A maize epimerase modulates cell wall synthesis and glycosylation during stomatal morphogenesis

Zhou, Yusen; Zhang, Tian; Wang, Xiaocui; Wu, Wenqiang; Xing, Jingjing; Li, Zuliang; Qiao, Xin; Zhang, Chunrui; Wang, Xiaohang; Wang, Guangshun; Li, Wenhui; Bai, Shenglong; Li, Zhi; Suo, Yuanzhen; Wang, Jiajia; Niu, Yanli; Zhang, Junli; Lan, Chen; Hu, Zhubing; Li, Baozhu; Zhang, Xuebin; Wang, Wei; Galbraith, David W.; Chen, Yuhang; Guo, Siyi; Song, Chun-Peng

doi:10.1038/s41467-023-40013-6

Download PDF

Article
Open access
Published: 20 July 2023

A maize epimerase modulates cell wall synthesis and glycosylation during stomatal morphogenesis

Nature Communications volume 14, Article number: 4384 (2023) Cite this article

4036 Accesses
3 Citations
12 Altmetric
Metrics details

Subjects

Abstract

The unique dumbbell-shape of grass guard cells (GCs) is controlled by their cell walls which enable their rapid responses to the environment. The molecular mechanisms regulating the synthesis and assembly of GC walls are as yet unknown. Here we have identified BZU3, a maize gene encoding UDP-glucose 4-epimerase that regulates the supply of UDP-glucose during GC wall synthesis. The BZU3 mutation leads to significant decreases in cellular UDP-glucose levels. Immunofluorescence intensities reporting levels of cellulose and mixed-linkage glucans are reduced in the GCs, resulting in impaired local wall thickening. BZU3 also catalyzes the epimerization of UDP-N-acetylgalactosamine to UDP-N-acetylglucosamine, and the BZU3 mutation affects N-glycosylation of proteins that may be involved in cell wall synthesis and signaling. Our results suggest that the spatiotemporal modulation of BZU3 plays a dual role in controlling cell wall synthesis and glycosylation via controlling UDP-glucose/N-acetylglucosamine homeostasis during stomatal morphogenesis. These findings provide insights into the mechanisms controlling formation of the unique morphology of grass stomata.

A plant NLR receptor employs ABA central regulator PP2C-SnRK2 to activate antiviral immunity

Article Open access 13 April 2024

Remote loop evolution reveals a complex biological function for chitinase enzymes beyond the active site

Article Open access 15 April 2024

Spatial co-transcriptomics reveals discrete stages of the arbuscular mycorrhizal symbiosis

Article Open access 08 April 2024

Introduction

Plant cell walls provide mechanical strength and modulate the directionality of cell growth, thereby determining cell shape and influencing their functions¹. The dynamic cell wall structure plays a key role in modulating plant growth and development and also responses to biotic and abiotic stresses^2,3,4. Wall formation involves synthesis, assembly, modification, and reorganization of different classes of wall polysaccharides and proteins, a process that is precisely coordinated and controlled^5,6.

Stomata are an excellent model for studying how alterations of cell walls modulate coordinated cell movement in response to environmental changes. Stomata provide the “gateway” for gas exchange between plants and their environment, regulating photosynthesis and transpiration and thereby influencing global plant carbon and water cycles^7,8. Unlike other epidermal cells, the unique functions of stomata require substantial and repeated wall extension during their lifetime. The mechanical strength of the stomatal wall needs to withstand the high turgor pressure (0.7–5 MPa)^9,10, much larger than that of other plant cells (0.1–1 MPa). How stomatal walls are strong yet reversibly extensible, ensuring correct stomatal formation and precise movement, is a fascinating research question in the field. Compared to the kidney-shaped guard cells typical of eudicots, the dumbbell-shaped graminoid guard cells (GCs) are flanked by two subsidiary cells (SCs)^11,12,13. Their presence allows the stomatal aperture to open and close more rapidly than that of non-graminoid stomata, increasing gas exchange and water-use efficiencies that are characteristic of grasses^14,15,16.

In both the kidney-shaped dicot GCs and the dumbbell-shaped grass GCs, turgor pressure arising from water influx and wall mechanics drives volume changes of the GCs, while the uneven thickening of the GC cell walls and the arrangement of cellulose microfibrils within these walls act together to control the direction of GC deformation¹⁷. In grasses, the elongated median regions of the dumbbell-shaped GCs have locally thickened cell walls at the region of the central canal, whereas the two bulbous ends have thinner walls. Cellulose microfibrils within the central canal region are arranged parallel to the GC long axis, and radiate from the central canal to the bulbous ends^18,19,20,21. When GC turgor rises, the bulbous ends expand and push the two GCs apart in the plane of the leaf epidermis, resulting in a stomatal opening. Grass stomatal development can be divided into six developmental stages, characterized by the sequential presence of epidermal progenitor cells, guard mother cells (GMCs), subsidiary mother cells (SMCs), subsidiary cells (SCs), young stomata (YS), and mature stomatal complexes^14,22,23. It remains poorly understood how changes in the composition and structure of the stomatal walls give rise to the unique shape and functions of grass GCs.

In higher plants, it is generally established that cellulose microfibrils are extruded extracellularly by cellulose synthase complexes (CSCs) moving on the plasma membrane. In contrast, matrix polysaccharides (hemicelluloses and pectins) are synthesized by glycosyltransferases (GTs), and are secreted to cell walls via vesicular transport⁵. Cell wall synthesis requires coordinated activities of sugar transporters and GTs in the Golgi apparatus²⁴. Activated nucleotide sugars are the fundamental building blocks for the synthesis of cell wall carbohydrate polymers, their amounts being regulated by the activities of nucleotide sugar interconversion enzymes (NSEs). In plants, the presence of eleven NSE families has been confirmed, providing up to 16 different nucleotide sugar donors^25,26. NSE family members include the UDP-glucose 4-epimerase (UGE) family, the members of which catalyze the reversible interconversion of UDP-glucose (UDP-Glc) and UDP-galactose (UDP-Gal). The UGE family is highly conserved in protein sequence and structure²⁷. UGEs in Arabidopsis function in cell wall synthesis, and are involved in vegetative and reproductive growth and development²⁸. In rice, mutation of OsUGE2 resulted in brittle leaves and culms²⁵. It is yet unknown whether the activities of UGEs also affect stomatal wall synthesis.

To address this question, we screened for maize mutants having abnormal stomatal function and morphology (bizui, bzu) retrieved from an ethyl methanesulfonate (EMS) mutagenesis population, and from the maizeGDB mutagenesis database (https://www.maizegdb.org/)²³. In this study, we identified bizui3 (bzu3), a maize mutant having collapsed GC walls and defects in local wall thickening. BZU3 encodes a UDP-Glc 4-epimerase and plays a key role in controlling cell wall synthesis during the morphogenesis of maize stomata. It also catalyzes epimerization of UDP-N-acetylgalactosamine (UDP-GalNAc) to UDP-N-acetylglucosamine (UDP-GlcNAc), regulating N-glycosylation patterns of many proteins that may be involved in cell wall synthesis and signaling pathways. Our results suggest a bifunctional role of maize epimerase in cell wall synthesis and glycosylation during stomatal morphogenesis.

