Wild-type lines of rice (Japonica, Hwayoung), T-DNA insertion mutant cipk18, and three complementary strains of cipk18 (com1, com2, com3) were used in this study. The OsCIPK18 T-DNA insertion line (1C-05857) in the Hwayoung background was obtained from Kyung Hee University, Korea (Jeon et al., 2000). To generate the complemented lines, the 35S::OsCIPK18 fragment was constructed and transformed into Agrobacterium tumefaciens strain EHA105. Calli from cipk18 were transformed as described by Cheng et al. (1998). Seeds used in this study were surface-sterilized with 1% sodium hypochlorite for 10 min, washed extensively with distilled water, and then incubated at 37°C in the dark for 2 days to break dormancy. Consistently growing germinating seeds were, respectively, transferred to two groups of hydroponic media for growth, namely, control (pure water) and ammonium toxicity treatment (4 mM NH₄Cl solution) with no other nutrients present in the hydroponic media. All seedlings subjected to hydroponic treatments were grown in a temperature-controlled incubator set at 28°C, 14 h light/22°C, 10 h dark. All mature plants were grown and harvested in Wuhan, Hubei, and Lingshui, Hainan, China.

The roots of seedlings of WT, cipk18, com1, com2, and com3 cultured for 3 days under ddH₂O and 4 mM NH₄Cl, respectively, were photographed at high resolution with a Nikon D7100 digital SLR camera, and root length data were obtained using smartRoot in ImageJ (Lobet and Draye, 2011). Seedlings of WT and cipk18 that had been cultured for 7 days under control and NH₄Cl treatments were selected and divided into groups of five plants to measure biomass, and the average value of individual plants was calculated and repeated three times. The roots of seedlings were drained and weighed directly to obtain fresh weight data, dried in an oven at 70°C for 3 days until their constant weight was obtained, and then weighed again to obtain dry weight data. The roots of 7-day-old rice seedlings were spread as far as possible, and high-resolution photographs were taken with a Nikon D7100 digital SLR camera to obtain root length, diameter, and number data using smartRoot in ImageJ.

The roots of 7-day-old rice seedlings were removed from the culture medium, rinsed three times with water, dried with paper towels to absorb the water, ground to powder with liquid nitrogen, weighed 0.1 g of the powder, and treated with 1 ml of 10 mM formic acid solution. After mixing thoroughly, they were centrifuged at 10,000 × g at 4°C for 10 min. Then, 1 ml OPA (100 mM KH₂PO₄, 100 mM K₂HPO₄, 3.75 mM o-phthalaldehyde, 2 mM β-ME) was added to 250 μl supernatants and reacted in the dark for 30 min, the absorbance was measured at 410 nm using infinite M200 Enzyme Scale. A standard curve was plotted with different concentrations of NH₄Cl solution, and then it was used to calculate the concentration of tissue NH4+.

The roots of WT and cipk18 seedlings cultured for 7 days under two treatments described above were ground to powder with liquid nitrogen. Then, 0.1 g of tissue was used to measure GS and GOGAT activity according to the method described on the kit (cominbio). Absorbance value measurement was done using infinite M200 Enzyme Scale.

To verify the expression changes of genes in WT and cipk18 under ammonium toxicity, we extracted total RNA from roots of rice seedlings under ddH₂O and 4 mM NH₄Cl treatment and performed qRT-PCR using specific primers. RNA was extracted as one biological replicate from the roots of approximately 5–8 seedlings grown for 7 days in the same growth state under both treatments. We rapidly ground the treated roots in liquid nitrogen, extracted total RNA from rice samples using TRIzol (Invitrogen), removed genomic DNA from total RNA using DNase I (Thermo Scientific), and performed cDNA first-strand synthesis using M-MLV reverse transcriptase (Promega). qRT-PCR was done using 2× ChamQ SYBR qPCR Master Mix (UE), specific primers P+/P–, and cDNA using the Bio-Rad CFX96TM Real-Time System. The relative expressions were calculated using the 2^−ΔΔCT method. Three technical replicates were used to calculate the mean of the expression levels for each biological replicate, and three biological replicates were used to generate the mean of the expression levels for each sample. Graphing and significance analysis were performed using GraphPad Prism 6.

