Rice bZIP11 acts as a transcriptional repressor that negatively regulates photosynthesis in response to nitrogen deficiency · New Phytologist · El Autómata
Lunes, 16 de noviembre de 2026
Portada de las 21:20
Todos los periódicos y la radio, en unoEscrito por IA a partir de 26 medios · Cómo se hace
Biología y genética
New Phytologist · texto original
Rice <fc>bZIP11</fc> acts as a transcriptional repressor that negatively regulates photosynthesis in response to nitrogen deficiency
Phenotypes of 43-d-old wild-type (WT), Osbzip11-KO, and OsbZIP11-OX seedlings grown under control N and low N conditions.
Nitrogen (N) is an essential macronutrient for plants. Plants primarily acquire N in the forms of nitrate and ammonium from the soil via an effective transport system that includes both high- and low-affinity transporters (Li et al., 2017). Once absorbed, these N compounds are assimilated into glutamine, which, in turn, is used to synthesize all other N-containing organic compounds, including nucleic acids, amino acids, and Chl pigments (Xu et al., 2012). A significant portion of the N acquired from the soil is devoted to the biosynthesis of photosynthesis-related proteins, such as ribulose 1,5-bisphosphate carboxylase/oxygenase (RubisCO). Consequently, a N deficit in the soil leads to a decline in photosynthetic activity, resulting in decreased biomass and reduced crop yield (Mu & Chen, 2021). As N availability varies temporally and spatially across environments, plants, which are sessile organisms that must survive in the location in which they germinate, need to modulate their gene expression patterns, metabolism, root structure, and N source uptake activity to optimize adaptation to the N environment (Kiba & Krapp, 2016; Ueda et al., 2017; Jia & von Wirén, 2020).
Over the past few decades, many studies have advanced our understanding of the adaptative responses of land plants to changing levels of nitrate, their major N source (Delgado et al., 2024). NODULE INCEPTION (NIN)-LIKE PROTEIN (NLP) transcription factors (TFs) act as nitrate sensors and are thus pivotal transcriptional activators that are responsible for nitrate-responsive changes in gene expression, hormone balance, and metabolism (Yan et al., 2016; K.-H. Liu et al., 2017; Liu et al., 2022; Maeda et al., 2018; Abualia et al., 2022; Saito et al., 2022; Durand et al., 2023). Our knowledge of the mechanisms that underlie plant adaptations to N deficiency, known generally as N-deficiency responses (NDRs), remains incomplete; however, even though a full understanding is essential for improving plant biomasses and crop yields, particularly under the N-deficient conditions prevalent across most terrestrial areas (Du et al., 2020). NDRs include reductions in photosynthetic activity and leaf chlorosis accompanied by the degradation of high molecular weight, N-containing organic compounds, such as proteins and Chl pigments, the translocation of N sources from old leaves to developing tissues, enhanced uptake of N sources due to increased expression of genes encoding nitrate transporters (NRTs) and ammonium transporters (AMTs), and changes in root architecture (Kiba & Krapp, 2016; Jia & von Wirén, 2020). The regulation of these responses varies between plant tissues to optimize growth and survival in specific environments (Wang et al., 2024).
NDRs consist of local responses in particular tissues or cells and systemic responses controlled by long-distance signals (Li et al., 2014; Luo et al., 2026). Local responses include N deficiency-induced decreases in expression of NITRATE-INDUCIBLE GARP-TYPE TRANSCRIPTIONAL REPRESSOR1 (NIGT1) genes (Sawaki et al., 2013), which encode transcriptional repressors that downregulate nitrate-induced gene expression (Maeda et al., 2018). This leads to the increased expression of several NRT2 genes that encode high-affinity NRTs, which, in turn, enhances nitrate uptake in Arabidopsis (Arabidopsis thaliana) roots (Kiba et al., 2018; Zhuo et al., 2024). In rice (Oryza sativa), OsLBD37/38/39, which are homologs of Arabidopsis LATERAL ORGAN BOUNDARY DOMAIN37/38/39 (Rubin et al., 2009), encode transcriptional repressors of NRT2 genes (Zhu et al., 2022). Meanwhile, a mechanism involving C-TERMINALLY ENCODED PEPTIDE (CEP) polypeptides regulates systemic responses to N deficiency in Arabidopsis (Tabata et al., 2014; Ohkubo et al., 2017; Ota et al., 2020). Despite these insights, the pathways that govern NDR regulation in individual tissues, which may be linked to different physiological processes, are not yet fully elucidated.
Our previous study, which utilized weighted gene co-expression network analysis and focused on a range of rice cultivars that displayed differences in nutrient responses, identified four TFs, HYPERSENSITIVITY TO LOW PI-ELICITED PRIMARY ROOT SHORTENING1 HOMOLOG3 (OsHHO3), OsHHO4, REGULATOR OF LEAF INCLINATION1 (RLI1), and OsbZIP11 (Os01g0859500; Ueda et al., 2020) that play pivotal roles in the regulatory network that controls NDRs in rice roots. One of these TFs, OsHHO3, a member of the NIGT1/HHO subfamily of the GARP/G2-like TF family (Ohama & Yanagisawa, 2024), was subsequently identified as a transcriptional repressor of three AMMONIUM TRANSPORTER 1 (OsAMT1) genes, OsAMT1;1, OsAMT1;2, and OsAMT1;3 (Liu et al., 2023), which encode high-affinity transporters responsible for most ammonium uptake activity under N-deficient conditions (Konishi & Ma, 2021). The reduction in OsHHO3 expression induced by N deficiency thus facilitates uptake of ammonium, the primary N source for rice and other paddy plants (Liu et al., 2023). The roles played by the other three TFs in NDR remain to be clarified.
In this study, we focused on OsbZIP11, a basic region/leucine zipper (bZIP) protein, to explore new aspects of the regulatory mechanism governing NDRs in rice. Genome-edited rice lines that lacked functional OsbZIP11 protein (henceforth, Osbzip11-KO lines) and transgenic rice lines that overexpressed OsbZIP11 (henceforth, OsbZIP11-OX lines) were generated and analyzed to elucidate the functions of OsbZIP11 in NDRs, allowing us to make an unexpected discovery: OsbZIP11 is a transcriptional repressor whose primary role is to repress numerous photosynthetic genes and reduce photosynthetic activity in response to N deficiency.
Materials and Methods
Plant materials and growth conditions
The rice (Oryza sativa L.) japonica cultivar Nipponbare was used as the wild-type (WT) line. Surface-sterilized seeds were germinated in distilled water in the dark at 30°C for 3 d. Following germination, seedlings were grown in Ca/Fe solution (0.1 mM CaCl2 and 12 μM Fe(III)-EDTA) for 3 d in a growth chamber under an aluminum foil cover. The resulting 6-d-old seedlings were then grown under various nutrient regimens in the same growth chamber, with 14 h : 10 h, 28°C : 25°C, light : dark photoperiod (200 μmol m−2 s−1 light intensity) at 70% relative humidity, or in a glasshouse maintained at 14 h : 10 h, 28°C : 25°C, light : dark photoperiod (with supplemental lighting). The 6-d-old seedlings prepared as above were used for all analyses in this study.
