Sugar transporter modulates nitrogen-determined tillering and yield formation in rice

Nitrogen (N) is an essential macronutrient that largely determines crop yield in agricultural production. The widespread adoption of semi-dwarf, high-yielding gramineous crop varieties during the Green Revolution enhanced lodging resistance and encouraged higher N fertilizer inputs to boost yield. However, excessive N fertilizer use imposes energy costs and environmental risks (soil acidification, eutrophication, greenhouse gas emissions), highlighting the need to identify genetic resources and germplasm with high N use efficiency (NUE) and key genes for NUE improvement. In rice, grain yield is mainly determined by tiller (panicle) number, grains per panicle, and grain weight. Because semi-dwarf varieties have defects in gibberellin (GA) biosynthesis, plant height is less sensitive to increased N, whereas promoting tillering is a promising approach to improve yield under high N. Enhanced tillering contributes substantially to improved rice yield and NUE in response to higher N inputs. OsTCP19 was identified as important for the tillering response to N (TRN), with an elite allele OsTCP19-H largely present in wild rice but lost in modern cultivated rice; nonetheless, many modern cultivars still show strong N-induced tillering, implying an OsTCP19-independent pathway. Tiller development begins from axillary buds and is strongly influenced by environmental cues, especially soil N availability. Elevated N promotes tiller outgrowth rather than initiation, improving effective panicle number. External N cues are integrated into internal tillering programs via hormones including auxin, strigolactone, and GA; GA acts as an inhibitor of tiller development. NGR5, a downstream component of GA signaling, is required for N-induced tillering: stabilized by DELLA via GID1 in response to N, NGR5 recruits PRC2 to suppress tillering inhibitors through H3K27 methylation, promoting tillering and yield. How N availability is integrated into the GA–NGR5 module remained unclear. Here, the authors identify a sugar transporter, OsSTP28, as a key regulator of N-responsive tillering and yield formation. N supply represses OsSTP28 to fine-tune apoplastic glucose levels in tiller buds, coordinating GA catabolism to boost tillering. An elite allele of OsSTP28 more efficiently promotes N-responsive tillering and yield formation in modern rice, offering a valuable germplasm for improving yield and NUE.

Prior work established that tillering responses to N involve hormonal pathways, notably GA signaling via NGR5, which is stabilized under elevated N and recruits PRC2 to repress tillering inhibitors through H3K27 methylation. GA generally inhibits tiller development. OsTCP19, a TCP transcription factor, contributes to geographical adaptation of TRN; its elite OsTCP19-H allele is enriched in wild and aus rice but mostly lost in modern cultivars, suggesting additional pathways govern TRN in current varieties. Sugar status at the shoot base is known to promote axillary bud outgrowth, and apoplastic sucrose hydrolysis with monosaccharide uptake by STP transporters is a key component of source–sink dynamics. An orthologous KNOX factor in maize (KNOTTED1) directly represses GA metabolism genes, hinting that KNOX–GA2ox regulation could be conserved. Recent studies showed glucose can activate TOR–PRC2 signaling to drive H3K27me3-mediated silencing, providing a potential link between sugar signaling and epigenetic regulation relevant to tillering.