The roots of WT and cipk18 seedlings treated for 7 days under two treatments as described previously were divided into three replicates to prepare specific RNA-seq libraries, respectively, for a total of 12 libraries. The libraries were submitted to BGI for sequencing using the Illumina HiSeq 2000 platform. The raw data were filtered using trimmomatic (Bolger et al., 2014) to remove the low-quality reads, and the resulting high-quality reads were aligned to the NIP reference genome (ftp://ftp.ensemblgenomes.org/pub/plants/release-44/fasta/oryza_sativa/dna/) using STAR software (Dobin et al., 2013). After alignment, the raw counts were normalized to trimmed mean of M value (TMM) using RSEM (Li and Dewey, 2011). Difference analysis between the two samples was performed using DESeq2, where genes with p_adj < 0.05 and |log2FC| > 1 were identified as differential genes. Gene Ontology (GO) enrichment analysis was performed using TBtools, and the GO background was provided on the AgriGO V2.0 website (Tian et al., 2017; Chen et al., 2020). Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis was done using the R package from clusterprofiler (Yu et al., 2012).

Transcription factor prediction was performed on the plantregmap website (http://plantregmap.gao-lab.org/), and the predicted results were filtered. TF network visualization was done by Cytoscape.

All statistical analyses were conducted using GraphPad Prism 6. The error bars in all the charts represent the standard deviation of the mean. The different letters above the error bar represent significant differences between groups. The comparison method uses one-way ANOVA combined with Duncan's post-hoc multiple test method. p < 0.05 was set as the significance cutoff.

To screen the specific CIPK associated with ammonium toxicity in rice, we compared the ammonium resistance of WT with the mutant of OsCIPK T-DNA insertion. Ammonium resistance was quantified based on the phenotype of the relative root growth (i.e., root growth in the presence of high NH4+ compared with root growth in water as control). Growth inhibition but relatively enhanced ammonium resistance was identified in a mutant (1C-05857), in which ammonium toxicity only shortened its root length by about 20%, but by about 40% in WT (Figures 1A–C). This mutant carried a homozygous T-DNA insertion in the 5′ untranslated region (UTR) of OsCIPK18 (Os05g0332300), leading to a knockdown of the OsCIPK18 transcript (Figures 1D,E, Supplementary Figure 1). We obtained three OsCIPK18 complementation lines (com1, com2, and com3) and confirmed that the increased ammonium resistance was attributed by the mutation of OsCIPK18 (Figures 1A–C,E). With increasing incubation time, cipk18 consistently showed resistance to excess NH4+ in terms of fresh weight and root length compared to WT and complementary lines (Supplementary Figure 2).

Furthermore, we investigated the development of root system between WT and the mutant cipk18 under control and ammonium stress conditions. Exposing to excess NH4+, WT exhibited a significant decrease in biomass, root length and root number, and an increase in root diameter, which indicated that external high NH4+ severely affects rice root development of WT (Figures 2A–H). However, ammonium toxicity did not cause the same changes in root diameter and root number in the mutant cipk18. The effects of ammonium toxicity on root length of cipk18 were diminished, and the fresh and dry weight also showed stronger ammonium resistance compared with WT (Figures 2A–H). Together, these results suggest that cipk18 was more resilient to ammonium toxicity than WT, and protein kinase OsCIPK18 might be involved in the adjustment of the root structure of rice seedlings by ammonium toxicity.