For the phenotypic analysis, 6-d-old seedlings were grown hydroponically for 17 d in half-strength (0.5×) Yoshida nutrient solution (Yoshida et al., 1976) that contained 0.71 mM NH4NO3 (1.42 mM N) and then for 20 d either in 0.5× Yoshida nutrient solution (control conditions) or in modified Yoshida nutrient solution (low N conditions) in a glasshouse. The NO3NH4 concentration in the modified 0.5× Yoshida nutrient solution was 0.0142 mM (0.0284 mM N). All nutrient solutions were supplemented with MES-KOH to a final concentration of 2 mM, and the pH was adjusted to 5.7. To investigate growth at the reproductive stage, 6-d-old seedlings were grown in 0.5× Yoshida nutrient solution for 8 d and then transplanted either to regular, N-sufficient soil (Honenagri Co., Niigata, Japan), which provides 0.46 g each of N, P, and K per plant, or to N-deficient (N-, P-, and K-free) soil (Inaho-Kako Co., Toyama, Japan) supplemented with 0.03 g NH4NO3, 0.45 g NaH2PO4, and 0.35 g K2SO4 per plant. The seedlings were grown in the glasshouse for the durations indicated in figure legends. All samples were harvested during daylight hours.
Generation of Osbzip11-KO and OsbZIP11-OX plants
Osbzip11-KO plants were generated using the CRISPR-Cas9 system, and OsbZIP11-OX plants were generated through transformation of WT rice with a binary vector (see Supporting Information Methods S1 for details). T2 or T3 progenies of biallelic homozygous mutants for OsbZIP11 and transgenic lines harboring the ZmUbi::OsbZIP11:MYC construct at a single locus were used as Osbzip11-KO and OsbZIP11-OX lines, respectively, in all analyses. DNA sequencing of target sites in Osbzip11-KO plants, as well as the top three possible off-target sites predicted using the CRISPR-P 2.0 website (H. Liu et al., 2017), confirmed both that genome editing had occurred at the target site and that there was no undesired genome editing at off-target sites.
Transcriptome analysis
Six-day-old WT and Osbzip11-KO seedlings were transferred to 0.5× Yoshida nutrient solution for 3 d, and then grown for 4 d either in 0.5× Yoshida nutrient solution or in another modified 0.5× Yoshida nutrient solution in which the NO3NH4 concentration was reduced to 0.0355 mM (0.071 mM N); this version of the modified 0.5× Yoshida nutrient solution was used only for transcriptomic analysis to ensure consistency with our previous RNA-seq analysis of the gene regulatory network associated with NDR (Ueda et al., 2020). Details regarding DNA microarray analysis using root RNA samples are provided in Methods S1. Welch's approximate Student's t-test was used to compare the different groups; differences were considered significant when P < 0.05. A heatmap was constructed using the pheatmap package of R (v.4.0.5). Gene Ontology (GO) enrichment analysis was performed using the PANTHER Classification System (PANTHER19.0).
RT-qPCR analysis
RNA samples were prepared from plant tissues using the Maxwell® RSC Plant RNA Kit (Thermo Fisher Scientific, Waltham, MA, USA) and reverse transcribed to cDNA using SuperScript II reverse transcriptase (Thermo Fisher Scientific) and the (dT)15 primer. cDNA samples were analyzed by qPCR using the StepOnePlus™ instrument (Thermo Fisher Scientific) with gene-specific primers (Table S1) and the KAPA SYBR FAST qPCR Kit (KAPA Biosystems, Wilmington, MA, USA). The level of transcript of each target gene was normalized against the expression of rice UBIQUITIN5 (OsUBQ5, Os01g0328400).
Co-transfection assays in rice protoplasts
DNA fragments corresponding to the promoters and 5′ untranslated regions (5′ UTR) of OsLhcb2.1 (−1978 to +50 base-pairs (bp), where +1 bp is the transcription start site), OsLhca1 (−1913 to +87 bp), and OsPsaO (−1941 to +69 bp) were amplified by PCR from WT rice gDNA using gene-specific primers (Table S1). The amplified fragments were cloned into the pJD301 vector (Luehrsen et al., 1992) to generate reporter plasmids containing the firefly luciferase (LUC) gene under the control of each promoter. The expression vector that harbored the OsbZIP11 coding sequence between the constitutive strong HBT promoter (a hybrid promoter in which the constitutive 35S enhancer is fused to the maize C4PPDK basal promoter) and the MYC epitope tag (EQKLISEEDL)-encoding DNA sequence was used as an effector plasmid (Ueda et al., 2020). An empty vector served as a control effector plasmid. The ZmUbi::GUS plasmid served as an internal control. Preparation of rice protoplasts and polyethylene glycol (PEG)-mediated co-transfection of reporter, effector, and internal control plasmids into rice protoplasts were performed as described previously (Ueda et al., 2020). LUC and GUS activity levels were determined as described previously (Ueda et al., 2020).
Fluorescence microscopy of OsbZIP11-GFP
To express the OsbZIP11-GFP fusion protein in rice protoplasts, the pHBT::OsbZIP11::GFP::NOS plasmid was constructed by replacing the EIN3 cDNA in the pHBT::EIN3::GFP::NOS plasmid (Yanagisawa et al., 2003) with the OsbZIP11 cDNA. Rice protoplasts were transfected with the resulting plasmid and incubated for 16 h in the dark to enable protein synthesis. Nuclei of transfected protoplasts were stained using 5 μg ml−1 4′,6-diamidino-2-phenylindole (DAPI). Green fluorescent protein (GFP) fluorescence emitted by OsbZIP11-GFP was observed using a BX51 fluorescence microscope (Olympus Co., Tokyo, Japan).
Chromatin immunoprecipitation assay
Six-day-old WT and OsbZIP11-OX seedlings were grown in 0.5× Yoshida nutrient solution for 4 d in the growth chamber. Following the method described by Saleh et al. (2008), chromatin immunoprecipitation (ChIP) assays were performed using shoot tissue and an anti-MYC polyclonal antibody (cat. no.: ab32; Abcam Ltd, Cambridge, UK), as described previously (Liu et al., 2023). DNA fragments were purified using the DNeasy Plant Mini Kit (Qiagen N.V., Venlo, the Netherlands) and analyzed using qPCR with gene-specific primers (Table S1). OsUBQ5 served as an internal reference. To calculate fold enrichments of specific genomic regions, values obtained from OsbZIP11-OX leaves were normalized with those from WT leaves.
Electrophoretic mobility shift assay
Electrophoretic mobility shift assay (EMSA) was performed as described previously (Konishi & Yanagisawa, 2008) using His-tagged OsbZIP11 recombinant protein and biotin-labeled DNA probes corresponding to an OsLhcb2.1 promoter region (−1978 to −1610), an OsLhca1 promoter region (−1531 to −1169), and an OsPsaO promoter region (−1642 to −1287; see Methods S1 for details). In some assays, synthetic DNAs generated by annealing two oligonucleotides (Table S1) were used as competitors.
Measurements of Fv/Fm, CO2 exchange ratios, and stomatal conductance
Six-day-old seedlings were grown in 0.5× Yoshida solution for 21 d and then either in control or low N conditions for the indicated period in figure legends. Fv/Fm, CO2 exchange ratios, and stomatal conductance of leaf blades were measured using the Kinetics multispectral fluorescence imaging system (FluorCam 800 MF; Photon System Instruments, Brno, Czech Republic) and the LI-6400 Portable Photosynthesis system (LI-COR; Lincoln, NE, USA), according to the manufacturers' instructions.