Results

bzu3 plants display abnormal GCs and severely impaired stomatal function

Among stomatal deficient mutants emerging from the screen, two recessive single mutants were identified that had a translucent leaf appearance and that would die after growing for about 21 days (Fig. 1a, Supplementary Fig. S1a, and Supplementary Tables S1, S2). These two mutants had similar abnormal GC phenotypes (Fig. 1b and Supplementary Fig. S1b), and were verified as allelic mutants (Supplementary Table S3). We named these mutants bzu3-1 (5808E from the maizeGDB database), and bzu3-2 (an EMS mutant in the Zheng58 background). Intriguingly, the guard cells in bzu3 mutants can be divided into two classes. Most (91.46 ± 5.03%) of the guard cells in bzu3-1 mutants were antenna-shaped, with the central canals having collapsed onto a line, leaving the remains of the two bulbous ends. A small percentage (8.54 ± 5.03%) of the guard cells were rod-shaped, but the local wall thickening at the central canals was still absent. The ratios of abnormal GCs were similar in mutant bzu3-2 (93.41 ± 3.73% antenna-shaped, and 6.59 ± 3.73% rod-shaped) (Supplementary Fig. S1c). By contrast, we counted 100% dumbbell-shaped GCs in the wild-type (WT) plants (Fig. 1b, c and Supplementary Fig. S1b–d). No differences in stomatal density were apparent between WT and bzu3-1, Zheng58, and bzu3-2 plants (Supplementary Fig. S1e). Since plant cell walls provide the prime constraint on cell shape, we examined the cell wall structure using transmission electron microscopy. In bzu3-1 plants, the GC walls collapsed and were much thinner (Fig. 1d). No significant differences were observed in the wall thickening of vascular bundles in leaves (Supplementary Fig. S2), implying the changes in cell wall thickening in the mutant were GC-specific. To define the stage of stomatal development of GCs at which the deformation occurred, we examined all stages of GC development of the stomatal lineage cells in the leaf base. As compared to the WT, the stomatal lineage cells in the mutant appeared normal until the later stages of stomatal elongation. Formation of the SC and of the pore aperture was not affected in the mutant (Fig. 1e). To further analyze the bzu3 phenotype, we tested stomatal conductances under different light conditions and levels of CO₂. For the wild-type plants, strong lighting and low CO₂ levels induced stomatal opening, whereas darkness and high CO₂ levels induced stomatal closing. In contrast, the stomata of bzu3 mutants remained closed and did not respond to changes in lighting and CO₂ levels. Thus, the abnormal GCs of bzu3 resulted in stomatal malfunctioning (Fig. 1f, g).

**Fig. 1: Mutant *bzu3* displays abnormal guard cell shapes and is defective in local wall thickening.**

BZU3 encodes a UDP-Glc 4-epimerase that specifically functions during GC maturation

To identify the gene responsible for the bzu3 phenotype, we generated parental populations between inbred line B73 and plants carrying the mutation. Because homozygous bzu3 mutant plants are lethal, we employed heterozygous bzu3-1+/- populations for crossing to B73. Map-based cloning using F₂ segregation populations identified the mutation in the 60.67-60.77 M region on the long arm of Chromosome 1, a region containing only one gene (Zm00001d029151; Fig. 2a). Genomic PCR amplification revealed a long transposon insertion in the second intron of Zm00001d029151 of bzu3-1 mutant plants (Fig. 2b and Supplementary Fig. S3a–c). With primers designed for several regions of Zm00001d029151, RT-qPCR failed to detect any measurable transcripts from the sequences following the site of transposon insertion (Supplementary Fig. S3b, d). Sanger sequencing of bzu3-2 mutant plants revealed a 1-base pair deletion in the 7^th exon of Zm00001d029151, the same gene identified as defective in bzu3-1 mutants (Fig. 2b and Supplementary Fig. S4a). The BZU3^bzu3-2 mutation results in a frame shift in the C terminus of the protein, and is predicted to severely impact protein stability (Supplementary Fig. S4b).

**Fig. 2: Identification and characterization of *BZU3* as the gene responsible for normal guard cell development in maize.**

Three additional independent bzu3 mutants (bzu3-3, bzu3-4, and bzu3-5) were generated in the B73-329 background using CRISPR-Cas9. All recapitulated the stomatal phenotype of bzu3 mutants (Fig. 2b, c and Supplementary Fig. S5). To confirm the identity of the gene responsible for the phenotype, genetic complementation using the full-length CDS of Zm00001d029151 under its native promoter fully rescued the stomatal phenotype in bzu3-2, resulting in 100% dumbbell-shaped GCs, and also completely rescued the growth phenotypes of bzu3-2 (Fig. 2d and Supplementary Figs. S6a, S7a). These results strongly support the proposition that the stomatal phenotype of bzu3 mutant plants is due to defects in Zm00001d029151, which we renamed BZU3. BLAST searches and phylogenetic analyses indicated that BZU3 encodes a UDP-glucose 4-epimerase (UGE) that shares homologs with Arabidopsis thaliana, rice and many other species across the plant kingdom (Supplementary Fig. S8).

To study the subcellular localization of BZU3, we performed transient expression of the BZU3-YFP fusion protein in maize mesophyll protoplasts. The results indicated that BZU3 was localized to the cytoplasm (Supplementary Fig. S9). The expression pattern of BZU3 in vivo was also observed in BZU3pro:BZU3-YFP;bzu3-2 transgenic plants. During the early stages of stomatal development, BZU3 was expressed in all epidermal cells. At later stages of stomatal development, the BZU3 signal gradually increased in GCs, being completely restricted to the GCs at the final stage of maturation (Fig. 2e and Supplementary Fig. S10). Statistical analysis of the BZU3-YFP signal indicated that the expression of BZU3 notably increased from the point of GMC division to the stomatal elongation stage (Supplementary Fig. S10). To examine whether BZU3 functions specifically in GCs and to determine the precise stage of GC development at which BZU3 function is required, two stomatal-specific promoters, BZU2/ZmMUTEpro and ZmFAMApro²³, were used to drive BZU3 for complementation testing. ZmMUTEpro:BZU3-YFP was expressed in GMCs and young guard cells until the elongation stage, but was barely detectable in mature GCs (Fig. 2f). ZmFAMApro:YFP-BZU3 was expressed in GCs from the symmetrical division of GMCs to the point of GC maturation, but not during early stomatal development (Fig. 2g). The expression patterns of these two stomatal-specific promoters are temporally different but overlapping. BZU3 driven by either promoter fully rescued the stomatal phenotype in bzu3-2 (Supplementary Fig. S6b, c). Importantly, ZmMUTEpro:BZU3-YFP and ZmFAMApro:YFP-BZU3 could completely rescue the growth phenotypes of bzu3-2 (Supplementary Fig. S7b, c). Considering that abnormal bzu3 GCs appeared only at the later stage of stomatal elongation (Fig. 1e), our results suggest that BZU3 functions in GC morphogenesis during the specific time window at which the young guard cells elongate.

BZU3 catalyzes interconversions of UDP-Gal/GalNAc and UDP-Glc/GlcNAc, affecting stomatal wall synthesis

Previous studies have shown that UDP-glucose 4-epimerases catalyze reversible conversions of UDP-Gal and UDP-Glc, and UDP-GalNAc and UDP-GlcNAc^26,29. To characterize the enzymatic activity of BZU3, the recombinant protein was purified (Supplementary Fig. S11). With UDP-Gal as substrate, the K_m (Michaelis–Menten constant), V_max (the maximum reaction velocity), and K_cat/K_m were 1.45 mM, 1.41 mM·min⁻¹, and 650.3 s⁻¹·mM⁻¹ respectively. K_m, V_max, and K_cat/K_m were 2.67 mM, 0.69 mM·min⁻¹, and 172.4 s⁻¹·mM⁻¹, respectively, when using UDP-Glc as substrate (Supplementary Fig. S12a, b and Supplementary Table S4). BZU3 also reversibly catalyzed the conversion of UDP-GalNAc to UDP-GlcNAc. K_m, V_max, and K_cat/K_m were 0.50 mM, 0.36 mM·min⁻¹, and 474.7 s⁻¹·mM⁻¹, respectively, when using UDP-GalNAc as the substrate, whereas K_m, V_max, and K_cat/K_m were 0.40 mM, 0.10 mM·min⁻¹, and 168.6 s⁻¹·mM⁻¹ with UDP-GlcNAc as the substrate (Supplementary Fig. S12c, d and Supplemental Table S4). BZU3 converted UDP-Gal/UDP-GalNAc to UDP-Glc/UDP-GlcNAc faster in the reversible isomerization reaction. When the reaction reached equilibrium, the ratio of UDP-Glc:UDP-Gal or UDP-GlcNAc:UDP-GalNAc was approximately 3:1 (Fig. 3a, b and Supplementary Fig. S13). Previous studies have shown that AtUGE1 and AtUGE3 can interconvert UDP-Xyl/UDP-L-Ara³⁰. We employed a similar approach to test whether BZU3 interconverts UDP-Xyl/UDP-L-Ara. Our data indicate that whereas BZU3 can convert about 45% of UDP-Xyl to UDP-L-Ara, BZU3 was almost completely incapable of conversion of UDP-L-Ara to UDP-Xyl under our experimental conditions (Supplementary Fig. S14).