The excessive accumulation and assimilation of NH4+ in plants are considered as the important causes accounting for ammonium toxicity (Chen et al., 2013). To estimate the accumulation and assimilation of NH4+ in the roots of both WT and cipk18, we examined NH4+ content and enzymatic activity of GS/GOGAT under the control and ammonium stress conditions (Figures 3A–C). When only water was provided, the roots of both WT and cipk18 had low and comparably basal amounts of NH4+ (Figure 3A). When excessive NH4+ was provided, a significant accumulation of NH4+ was observed in both lines (WT and cipk18). But the amount of NH4+ in the cipk18 roots was significantly lower than that in the WT (Figure 3A). The enzymatic activities of GS and GOGAT for NH4+ assimilation in roots of both WT and cipk18 were significantly increased due to external high NH4+. But there was no significant difference in the enzymatic activity between WT and cipk18 (Figures 3B,C). These data indicated that the observed differential ammonium resistance between WT and cipk18 lines could be a result of differences in root NH4+ accumulation, which was not associated with GS/GOGAT-dependent assimilation.

The difference in NH4+ accumulation in roots prompted us to investigate the changes in transcript abundance of OsAMT1s that encoded ammonium transporters. The expression of OsAMT1;1 and OsAMT1;3 was significantly suppressed to the same extent in both WT and cipk18 by external high NH4+ (Figures 3D,F). The expression of OsAMT1;2 stayed unchanged in WT but was significantly downregulated in cipk18 upon the external high NH4+ (Figure 3E). These findings suggest that the differences in NH4+ accumulation between WT and cipk18 lines may be the result of direct or indirect involvement of OsCIPK18-mediated signaling in the transcriptional regulation of OsAMT1;2. Furthermore, this correlation prompted us to focus on the transcript changes in WT and cipk18 during high NH4+ stress and identify the key genetic elements involved in differential ammonium resistance between the two lines.

Transcriptome comparisons were performed between the control and ammonium stress treatments. In the WT line, a total of 2,538 excess NH4+-induced differentially expressed genes (DEGs) were identified, of which 674 DEGs were upregulated and 1,864 DEGs were downregulated in the presence of excess NH4+ (Figure 4A). In mutant cipk18, 4,066 excess NH4+-induced DEGs were identified, of which 43.6% (303 + 1,468 out of 4,066) had the same expression change trend in WT (Figure 4A). These overlapping DEGs most likely reflect regulatory networks induced by excess NH4+ independent of OsCIPK18. Notably, there were 730 (337 + 393) genes differentially expressed only in WT but not in cipk18, named WT-specific DEGs, reflecting ammonium toxicity-induced changes in gene expression dependent on OsCIPK18 (Figure 4A, Supplementary Table 1). These transcript changes may be one of the reasons for the phenotypic differences between WT and cipk18. In addition, 2,258 (727 + 1,531) DEGs were uniquely induced by high NH4+ in cipk18 but not in WT, i.e., cipk18-specific DEGs, possibly reflecting an additional mechanism of ammonium resistance caused by knockdown of OsCIPK18 (Figure 4A, Supplementary Table 2). It is necessary to be concerned that there were 270 DEGs between WT and cipk18, even in the control conditions (Supplementary Figure 3A). These DEGs demonstrate important regulatory roles for OsCIPK18 in various signaling and physiological pathways in rice roots (Supplementary Figures 3C,D), which could explain the phenotypic differences observed in cipk18 and WT lines (Figures 1A, 2A). There were 4.9% WT-specific DEGs (36 out of 730) and 2.1% cipk18-specific DEGs (48 out of 2258) overlapped with DEGs between WT and cipk18 (Supplementary Figure 3B), and these overlapped DEGs may constitute factors that influence the response of cipk18 to ammonium toxicity.