- Genetic mapping: Constructed a four-parent rice MAGIC population (120 lines) to assess TRN under field LN (150 kg N/ha) and HN (350 kg N/ha). Performed GWAS (MLM, TASSEL v5.0) identifying a locus on chromosome 11 (peak SNP at chr11:22627320; −log10 P = 5.20) linked to tiller number under HN. - Candidate gene validation: Assessed expression of genes in the LD block; only OsSTP28 (Os11g0594000) showed N-responsive expression negatively correlated with TRN. Haplotype analysis of promoter InDels (two indels near positions 22624546 and 22625579) distinguished Hap.C (common) and Hap.H (high-response) alleles; transient dual-luciferase assays implicated the P2 2-bp indel as causal for N responsiveness. - Gene expression analyses: qRT-PCR for tissue specificity and N forms (NH4+, NO3−, NH4NO3, Gln) responses; inhibition of N assimilation with MSX to test regulatory dependence. GUS reporter and in situ hybridization to localize OsSTP28 expression to axillary buds. - Genetics and complementation: Generated CRISPR/Cas9 OsSTP28 knockout lines (stp28-1/2/3) in Nipponbare; monitored tiller numbers over development in field and hydroponics. Complemented stp28-3 with OsSTP28 CDS driven by Hap.C or Hap.H promoters (promoter swapping) to test TRN restoration and allele performance. - Transporter characterization: Subcellular localization of OsSTP28-eGFP in rice protoplasts. Heterologous transport assays in yeast EBY.VW4000 and Xenopus laevis oocytes to determine hexose specificity, directionality (influx), and kinetics (14C-glucose uptake; Km ≈ 76.15 ± 5.54 µM). - Sugar profiling: Time-course (every 4 h, 2 d) diel measurements of sucrose, glucose, fructose in shoot bases. Apoplastic versus intracellular sugar quantification in shoot base and leaves under LN/HN using vacuum infiltration–centrifugation and UPLC/HPLC. 14C-glucose feeding to quantify blade-to-shoot base translocation. - Transcriptomics: RNA-seq of shoot base (end of night) in WT and stp28 under LN/HN; DEGs via DESeq2; GO enrichment and GSEA to identify affected pathways. - GA metabolism assays: qRT-PCR for GA2ox genes (OsGA2ox3/5/8/9); GA2ox enzymatic activity assays; UPLC–MS/MS quantification of GA isoforms (bioactive GA1 and catabolic GA9/GA34). Generated ga2ox5 mutants and stp28/ga2ox5 double mutants to test genetic epistasis. - Transcriptional regulation: Identified OSH15 (KNOX) as a DEG; validated OSH15 expression changes; EMSA to test OSH15 binding to TGAC motifs in GA2ox promoters; transient dual-luciferase assays for repression activity. Created osh15 mutants to assess effects on GA2ox expression/activity and tillering; tested genetic placement relative to OsSTP28 by complementation. - Epigenetics and glucose signaling: ChIP-PCR for H3K27me3 at OSH15 under LN/HN and in stp28. Examined transcript levels of glucose signaling components (HKs, TOR, SnRKs). Applied external glucose (2%) and 2-deoxy-D-glucose to probe glucose signaling effects on OSH15–GA2oxs and tillering. - Field trials and NUE: Multi-year field experiments in Nanjing under LN/HN measuring panicle number, grains per panicle, 1000-grain weight, yield per plant, and NUE in WT, stp28, and complementation lines; genome editing of OsSTP28 in elite cultivars (MH63, GLA4) to test generality. Assessed independence from OsTCP19 via expression analyses and allele surveys in modern japonica varieties.

- GWAS in a 120-line MAGIC population identified a locus on chromosome 11 (peak SNP at chr11:22627320; −log10(P)=5.20) associated with tiller number under high N; OsSTP28 within a 30-kb LD block emerged as the causal gene. - OsSTP28 is preferentially expressed in axillary buds, negatively regulated by internal N status (repressed by NH4+, NO3−, NH4NO3, and Gln; MSX alleviates repression), and its expression response to N (ERN) negatively correlates with TRN across lines. - Promoter polymorphisms define Hap.H (high N-responsive) and Hap.C (common) alleles; a critical 2-bp indel (P2) in Hap.H confers stronger N-responsive transcription and TRN in transient assays and field phenotypes. Hap.H is frequent across indica and japonica accessions. - Functionally, OsSTP28 knockout (stp28) increases tiller outgrowth; at ripening, tiller number increased by ~49% (LN) and ~35% (HN) vs WT. TRN (HN/LN tiller ratio) is less sensitive in stp28 than WT, showing OsSTP28 is required for full N-responsiveness while acting as a repressor of tillering. - OsSTP28 encodes a plasma membrane monosaccharide transporter mediating glucose influx: yeast and oocyte assays show high affinity for glucose (Km ≈ 76.15 ± 5.54 µM), influx not efflux, with minor transport of mannose/galactose and none for fructose. - Diel sugar dynamics: OsSTP28 expression peaks at night; glucose exhibits rhythmic oscillation. Knockout elevates glucose particularly at night without shifting phase. Under HN or stp28, apoplastic glucose in shoot base and leaf blades increases at night, with little change in apoplastic sucrose/fructose or intracellular sugars. - RNA-seq and enrichment analyses show activation of carbohydrate metabolism pathways and GA catabolism in stp28; GA2ox genes (OsGA2ox3/5/8/9) are upregulated in stp28 and in WT under HN. - GA metabolism: GA2ox activity increases in stp28 and under HN; bioactive GA1 decreases while catabolic GA9 increases in stp28 shoot bases; GA34 and GA3 largely undetected in this setup. - Genetic epistasis: ga2ox5 knockout reduces GA2ox activity and tillering/TRN; the enhanced tillering of stp28 is abolished in stp28/ga2ox5 double mutants, placing OsSTP28 upstream of GA2ox-mediated GA catabolism. stp28/ngr5 double mutants resemble ngr5, placing OsSTP28 upstream of NGR5. - OSH15, a KNOX transcriptional repressor, is downregulated in stp28 and by HN; OSH15 directly binds TGAC motifs in GA2ox promoters and represses their transcription. osh15 mutants show increased GA2ox expression/activity and enhanced tillering, mimicking stp28. Complementation of OsSTP28 restores OSH15 and GA2ox expression/activity. - Mechanism: Elevated apoplastic glucose (by HN or stp28) activates glucose signaling (HKs, TOR) and represses SnRKs; ChIP-PCR shows increased H3K27me3 at OSH15 under HN and in stp28, consistent with epigenetic silencing. External glucose suppresses OSH15, upregulates GA2oxs, increases GA2ox activity, and promotes tillering; effects persist with 2-deoxyglucose, indicating a signaling role for glucose. - Agronomic outcomes: In field trials, stp28 mutants have higher panicle number; grain number per panicle and 1000-grain weight are unchanged or reduced under LN, reflecting tradeoffs. Yield per plant and NUE increase by ~20–21% (LN) and ~10–13% (HN). Complementation with OsSTP28 Hap.H yields higher panicle number, yield, and NUE than Hap.C under HN. CRISPR knockout in elite cultivars (MH63, GLA4) enhances N-responsive tillering and yield. OsSTP28 acts independently of OsTCP19 and likely complements its loss in modern varieties.

