Maize STEM BENDING1 controls plant architecture through cell wall and auxin pathways Authors: Shiya Zhao, Babar Farid, Xintong Liu, Liang Dong, Fang Yang Published: 2026 ABSTRACT Crop yield is largely determined by plant architecture. Identifying the key genes responsible for maize architecture can help optimize plant architecture for high-density planting. Here, we characterized a dominant stem bending mutant (Stb1) in maize, which exhibited bent stems and curved leaves leading to a slanted architecture. Genetic mapping delimited Stb1 to a 3.7 Mb interval on chromosome 1. Transcriptome analysis suggested that Stb1 mutation impaired the molecular pathways related to cell wall components and environmental stimuli. Consistently, lignin and hemicellulose contents were altered between the convex side and concave side of Stb1 stems. In addition, the shoot gravitropism regulator LAZY1 transcripts and endogenous auxin levels were differentially accumulated between the two sides of Stb1 stems. Our study reveals that STB1 plays a role in maintaining upright stems through cell wall and auxin-related pathways, providing a new target for optimizing maize shoot architecture.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1007/s44307-026-00136-5. Introduction Maize (Zea mays L.), as one of the most vital cereal crops worldwide, plays a crucial role in ensuring food security by serving as a major source of human food and animal feed. Driven by the pursuit of higher yield through increasing planting density accompanied by thinner and taller stems, stalk lodging has emerged as a major threat to sustainable maize production (Cui et al. 2022). To address this constraint, improving stem mechanical strength can not only effectively prevent yield loss from stalk lodging but also enhance maize tolerance to high-density planting. Stem mechanical bending strength is primarily determined by the cell wall structure components containing cellulose and lignin, and also the spatial morphology of stem, including stalk diameter, rind thickness (Robertson et al. 2017). Recent studies have identified several key genes regulating stem mechanical rigidity and stalk lodging resistance in maize. For instance, the F-box domain protein STIFF1 underlying a major stalk-strength QTL increases the cellulose and lignin accumulation in rind sclerenchyma cells of stem through activating the NAC-MYB expression (Zhang et al. 2020). Another stalk lodging major QTL, stiff2, which corresponds to the flowering-time gene ZmCCT, significantly enhances stalk bending strength by indirectly regulating a set of downstream genes involved in stalk elongation, stalk thickening, and lignin biosynthesis (Zhong et al. 2026). One class III peroxidase gene ZmPRX38 acts downstream of the phenylpropanoid pathway to catalyze monolignol oxidation, and its favorable haplotype significantly improves stalk mechanical strength and yield performance (Wang et al. 2026). Identification of key genes for stem rigidity provide useful genetic resources for stem lodging resistance improvement. For example, targeted genetic modifications of cellulose synthase (CESA) genes and lignin pathways can effectively improve stem bending strength (Polko and Kieber 2019). Stem bending is also connected with shoot gravitropism. LAZY1 (LA1) acts as a conserved master regulator of shoot gravitropism across rice, maize, and Arabidopsis, to control tiller/stem upright architecture through auxin signaling and gradient regulation (Li et al. 2007; Dong et al. 2013; Yoshihara et al. 2007). Mutating maize la1 disrupts polar auxin transport (PAT) and distribution, leading to altered growth orientations and consequent stem lodging (Dong et al. 2013). However, while the roles of cell wall rigidity on stem lodging resistance have been documented, its molecular basis remains poorly understood. New genetic loci modulating stalk growth would expand our understanding on the mechanisms of stem strength and provide new targets for enhancing crop lodging resistance. Here, we identify a natural maize mutant, Stem bending 1 (Stb1), which displayed distinct stem bending, reduced plant height and curved leaves post-floral transition. Our results demonstrate that the Stb1 locus is associated with altered stem cell wall structure and asymmetric auxin distribution, thereby providing new insights into the genetic basis of slanted plant