Immunoblot analysis
Proteins were extracted from leaves of 150-d-old soil-grown rice plants using SDS-PAGE loading buffer [50 mM Tris–HCl (pH 8.0), 2% (w/v) SDS, 10% (w/v) glycerol, 6% (v/v) 2-mercaptoethanol, and 0.003% (w/v) bromophenol blue], and separated on a 12% (w/v) SDS-polyacrylamide gel. Proteins were transferred to a membrane, detected using anti-RbcL (cat. no.: AS03037), anti-Lhca1 (cat. no.: AS01005), anti-Lhcb1 (cat. no.: AS09522), and anti-CP43 (cat. no.: AS04045) antibodies (all from Agrisera AB, Vännäs, Sweden), and visualized using an anti-rabbit IgG HRP-linked antibody and Supersignal West Dura Extended Duration Substrate (Thermo Fisher Scientific).
Measurements of ammonium uptake activity, C and N contents, and glucose and sucrose contents
For measurements of ammonium uptake activity and C and N contents, 6-d-old seedlings were grown in 0.5× Yoshida nutrient solution for 4 d and then for 3 d either in 0.5× Yoshida nutrient solution or in 0.5× modified Yoshida nutrient solution. Seedling roots were placed in 0.1 mM CaSO4 solution for 1 min and then in 0.2 mM (15NH4)2SO4 solution for 30 min. Roots were washed in 0.1 mM CaSO4 solution for 1 min and dried at 37°C for 3 d. The 15N:14N ratio and C and N contents in entire seedlings were analyzed by Shoko Scientific Co. Ltd (Yokohama, Japan). For measurements of glucose and sucrose contents, 6-d-old seedlings were grown in 0.5× Yoshida nutrient solution for 4 d and then either grown under control or low N conditions for 5 d. Total sugars were extracted from ground leaf tissue samples with an extraction buffer [200 mM HEPES-KOH (pH 7.7) and 80% ethanol], and glucose and sucrose contents were measured using the Glucose (HK) assay kit (Merck, Darmstadt, Germany). The sucrose content was calculated based on the amount of glucose produced by invertase treatment.
Results
Removal of the functional OsbZIP11 gene and its overexpression have opposite effects on seedling growth under N-deficient conditions
To investigate the role of OsbZIP11, we generated Osbzip11-KO rice plants in the WT background using CRISPR/Cas9-mediated gene editing. We also created OsbZIP11-OX plants by overexpressing OsbZIP11 under the control of the maize (Zea mays) ubiquitin promoter in WT rice. Sequence analysis of three independent gene-edited lines, Osbzip11-1, Osbzip11-2, and Osbzip11-3, revealed that their genomes encoded only truncated forms of OsbZIP11 due to the frameshifts caused by DNA deletions or insertions (Fig. S1), confirming that all three lines were Osbzip11-KO plants. By contrast, expression levels of OsbZIP11 were > 15-fold higher than in WT plants in two independent transgenic lines, L1 and L2, confirming that both were OsbZIP11-OX lines (Fig. S2).
A comparison between WT, Osbzip11-KO, and OsbZIP11-OX seedlings found no significant differences in their growth under control (0.71 mM NH4NO3) conditions. Under low N (0.0142 mM NH4NO3) conditions, however, although seedling height, primary root length, and dry root weight did not differ notably between groups, the tiller number differed significantly, resulting in changes in dry shoot weight (Figs 1, S3). All three lines of Osbzip11-KO rice showed an increase in tiller number (Fig. S3), whereas a decreased tiller number was observed in both OsbZIP11-OX lines (Fig. 1). These results suggested that OsbZIP11 had a negative impact on growth under N-deficient conditions.
Phenotypes of 43-d-old wild-type (WT), Osbzip11-KO, and OsbZIP11-OX rice seedlings grown under control nitrogen (N) and low N conditions. (a) Growth. (b) Tiller number. (c) Plant height. (d) Primary root length. (e) Dry shoot weight. (f) Dry root weight. Six-day-old seedlings were initially grown in 0.5× Yoshida solution for 17 d and then under either control N or low N conditions for 20 d. Photographs in (a) show WT, Osbzip11-1, a representative Osbzip11-KO line, and two independent lines, L1 and L2, of OsbZIP11-OX. Bars, 15 cm. (b–f) Means and SD of six biological replicates are shown; different letters indicate significant differences between means determined using Tukey's multiple comparisons test at P < 0.05 (one-way ANOVA).
OsbZIP11 is involved in a gene regulatory network that controls NDR in roots
The relationship between OsbZIP11 and the regulatory network governing NDR was investigated by a transcriptomic analysis of genes expressed in root tissues of Osbzip11-KO (Osbzip11-1) and WT seedlings grown under control or low N conditions. We found that 397 and 70 genes met our criteria for upregulated genes (fold change [FC] > 2; P < 0.05) under control and low N conditions, respectively, whereas 183 and 3 genes met our criteria for downregulated genes (FC < 0.5; P < 0.05) under control and low N conditions, respectively (Fig. 2a). Ueda et al. (2020) previously identified 29 genes encoding TFs that were likely to be influential members of the regulatory network controlling NDRs. A heatmap analysis indicated that most of these 29 genes were differentially expressed between WT and Osbzip11-KO roots (Fig. 2b). Levels of OsHHO3, OsHHO4, and OsRLI1 expression were higher in the roots of Osbzip11-KO seedlings. This result was consistent with the mutual regulatory relationships observed previously between OsHHO3, OsHHO4, OsRLI1, and OsbZIP11, the four most influential TFs in this network (Ueda et al., 2020). In addition, our results suggested that OsbZIP11 negatively regulated OsNIGT1, a member of the NIGT1/HHO family involved in nitrate utilization (Sawaki et al., 2013). An reverse transcription quantitative polymerase chain reaction (RT-qPCR) analysis not only confirmed the effects of removing functional OsbZIP11 protein on the expression of these genes but also revealed that overexpression of OsbZIP11 produced opposite effects (Fig. 2c).
Upregulated and downregulated genes in roots of Osbzip11-KO rice seedlings. (a) Venn diagrams showing the overlap between upregulated (fold change (FC) > 2; P < 0.05) and downregulated (FC < 0.5; P < 0.05) genes in Osbzip11-1 roots, relative to their expression in wild-type (WT), under control and low N conditions. The numbers of genes in each section are shown. Purple text indicates genes involved in N metabolism. Green text indicates genes putatively identified as influential transcription factors (TFs) in nitrogen-deficiency responses (NDRs). (b) Heatmaps showing changes in expression of genes encoding (left) putative influential TFs in the NDR network, (middle top) nitrate transporter genes, and (middle bottom) genes involved in N signaling, (right top) ammonium transporter genes, and (right bottom) genes involved in N assimilation in the roots of Osbzip11-1, relative to expression in WT roots, under control (CN) and low N (LN) conditions. Data for 27 genes encoding putative influential TFs are presented because the DNA microarray did not contain probes corresponding with a homeodomain-leucine zipper TF gene (Os03g0231150) and NLP4 (Os09g0549450). (c) Reverse transcription quantitative polymerase chain reaction (RT-qPCR) analysis of levels of OsHHO3 (Os03g0764600), OsHHO4 (Os07g0119300), OsRLI1 (Os07g0119300), OsNIGT1 (Os02g0531600), OsNPF6.5 (Os10g0554200), OsNPF7.2 (Os02g0699000), OsNRT2.3 (Os01g0704100), OsNAR2.1 (Os02g0595900), AMT1;1 (Os04g0509600), AMT1;2 (Os02g0620600), AMT1;3 (Os02g0620500), OsGS1;1 (Os02g0735200), and OsACTPK1 (Os02g0120100) transcripts in the roots of WT, Osbzip11-KO (Osbzip11-1), and OsbZIP11-OX seedlings. Six-day-old seedlings were grown in 0.5× Yoshida solution for 3 d and then under either control or low N conditions for 4 d. Means and SD of four biological replicates are shown; asterisks indicate significant differences between WT seedlings and other lines determined using Student's t-test; *, P < 0.05; **, P < 0.01.