Ammonium could be transported through simple osmotic diffusion, non-selective cation channels, and potassium transport channels (Britto et al., 2014; Bittsánszky et al., 2015). Our previous expression analysis of AMT1 subfamily members had demonstrated that OsCIPK18 regulated NH4+ influx by altering the expression of OsAMT1;2 (Figure 3E). Using RNA-seq, we further identified four K⁺ transporters, OsHAK7 (Os07g0669700), OsHAK18 (Os09g0563200), OsHAK22 (Os07g0102100) (Banuelos et al., 2002), and OsHKT1 (Os06g0701700) (Yao et al., 2010), a plasma membrane H⁺-ATPase OsA1 (Os03g0689300), as well as an outward-rectifying shaker-like potassium channel OsSKOR (Os04g0445000) (Kim et al., 2015) in cipk18-specific DEGs (Figure 5). In cipk18, four K⁺ transporters were uniquely downregulated by excess NH4+ to reduce NH4+ uptake; meanwhile, OsSKOR was uniquely upregulated to increase NH4+ efflux (Figure 5). OsA1 could cooperatively improve N and C utilization and facilitates ammonium absorption in rice (Zhang M. et al., 2021). Its expression was significantly downregulated by excess NH4+ in cipk18, which may be one of the pathways through which OsCIPK18 regulates NH4+ uptake (Figure 5). In summary, we considered that the lower amount of NH4+ in cipk18 could be the result of the simultaneous downregulation of OsAMT1;2, OsA1, OsHAK7/18/22, OsHKT1, and upregulation of OsSKOR. To further investigate how OsCIPK18 functions in the inhibition of root growth by NH4+, we investigated the functional distribution of WT-specific and cipk18-specific DEGs under ammonium toxicity, respectively, and tried to find the key genes.

WT-specific DEGs induced by excessive NH4+ differentially expressed only in WT represent an OsCIPK18-dependent ammonium toxicity response process. We explored the function of WT-specific DEGs based on GO and KEGG classification (Figures 4B,D). A total of 25 GO pathways and 10 KEGG pathways were enriched, mainly including transmembrane transporter activity, ion binding, vesicle, oxidoreductase activity, starch and sucrose metabolism, amino acid metabolism, and biosynthesis of small molecules, such as isoquinoline alkaloid and ubiquinone (Figures 4B,D).

Transcription factors, as key factors coordinating the expression of downstream genes involved in metabolic and developmental pathways, are important players in the response to ammonium toxicity (Kikuchi et al., 2000; Huang et al., 2015; Gu et al., 2017). Consistent with this, we found that multiple TF family members, such as AP2/ERF, MYB, WRKY, and NAC, were induced to be upregulated or downregulated upon exposure to excess NH4+ (Supplementary Table 1). In WT-specific DEGs, according to the predicted TF network using PlantRegMap, we found that five core TFs, OsERF130 (Os05g0497200), OsWRKY74 (Os09g0334500), OsWRKY16 (Os01g0665500), OsNIGT1 (Os02g0325600), and OsGLK1 (Os06g0348800), had possible interactions with multiple genes among the WT-specific DEGs (Figure 6A, Supplementary Figure 4, Supplementary Table 3). Among them, OsWRKY74 (Dai et al., 2016), whose expression was downregulated by NH4+ in WT, affected the elongation of roots and the increase of crown root number. This led us to speculate that ammonium toxicity signals might be transmitted through OsCIPK18 to OsWRKY74 to influence root development. Overall, this network revealed some key TFs that may be involved in root growth regulated by OsCIPK18 under NH4+ stress and regulatory relationships in WT-specific DEGs.

In addition, genes associated with ABA signaling in WT-specific DEGs attracted our attention. OsRePRP2.1 (Os07g0418700) and OsRePRP2.2 (Os07g0418600) were reported to be repressors of root cell expansion and were induced by ABA to be expressed in the elongation zone of roots (Tseng et al., 2013). In this study, their expression was significantly upregulated in WT (Figures 7A,B). Meanwhile, OsSAPK6 (Os02g0551100), encoded stress-activated protein kinase and involved in ABA signaling (Kobayashi et al., 2004), was uniquely downregulated in cipk18 (Figures 7A,B). The expression changes of these genes implied that excess NH4+ activated ABA signaling and increased the expression of OsRePRP2.1/2.2 to inhibit root elongation. In contrast, due to the deletion of OsCIPK18, the expression of the ABA signaling factor OsSAPK6 was downregulated and the expression of OsRePRP2.1/2.2 was no longer upregulated, allowing the mutant cipk18 to show greater resistance to ammonium.