architecture. Materials and methods Plant materials and phenotypic analysis The dominant maize Stb1 mutant was identified from natural variations in the B104 background. F2 mapping populations were generated by crossing Stb1 (in B104) with the inbred line KN5585. Agronomic traits and yield components were evaluated using 10–20 plants per genotype. Histological observation The 3rd to 4th stem internodes and ear leaves from WT and Stb1 plants were fixed in FAA, embedded in paraffin and sectioned. Sections were stained with Toluidine Blue O (TBO) and imaged using a Nikon ECLIPSE Ni-U microscope. Cell size (average area of ≥ 60 cells across six independent microscopic fields of view from three biological replicates) and stomatal density were quantified using ImageJ. Cell number was enumerated per 1 mm2 area. Genetic analysis and BSR-seq Genetic segregation in the F2 population (n = 245) was validated using the chi-square (χ2 test, df = 1) goodness-of-fit test. For Bulk Segregant RNA-seq (BSR-seq), equal amounts of total RNA from leaves of 20 extreme mutants and 20 WT F2 individuals were pooled. Sequencing reads were aligned using TopHat2 (Kim et al. 2013), and the locus was initially mapped via DeepBSA (Li et al. 2022). Fine mapping was performed using 12 polymorphic InDel and KASP markers (Supplemental Table 1) designed within the target region on chromosome 1 (252.8–289.2 Mb) using TBtools (Chen et al. 2020). Transcriptome and RT-qPCR analyses Stem tissue was collected from approximately 30-day-old seedlings, with two biological replicates, and immediately frozen in liquid nitrogen. Total RNA was extracted using RNA extraction kit (Genestonebio, TR205, China), and libraries were constructed using Illumina Standard Library Preparation Kit. Each library generated approximately 20 M raw reads. RNA-seq reads from the 3rd to 4th internodes of WT and Stb1 stems were mapped to the B73 reference genome (B73ref_v4) using TopHat2 (Kim et al. 2013). Differentially expressed genes (DEGs) were identified using DESeq2 (Love et al. 2014). To ensure statistical rigor, DEGs were filtered based on two key metrics: Log2 Fold Change (Log2FC), which quantifies the magnitude of expression ratio changes between Stb1 and WT (with |Log2FC|> 1 indicating a greater than twofold difference in expression), and the False Discovery Rate (FDR), which adjusts raw p-values to control for false positives in multiple hypothesis testing using the Benjamini–Hochberg method (FDR ≤ 0.05). The identified DEGs were subsequently subjected to Gene Ontology (GO) enrichment analysis via TBtools (Chen et al. 2020). Relative expression of ZmLA1 was quantified by RT-qPCR using elongation factor 1α as the internal control with three biological replicates (primers listed in Supplemental Table 2). Results A natural maize mutant, Stem bending 1 (Stb1), isolated from the B104 inbred line, exhibited distinct stem bending (~42° vs. ~ 10° in WT) and curved leaves post-floral transition (Fig. 1a-c). Stb1 plants displayed reduced plant and ear heights due to shorter internodes, though internode number remained unchanged (Fig. 1d). Cellular analysis of Stb1 stems revealed that while cells on both concave and convex sides were significantly smaller than those of WT, cell sizes between the two sides were comparable (Fig. 1e-h). Notably, cell number per unit area on the convex side was significantly higher than that on the concave side (Fig. 1i), suggesting that the accelerated cell division on the convex side could drive the differential stem growth and the subsequent bending. Additionally, Stb1 mutants developed smaller ears with a ~ 32% reduction in ear length and a ~ 20% decrease in kernel number per row (KNPR) (Fig. 1j-l). The leaves above ear in Stb1 were shorter and wider, displaying a ~ 9.67% decrease in length and a ~ 9.28% increase in width compared to WT (Fig. 1m, n). Cellular investigation revealed that the stomata number per unit area on the abaxial side of Stb1 leaves was significantly increased compared to WT (Fig. 1o-q). Thus, the Stb1 mutation triggers pleiotropic developmental and agronomic phenotypes in maize.Fig. 1Phenotypic and cellular characterization of the Stb1 mutant. a Gross morphology of wild-type (WT) and Stb1 plants in the field. Bar = 10 cm. b Stem bending phenotype, with arrows indicating the inner concave side (Stb1-Concave) and outer convex side (Stb1-Convex). Scale bar = 10 cm. c Stem bending angle of WT and Stb1. d Comparison of internode length between WT and Stb1 in maize mature period. 