**Fig. 3: BZU3 catalyzes conversion of UDP-Gal to UDP-Glc in vitro, and *bzu3* mutants show defective cell wall synthesis.**

Since bzu3 GCs showed defects in local wall thickening and BZU3 catalyzes the conversion of UDP-Gal to UDP-Glc, the main nucleotide sugar for cell wall synthesis, we examined different wall components present in maize stomata. Stomata at later stages of development were stained with Direct Red 23, a dye that specifically identifies cellulose^31,32. In WT, the red fluorescent signal in GC walls gradually increased, indicating deposition of cellulose and GC wall thickening (Fig. 3c, d. P < 0.001). No gradual accumulation of cellulose signals was observed in bzu3 plants. Immunofluorescence labeling suggested that the amount of mixed-linkage glucan (MLG) in the GC wall of bzu3 was also significantly reduced as compared to the WT (Fig. 3e, f. P < 0.001). The cell wall composition of mature stomata was also analyzed using stimulated Raman scattering microscopy (SRS). The intensity of the Raman signals for cellulose and hemicellulose were significantly lower in GCs of bzu3 than that of the WT (Fig. 3g, h. P < 0.001). Immunofluorescence labeling indicated that the amount of homogalacturonan in the GC wall was very low, and it was difficult to determine whether there was a difference between the wild type and the mutant (Supplementary Fig. S15). Collectively, these results suggest that severely impaired cellulose and MLG synthesis leads to defects in local wall thickening and collapse of maize GCs in bzu3 plants.

Structures of BZU3 docked with NAD⁺/UDP-Glc and NAD⁺/UDP-GlcNAc

To determine whether BZU3 also catalyzes the conversion of UDP-Gal to UDP-Glc in vivo, we performed UPLC-MS analysis of nucleotide sugars in the maize leaf base, finding that levels of UDP-Gal were remarkably increased, and UDP-Glc levels reduced, in bzu3 as compared to the WT (Fig. 4a. P < 0.001). This suggests that mutation of BZU3 results in a disturbed balance between the cellular concentrations of UDP-Gal and UDP-Glc. To test whether a reduced level of UDP-Glc led to the stomatal phenotype in bzu3, exogenous UDP-Glc was applied at the maize leaf base. Some stomata (~10%) recovered the normal dumbbell shape, indicating that UDP-Glc plays a crucial role in maize GC development (Fig. 4b).

**Fig. 4: BZU3 catalyzes UDP-Gal epimerization to UDP-Glc in vivo, and UDP-Glc is crucial for maintaining the dumbbell-shape of maize guard cells.**

To further address the molecular functions of BZU3, we solved the 3-D structures of the BZU3 protein, both in apo form and in complexes with NAD⁺/UDP-Glc and NAD⁺/UDP-GlcNAc, at resolutions of 2.6 Å, 1.25 Å, and 2.15 Å respectively (Supplementary Table S5). BZU3 formed a homo-dimer (Supplementary Fig. S16a, b), with each protomer folded into two distinct structural units, the N-terminal NAD⁺ binding domain (residues 1-189) and the C-terminal sugar-binding domain (residues 190-355). The catalytic site was located in the cleft between these two structural domains (Fig. 4c). The homo-dimerization structure was consistent with in vivo bimolecular fluorescence complementation (BiFC) results (Supplementary Fig. S16c). BZU3^bzu3-2 failed to form homo-dimers, probably due to the frameshift which would be expected to alter the structural folding of its C-terminal domain (Supplementary Fig. S16d). The flexible sugar moiety was held in place in the vicinity of the catalytic triad of Ser-Tyr-Lys (S133/Y157/K161). S133 forms a bifurcated contact with the sugar moiety of the substrate, whereas both Y157 and K161 directly interact with the ribose moiety of NAD⁺ (Fig. 4c). The maize BZU3 shared similar structures with the GALE proteins from humans and E. coli (Supplementary Fig. S17). The highly conserved NAD⁺ binding motifs (G¹³XXG¹⁶XXG¹⁹ and Y¹⁵⁷XXXK¹⁶¹) of the UGE family were indicated in the aligned sequences. The structure of BZU3 in complex with NAD⁺/UDP-GlcNAc was similar to that of BZU3/NAD⁺/UDP-Glc, except that the sugar moiety of the substrate adopts a distinct orientation (Fig. 4d).

Drawing on the structural model, we designed mutations within the conserved catalytic triad and found that S133A significantly reduced catalytic activity, while Y157F completely abolished the inter-conversion between UDP-Glc and UDP-Gal, and UDP-GlcNAc and UDP-GalNAc (Fig. 3a, b, Supplementary Fig. S13, and Supplementary Table S6). Consistent with this result, the transgene of BZU3pro:BZU3^Y157F-YFP failed to rescue the stomatal phenotype in bzu3-2 (Fig. 4e and Supplementary Figs. S6d, S7d). Taken together, these observations imply that the catalytic activity of BZU3 is required for the morphogenesis of maize GCs.

Mutation of BZU3 alters N-glycosylation pattern

Previous studies have shown that UDP-GlcNAc is the main donor for N-glycosylation³³. Considering that BZU3 also catalyzes UDP-GalNAc to UDP-GlcNAc in vitro, we undertook glycoproteomics analyses of the N-glycans. Based on the libraries of maize protein group and plant N-glycosylation modification in the Uniprot database, we identified 2,194 intact N-glycopeptides, and quantified 181 differentially expressed intact N-glycopeptides (DEGPs) (two out of the three technical replicates, ≥ 1.5-fold change and p-value < 0.05) (Supplementary Fig. S18). A total of 173 proteins decorated by all three N-glycan types were down-regulated in bzu3-2, while only 8 proteins decorated by the core N-glycan type were up-regulated (Fig. 5a). Consistent with this result, down-regulated N-glycan species in the mutants were also much more abundant as compared to their up-regulated counterparts. As expected, N-glycan species containing xylose and fucose were largely reduced in the mutants as compared to the wild-type (Fig. 5b). These two monosaccharide modifications play important roles in a wide variety of biological processes, including immune responses, signal transduction, and cell development^34,35,36. Gene ontology analysis revealed that these affected N-glycoproteins were enriched in categories related to metabolism and biosynthesis of cell walls, cell-surface receptors, signaling, etc (Fig. 5c). Mass spectrometric analysis of the differentially expressed proteins also identified many glycosyltransferases and leucine-rich repeat receptor-like kinases (Supplementary Data S1). Within the differentially expressed proteins, e.g. PAN2 (PANGLOSS2), GT14 (Glycosyltransferase 14), and FUT (Fucosyltransferase, Zm00001d014505), RPLC-MS/MS analysis demonstrated that the N-glycosylation sites of PAN2/GT14/FUT were N554/N283/N121, respectively (Fig. 5d–f). Furthermore, we performed a gel shift assay of proteins generated by site-directed mutagenesis. PAN2^N554Q/GT14^N283Q/FUT^N121Q proteins migrated faster than PAN2/GT14/FUT proteins (Supplementary Fig. S19a–c), indicating that PAN2, GT14, and FUT can indeed be N-glycosylated.