In this study, cipk18-specific DEGs induced by excess NH4+ in cipk18 were considered as components of the ammonium resistance mechanisms. Based on the GO and KEGG classification, the function of cipk18-specific DEGs is mainly related to ion transport, transmembrane transporter activity, cell wall macromolecule metabolic process, plant hormone signal transduction, cellular catabolic process, glutathione metabolism, sugar catabolic process, and photosystem (Figures 4C,E).

In cipk18-specific DEGs, we identified six key TFs that are likely to be the regulatory center of this part network, including OsERF54 (Os01g0657400), OsMYB2P-1 (Os05g0140100), OsWRKY96 (Os12g0507300), OsNAC29 (Os08g0115800), OsNAC73 (Os01g0672100), and ARF5 (Os04g0664400) (Figure 6B, Supplementary Figure 4, Supplementary Table 4). Among these genes, OsMYB2P-1, an R2R3 MYB TF engaged in phosphorus starvation response and regulation of root architecture in rice, was uniquely downregulated by NH4+ in cipk18 (Dai et al., 2012). OsNAC29, a top-layer TF for secondary wall formation (Huang et al., 2015), was uniquely downregulated by NH4+ in cipk18, which was consistent with the cellulose synthesis-related pathway was enriched by GO classification. We, therefore, hypothesized that OsCIPK18 is associated with cell wall formation under ammonium toxicity.

Notably, OsARF5 is an auxin response factor whose expression is upregulated by excess NH4+ in the mutant cipk18 (Figures 7A,B). This reminds us to be concerned about the role of auxin signaling in ammonium toxicity response. As we all know, ammonium feeding has been shown to suppress root auxin content (Kudoyarova et al., 1997; Britto and Kronzucker, 2002). In cipk18, we revealed that exposure to high NH4+ resulted in the upregulation of auxin synthesis and response gene expression, including the auxin influx transporter OsABCB14 (Os04g0459000) (Xu et al., 2014), the IAA synthetic pathway gene OsYUCCA6 (Os07g0437000) (Yamamoto et al., 2007), and three auxin-responsive SAUR gene family members [OsSAUR12 (Os02g0769100), OsSAUR21 (Os04g0617050), OsSAUR58 (Os12g0626200)] (Figures 7A,B) (Jain et al., 2006; Zhang T. et al., 2021). These results implied that knockdown of OsCIPK18 may have prevented the inhibition of auxin synthesis in roots by excess NH4+, allowing upregulation of auxin synthesis and response genes, thereby enhancing the resistance to ammonium toxicity in cipk18. Meanwhile, we found that OsARF3 (Os01g0753500) (Wang et al., 2007) was uniquely downregulated by excess NH4+ in WT-specific DEGs (Figures 7A,B), further supporting this hypothesis. Therefore, we hypothesized that seedling roots inhibit auxin signaling via OsCIPK18 in response to ammonium toxicity signals, resulting in inhibition of root growth in rice.

As a result of human intervention in the nitrogen cycle, including increasing soil nitrogen input and irrational fertilization practices in the biosphere, plants have to deal with unprecedented NH4+ stress (Gerendás et al., 1997; Britto and Kronzucker, 2002). Even though rice is a recognized ammonium-tolerant species and highly adapted to NH4+ as a nitrogen source, it is still threatened by ammonium toxicity (Balkos et al., 2010). Our study showed strong evidence that WT exhibited significant growth inhibition when it was grown under excessive NH4+ treatment, with few and short thick roots and reduced biomass (Figures 1, 2). However, a T-DNA insertion mutant cipk18 exhibits enhanced resistance in response to ammonium toxicity as both biomass and root architecture showed that the inhibition of high NH4+ on roots growth in cipk18 was less than that in WT (Figures 1, 2).