6i-18i indicates the number of internodes above the ground from bottom to top. e–g Stem cell observation of WT (e), concave side of Stb1 (f), convex side of Stb1 (g). Scale bar = 50 μm. h, i Comparison of cell size (h) and cell density (number per mm2) (i) in stem longitudinal sections (n = 3 biological replicates). j Comparison of mature ear of WT and Stb1. Scale bar = 10 cm. k Ear length of WT and Stb1. l Number of grains in rows of WT and Stb1. m, n Leaf width (m) and leaf length (n) of the leaf immediately above the uppermost ear (n = 20). o, p Abaxial leaf epidermal cells of WT (o) and Stb1 (p), showing cell shrinking in the mutant. Scale bar = 100 μm. q Stomatal density on the abaxial leaf surface of WT and Stb1 (n = 3 biological replicates). Data are presented as mean ± SD. In (c, d, k-n, q), statistical significance was determined by Student's t-test (*p < 0.05, **p < 0.01, ***p < 0.001). In (h, i), different lowercase letters indicate significant differences (p < 0.05) determined by one-way ANOVA with Tukey's multiple comparisons test Phenotypic and cellular characterization of the Stb1 mutant. a Gross morphology of wild-type (WT) and Stb1 plants in the field. Bar = 10 cm. b Stem bending phenotype, with arrows indicating the inner concave side (Stb1-Concave) and outer convex side (Stb1-Convex). Scale bar = 10 cm. c Stem bending angle of WT and Stb1. d Comparison of internode length between WT and Stb1 in maize mature period. 6i-18i indicates the number of internodes above the ground from bottom to top. e–g Stem cell observation of WT (e), concave side of Stb1 (f), convex side of Stb1 (g). Scale bar = 50 μm. h, i Comparison of cell size (h) and cell density (number per mm2) (i) in stem longitudinal sections (n = 3 biological replicates). j Comparison of mature ear of WT and Stb1. Scale bar = 10 cm. k Ear length of WT and Stb1. l Number of grains in rows of WT and Stb1. m, n Leaf width (m) and leaf length (n) of the leaf immediately above the uppermost ear (n = 20). o, p Abaxial leaf epidermal cells of WT (o) and Stb1 (p), showing cell shrinking in the mutant. Scale bar = 100 μm. q Stomatal density on the abaxial leaf surface of WT and Stb1 (n = 3 biological replicates). Data are presented as mean ± SD. In (c, d, k-n, q), statistical significance was determined by Student's t-test (*p < 0.05, **p < 0.01, ***p < 0.001). In (h, i), different lowercase letters indicate significant differences (p < 0.05) determined by one-way ANOVA with Tukey's multiple comparisons test To determine the genetic cause of the Stb1 mutation, an F2 population was developed by crossing Stb1 with KN5585 and further self-crossing. The F2 population segregated in a 3:1 (Stb1: WT) ratio, indicating that the stem bending phenotype was controlled by a single dominant locus (χ2 test, Fig. S1a). Bulk segregant RNA-seq (BSR-seq) initially mapped the Stb1 mutation to a ~ 37 Mb region on chromosome 1 with an allele frequency difference of 0.99 between pools (Fig. S1b-d). Subsequent fine mapping with 320 F2 individuals narrowed the locus to a 7.3 Mb interval (Chr1: 263.5–270.8 Mb) between markers M5 and M6 (Fig. 2a, Fig. S2a, b). Utilizing polymorphic InDel markers M10 and M11 (Fig. S2c, d) on 210 segregating plants further refined the candidate region to a 3.7 Mb interval (Chr1: 263.8–267.5 Mb) containing 69 annotated genes (Fig. 2a, Supplemental Table 3). Among these, three genes were differentially expressed based on the stem transcriptome of Stb1 (Fig. 2b). Notably, Zm00001d033543 encoding a phosphoglycerate mutase exhibited high stem expression in public RNA-seq datasets (Fig. 2c) (Yi et al. 2019; Yu et al. 2022; Sun et al. 2024), identifying it as a prime Stb1 candidate requiring further functional validation.Fig. 2Fine mapping of Stb1 and multi-dimensional analysis of downstream pathways. a Fine mapping of the Stb1 locus on chromosome 1, narrowing the candidate interval to 3.7 Mb between markers M10 and M11 (n, number of F2 segregating individuals). b Expression changes (log2 fold change) of the three candidate genes within the localized genomic window. c Tissue-specific expression profiles of the three candidate genes based on public datasets. d Venn diagram showing overlapping and distinct DEGs in Stb1-Concave vs. WT and