**Fig. 5: Glycoproteomics analysis identifies altered glycosylation in *bzu3* mutants.**

Discussion

In this study, we provide cellular and structural evidence that BZU3, a maize UDP-glucose 4-epimerase, plays a key role in modulating stomatal morphogenesis. Firstly, we demonstrated that BZU3 is responsible for supplying UDP-Glc for stomatal wall formation, thus ensuring normal stomatal function. Secondly, in vitro enzymatic studies and crystallography results indicate that BZU3 catalyzes the conversion of UDP-Gal/GalNAc to UDP-Glc/GlcNAc. Lastly, our glycomics data suggest that BZU3 affects N-glycosylation of many proteins that are involved in wall formation and signaling.

Grass stomata can respond to environmental changes more rapidly, which may contribute to their widespread and diversification during a global aridification 30-40 million years ago¹³. Their characteristic dumbbell shape, along with the uneven pattern of cell wall thickening, raises questions as to the building blocks required for graminoid stomatal wall synthesis, and how cell wall composition and structure modulate grass stomatal shape and function. Our biochemical data indicated that the epimerase activity of BZU3 prefers to catalyze conversion of UDP-Gal to UDP-Glc (Fig. 3). In bzu3 mutants, local wall thickening at the central canal region of the dumbbell-shaped stomata is severely impaired (Fig. 1). Further analysis revealed reduced levels of UDP-Glc and reduced labeling of cellulose and mixed-linkage glucan in the stomatal wall (Fig. 3). These results suggest that BZU3 is the key player in providing UDP-Glc for maize stomatal wall formation, which is essential for stomatal morphogenesis and function. In most cell types, UDP-Glc is predominantly supplied to plasma membrane-associated CSCs through pathways involving sucrose synthases (SuSy) and UDP-glucose pyrophosphorylase (UGPase)^24,37. The alternative or compensatory supply of UDP-Glc provided by BZU3 in maize guard cells suggests that these precisely spatiotemporal-regulated epimerase activities evolved to determine the special shapes and functions of maize stomates. Cell wall composition and structure, with stiff cellulose microfibrils embedded in a soft matrix, are suggested to be very important for normal stomatal functioning in many other plants, including Arabidopsis, Commelina, and mosses^17,19,38. Defects in cellulose synthesis due to inhibition or mutation of CSCs or degradation of cellulose result in significantly larger stomatal apertures³⁹.

BZU3 belongs to the UGE superfamily, whose members are highly conserved in gene sequence and structure (Supplementary Fig. S8)²⁷. In Arabidopsis, there are five UGEs that are suggested to affect vegetative growth and pollen development. No defects related to stomatal morphology were found in atuge2, atuge4, and atuge2 atuge4 mutants (Supplementary Fig. S20)²⁸. There are at least four putative UGEs in rice²⁸. Previous studies have suggested that the function of OsUGE2 is related to cell wall synthesis²⁵, but no stomatal defects were observed in osuge2 mutants (Supplementary Fig. S21). Our study in rice found that mutation of OsUGE1, which has the closest phylogenetic relationship with BZU3, also did not result in stomatal defects (Supplementary Fig. S21). Recently, Wang et al. found that OsUGE1, as a transcriptional activation factor, is crucial for the degradation of tapetum and male fertility in rice⁴⁰. Meanwhile, Yang et al. found that OsUGE1 is negatively regulated in rice root hair elongation by targeted regulation of OsGRF6⁴¹. Intriguingly, mutagenesis of rice OsUGE3 also led to abnormal stomata (antenna-shaped and rod-shaped) (Supplementary Fig. S21). Although there are four additional UGE homologs in maize, we identified only two GC UGEs (including BZU3) in the single-nucleus transcriptome⁴². Compared to Arabidopsis UGEs, BZU3 is specifically required at the later stage of stomatal elongation, a time when the GC wall becomes locally thickened at the central canal, a process that would require copious amounts of UDP-Glc to accommodate the increased activity in cell wall synthesis. The isomerization of UDP-Gal to UDP-Glc by BZU3 might conveniently satisfy this requirement. Considering the differences between grasses and dicots in stomatal shape, wall composition, and structure, the spatiotemporal-regulated epimerase activities of maize BZU3 may have evolved to accommodate the highly specialized shape and function of grass stomata.

Upon exposure to exogenous UDP-Glc, the dumbbell-shaped stomata of the BZU3 mutants were not completely recovered. One simple explanation is that the exogenous supply of UDP-Glc was not sufficiently taken up by the cells, but since exogenous UDP-Glc was capable of restoring the growth phenotype of at least some UDP-Glc deficient plants, we examined other possibilities^43,44. An alternative explanation for the partial recovery of the stomatal phenotype may be that BZU3 controls other regulatory events, such as post-translational modifications of proteins, during stomatal development. Post-translational modification, for example, N-glycosylation, ensures correct folding and normal functions of proteins. Thus, an alternative role of BZU3 may relate to its catalytic activity regulating the level of UDP-GlcNAc, an important donor for N-glycosylation. Indeed, crystallized BZU3 protein bound to NAD⁺/UDP-GlcNAc (Fig. 4), and our glycoproteomics data identified many proteins whose glycosylation patterns were affected by BZU3 mutation, including glycosyltransferases (GT14 and FUT), leucine-rich receptors PAN2, and wall-associated kinases as suggested by GO analysis (Fig. 5 and Supplementary Data S1). Therefore, an alternative hypothesis is that a key developmental step prior to extensive cell wall modification, maturation, and thickening are affected.

The glycosylation site of PAN2 at N554 was further confirmed by gel-shift assays (Supplementary Fig. S19a). We hypothesized that, similar to the pan2 mutant⁴⁵, bzu3 mutants with abnormal glycosylation patterns of PAN2 may show similar phenotypes in the shape of subsidiary cells and inter-stomatal cells. As shown in Supplementary Fig. S19d, a typical triangular subsidiary cell has the following wall segments, one is the inter-stomatal cell shared with subsidiary cells (marked in yellow and named as x), and the other one is flanking non-stomatal row cells which form one of three points becoming the triangular subsidiary shape (marked in red and named as y). The bzu3 mutant subsidiary cell shapes differ slightly from the wild-type (Zheng58), with a significantly shorter wall (x) of subsidiary cells in bzu3-2 (14.9 ± 0.1 μm) compared with the wall segments of wild-type (16.4 ± 0.3 μm) (Supplementary Fig. S19e). By contrast, the length of wall y in the bzu3-2 mutant was not significantly different from Zheng58 (13.2 ± 0.1 μm versus 13.4 ± 0.4 μm) (Supplementary Fig. S19f). These observations showed similar a phenotype in the shape of subsidiary cells in bzu3 and pan2 mutants, implying that the glycosylation of PAN2 was altered in bzu3 which, in affecting the function of PAN2, results in a phenotype in bzu3 that is partially similar to the shape of subsidiary cells seen with pan2. It warrants further investigation to provide details concerning the regulatory role of BZU3 during stomatal morphogenesis via signaling and other mechanisms that are uncovered.

In summary, our results demonstrate a key role for BZU3, as an epimerase, in modulating the level of UDP-Glc for maintaining the dumbbell shape of maize GCs. In addition, BZU3 also influences the N-glycosylation patterns of many glycosyltransferases and cell-surface receptors that may be involved in the synthesis of stomatal wall matrix or signaling events during stomatal morphogenesis. Thus, the mutation of BZU3 leads to defects in stomatal walls and affects the morphogenesis and functions of maize GCs (Fig. 6). Since the unique dumbbell-shaped grass stomata are more efficient in regulating gas exchange, the genetic basis and regulatory mechanisms of grass stomatal morphogenesis will provide valuable insights for improving the efficiency of stomata and facilitate targeted-molecular breeding in a changing climate.