The CIPK family interacts with members of the CBL family in response to the Ca2+-mediated signaling pathway (Straub et al., 2017). OsCIPK18, as a member of the CIPK family, must also play an important role in rice growth and development, but this has not been reported yet. In this study, the mutant cipk18 exhibited significant growth inhibition compared with WT (Figures 1, 2). Using GO and KEGG analysis, the 270 DEGs that were identified between WT and cipk18 were associated with biological functions, such as signaling, response to stimuli, ion binding, and nitrogen metabolism (Supplementary Figures 3A,C,D). This discovery facilitates the study of the effects of OsCIPK18 in rice, but this study focuses on the role of OsCIPK18 in response to ammonium toxicity. The mutant cipk18 shows NH4+-resistance phenotype, but mutants of two other members of the CIPK family, OsCIPK9 and OsCIPK23, were sensitive to NH4+ (Xuan et al., 2019), which suggested that the function of OsCIPK18 in response to ammonium toxicity might be opposite to that of OsCIPK9 and OsCIPK23.

Traditional theories suggest that high intracellular ammonium concentration is one of the major causes of ammonium toxicity in higher plants, and there is also emerging evidence that acidic stress caused by excessive NH4+ assimilation is the primary cause (Hachiya et al., 2012, 2021; Esteban et al., 2016). In this study, intracellular NH4+ accumulation and GS/GOGAT-dependent NH4+ assimilation in roots were significantly increased by excess NH4+ in both WT and the mutant, but relatively low NH4+ accumulation was observed in cipk18 (Figures 3A–C), suggesting that OsCIPK18 is involved in the NH4+ uptake without affecting NH4+ assimilation. In the presence of elevated external NH4+, NH4+ influx dependent on the high-affinity transport system is downregulated to prevent ammonium toxicity (Kronzucker et al., 2001; Beier et al., 2018; Kumar et al., 2020). For example, CBL1-CIPK23 phosphorylates and inactivates AMT1;1/1;2 in response to ammonium stress, thereby reducing NH4+ transport (Straub et al., 2017). In this study, expression of OsAMT1;1 and OsAMT1;3 was significantly downregulated by excess NH4+ in both WT and mutant (Figures 3D,E). However, in cipk18, the expression of OsAMT1;2, OsHAK7, OsHAK18, OsHAK22, OsHKT1, and OsA1, which are reported to transport NH4+, were all inhibited by excess NH4+ (Figures 3E, 5A,B). Changes in the expression of these genes coincided with reduced ammonium accumulation in the mutant, suggesting that OsCIPK18 plays an important role in maintaining the expression of ammonium transporters at the transcriptional level during the response to ammonium toxicity.

Apart from the regulation of NH4+ uptake, it is not known how OsCIPK18 regulates root growth under ammonium stress. Transcriptome-wide analysis facilitated our observation of high NH4+-induced changes in gene expression associated with root growth at the genome-wide level. Sun et al. used RNA-seq to reveal the spatiotemporal specificity of gene expression in rice after high ammonium treatment for 4 and 12 h, highlighting the role of TFs and phytohormones in ammonium resistance (Sun et al., 2017). In our data, WT-specific DEGs and cipk18-specific DEGs were identified as differential genes involved in response to ammonium toxicity and located downstream of OsCIPK18 (Figure 4A). Strategies for coping with salt stress in plants include adjusting ATP formation and enhancing energy metabolism (Zhao et al., 2013). Genes related to amino acid metabolism, starch, and sucrose metabolism, as well as oxidoreductase, lysozyme, and many other enzymatic activities are regulated by OsCIPK18 under ammonium toxicity (Figures 4B–E). Recent evidence suggests that the biosynthesis of vitamin B6 significantly improves root tolerance to ammonium (Liu et al., 2022). Our data show that genes associated with vitamin B6 binding are also regulated by OsCIPK18 (Figure 4B). The above evidence highlights the fundamental function of OsCIPK18 in the ammonium toxicity response.