Although only a few N-related genes were among the top 50 genes upregulated or downregulated in Osbzip11-KO roots under either control or low N conditions (Tables S2–S5), the heatmap analysis suggested that several genes critical for modulating N use efficiency were differentially expressed in Osbzip11-KO roots (Fig. 2b). These genes included OsNPF6.5/OsNRT1.1B, which encodes an NRT intimately associated with N use efficiency (Hu et al., 2015), and OsNPF7.2, which encodes a low-affinity NRT that positively regulates tiller number and grain yield (Wang et al., 2018). We therefore conducted additional RT-qPCR analyses of OsNPF6.5 and OsNPF7.2 expression, as well as OsAMT1;1, OsAMT1;2, OsAMT1;3, OsNRT2.3, OsNAR2.1, and OsGS1;1, which are involved in N acquisition or assimilation. In addition, we analyzed the expression of OsACTPK1, a potential target of OsbZIP11 (Ueda et al., 2020) that encodes a serine/threonine/tyrosine kinase known to reduce high-affinity ammonium uptake under ammonium-sufficient conditions (Beier et al., 2018). Although these analyses found that knockout or overexpression of OsbZIP11 had only minimal effects on the expression of most of these genes, OsNPF7.2 expression was significantly upregulated in Osbzip11-KO roots but significantly downregulated in OsbZIP11-OX roots under low N conditions (Fig. 2c).
OsbZIP11 is most closely associated with Module 16 in the root NDR network
The relationship between OsbZIP11 function and the 16 modules in the root NDR regulatory network was examined by generating heatmaps that illustrate OsbZIP11 knockout-dependent changes in the root NDR regulatory network (Fig. S4). Although Modules 1, 3, 5, 9, 11, 15, and 16 responded to N deficiency, Modules 1, 3, and 5 were suggested to contribute largely to NDRs in roots (Ueda et al., 2020). Modules 1, 3, and 5 contained upregulated, downregulated, and unaffected genes in Osbzip11-KO roots, indicating that OsbZIP11 knockout did not cause uniform changes in gene expression within each module (Fig. S4). Thus, we could not infer the physiological role of OsbZIP11 in roots solely from these changes. By contrast, knocking out OsbZIP11 consistently had a notable positive effect on the expression of each gene in Module 16 (Figs 3a, S4). Most genes in Module 16 were upregulated in Osbzip11-KO lines under low N conditions, but not under control conditions. Furthermore, the repressive effects of N deficiency on genes of Module 16 were largely mitigated in Osbzip11-KO roots (Fig. 3b). Scrutiny of the top 50 upregulated or downregulated genes in Osbzip11-KO roots (Tables S2–S5) found that 23 of the 35 genes in Module 16 were among the top 50 upregulated genes in Osbzip11-KO roots under low N conditions (Table S2), suggesting that OsbZIP11 function is most closely associated with Module 16 genes in the root network. This was surprising because Module 16 is a small module containing 28 genes involved in photosynthesis; therefore, this module was unlikely to play a critical role in the root. Thus, we hypothesized that the primary role of OsbZIP11 is to synchronously suppress numerous photosynthetic genes in shoots and that this negative regulation is also detectable in roots, despite the extremely low expression levels of photosynthetic genes in these tissues.
Changes of expression levels of genes belonging to Module 16 in shoots and roots of Osbzip11-KO and OsbZIP11-OX rice seedlings. (a) Heatmap analysis of nitrogen (N) deficiency-induced changes and OsbZIP11 knockout-induced changes in expression levels of genes in Module 16 in roots under control (CN) and low N (LN) conditions. Data for two genes belonging to Module 16 are missing due to the absence of corresponding probes from the DNA microarray. (b) Correlation analysis showing the relationship between N deficiency-induced decreases and the OsbZIP11 knockout-induced increases under LN conditions in expression of genes in Module 16. (c) A bar chart showing Gene Ontology (GO) biological process terms that were overrepresented (enrichment score > 2; false discovery rate (FDR) < 0.001). Enrichment scores are shown as horizontally oriented bars, color-coded by FDR. The number of differentially expressed genes for each term is indicated using the size of the circle accompanying each bar. (d) Reverse transcription quantitative polymerase chain reaction (RT-qPCR) analyses of transcript levels of OsLhcb2.1 (Os03g0592500), OsLhca1 (Os06g0320500), OsLhca4 (Os08g0435900), OsPsaO (Os04g0414700), OsPsbO (Os01g0501800), OsPsbP (Os07g0141400), OsPsbR3 (Os08g0200300), and OsRbcS3 (Os12g0291100) in wild-type (WT), Osbzip11-KO (Osbzip11-1), and OsbZIP11-OX seedlings. Six-day-old seedlings were grown in 0.5× Yoshida solution for 3 d and then under either control or low N conditions for 4 d. Means and SD of four biological replicates are shown; asterisks indicate significant differences between WT seedlings and other lines determined using Student's t-test; *, P < 0.05; **, P < 0.01.
Consistent with this hypothesis, GO enrichment analysis also showed that OsbZIP11 function is more clearly and more strongly associated with the expression of photosynthetic genes, especially Photosystem I (PSI)-related genes, than with genes involved in any other physiological process, including control of N-related gene expression (Fig. 3c; Tables S6, S7). OsbZIP11 function was additionally linked only to vague GO biological process terms, including responses to chemicals, stress, and stimuli, further supporting our hypothesis.
OsbZIP11 negatively regulates numerous photosynthetic genes in shoots and roots
To clarify the regulatory role of OsbZIP11 in photosynthetic gene expression, we determined the expression levels of photosynthetic genes belonging to Module 16 in shoots and roots of WT, Osbzip11-KO, and OsbZIP11-OX seedlings under N-deficient and N-sufficient conditions (Fig. 3d). We selected eight photosynthetic genes: OsLhcb2.1, OsLhca1, and OsLhca4, which encode Chla/b binding (CAB) proteins in light-harvesting complex I or II, OsRbcS3, which encodes the small subunit of RuBisCO, OsPsaO, and OsPsbO, which encode the PSI-O subunits of PSI and photosystem II (PSII), respectively, and OsPsbP and OsPsbR3, which encode subunits of PSII. All eight genes showed higher expression in Osbzip11-KO shoots than in WT shoots; by contrast, their expression was lower in OsbZIP11-OX shoots than in WT shoots. Furthermore, although N deficiency reduced expression of these genes in WT shoots, this effect was alleviated in Osbzip11-KO seedlings and enhanced in OsbZIP11-OX lines. Similar changes in expression were observed in roots (Fig. 3d). The upregulation of photosynthetic genes observed in Osbzip11-KO roots is probably a physiologically negligible side effect resulting from the fact that photosynthetic genes were still expressed in roots, albeit at very low levels.