**Fig. 6: A proposed working model for BZU3.**

Methods

Plant materials and growth conditions

The bzu3-1 mutant in the 5808E background was obtained from the Maize Genetics Cooperation Stock Center (http://maizecoop.cropsci.uiuc.edu/). The bzu3-2 mutant in the Zheng58 background was isolated from an EMS (ethyl methanesulfonate) mutagenized database. Both bzu3-1 and bzu3-2 mutants were identified from self-crossed heterozygotes. Seedlings of WT, bzu3-1, Zheng58, and bzu3-2 were planted in the greenhouse at 16 h/8 h and 32 °C/25 °C (day/night). For all field trials, WT, bzu3-1+/-, Zheng58, bzu3-2+/- plants were planted in Kaifeng, Henan in summer, and in Sanya, Hainan in winter.

Stomatal phenotyping by microscopy

For stomatal density and developmental stage analyses, the second leaf of 7-day-old seedlings was collected and treated overnight with ethanol/acetic acid solution (7:1, v/v). The stomata on the abaxial side of the leaf epidermis were observed using an Olympus CX43 microscope. Under the 10× objective, we counted stomata from each leaf of 10 plants (5–7 vision fields per leaf). The visual field under the 10× objective was fully occupied by the leaf epidermis, and its area was 1.142 mm². The stomatal density was calculated as the number of stomata per area of the visual field. Different types of stomata were identified using the 20× objective, and we counted ~1000 stomata from 10 plants. To observe the lineages of the stomatal cells throughout stomatal development, young leaf segments taken from the bottom 2–3 cm of the third leaf of 7-day-old seedlings were cut into 0.2 cm × 0.5 cm pieces and stained for 5 minutes with 0.01 mg/mL propidium iodide (PI). After rinsing with ddH₂O, the samples were observed by a Zeiss LSM710 confocal microscope with excitation at 561 nm.

Transmission electron microscopy was mainly based on Yadav et al. with slight modifications⁴⁶. The middle part of the second leaf was collected from the 7-day-old seedlings. The leaf samples were fixed immediately in 5% glutaraldehyde and 4% paraformaldehyde buffered with 0.1 M sodium phosphate, pH 7.2. After being rinsed in the same buffer, the samples were postfixed with 1% osmium tetroxide in 0.1 M sodium phosphate, pH 7.2 for 3 h at 4 °C. Samples were dehydrated in an ethanol gradient series, then transferred to acetone, and finally infiltrated and embedded with Epon812 epoxy resin. Ultra-thin sections were obtained using a Leica EM UC7 ultramicrotome, and were stained with uranyl acetate and lead citrate. The sections were observed under a transmission electron microscope (FEI, TECNAI G2 Spirit Bio TWIN).

Stomatal conductivity

Seedlings were grown to the 3-leaf stage, the second leaf being used to measure stomatal conductivity with a Li-COR LI-6800 photosynthetic apparatus. The parameter settings were: (1) Flow: On; Pump Speed: Auto; Flow Setpoint: 500 μmol s⁻¹; Press. Valve: 0.1 kPa. (2) H₂O: On; RH_air: 55%. (3) CO₂ injector: On; CO₂_s: 400 μmol mol⁻¹ (adjusted according to experimental requirements). (4) Mixing fan: On; Fan Speed: 10,000 rpm. (5) Temperature: On; Tleaf: 25 °C. (6) Light: 1,500 μmol m⁻² s⁻¹ (adjusted according to experimental requirements). (7) Select Auto Log program: 3 min acquisition interval, and the total acquisition time is 2-3 h.

Map-based cloning for bzu3

Since the mutation is lethal in homozygotes, we used bzu3-1+/- × B73 to generate parental populations for map-based cloning. For preliminary mapping of bzu3-1, we employed a total of 384 SSR markers (SIGMA catalog no. M4193) selected from the maize genetic and genomic database. The entire genome was covered with an average of 20 cM units of mapping distance between two SSR markers. The mutation of bzu3-1 was localized to the long arm of the 1^st chromosome. Next, 60 pairs of SSR markers covering 20–110 M of chromosome 1 were employed, and bzu3-1 was localized to 53.7–63.7 M. 7000 F₂ plants were used for fine mapping, and finally, bzu3-1 was localized at the region of 60.67–60.77 M on Chromosome 1. Searching the MaizeGDB database revealed only one gene (Zm00001d029151) in this region. With Sanger sequencing, we found that a viral retrotransposon, having a full length of 5–6 kb and containing forward long terminal repeats, was inserted into the second intron in Zm00001d029151 in bzu3-1 mutant plants.

Generation of CRISPR lines

To generate the CRISPR-Cas9 mutant of BZU3, the sgRNAs of BZU3 were designed and evaluated using the CRISPOR online server (http://crispor.tefor.net/). The sgRNAs were inserted into the pBUE411 vector and verified by Sanger sequencing. The pBUE411-BZU3-sgRNA1-sgRNA2 construction was transformed via agrobacteria (strain EHA105) into inbred line B73-329 as previously described²³. Positive transgenic lines were confirmed by PCR and verified by sequencing. Three stably-inherited alleles were selected for further phenotypic investigation. The BZU3 CRISPR-Cas9 mutants were produced in the Plant Transformation Facility, at Henan University. The primers were listed in Supplementary Table S7.

Genetic complementation

The promoter sequences of BZU3, ZmMUTE, and ZmFAMA were amplified from the maize genome, and the BZU3 sequence was amplified from cDNAs. We constructed BZU3pro:BZU3-YFP, ZmMUTEpro:BZU3-YFP, and ZmFAMApro:YFP-BZU3 vectors. Vector BZU3pro:BZU3^Y157F-YFP was constructed using BZU3pro:BZU3-YFP as a template. Maize plants of the inbred line B73-329 were transformed using agrobacteria EHA105 carrying the vectors described above. BZU3pro:BZU3-YFP transgenic plants were crossed with bzu3-2 +/- to produce the F₁ generation. F₂ plants were segregated from self-crossed BZU3pro:BZU3-YFP (♂) × bzu3-2 +/- (♀) F₁ generation and BZU3pro:BZU3-YFP;bzu3-2 plants were isolated. Similarly, ZmMUTEpro:BZU3-YFP;bzu3-2, ZmFAMApro:YFP-BZU3;bzu3-2, and BZU3pro:BZU3^Y157F-YFP;bzu3-2 plants were obtained. The stomatal phenotype of this complemented transgenic maize was observed and compared to the WT. The primers were listed in Supplementary Table S7.

Purification of BZU3 and enzyme kinetics analyses of BZU3

The coding sequence of BZU3 was cloned into the pET28a vector to construct T7pro:6×His-BZU3, from which BZU3^S133A and BZU3^Y157F were made using site-directed mutagenesis. The primers were listed in Supplementary Table S7. The plasmids were transformed into E.coli Rosetta (DE3), and proteins were expressed following 0.3 mM IPTG induction at 23 °C for 16 h. The cells were collected and resuspended in 1×PBS buffer containing 1 mM PMSF. After cell lysis, proteins were purified on a nickel column and examined using 10% PAGE.

To measure BZU3 enzymatic kinetics, 25 nM BZU3 was incubated with different concentrations of UDP-Glc or UDP-Gal dissolved in 10 mM Tris-HCl (pH 8.0) buffer at 30 °C for 1 min. To measure the ratio between UDP-Glc and UDP-Gal after the reactions reached equilibrium, 2 μM BZU3 was incubated with 2 mM of UDP-Glc or UDP-Gal in 10 mM Tris-HCl (pH8.0) at 30 °C for 30 min. The reactions were stopped at 70 °C for 10 min. After the reaction solutions were filtered with a 0.45 µm filter, the surplus of UDP-Glc and products of UDP-Gal were measured using HPLC. Similar procedures were performed with BZU3^S133A and BZU3^Y157F to examine the effects of mutation at the active sites. Catalysis between UDP-GlcNAc and UDP-GalNAc was followed using a similar protocol.

