by the reduction of free NH 4 + and ROS contents to the normal levels as observed in the control (1 mM NH 4 + ) conditions upon a 24 h of sucrose feeding (Fig. 5b,   6a). With the tranquilness of ROS burst, enhanced GSH content and GST activity are no longer essential (Fig. 6b, c) and the Rubisco CO 2 xation enzymes restore to act at normal (or even facilitated) activity ( Fig. 6g). Thus, the NH 4 + excess stress caused growth inhibition as evidenced elsewhere could be interpreted by the carbon scarcity resulted from the reduction of carbon synthesis and violent carbon consumption for energy production.

Conclusions
The present work reveals that the essential nature of internal NH 4 + excess stresses in rice plant is closely correlated to its accompanying ROS bursts. Coping with ROS bursts by energy and carbon consumption processes leads to carbon scarcity that accumulatively inhibits plant growth. Sucrose feeding alleviates stress responses, indicating that enhancement of carbon income processes could be a useful approach for further improving nitrogen performances of rice under current farming practice of N fertilizer overuse.

Plant growth and treatments
Rice seeds of Oryza. sativa ssp. Japonica Nipponbare were obtained from Prof. Yingguo Zhu's group, College of Life Sciences, Wuhan University. The seeds were sterilized, germinated and seedlings were grown in a growth chamber according to previously described [31]. The growth chamber was set with 16/8 h day/night, 27/25℃,day/night; light intensity was 400 mol m -2 s -1 , relative humidity was set at 70%. Seedlings were grown in the IRRI solution [31]. The pH of hydroponics was buffered to 5.7 with 10 mM MES and renewed every 48h. For long-term growth tests, uniform seedlings of 7 d were treated with either 1 mM (control) or 20 mM (high NH 4 + ) NH 4 Cl supplemented to nitrogen-free IRRI solutions for further 14 days with daily refreshment of the culture solutions. To achieve a rapid NH 4 + excess in rice plants without causing a growth discrepancy, so that NH 4 + excess-responsive genes could be analyzed at the early stages of responses, 10-d old seedlings were promptly treated with 'control' or 'high NH 4 + ' (see above) for 4 h in the presence of 1 mM methionine sulfoximine (MSX, a potent inhibitor of glutamine synthetase) to block the assimilation of NH 4 + .
For carbon source feeding experiments, 14-d old seedlings were treated with control (1 mM NH 4 + ) or high NH 4 + (20 mM NH 4 + ) in IRRI solution in the presence of 1% (w/v) sucrose for 24 h. To avoid the burst of microbes associated with sucrose-containing hydroponics, antibiotics penicillin (50 mg L -1 ) and chloramphenicol (25 mg L -1 ) were included to the culturing solution according to Lejay's description [46]. Also in order to prevent possibly undesired impacts, the treatment was limited to within 24 h.

RNA-Seq and quantitative real-time PCR analyses
Total RNAs from treated root or shoot samples was extracted with TRIzol total RNA extraction kit (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. For RNA-Seq analysis, RNAs from control (1mM NH 4 + or high NH 4 + plus 1 mM MSX treated (4 h) tissue samples were used for library construction and sequencing. Data extraction, identi cation of differentially expressed genes (DEGs) and functional annotation were analyzed according to our previous work [31]. DEGs were designated with expression fold changes greater than 2 (p < 0.05) between the rapid NH 4 + accumulation (high NH 4 + + MSX) and the control conditions. Quantitative real-time PCR (qRT-PCR) analyses was carried out to reveal possible responses at the gene expression level related to special conditions such as NH 4 + excess stress or sucrose feeding treatments.
About 1 μg of total RNA was used to synthesize first-strand cDNA using thePrimeScript™ RT Master Mix (Perfect Real Time, TaKaRa, Japan) according to the manufacturer's description. Primer sequences used for qRT-PCR were listed in Additional le4. Thermocycling and fluorescence detection were performed with C1000 Thermal Cycler CFX96 Real-Time System (Bio-Rad) using the SYBR Premix Ex Taq (TaKaRa, Japan) as indicated by the manufacturer's protocol. The reaction was performed under the following conditions: 95 °C for 30 s, followed by 44 cycles of 95 °C for10 s, 60 °C for 15 s and 72 °C for 15 s. For fold change analysis, gene expression abundance was quantified with -2 ΔΔCt and normalized against the internal OsActin gene. PCR amplifications were repeated three times using cDNA templates synthesized from three independent plant samples.
Tissue free NH 4 + , GSH and sucrose content assays For antioxidative enzyme activity analyses, 0.2 g of fresh root or shoot samples were ground in liquid N 2 , homogenized and crude extracts were used for the measurements of CAT, POD and SOD activities as previously described [32]. The speci c activity of GST was assayed in the supernatant by following the increase of absorbance at 340 nm using GST Assay Kit according to the manufacturer instructions (CS0410, Sigma, USA). One unit of activity was defined as the amount of enzyme required to form 1 M product per minute at 30℃.Enzyme activities were expressed as U. mg -1 FW.

Measurement of GS, GOGAT and Rubisco activities
To prepare the crude enzyme extracts, roots or shoots of each sample were ground into ne powder with liquid N 2 and homogenized with 50 mM Tris-HCl buffer (pH 7.6, containing 10 mM MgCl 2 , 1mM EDTA, 1mM β-mercaptoethanoland 4% (w/v) polyvinylpolypyrrolidone-40) using a chilled pestle and mortar. The homogenate was centrifuged at 15000 g for 30 min at 4℃and the supernatants were used for the determination of enzyme activities. The glutamine synthetase (GS) activity was measured according to Sakurai's description [54]. One unit of GS activity was expressed as the amount of enzyme catalyzing the formation of 1 mol γ-glutamyl hydroxamate per min at 37℃ [55]. The glutamate synthase (GOGAT) activity in the supernatants was determined by the conversion of 2-ketoglutarate to glutamate in a reaction mixture containing 200 mM KH 2 PO-KOH pH 7.5,10 mM glutamine (Gln), 10 mM 2-ketoglutarate, 0.14 mM NADH [56], One unit of GOGAT activity was defined as the oxidation rate of 1 nmol NADH per min at 30℃. And the Rubisco activity was measured according to [57]. One unit of Rubisco activity was de ned as the oxidation rate of 1 nmol NADH per min at 25℃.

Statistical analysis
Experiment data were expressed as means ± S. E. M. of 3 independent replicates. Statistical differences were evaluated by Duncan's or t-test with SPSS 13.0 and the level of statistically signi cant difference was set at p 0.05.

Supplementary information
Additional les Additional le 1: Table S1. Summary of total DEGs identi ded in rice roots and shoots following a 4 h rapid NH 4 + accumulation treatment.