Several core TFs were identified in WT-specific and cipk18-specific DEGs that could be directly or indirectly regulated by OsCIPK18 under ammonium stress (Figures 6A,B). Members of the AP2/ERF family are associated with plant defense programs against abiotic stresses and share a conserved DNA-binding domain that activates downstream gene expression by binding specifically to cis-acting elements in the promoters of abiotic stress-responsive genes (Mizoi et al., 2012). In this study, two ERF family genes (OsERF54, OsERF130) were found to be at the center of the transcriptional regulatory network involved in ammonium toxicity by OsCIPK18 (Figures 6A,B), suggesting that OsCIPK18 might regulate the expression of ERF genes to further regulate downstream genes in response to ammonium toxicity (Figure 8). WRKY TFs in rice are rapidly induced in response to abiotic stresses, such as salinity, aluminum, drought, and osmotic stress, to regulate developmental processes, such as seed development, root growth, and leaf senescence (Ross et al., 2007). Three key genes (OsWRKY16, OsWRKY74, and OsWRKY96) belonging to the WRKY family were identified in WT-specific DEGs or cipk18-specific DEGs (Figures 6A,B). Among them, OsWRKY74 was reported to be involved in the response of Pi starvation, Fe starvation, as well as cold stress, and promote the elongation of primary and adventitious roots (Dai et al., 2016). In this study, inhibition of OsWRKY74 by excess NH4+ was lost in the mutant, suggesting that OsWRKY74 lies downstream of OsCIPK18 in regulating root growth under ammonium toxicity. In addition, members of the NAC, MYB, and other families were identified to be regulated by OsCIPK18 and involved in the inhibition of root growth in response to ammonium toxicity. These TFs and their interacting functional genes together form a response network of excess NH4+ regulated by OsCIPK18, which will contribute to the understanding of the process of ammonium toxicity production.

Phytohormones play an integral role in the regulatory network of response to ammonium toxicity in plants (Zou et al., 2013; Lei et al., 2016; Di et al., 2018; Sun et al., 2020). Excessive ammonium supply reduced free IAA content in the roots and significantly accelerated tissue ABA accumulation (Di et al., 2018; Sun et al., 2020). In the mutant cipk18, an auxin biosynthetic gene OsYUCCA6, an auxin influx transporter OsABCB14, and auxin-related response genes (OsARF5, OsSAUR12, OsSAUR21, and OsSAUR58) were specifically upregulated by excess NH4+ (Figures 7A,B). This demonstrated that both auxin synthesis and response pathway were activated in cipk18, suggesting that excessive ammonium suppressed root growth through direct or indirect regulation of the auxin synthesis gene by OsCIPK18, reducing IAA content and inhibiting the auxin response pathway (Figure 8). In the ABA signaling pathway, OsRePRP2.1 and OsRePRP2.2 were able to be activated by ABA and inhibited cell expansion (Tseng et al., 2013), and OsSAPK6 expression was significantly increased by ABA treatment (Chae et al., 2007). Under ammonium stress, OsSAPK6 was uniquely downregulated, and OsRePRP2.1/2.2 were no longer upregulated in cipk18, implying that knockdown of OsCIPK18 results in blocked ABA signaling, thereby reducing the inhibitory effect of ammonium toxicity on root growth (Figures 7A,B). These data reveal that OsCIPK18 is a key node in hormone signaling under ammonium stress, which provides great insight into the involvement of key phytohormones in rice in response to NH4+ stress.