Cellulose staining

To check the cellulose content in the third leaf base of 7-day-old seedlings, the tissues were decolorized in 12.5% acetic acid for 2 h before being dehydrated in 100% and 50% ethanol for 4-5 h. After rinsing in ddH₂O for 20 min, the samples were stained with 0.02% (w/v) Direct Red 23 for 5 h and then rinsed with dd H₂O. The samples were observed using a Zeiss LSM710 confocal microscope with excitation at 561 nm. The method is based on that of Riglet et al.³¹.

Immunofluorescence analysis

Mixed-linkage glucans were detected by immunofluorescence labeling using methods adapted from Giannoutsou et al.⁴⁷. Briefly, the 2-3 cm leaf base of WT and bzu3 mutant plants were cut into thin strips, 1-2 mm wide and 3-5 mm long, and were fixed in 2% (v/v) paraformaldehyde and 0.1% (v/v) glutaraldehyde in PEM buffer (pH 6.8) for 1.5 h at 4 °C. After rinsing in PEM buffer, the fixed samples were dehydrated in gradient ethanol series and post-fixed in 0.25% (w/v) osmium tetroxide. The samples were infiltrated and embedded with LR White acrylic resin (LRW, Sigma-Aldrich). Sections (1 μm thickness) were transferred onto poly-lysine glass slides and were blocked with 5% BSA (v/v) in 1×PBS solution for 5 h. To detect the mixed-linkage glucans, BG-1 primary antibodies (Biosupplies, Australia) were diluted 1:100 in 1×PBS buffer containing 2% BSA, and were incubated with the samples at 4 °C overnight. After rinsing three times with 1×PBS buffer, the samples were incubated with secondary antibodies (anti-mouse IgGs conjugated with FITC) diluted 1:500 in the same buffer at 37 °C for 1 h. After rinsing a further three times, the slides were mounted with anti-fade medium containing p-phenylenediamine, and were observed using a confocal microscope (Zeiss, LSM710) with excitation at 561 nm.

Stimulated Raman scattering microscopy (SRS)

Label-free imaging of cellulose and hemicellulose was performed with a home-built stimulated Raman scattering microscope⁴⁸. A tunable picosecond two-color laser (picoEmerald-S, APE GmbH, Germany) served as the light source. It generated two synchronized 80 MHz picosecond laser beams. The 1032 nm Stokes beam was modulated at 20 MHz. The pump beam was tuned to 926.5 nm to detect cellulose and hemicellulose. Stokes and pump beams were coupled into a laser-scanning upright microscope (FVMPE, Olympus, Japan) equipped with a 25X water-immersion objective lens (XLPLN25XWMP2, Olympus, Japan). SRS signals were collected by a photodiode and lock-in amplifier module (customized from APE GmbH, Germany). Fresh maize leaf epidermal strips were placed on a glass slide with water and covered with a cover glass. SRS images (512 × 512 pixels) were acquired. The pixel dwell time was 2 to 4 µs. Each image was line-averaged 3 to 5 times to reduce noise. The signal intensity corresponds to the content of cellulose and hemicellulose.

UPLC-MS analysis of endogenous UDP-Glc and UDP-Gal contents

The third leaf base of 7-day-old seedlings (Elongation stage of GC morphogenesis) was collected and ground to a fine powder in liquid nitrogen. 100 ± 2 mg of plant materials was weighed and extracted with 1 mL of pre-chilled methanol/water (3/1, v/v) including 0.1% formic acid. The samples were vortexed for 1–2 min and sonicated for 15 min at room temperature before being centrifuged at 8,500 g for 10 min at 4 °C. The supernatant was filtered using a 0.22 µm membrane and transferred to glass vials for analysis.

UDP-Glc and UDP-Gal were quantified by ultra-high-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS). Samples were analyzed using a Xevo TQ-XS system (Waters, USA) with an ESI ion source. The data were collected under the negative ion mode and multiple reaction monitoring mode. Precursor and fragment ions are: UDP-Glc (m/z 564.8–322.9) and UDP-Gal (m/z 564.8–322.9). Chromatographic separation was conducted with a BEH Amide column (2.1 × 100 mm, 1.7 µm) maintained at 30 °C. The mobile phase includes solvent A (50 mM ammonium formate, adjusted to pH 3.6 with formic acid) and solvent B (acetonitrile), with a flow rate of 0.4 ml·min^-1. The isometric gradient system was set to 0–35 min, 21% solvent B. The auto-sampler temperature was set to 4 °C and 5 μL of the sample was injected.

Subcellular localization of BZU3

To observe the expression pattern of BZU3, BZU3pro:BZU3-YFP, and the nuclear marker 35S:H2B-mCherry were transformed transiently into maize protoplasts. The localization of the proteins was observed by confocal microscopy (Zeiss, LSM710). The methods were based on Zhang et al.⁴⁹. Maize protoplasts were isolated from a 21-day-old etiolated seedling of inbred line B73 as previously described. 10 µg of each plasmid was used and the transformed protoplasts were incubated overnight.

Bimolecular fluorescence complementation assays

The coding sequences of BZU3 and BZU3^bzu3-2 were amplified from the genome of the wild-type Zheng58 and mutant bzu3-2 plants. Constructs of 35S:BZU3-cYFP, 35S:nYFP-BZU3, 35S:BZU3^bzu3-2-cYFP, 35S:nYFP-BZU3^bzu3-2 were respectively made. The primers were listed in Supplementary Table S7. Corresponding pairs of plasmids were transformed into tobacco cells using Agrobacterium tumefaciens GV3101. Two or three days after inoculation, fluorescence patterns were observed using a confocal microscope (Zeiss LSM710).

Crystallization and structure determination

The purified BZU3 protein (~ 5.0 mg/ml) was crystallized using the sitting-drop vapor diffusion method against a reservoir solution containing 100 mM sodium malonate, pH 5.0 and 12 ~ 16% (w/v) PEG3350. To prepare the enzyme complex bound with UDP-Glc or UDP-GlcNAc, the long-rod shape BZU3 crystals were soaked by the addition of 5 mM UDP-Glc or UDP-GlcNAc in their mother-liquors overnight before harvest. Cryoprotection was obtained by the sequential addition of mother-liquors supplemented with 25% (w/v) glycerol, followed by flash freezing in liquid N₂. The X-ray diffraction datasets were collected at the beamline BL19U1 of the Shanghai Synchrotron Radiation Facility (SSRF). Data processing and reduction were carried out using the XDS package⁵⁰. The structures were solved by molecular replacement using Phaser-MR using human GALE (PDB 1HZJ, sharing 58% sequence identity) as the search model⁵¹. The models were completed using the COOT program, and further refinement was done with the Phenix package⁵². Structure validation was performed with Molprobity⁵³. Details for crystallization, data collection, structure refinement, and analyses are shown in Supplemental Table S5. The pictures shown in the paper were prepared using PyMOL⁵⁴.

UDP-Glucose treatment assay

The old coleoptile and outer first leaf of bzu3 mutants (3-day-old seedlings) were carefully removed, leaving the second leaf supported by a toothpick. The leaf base was painted with 450 µM UDP-glucose in 0.1% Triton X-100 solution. The treatment was repeated 8 times/day for 3 days and the treated part was marked. After another 2-3 days when the leaf sheath was formed, the treated leaf bases were collected for observation.