Stb1-Convex vs. WT. e, f Expression changes (e) and tissue-specific expression profiles (f) of key DEGs involved in xylan transferase, cellulose synthesis, and lignin polymerization. g, h Quantification of hemicellulose (g) and lignin (h) contents in the stems of WT, Stb1-Concave, and Stb1-Convex. i FPKM values of DEGs associated with cell wall composition in Stb1 stems j, k Transmission electron microscopy (TEM) images (j) and quantification of secondary cell wall thickness (k) in WT, Stb1-Concave, and Stb1-Convex stem internodes (n = 3 biological replicates). Scale bars = 20 μm (left) and 5 μm. l Relative expression level of the ZmLA1 gene across genotypes and stem sides. m Endogenous IAA content in the stems of WT, Stb1-Concave, and Stb1-Convex. Stb1-Concave and Stb1-Convex represent the inner concave and outer convex sides of the bent stem, respectively. Data are presented as mean ± SD. Different lowercase letters indicate significant differences (p < 0.05) determined by one-way ANOVA with Tukey's multiple comparisons test Fine mapping of Stb1 and multi-dimensional analysis of downstream pathways. a Fine mapping of the Stb1 locus on chromosome 1, narrowing the candidate interval to 3.7 Mb between markers M10 and M11 (n, number of F2 segregating individuals). b Expression changes (log2 fold change) of the three candidate genes within the localized genomic window. c Tissue-specific expression profiles of the three candidate genes based on public datasets. d Venn diagram showing overlapping and distinct DEGs in Stb1-Concave vs. WT and Stb1-Convex vs. WT. e, f Expression changes (e) and tissue-specific expression profiles (f) of key DEGs involved in xylan transferase, cellulose synthesis, and lignin polymerization. g, h Quantification of hemicellulose (g) and lignin (h) contents in the stems of WT, Stb1-Concave, and Stb1-Convex. i FPKM values of DEGs associated with cell wall composition in Stb1 stems j, k Transmission electron microscopy (TEM) images (j) and quantification of secondary cell wall thickness (k) in WT, Stb1-Concave, and Stb1-Convex stem internodes (n = 3 biological replicates). Scale bars = 20 μm (left) and 5 μm. l Relative expression level of the ZmLA1 gene across genotypes and stem sides. m Endogenous IAA content in the stems of WT, Stb1-Concave, and Stb1-Convex. Stb1-Concave and Stb1-Convex represent the inner concave and outer convex sides of the bent stem, respectively. Data are presented as mean ± SD. Different lowercase letters indicate significant differences (p < 0.05) determined by one-way ANOVA with Tukey's multiple comparisons test To explore the downstream pathways of Stb1, we conducted transcriptome analysis on bolting-stage stems (|Log2 FC|> 1, FDR ≤ 0.05). We identified 463 differentially expressed genes (DEGs) on the concave side (162 up-regulated and 301 down-regulated) and 1,029 DEGs on the convex side (619 up-regulated and 410 down-regulated) compared to WT (Fig. 2d; Fig. S3a-b). Additionally, 840 DEGs were identified between the concave and convex sides (585 up-regulated and 255 down-regulated) (Fig. S3c). Gene Ontology (GO) analysis revealed significant enrichment in environmental stimulus responses, cell differentiation, and cell wall metabolism/biogenesis (Fig. S3d-f). Notably, eight cell wall-related genes were downregulated in Stb1 stems, including five xylan transferase, one cellulose synthase, and two lignin polymerization-related genes (Fig. 2e). Five of these genes (Zm00001d047750, Zm00001d024386, Zm00001d027313, Zm00001d043477, and Zm00001d052324) exhibited high stem-specific expression in public datasets (Fig. 2f), suggesting their potential roles in stem development. To evaluate cell wall properties, we quantified structural components in Stb1 stems. Hemicellulose content was elevated overall, with the concave side containing significantly more hemicellulose than the convex side (Fig. 2g). Compared to WT, lignin content was significantly higher on the convex side (Fig. 2h), providing structural rigidity for the rapidly expanding tissue. This was consistent with the upregulated expression of xylan transferase genes on the concave side (Fig. 2i), whereas cellulose content remained unchanged (Fig. S4a). Microscopic analysis further revealed that Stb1 developed significantly thicker secondary cell walls than WT (Fig. 2j, k). Collectively, these asymmetric alterations in cell wall composition