The study uncovers OsSTP28 as a key integrator linking environmental N availability to hormonal control of tillering. Elevated N represses OsSTP28, increasing apoplastic glucose in tiller buds at night. Glucose acts as a signal that promotes PRC2-dependent H3K27me3 at OSH15, silencing this KNOX repressor. Reduced OSH15 derepresses GA2ox genes, enhancing GA catabolism, lowering bioactive GA levels in the shoot base, and thus promoting tiller outgrowth. Genetic analyses position OsSTP28 upstream of GA2ox-mediated GA metabolism and the NGR5 pathway, providing a mechanism to feed N status into the GA–NGR5 regulatory axis that is crucial for N-responsive tillering. The identification of an elite promoter allele (Hap.H) of OsSTP28 that responds more strongly to N offers a practical route to enhance TRN, yield, and NUE in modern rice. The prevalence of Hap.H across indica and japonica and independence from OsTCP19 indicate parallel, complementary pathways for N-responsive tillering, explaining why modern cultivars retain high TRN despite loss of OsTCP19-H. The conservation of KNOX regulation of GA2ox suggests broader applicability across grasses. These findings are significant for sustainable agriculture: manipulating OsSTP28 expression or deploying Hap.H can fine-tune carbohydrate–hormone crosstalk to increase planting density via tillering and improve yield/NUE, particularly under limited N inputs.

This work identifies OsSTP28 as a plasma membrane glucose influx transporter whose N-repressed expression elevates apoplastic glucose in tiller buds, triggering epigenetic silencing of OSH15 and activation of GA2ox-mediated GA catabolism to promote tillering. The OsSTP28–OSH15–GA2oxs module mechanistically connects N availability, sugar signaling, and GA metabolism upstream of NGR5 to drive N-responsive tillering and yield formation. An elite OsSTP28 promoter allele (Hap.H) enhances N-responsiveness and improves yield and NUE, representing a valuable breeding target. Future research should elucidate the molecular sensors of apoplastic glucose, refine spatial–temporal sugar measurements inside cells, test conservation across other cereals, and optimize allele deployment or genome editing strategies to balance tiller number with panicle size/seed weight for maximal agronomic benefit.

- The precise mechanism by which plants sense and transduce apoplastic glucose signals remains unresolved. - Intracellular (cytosolic) glucose levels were difficult to quantify accurately; despite no detected differences, technical limitations may mask relevant changes. - Some tradeoffs were observed: increased panicle number sometimes coincided with unchanged or reduced grains per panicle and 1000-grain weight under limited N, which may affect overall yield stability depending on environments. - Most validation was in selected genetic backgrounds and environmental conditions; broader multi-location, multi-year trials and testing in diverse germplasm are needed to generalize findings. - GA profiling detected limited sets of GA isoforms; comprehensive hormone profiling across development could refine mechanistic insights.