To further clarify the physiological significance of OsbZIP11-mediated regulation in photosynthetic gene expression, we conducted additional RT-qPCR analyses. As the top 50 upregulated genes in Osbzip11-KO seedlings grown under low N conditions included 6 (OsPsaD, OsPsaE, OsPsaG, OsPsaH, OsPsaN, and OsPsaO) of nine nuclear genes encoding PSI core subunits (Table S4), the expression levels of OsPsaD-H, K, and L in Osbzip11-KO and OsbZIP11-OX shoots were analyzed (Fig. S5), indicating that OsbZIP11 is responsible for synchronous N deficiency-induced suppressions of all nuclear PSI core subunit genes in shoots. In particular, N deficiency barely reduced OsPsaE and OsPsaG expression in Osbzip11-KO shoots, suggesting that N deficiency-induced downregulation of these genes is primarily mediated by OsbZIP11. Similarly, OsbZIP11 was involved in N deficiency-induced downregulation of major CAB genes (OsLhca2 and OsLhca3 and OsLhcb1a, OsLhcb1b, OsLhcb3, OsLhcb4, and OsLhcb6; Figs 3d, S6) as well as all rbcS genes (OsRbcS1-5) in shoots (Figs 3d, S7). These results suggest that OsbZIP11 is a key regulator of synchronous declines in photosynthetic gene expression under N deficiency.
N deficiency produces opposite effects on OsbZIP11 expression in shoots and roots
The negative regulation of photosynthetic genes by OsbZIP11 was characterized by examining the tissue-specificity of OsbZIP11 expression and the impacts of N deficiency on OsbZIP11 expression. An RT-qPCR analysis of OsbZIP11 expression in leaves, the lamina joint, stems, primary roots, and adventitious roots of WT plants grown under control conditions (Fig. 4a) revealed that OsbZIP11 expression was higher in roots than in leaves and higher in older leaves than in younger ones (Fig. 4b). On the contrary, N deficiency produced opposite effects in shoots and roots, as OsbZIP11 expression increased dramatically in the shoots but decreased moderately in the roots (Fig. 4c). These results suggested that N deficiency induced OsbZIP11 expression in shoots, resulting in the suppression of photosynthetic genes, but caused it to decline in roots, leading to altered expression of OsbZIP11 target genes, which included an upregulation of photosynthetic genes as a physiologically negligible side effect (Figs 3a, S4). Furthermore, comparison of the N-deficiency responsiveness of OsbZIP11 expression in various aerial and subterranean tissues showed a stronger response in older leaves, whereas decreases in OsbZIP11 expression were slightly greater in adventitious roots and root apical parts than in root basal parts (Fig. 4d).
Reverse transcription quantitative polymerase chain reaction (RT-qPCR) analysis of tissue-dependent and nitrogen (N) deficiency-responsive OsbZIP11 expression in rice. (a) Locations of the sampling sites of the wild-type (WT) tissues used in the analysis shown in (b). The sampled tissues were the leaf blades (L1-L4, ordered by leaf age), the lamina joint (LJ), the stem (S), the root (R), and the adventitious root (AR) of 35-d-old WT seedlings grown under control conditions. (b) Tissue-dependent expression of OsbZIP11 in WT seedlings. Expression in each tissue type is shown relative to the expression level in the L1 leaf blade. (c) N deficiency-responsive OsbZIP11 expression in the most fully expanded leaf blade and roots. Six-day-old WT seedlings were grown in 0.5× Yoshida solution for 3 d and then under low N conditions for the period indicated. Expression levels shown for each sample are relative to the levels in the leaf blade at the start of N-deficiency treatment (time point 0). (d) N deficiency-responsive OsbZIP11 expression in leaf blades (L1-L3, ordered by leaf age), LJ, S, the root basal part (RB), and the root apical part (RA), and AR. Six-day-old WT seedlings were grown in 0.5× Yoshida solution for 8 d and then under low N conditions for the indicated period. Expression levels shown for each sample are relative to the levels in the youngest leaf blade (L1) at the start of N-deficiency treatment. Means and SD from four biological replicates are shown. In (b, c), different letters indicate significant differences between means determined using Tukey's multiple comparisons test at P < 0.05 (one-way ANOVA). In (d), asterisks indicate significant differences between the sample at the start of N-deficiency treatment and others determined using Student's t-test; *, P < 0.05; **, P < 0.01.
OsbZIP11 is a transcriptional repressor of photosynthetic genes
The possibility that OsbZIP11 directly repressed photosynthetic genes was investigated using co-transfection experiments, ChIP assays, and EMSA (Fig. 5). The nuclear localization of OsbZIP11 was first confirmed by transiently expressing the OsbZIP11-GFP fusion protein in rice protoplasts (Fig. 5a). Next, co-transfection assays were conducted using reporter plasmids, which contained the LUC gene under the control of a promoter from one of three photosynthetic genes (OsLhcb2.1, OsLhca1, and OsPasO). Additionally, an expression vector was used to produce OsbZIP11-MYC protein. A plasmid containing LUC driven by the 35S promoter and an empty vector served as negative controls for the reporter and effector, respectively (Fig. 5b). LUC activity was strongly reduced in protoplasts co-transfected with any of the reporter constructs containing photosynthetic gene promoters plus the OsbZIP11-MYC expression vector, but not in protoplasts co-transformed with the 35S promoter-driven LUC construct plus the OsbZIP11-MYC expression vector (Fig. 5c). By contrast, co-transfection with the empty vector did not affect the activity of any of the promoters tested (Fig. 5c).
OsbZIP11 directly represses the promoters of photosynthetic genes. (a) Nuclear localization of OsbZIP11. Images of a rice protoplast expressing green fluorescent protein (GFP)-tagged OsbZIP11 (OsbZIP11-GFP) captured under excitation (GFP; top panel) or visible (bright field; bottom panel) light; a 4′,6-diamidino-2-phenylindole (DAPI)-stained image is also shown (middle panel). Bar, 10 μm. (b) Reporter and effector plasmids used in co-transfection assays. The reporter plasmids contained the firefly luciferase (LUC) gene under the control of the OsLhcb2.1 (−1978 to +50 bp, relative to the transcription start site), OsLhca1 (−1913 to +87 bp), OsPsaO (−1978 to +50 bp), or 35S (negative control) promoters. The expression vector that contained the OsbZIP11 cDNA between a derivative of the 35S promoter (HBT) and an MYC epitope tag-encoding sequence (MYC) and an empty vector were used as effector plasmids. All plasmids also contained the nopaline synthase gene terminator (NOS). (c) Co-transfection assays. Rice protoplasts were co-transfected with the effector plasmid (OsbZIP11-MYC) or an empty vector together with a reporter plasmid and an internal control plasmid, pZmUbi-GUS. Means and SD of four biological replicates are shown. (d) Structures of the OsLhcb2.1, OsLhca1, and OsPsaO promoters. Black squares: 5′ untranslated region; open squares: coding regions. Pale orange boxes: designated promoter regions; green boxes: regions amplified in chromatin immunoprecipitation (ChIP) assays. Blue bars: G-boxes; yellow bars: G/C hybrid boxes; red bars: TGA-binding motifs. (e) ChIP analysis of samples extracted from OsbZIP11-OX (L2) seedlings. Different regions of each promoter indicated in (d) were amplified by qPCR using DNA immunoprecipitated with anti-MYC antibodies. Fold enrichments of specific genomic regions, which were determined by normalizing values obtained from OsbZIP11-OX leaves with those from wild-type (WT) leaves, are shown. The promoter region of UBQ5, an unrelated gene, was amplified as a control. Means and SD of four biological replicates are shown. In (c), asterisks indicate significant differences between LUC activity in protoplasts transfected with the OsbZIP11-MYC expression vector and that observed in plasmids transfected with the empty vector; in (d), asterisks indicate significant differences between the relative enrichment of the control region of UBQ5 and the regions analyzed. Statistical significance was determined using Student's t-test; *, P < 0.05; **, P < 0.01. (f) Electrophoretic mobility shift assay (EMSA). OsbZIP11 protein was incubated with biotin-labeled DNA probes containing the DNA sequence of region a of the OsLhcb2.1 promoter, region b of the OsLhca1 promoter, or region b of the OsPsaO promoter, which included a G-box motif, or both a G-box and a TGA-binding motif. For competition assays, 30-bp synthetic DNAs containing either a G-box or a TGA-binding motif and their adjacent sequences, as well as unlabeled DNA probes, were added to reaction mixtures in a 50-fold molar excess compared with the labeled probe. Mutant competitors (Mut), which carried mutations in the motif, were also used. The positions of free DNA and the DNA-OsbZIP11 complex are indicated.