and structure disrupt tissue mechanical balance, driving the stem bending phenotype. Our transcriptome analysis revealed that LAZY1 (LA1), a classical regulator of shoot gravitropism (Dong et al. 2013), was upregulated in Stb1 stems. Real-time qPCR confirmed that ZmLA1 transcripts were significantly more abundant on the concave side than those on the convex side (Fig. 2l). Given that ZmLA1 regulates auxin pathway, we quantified endogenous auxin levels and found that IAA significantly accumulated at a higher level on the convex side of Stb1 stems compared to those on the concave side and in the segregated WT (Fig. 2m). Consistently, the stem-preferential auxin-responsive gene SAUR71 (Zm00001d026308) was upregulated in Stb1, showing more pronounced induction on the convex side (Fig. S4b). These results demonstrate that Stb1 maintains upright stem growth by modulating ZmLA1 expression and proper auxin gradients across opposing stem sides. Discussion Stalk lodging represents a primary constraint to sustainable maize production under high-density cultivation. In this study, we characterized the stem-defective mutant Stb1, which displayed pleiotropic phenotypes, including stem bending, reduced plant height with shortened internodes, and smaller ears, etc. (Fig. 1). The asymmetric mechanical properties driving Stb1 stem bending likely originate from cell wall compositional imbalance. Transcriptome analysis revealed widespread dysregulation of genes associated with cell wall metabolic processes (Fig. 2e), with significant alterations in cellulose synthases and xylan biosynthetic genes critical for secondary cell wall assembly. Biochemical quantification showed that hemicellulose content was significantly higher on the concave side, whereas lignin content was elevated on the convex side. Hemicellulose normally provides structural flexibility via cross-linking with cellulose microfibrils, while lignin acts as a hydrophobic filler that rigidifies the cell wall matrix, restricting cell expansion and flexibility (Mengistie and McDonald 2024). This unbalanced composition likely alters the local elastic properties of opposing stem sides, which may in turn contribute to bending. Consistent with our observations, alterations in cell wall composition have been associated with stem bending phenotypes in several plant species. In Arabidopsis, overexpression of a pectin methylesterase inhibitor (PMEI) causes stems to develop twists and loops, accompanied by altered cell wall sugar composition and elasticity (Müller et al. 2013). In rice, THE FLEXIBLE CULM 1 (FC1) gene encodes a cinnamyl-alcohol dehydrogenase (CAD), a key enzyme in the lignin biosynthesis pathway. Loss of FC1 function leads to reduced lignin content and altered monomer composition, resulting in decreased cell wall thickness, which may compromise stem mechanical support and ultimately cause a flexible and bending stem phenotype (Li et al. 2009; Hirano et al. 2012). This observation echoes the asymmetric lignin deposition pattern on the convex and concave sides of Stb1 stems, suggesting that alterations in cell wall components, particularly lignin metabolism, may contribute to the bending phenotype by affecting the local mechanical properties of opposing stem sides. Together, these observations suggest that changes in cell wall composition and structure may disrupt tissue mechanical balance and flexibility, thereby contributing to stem bending or morphological abnormalities across diverse plant species. Asymmetric auxin distribution is a prerequisite for organ bending upon environmental stimuli (Rashotte et al. 2001). This process relies heavily on polar auxin transport (PAT) mediated by LAZY genes (Dong et al. 2013; Li et al. 2007). The localized upregulation of ZmLA1 on the concave side of Stb1 may impair lateral or basipetal PAT, potentially leading to hyper-accumulation of auxin on the convex side and contributing to the bending phenotype. Interestingly, cell wall integrity signaling and auxin have been shown to interact in regulating asymmetric growth (Walia et al. 2024). We propose that in Stb1, cell wall-derived mechanical cues and auxin-dependent growth signals may function in a coordinated manner to influence differential growth on opposing stem sides. Although the Stb1 locus has been successfully delimited to a 3.7 Mb candidate interval containing 69 annotated genes, identifying