Glycomics analysis

A total of 1 g samples of maize seedling leaf bases were collected from the wild-type plants and the mutants. The samples were ground in liquid nitrogen, and then dissolved in 2 mL 8 M urea followed by protein alkylation and digestion. The enriched intact N-glycopeptides were then isotopically labeled with diethyl. The samples of each group were mixed in a 1:1 ratio and analyzed online by C18-RPLC-MS/MS⁵⁵, as follows: the enriched N-glycome samples were injected into a C18 precolumn (Agilent ZORBAX 300SB, 360 μm o.d. × 200 μm i.d., 7 cm long) for sample loading. Chromatographic separation was performed using a 75 cm-long analytical column (360 μm o.d. × 75 μm i.d.) packed with the same C18 precolumn. The buffers were as follows: Buffer A contains 99.9% H₂O and 0.1% FA, and Buffer B contains of 99.9% ACN and 0.1% FA. The flow rate of the mobile phase was 0.2 mL/min, with a multi-step gradient: 1% B 10 min, 1 − 25% B 20 min, 25 − 60% B 190 min, 60 − 95% B 10 min, 95 − 95% B 10 min, and 95 − 1% B 5 min. MS spectra were detected as follows: m/z range 500 − 2500, mass resolution 35 k, automatic gain control (AGC) target 5 × 10⁵, max ion injection time 200 ms. MS/MS spectra were obtained using the Top20 data-dependent mode, with the following settings: mass resolution 17.5 k, AGC target 1 × 10⁵, max ion injection time 200 ms, dynamic exclusion 50 s, HCD normalized collision energy 10%, isolation window 3 Th. The ESI conditions were as follows: spray voltage 2.8 kV, capillary temperature 250 °C, and S-lens RF level 75. Based on the libraries of the Corn protein group and of the plant N-glycosylation modifications in the Uniprot database, we identified 2194 intact N-glycopeptides in qualitative analysis and quantified 181 differentially expressed intact N-glycopeptides (DEGPs) (two out of three technical replicates, ≥ 1.5-fold change and p-value < 0.05). Gene ontology enrichment analysis was conducted using the online tool ShinyGO (http://bioinformatics.sdstate.edu/go/) with a FDR adjusted P-value cutoff of 0.05.

Verification of N-Glycosylation site

The full-length sequence of PAN2, GT14, and FUT was amplified from maize cDNAs and inserted into the pCAMBIA2300 vector (35S:PAN2-GFP, 35S:GT14-GFP, and 35S:FUT-GFP). 35S:PAN2^N554Q-GFP, 35S:GT14^N283Q-GFP, and 35S:FUT^N121Q-GFP were constructed using site-directed mutagenesis. The primers were listed in Supplementary Table S7. The plasmids were transformed into Agrobacterium tumefaciens GV3101. Subsequently, 35S:PAN2-GFP and 35S:PAN2^N554Q-GFP, 35S:GT14-GFP and 35S:GT14^N283Q-GFP, 35S:FUT-GFP and 35S:FUT^N121Q-GFP were expressed transiently in tabacco leaves. Finally, the total proteins were extracted and analyzed by Western blotting, using anti-GFP antibodies (Transgen, HT801, diluted 1:10000).

Statistics and reproducibility

The maize plants used in this study were grown in the greenhouse at 16 h/8 h and 32 °C/25 °C (day/night). The samples were obtained from 7-day-old seedlings unless noted otherwise. For analysis, plants of the same genotype were randomly picked for experiments to avoid bias. Sufficient sample sizes were determined by pilot experiments as indicated in the main text, in the figures, and in the figure legends. No data were excluded from the study. All experiments were repeated at least three times, with at least three independent biological replicates.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The coordinates of the atomic model of UDP-Glc/GlcNAc 4-Epimerase in this paper have been deposited in the Protein Data Bank (PDB). The accession codes for UDP-Glc/GlcNAc 4-Epimerase with NAD, UDP-Glc/GlcNAc 4-Epimerase with NAD/UDP-Glc, and UDP-Glc/GlcNAc 4-Epimerase with NAD/UDP-GlcNAc are 7XPP, 7XPO, and 7XPQ, respectively. The data of glycomics analysis are available via ProteomeXchange with identifier PXD041642. Source data are provided with this paper.