Plant bZIP TFs bind to four specific sequence motifs, the G-box (CACGTG), the C-box (GACGTC), the A-box (TACGTA), and the G/C hybrid (GTCGTG; Foster et al., 1994; Izawa et al., 1993), while TGACG-BINDING FACTOR (TGA)-class bZIP TFs bind to the TGACG motif (Tomaž et al., 2022). OsbZIP11 is classified into the TGA family of bZIP proteins (Ueda et al., 2020). ChIP analysis of samples from OsbZIP11-OX seedlings revealed that OsbZIP11 specifically bound to region a of the OsLhcb2.1 promoter, region b of the OsLhca1 promoter, and region b of the OsPsaO promoter, which contained the G-box motif recognized by bZIP proteins (Fig. 5d,e), although the OsLhcb2.1, OsLhca1, and OsPsaO promoters also contain the TGA-binding motif. Thus, it is likely that OsbZIP11 binds to DNA in a sequence-specific manner in planta. To characterize the interaction between OsbZIP11 and DNA, EMSA was performed using recombinant OsbZIP11 protein and three types of biotin-labeled DNA probes that corresponded to the OsbZIP11-binding regions identified within the OsLhcb2.1, OsLhca1, and OsPsaO promoters (Fig. 5f). OsbZIP11 bound to all tested DNA probes, and the presence of an unlabeled DNA probe reduced the binding of OsbZIP11 to the labeled DNA. To further assess the sequence-specific recognition of OsbZIP11, competition assays were also performed using 30-bp synthetic DNAs containing either a G-box or TGA-binding motif, which was identified in the OsbZIP11-binding regions of OsLhcb2.1, OsLhca1, and OsPsaO promoters, along with the sequences adjacent to the motif. DNAs containing a native motif, except for the one with the G-box motif within region b of the OsLhca1 promoter, competed with the labeled DNA probe. By contrast, synthetic DNAs in which the G-box or TGA-binding motif was mutated lost the ability to compete with the labeled DNA probe. These results indicated that OsbZIP11 directly and selectively binds to several G-box and TGA-binding motifs in the OsLhcb2.1, OsLhca1, and OsPsaO promoters.
These results, taken together, indicated that OsbZIP11 was a transcriptional repressor and acted to suppress the expression of OsLhcb2.1, OsLhca1, and OsPsaO, as well as probably other photosynthetic genes.
OsbZIP11 suppresses photosynthesis under N-deficient conditions
The targets of OsbZIP11 included genes encoding proteins involved in both the light and dark reactions of photosynthesis (Figs 3, S5–S7). We therefore measured the maximum quantum yield of PSII (Fv/Fm) and the CO2 exchange ratio in WT, Osbzip11-KO, and OsbZIP11-OX seedlings that had been grown in 0.5× Yoshida solution for 21 d and then under control or low N conditions (Fig. 6). The Fv/Fm ratios in the second and third leaves of WT, Osbzip11-KO, and OsbZIP11-OX seedlings were comparable in plants grown under control conditions; however, when plants were grown under low N conditions, Fv/Fm was higher in Osbzip11-KO than in WT seedlings but lower than WT in OsbZIP11-OX seedlings (Fig. 6a,b). Furthermore, the decline in Fv/Fm was more pronounced in older (second) leaves. A time-course analysis of Fv/Fm across the period of N deficiency showed that the differences in Fv/Fm became more pronounced from Day 9 of treatment, which aligned with the apparent increase in OsbZIP11 expression noted from Day 6 of N deficiency (Fig. 4c). Similar results were obtained from measurements of the CO2 exchange ratio (Fig. 6c), while stomatal conductance was similar among WT, Osbzip11-KO, and OsbZIP11-OX seedlings (Fig. S8). These results are consistent with the expression levels of photosynthetic genes in WT, Osbzip11-KO, and OsbZIP11-OX seedlings (Figs 3, S5–S7), suggesting that N deficiency-induced OsbZIP11 expression (Fig. 4) decreases the expression of photosynthetic genes (Fig. 3) and thereby downregulates photosynthesis in response to N deficiency.
Enhanced or reduced levels of photosynthetic activity in Osbzip11-KO and OsbZIP11-OX rice seedlings in nitrogen (N)-deficient environments. (a) Fv/Fm images of second and third leaf blades. (b) Fv/Fm of second leaf blades. (c) CO2 exchange ratio in second leaf blades. Six-day-old wild-type (WT), Osbzip11-KO (Osbzip11-1), and OsbZIP11-OX seedlings were grown in 0.5× Yoshida solution for 21 d and then either in control or low N conditions for 17 d in (a) or for the indicated periods in (b, c). In (b, c), means and SD of six biological replicates are shown. Asterisks indicate significant differences between WT seedlings and other lines determined using Student's t-test; *, P < 0.05; **, P < 0.01.
Phenotypes of soil-grown Osbzip11-KO and OsbZIP11-OX plants at the late vegetative and reproductive growth stages
Soil-grown plants showed significant effects of OsbZIP11 knockout and OsbZIP11 overexpression at the late vegetative and reproductive growth stages as well as at the early vegetative growth stage (Fig. 7). Leaf senescence was earlier in OsbZIP11-OX plants than in WT plants, but delayed in Osbzip11-KO plants. This difference in leaf senescence timing was more pronounced in plants grown in N-deficient soil (Fig. 7a). In addition, the differences in Fv/Fm observed between WT, Osbzip11-KO, and OsbZIP11-OX plants increased over the period of cultivation and became notable even in plants grown in regular soil (Fig. 7b). Immunoblot analyses of photosynthesis-associated proteins, including the large subunit of RuBisCO (rbcL), Cab proteins (Lhcb1 and Lhca1), and Chl protein 43 kDa (CP43), were carried out using leaves from 150-d-old plants at the reproductive stage to compare the amounts of these proteins per fresh leaf weight among WT, Osbzip11-KO, and OsbZIP11-OX leaves. The results indicated that amounts of these proteins per tissue weight were significantly lower in Osbzip11-KO plants but higher in OsbZIP11-OX plants grown in N-deficient soil, although the protein levels were mostly comparable across all genotypes when plants were grown in regular soil (Fig. 7c). These results suggested that OsbZIP11 negatively regulates photosynthesis in response to low N availability during both the vegetative and reproductive stages. Given these results, we also investigated grain yields in Osbzip11-KO and OsbZIP11-OX plants. Neither the loss of functional OsbZIP11 nor overexpression of OsbZIP11 had any pronounced effect on panicle length, the number of grains per panicle, or fertility (Fig. 7e–g,i). OsbZIP11-KO plants, however, had more tillers and higher grain yield; conversely, OsbZIP11-OX plants had fewer tillers and reduced grain yield (Fig. 7d,h,j).