the exact functional candidate is the first focus of next work. The direct regulatory pathways of STB1 will then be further investigated after the causal gene is isolated. Nevertheless, our findings establish STB1 as a vital regulator of maize upright stem growth and shoot architecture. Elucidating the STB1-mediated coordination of cell wall properties and mechanical strength provides valuable insights into breeding lodging-resistant elite maize germplasm. Supplementary Information Supplementary Material 1: Supplementary Fig. S1 | Genetic analysis and BSR-seq gene mapping. a Genetic analysis of Stb1. b BSR-seq to obtain the primary localisation interval on maize chromosome 1. c Initial positioning results of Deep learning algorithm. The abscissa represents the length of chromosome 1-10 of maize, and the ordinate represents the calculated value of the algorithm. d Initial localization interval on chromosome 1 in maize. Left, peak, and right indicate the corresponding positions within the localization interval, and value represents the correlation score. Supplementary Fig. S2 | KASP and InDel markers for gene mapping. M5 represents KASP marker. M6, M10 and M11 represent Indel markers. a M5: 263.5Mb KASP; b M6: 270.8 Mb; c M10: 263.8Mb; d M11: 267.5Mb. W represents wild type (KN5585), and M represents Stb1 mutant (B104), F1 represents KN5585/B104. Supplementary Fig. S3 | Volcano plot of DEGs and GO enrichment in Stb1 a-c Volcano plots displaying differentially expressed genes (DEGs) between (a) Stb1 concave and WT (301 down-, 162 up-regulated), (b) Stb1 convex and WT (410 down-, 619 up-regulated), and (c) Stb1 concave and Stb1 convex (255 down-, 585 up-regulated). d-f GO enrichment analyses of DEGs for the respective comparisons in (a-c). Supplementary Fig. S4 | Cellulose content and expression profile of the auxin-responsive gene SAUR71 in Stb1 a Cellulose content in the stems of wild type (WT) and the concave and convex sides of Stb1 bending stems. Data are presented as mean ± SD. Different lowercase letters indicate statistically significant differences (P< 0.05). b Expression analysis of SAUR71 (Zm00001d026308) in WT and the concave and convex sides of Stb1 stems based on RNA-seq profiling. c Tissue-specific expression pattern of Zm00001d026308 across various developmental tissues in maize (data adapted from Han et al. 2023). Color scale represents relative expression levels.Supplementary Material 2. Supplementary Material 1: Supplementary Fig. S1 | Genetic analysis and BSR-seq gene mapping. a Genetic analysis of Stb1. b BSR-seq to obtain the primary localisation interval on maize chromosome 1. c Initial positioning results of Deep learning algorithm. The abscissa represents the length of chromosome 1-10 of maize, and the ordinate represents the calculated value of the algorithm. d Initial localization interval on chromosome 1 in maize. Left, peak, and right indicate the corresponding positions within the localization interval, and value represents the correlation score. Supplementary Fig. S2 | KASP and InDel markers for gene mapping. M5 represents KASP marker. M6, M10 and M11 represent Indel markers. a M5: 263.5Mb KASP; b M6: 270.8 Mb; c M10: 263.8Mb; d M11: 267.5Mb. W represents wild type (KN5585), and M represents Stb1 mutant (B104), F1 represents KN5585/B104. Supplementary Fig. S3 | Volcano plot of DEGs and GO enrichment in Stb1 a-c Volcano plots displaying differentially expressed genes (DEGs) between (a) Stb1 concave and WT (301 down-, 162 up-regulated), (b) Stb1 convex and WT (410 down-, 619 up-regulated), and (c) Stb1 concave and Stb1 convex (255 down-, 585 up-regulated). d-f GO enrichment analyses of DEGs for the respective comparisons in (a-c). Supplementary Fig. S4 | Cellulose content and expression profile of the auxin-responsive gene SAUR71 in Stb1 a Cellulose content in the stems of wild type (WT) and the concave and convex sides of Stb1 bending stems. Data are presented as mean ± SD. Different lowercase letters indicate statistically significant differences (P< 0.05). b Expression analysis of SAUR71 (Zm00001d026308) in WT and the concave and convex sides of Stb1 stems based on RNA-seq profiling. c Tissue-specific expression pattern of Zm00001d026308 across various developmental tissues in maize (data adapted from Han et al. 2023). Color scale represents relative expression levels. Supplementary Material 2.