Analysis of soil-grown wild-type (WT), Osbzip11-KO, and OsbZIP11-OX plants. (a) Photographs of shoots (left) and leaf tips (right) of 150-d-old WT, Osbzip11-KO (Osbzip11-1), and OsbZIP11-OX plants grown in regular soil (control N) or Nitrogen (N)-deficient soil (low N). (b) Fv/Fm ratios of WT, Osbzip11-1, and OsbZIP11-OX plants grown in regular or N-deficient soil for the indicated periods. (c) Levels of the photosynthesis-associated proteins OsRbcL, OsLhcb1, OsLhca1, and CP43 in second leaves of 150-d-old WT, Osbzip11-1, and OsbZIP11-OX plants grown in regular or N-deficient soil. Total proteins extracted from leaf tissues of equal weight from WT, Osbzip11-1, and OsbZIP11-OX plants were used to detect each protein. (d) Tiller number. (e) Panicle length. (f) Number of grains/panicle. (g) Fertility at harvest. (h) Yield/plant. (i) Representative photographs of panicles from 170-d-old WT, Osbzip11-KO, and OsbZIP11-OX soil-grown plants. (j) Representative photographs of rice grains collected from single soil-grown plants from the indicated lines. Data from soil-grown WT, Osbzip11-KO (Osbzip11-1), and OsbZIP11-OX plants are shown in (d–h). Means and SD of four biological replicates are shown in (b) and of six biological replicates in (d–h); asterisks indicate significant differences between WT seedlings and other lines determined using Student's t-test; *, P < 0.05; **, P < 0.01.
Ammonium uptake activity and C and N contents in Osbzip11-KO and OsbZIP11-OX plants
To further understand the role played by OsbZIP11 in roots, we measured ammonium uptake activity using 15N-labeled ammonium in seedlings grown under control or low N conditions (Fig. 8a). Ammonium uptake activity was elevated in Osbzip11-KO seedlings but reduced in OsbZIP11-OX seedlings, especially under low N conditions. This suggested that OsbZIP11 played a role in suppressing ammonium uptake in the root, which was consistent with the decreased expression of OsbZIP11 in roots under N-deficient conditions (Fig. 4). However, since OsbZIP11 activated AMT gene expression only very weakly (Fig. 2b,c), it might affect ammonium uptake activity indirectly in a post-transcriptional manner at least in part.
Ammonium uptake activity and carbon (C) and nitrogen (N) contents in wild-type (WT), Osbzip11-KO, and OsbZIP11-OX rice seedlings. (a) Ammonium uptake activity. (b) C content. (c) N content. (d) C : N ratio. Six-day-old WT, Osbzip11-KO (Osbzip11-1), and OsbZIP11-OX seedlings were grown in 0.5× Yoshida nutrient solution for 4 d and then under either control or low N conditions for 3 d. Means and SD of four biological replicates are shown; asterisks indicate significant differences between WT seedlings and other lines determined using Student's t-test; *, P < 0.05; **, P < 0.01.
As our results indicated that OsbZIP11 negatively regulated both photosynthesis and ammonium uptake, we measured the C and N contents of seedlings grown under control or low N conditions. Under low N conditions, a statistically significant increase in C content was observed in Osbzip11-KO seedlings, as was a corresponding decrease in C content in OsbZIP11-OX seedlings (Fig. 8b). We note that, in consistency with this result and altered photosynthetic activity in Osbzip11-KO and OsbZIP11-OX seedlings (Fig. 6), glucose and sucrose contents in leaves were significantly decreased in OsbZIP11-OX seedlings under low N conditions; whereas, although not statistically significant, there was a tendency for these contents to increase in Osbzip11-KO seedlings (Fig. S9). Furthermore, a statistically significant increase in N content and a statistically significant decrease in the C/N ratio were detected in OsbZIP11-KO seedlings grown under control conditions (Fig. 8c,d), although changes in C and N contents and the C : N ratio were noticeably smaller than those observed in growth. This may be partly due to the use of young seedlings to simultaneously measure C and N contents and ammonium uptake activity. Collectively, these results suggest that loss of functional OsbZIP11 results in synchronous activation of photosynthesis and N acquisition, while OsbZIP11 overexpression has an opposite effect, repressing both pathways.
Discussion
This study demonstrated that a main function of OsbZIP11, a rice protein first identified as a potential pivotal regulator of the gene network governing NDRs in roots, is to act as a key regulator of photosynthesis by synchronously downregulating many photosynthetic genes in response to N deficiency. This unexpected finding reveals a novel transcriptional regulatory mechanism by which rice suppresses photosynthetic activity and growth rates in an N deficiency-dependent manner, providing new insight into the coordinated regulation of C and N acquisition in environments with varying N source levels.
OsbZIP11 reduces photosynthetic activity under N-deficient conditions
Regulatory mechanisms governing photosynthetic activity are a critical research target in plant science. Extensive research, mostly in the model plant Arabidopsis, has identified several mechanisms that regulate photosynthetic genes. Studies have primarily focused on light-induced gene expression (Jiao et al., 2007). PHYTOCHROME-INTERACTING FACTORs (PIFs) and ELONGATED HYPOCOTYL5 (HY5) are well-known TFs that control light-regulated expression of photosynthetic genes (Oyama et al., 1997; Ang et al., 1998; Ni et al., 1998; Leivar & Monte, 2014), whereas three other bZIP TFs, bZIP16, bZIP68, and GBF1, regulate a subset of photosynthesis-related genes in response to blue light (Lindbäck et al., 2023). In addition, ATHB17, a homeodomain-leucine zipper protein, may repress several photosynthetic genes involved in the light reactions of photosynthesis and enhance plant tolerance to abiotic stresses, such as abscisic acid, drought, and high levels of NaCl (Zhao et al., 2017). Recently, we also showed that NLP-mediated nitrate signaling enhances expression of AtHB52 and AtHB54, which encode TFs involved in the quality control of photodamaged thylakoid membrane proteins in chloroplasts in Arabidopsis (Ariga et al., 2022). However, the mechanisms by which plants regulate photosynthesis in response to changes in N availability remain largely unknown, although downregulating photosynthetic genes under N-deficient conditions is likely a key and effective strategy for conserving energy due to the fact that a significant portion of N acquired from the soil is allocated to synthesize photosynthesis-related proteins. Therefore, the identification of OsbZIP11 as a key regulator that synchronously controls the expression of numerous photosynthetic genes in response to changes in N availability is a crucial step toward elucidating the mechanism that coordinately regulates C and N acquisition.
Importantly, OsbZIP11 suppressed many photosynthetic genes, some of which have been experimentally confirmed to be essential or critical for plant growth and survival. For instance, the Arabidopsis PsaD double-knockout mutant was seedling-lethal (Ihnatowicz et al., 2004). Similarly, disrupting one of the two PsaE genes in Arabidopsis caused a marked increase in light sensitivity and photoinhibition and a decrease of c. 50% in growth rate (Varotto et al., 2000). Although Arabidopsis possesses two copies each of PsaD and PsaE, rice has only one copy of each. Thus, OsbZIP11-mediated repression of OsPsaD and OsPsaE could significantly reduce photosynthetic activity and growth rate. Notably, OsbZIP11 accounted for most N deficiency-responsive repression of OsPsaE (Fig. 5s). Additionally, Ds insertion in the PsaF gene (OsPS1-F) resulted in semi-dwarf rice plants with reduced tiller number and grain yield (Ramamoorthy et al., 2018), and the essential role of OsLhcb3 in photosynthesis has been shown previously using knockdown lines (Wang et al., 2023). The essential roles played by the target genes of OsbZIP11 underscore the physiological significance of OsbZIP11-mediated regulation.
The downregulation of genes involved in photosynthesis under N-deficient conditions was significantly mitigated in Osbzip11-KO plants. These plants, however, still exhibited higher levels of photosynthesis-related gene expression under control conditions than in low N conditions (Figs 3d, S5–S7), suggesting that another TF, in addition to OsbZIP11, may be responsible for regulating the downregulation process of photosynthetic genes under N-deficient conditions. OsbZIP41 is a candidate for such a TF, as it is closely related to OsbZIP11 and is one of 29 putative influential TFs in the gene regulatory network governing NDR in the root (Ueda et al., 2020). On the other hand, intriguingly, OsbZIP11 is the rice homolog of LG2 in maize; the two proteins share c. 77% similarity (Fig. S10). In maize, LG2 is involved in establishing the leaf blade–sheath boundary (Walsh et al., 1997), as the ligule and auricles are often absent or mispositioned in the lg2 mutant, leading to a diffuse blade–sheath boundary. An OsbZIP11 mutant of rice also displays a ligule-less phenotype (Wang et al., 2021). Consistently, we observed a ligule-less phenotype in Osbzip11-KO seedlings and found that alterations in the angle between the leaf blade and sheath resulted in an upright form with reduced leaf blade spread in Osbzip11-KO seedlings; by contrast, OsbZIP11-OX seedlings exhibited the opposite phenotype (Fig. S11). Since the canopy structure of plants significantly influences their efficiency in harvesting light energy, the altered forms of the lg2 and Osbzip11-KO seedlings probably affect their photosynthetic efficiency. OsbZIP11 was expressed at higher levels in older leaves (Fig. 4), thus OsbZIP11-mediated regulation might be involved in controlling changes in photosynthetic activity in response to aging, development, or adaptation to variable light conditions, as well as to N availability. Further analyses of OsbZIP41, LG2, and AtTGA9 (one of the Arabidopsis homologs of OsbZIP11 and LG2; Fig. S10) as well as other OsbZIP11 homologs will be valuable in developing a full understanding of the mechanisms regulating photosynthetic activity.
The relationship between OsbZIP11 and the gene regulatory network controlling NDR in roots
Sucrose enhances expression of OsNRT2.1, OsNRT2.2, OsNRT2.3a, and OsNRT2.3b, resulting in enhanced nitrate uptake in roots (Li et al., 2024). Photosynthetic activity also regulates tillering bud elongation and N use efficiency in rice (Wen et al., 2024). Therefore, enhanced and reduced photosynthetic activity in Osbzip11-KO and OsbZIP11-OX seedlings might indirectly affect the gene regulatory network controlling NDR in roots. However, since levels of OsNRT2.3 expression were mostly comparable in WT, Osbzip11-KO, and OsbZIP11-OX roots (Fig. 2c), changes in the network controlling NDR and N acquisition activity in Osbzip11-KO and OsbZIP11-OX roots could not solely be attributed to changes in photosynthetic activity.
On the contrary, OsbZIP11 was expressed at a significantly higher level in roots than in shoots, and N deficiency decreased OsbZIP11 expression (Fig. 4c), probably leading to increases in ammonium uptake activity (Fig. 8a). Furthermore, the level of OsbZIP11 expression affected the expression of most of the genes encoding TFs considered influential members of the regulatory network governing NDR in roots, as well as the expression of OsNPF7.2, which is an important modulator of N use efficiency (Fig. 2). These findings suggest that OsbZIP11 also directly regulated gene expression in roots and thereby influenced the regulatory network governing NDR. However, the loss of functional OsbZIP11 had pleiotropic and complex effects on gene expression in roots; the affected modules included Module 1, which is associated with carbohydrate biosynthesis, response to hormone stimulus, and amino acid transport, Module 5, which is associated with N compound biosynthesis, carbohydrate metabolism, and transporter activity, and Module 9, which is associated with N compound metabolism and glucose metabolism (Ueda et al., 2020; Fig. S4). Furthermore, knockout of OsbZIP11 led to increased expression of OsHHO3, which encodes a transcriptional repressor of three critical AMT1 genes (Liu et al., 2023); however, the expression levels of AMT1 genes in Osbzip11-KO seedlings were comparable to, or even slightly higher than, those in WT seedlings (Fig. 2c). Similarly, although ammonium uptake activity was enhanced in Osbzip11-KO seedlings (Fig. 8a), expression of OsACTPK1, which encodes a protein that downregulates ammonium uptake activity (Beier et al., 2018), increased in Osbzip11-KO seedlings (Fig. 2c). These complicated changes in OsbZIP11-KO and OsbZIP11-OX roots may be attributed to the mutual negative regulation between the four pivotal TFs (OsHHO3, OsHHO4, RLI1, and OsbZIP11) in the network governing NDR (Ueda et al., 2020), suggesting that further analyses will be necessary to untangle the intricate gene regulatory network governing NDRs in roots.
OsbZIP11 is involved in the coordinated regulation of C and N assimilation
The photosynthetic and ammonium uptake activities were enhanced in Osbzip11-KO seedlings, while opposite effects were observed in OsbZIP11-OX plants (Figs 6, 8). Therefore, OsbZIP11 likely modulates C and N assimilation activities and growth rate in proportion to the level of available N by optimizing the expression of specific sets of genes in shoots and roots. In this context, it is important to note that N deficiency altered the expression levels of OsbZIP11 in shoots and roots in opposite directions, as OsbZIP11 negatively regulated both photosynthetic and ammonium uptake activities. Under N-sufficient conditions, plants do not need to expend a large amount of energy to N acquisition and thus can allocate a larger proportion of their energy to promoting shoot growth and enhancing photosynthesis. By contrast, under N-deficient conditions, plants must allocate more energy to acquiring N than to photosynthesis. Therefore, the observation that disrupting OsbZIP11 expression only had small positive effects on photosynthesis, ammonium uptake activity, seedling growth, and grain yield under N-sufficient control conditions may be because photosynthesis, rather than N acquisition, was the limiting process for growth under these conditions where expression levels of OsbZIP11 in shoots were low. Under N-deficient conditions, however, disruption of OsbZIP11 produced positive effects on photosynthesis, ammonium uptake activity, seedling growth, and grain yield. These changes were consistent with the idea that N acquisition was the limiting process for growth under these conditions. Further analysis of the regulatory mechanisms that oppositely control OsbZIP11 expression in shoots and roots in response to N deficiency may provide clues to understanding how rice plants balance N assimilation and photosynthesis under changing environmental conditions.
Acknowledgements
We thank Drs Seiichi Toki, Hiroaki Saika, and Masaki Endo (the National Agriculture and Food Research Organization (NARO), Tsukuba, Japan) for providing rice seeds, vectors, and technical advice. This study was partially funded by the Japan Society for the Promotion of Science (KAKENHI grant nos.: 22H04977 and 25K24516) to SY.
Competing interests
None declared.
Author contributions
SY conceived the study and designed the research. NO, YS, MY and YU performed the experiments. NO and SY wrote the manuscript.
Data availability
DNA microarray data are deposited in the Gene Expression Omnibus repository of the National Center for Biotechnology Information database under accession no.: GSE297147.
References
Disclaimer
The New Phytologist Foundation remains neutral with regard to jurisdictional claims in maps and in any institutional affiliations.