References

Somerville, C. et al. Toward a systems approach to understanding plant cell walls. Science 306, 2206–2211 (2004).
Article ADS CAS PubMed Google Scholar
Ezquer, I., Salameh, I., Colombo, L. & Kalaitzis, P. Plant cell walls tackling climate change: insights into plant cell wall remodeling, its regulation, and biotechnological strategies to improve crop adaptations and photosynthesis in response to global warming. Plants (Basel) 9, 212 (2020).
Article CAS PubMed Google Scholar
Sapala, A. et al. Why plants make puzzle cells, and how their shape emerges. Elife 7, e32794 (2018).
Article PubMed PubMed Central Google Scholar
Lampugnani, E. R., Khan, G. A., Somssich, M. & Persson, S. Building a plant cell wall at a glance. J. Cell Sci. 131, jcs207373 (2018).
Article PubMed Google Scholar
Anderson, C. T. & Kieber, J. J. Dynamic construction, perception, and remodeling of plant cell walls. Annu Rev. Plant Biol. 71, 39–69 (2020).
Article CAS PubMed Google Scholar
Voiniciuc, C., Pauly, M. & Usadel, B. Monitoring polysaccharide dynamics in the plant cell wall. Plant Physiol. 176, 2590–2600 (2018).
Article CAS PubMed PubMed Central Google Scholar
Jasechko, S. et al. Terrestrial water fluxes dominated by transpiration. Nature 496, 347–350 (2013).
Article ADS CAS PubMed Google Scholar
Hetherington, A. M. & Woodward, F. I. The role of stomata in sensing and driving environmental change. Nature 424, 901–908 (2003).
Article ADS CAS PubMed Google Scholar
Chen, Y., Li, W., Turner, J. A. & Anderson, C. T. PECTATE LYASE LIKE12 patterns the guard cell wall to coordinate turgor pressure and wall mechanics for proper stomatal function in Arabidopsis. Plant Cell 33, 3134–3150 (2021).
Article PubMed PubMed Central Google Scholar
Franks, P. J. & Farquhar, G. D. The mechanical diversity of stomata and its significance in gas-exchange control. Plant Physiol. 143, 78–87 (2007).
Article CAS PubMed PubMed Central Google Scholar
Stebbins, G. L. & Shah, S. S. Developmental studies of cell differentiation in the epidermis of monocotyledons. II. Cytological features of stomatal development in the gramineae. Dev. Biol. 2, 477–500 (1960).
Article Google Scholar
Rudall, P. J., Chen, E. D. & Cullen, E. Evolution and development of monocot stomata. Am. J. Bot. 104, 1122–1141 (2017).
Article CAS PubMed Google Scholar
Nunes, T. D. G., Zhang, D. & Raissig, M. T. Form, development and function of grass stomata. Plant J. 101, 780–799 (2020).
Article CAS PubMed Google Scholar
Raissig, M. T. et al. Mobile MUTE specifies subsidiary cells to build physiologically improved grass stomata. Science 355, 1215–1218 (2017).
Article ADS CAS PubMed Google Scholar
Kollist, H., Nuhkat, M. & Roelfsema, M. R. Closing gaps: linking elements that control stomatal movement. N. Phytol. 203, 44–62 (2014).
Article CAS Google Scholar
McAusland, L. et al. Effects of kinetics of light-induced stomatal responses on photosynthesis and water-use efficiency. N. Phytol. 211, 1209–1220 (2016).
Article Google Scholar
Woolfenden, H. C. et al. Models and mechanisms of stomatal mechanics. Trends Plant Sci. 23, 822–832 (2018).
Article CAS PubMed Google Scholar
Giannoutsou, E., Sotiriou, P., Nikolakopoulou, T. L., Galatis, B. & Apostolakos, P. Callose and homogalacturonan epitope distribution in stomatal complexes of Zea mays and Vigna sinensis. Protoplasma 257, 141–156 (2020).
Article CAS PubMed Google Scholar
Shtein, I. et al. Stomatal cell wall composition: distinctive structural patterns associated with different phylogenetic groups. Ann. Bot. 119, 1021–1033 (2017).
Article CAS PubMed PubMed Central Google Scholar
Galatis, B. & Apostolakos, P. The role of the cytoskeleton in the morphogenesis and function of stomatal complexes. N. Phytol. 161, 613–639 (2004).
Article Google Scholar
Spiegelhalder, R. P. & Raissig, M. T. Morphology made for movement: formation of diverse stomatal guard cells. Curr. Opin. Plant Biol. 63, 102090 (2021).
Article CAS PubMed Google Scholar
Wu, Z. et al. Multiple transcriptional factors control stomata development in rice. N. Phytol. 223, 220–232 (2019).
Article CAS Google Scholar
Wang, H. et al. BZU2/ZmMUTE controls symmetrical division of guard mother cell and specifies neighbor cell fate in maize. PLoS Genet 15, e1008377 (2019).
Article CAS PubMed PubMed Central Google Scholar
Verbancic, J., Lunn, J. E., Stitt, M. & Persson, S. Carbon supply and the regulation of cell wall synthesis. Mol. Plant 11, 75–94 (2018).
Article CAS PubMed Google Scholar
Zhang, Y. et al. BRITTLE PLANT1 is required for normal cell wall composition and mechanical strength in rice. J. Integr. Plant Biol. 63, 865–877 (2021).
Article CAS PubMed Google Scholar
Seifert, G. J. Nucleotide sugar interconversions and cell wall biosynthesis: how to bring the inside to the outside. Curr. Opin. Plant Biol. 7, 277–284 (2004).
Article CAS PubMed Google Scholar
Fushinobu, S. Molecular evolution and functional divergence of UDP-hexose 4-epimerases. Curr. Opin. Chem. Biol. 61, 53–62 (2021).
Article CAS PubMed Google Scholar
Rosti, J. et al. UDP-glucose 4-epimerase isoforms UGE2 and UGE4 cooperate in providing UDP-galactose for cell wall biosynthesis and growth of Arabidopsis thaliana. Plant Cell 19, 1565–1579 (2007).
Article CAS PubMed PubMed Central Google Scholar
Broussard, A. et al. Human UDP-galactose 4’-epimerase (GALE) is required for cell-surface glycome structure and function. J. Biol. Chem. 295, 1225–1239 (2020).
Article PubMed Google Scholar
Kotake, T. et al. Bifunctional cytosolic UDP-glucose 4-epimerases catalyse the interconversion between UDP-D-xylose and UDP-L-arabinose in plants. Biochem. J. 424, 169 (2009).
Article CAS PubMed Google Scholar
Riglet, L. et al. KATANIN-dependent mechanical properties of the stigmatic cell wall mediate the pollen tube path in Arabidopsis. Elife 9, e57282 (2020).
Article CAS PubMed PubMed Central Google Scholar
Anderson, C. T., Carroll, A., Akhmetova, L. & Somerville, C. Real-time imaging of cellulose reorientation during cell wall expansion in Arabidopsis roots. Plant Physiol. 152, 787–796 (2010).
Article CAS PubMed PubMed Central Google Scholar
Pattison, R. J. & Amtmann, A. N-glycan production in the endoplasmic reticulum of plants. Trends Plant Sci. 14, 92–99 (2009).
Article CAS PubMed Google Scholar
Stenitzer, D. et al. O-methylated N-glycans Distinguish Mosses from Vascular Plants. Biomolecules 12, 136 (2022).
Article CAS PubMed PubMed Central Google Scholar
Ruan, N. et al. FRAGILE CULM 18 encodes a UDP-glucuronic acid decarboxylase required for xylan biosynthesis and plant growth in rice. J. Exp. Bot. 73, 2320–2335 (2022).
Article CAS PubMed Google Scholar
Leonard, R. et al. Identification of an Arabidopsis gene encoding a GH95 alpha1,2-fucosidase active on xyloglucan oligo- and polysaccharides. Phytochemistry 69, 1983–1988 (2008).
Article CAS PubMed Google Scholar
McFarlane, H. E., Doring, A. & Persson, S. The cell biology of cellulose synthesis. Annu Rev. Plant Biol. 65, 69–94 (2014).
Article CAS PubMed Google Scholar
Jones, L., Milne, J. L., Ashford, D. & McQueen-Mason, S. J. Cell wall arabinan is essential for guard cell function. Proc. Natl Acad. Sci. USA 100, 11783–11788 (2003).
Article ADS CAS PubMed PubMed Central Google Scholar
Rui, Y. & Anderson, C. T. Functional analysis of cellulose and xyloglucan in the walls of stomatal guard cells of arabidopsis. Plant Physiol. 170, 1398–1419 (2016).
Article CAS PubMed PubMed Central Google Scholar
Wang N., et al. UDP-glucose epimerase 1, moonlighting as a transcriptional activator, is essential for tapetum degradation and male fertility in rice. Mol. Plant https://doi.org/10.1016/j.molp.2023.03.008 (2023).
Yang, S. et al. OsUGE1 is directly targeted by OsGRF6 to regulate root hair length in rice. THEOR APPL GENET 136, 108 (2023).
Article CAS PubMed Google Scholar
Sun, G. et al. The maize single-nucleus transcriptome comprehensively describes signaling networks governing movement and development of grass stomata. Plant Cell 34, 1890–1911 (2022).
PubMed PubMed Central Google Scholar
Anderson, R. L. & Ray, P. M. Labeling of the plasma membrane of pea cells by a surface-localized glucan synthetase. Plant Physiol. 61, 723–730 (1978).
Article CAS PubMed PubMed Central Google Scholar
Janse van Rensburg, H. C. & Van den Ende, W. UDP-Glucose: a potential signaling molecule in plants? Front Plant Sci. 8, 2230 (2017).
Article PubMed Google Scholar
Sutimantanapi, D., Pater, D. & Smith, L. G. Divergent roles for maize PAN1 and PAN2 receptor-like proteins in cytokinesis and cell morphogenesis. Plant Physiol. 164, 1905–1917 (2014).
Article CAS PubMed PubMed Central Google Scholar
Yadav, V. et al. ABCG transporters are required for suberin and pollen wall extracellular barriers in Arabidopsis. Plant Cell 26, 3569–3588 (2014).
Article CAS PubMed PubMed Central Google Scholar
Giannoutsou, E., Apostolakos, P. & Galatis, B. Spatio-temporal diversification of the cell wall matrix materials in the developing stomatal complexes of Zea mays. Planta 244, 1125–1143 (2016).
Article CAS PubMed Google Scholar
Saar, B. G. et al. Label-free, real-time monitoring of biomass processing with stimulated Raman scattering microscopy. Angew. Chem. Int Ed. Engl. 49, 5476–5479 (2010).
Article CAS PubMed Google Scholar
Zhang, X. et al. Genetic variation in ZmTIP1 contributes to root hair elongation and drought tolerance in maize. Plant Biotechnol. J. 18, 1271–1283 (2020).
Article CAS PubMed Google Scholar
Kabsch, W. Xds. Acta Crystallogr D. Biol. Crystallogr 66, 125–132 (2010).
Article CAS PubMed PubMed Central Google Scholar
McCoy, A. J. et al. Phaser crystallographic software. J. Appl Crystallogr 40, 658–674 (2007).
Article CAS PubMed PubMed Central Google Scholar
Adams, P. D. et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr D. Biol. Crystallogr 66, 213–221 (2010).
Article CAS PubMed PubMed Central Google Scholar
Chen, V. B. et al. MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr D. Biol. Crystallogr 66, 12–21 (2010).
Article CAS PubMed Google Scholar
DeLano W. L. Pymol: An open-source molecular graphics tool. CCP4 Newsletter On Protein Crystallography, 82-92 (2002).
Han, Y., Xiao, K. & Tian, Z. Comparative glycomics study of cell-surface n-glycomes of hepg2 versus lo2 cell lines. J. Proteome Res 18, 372–379 (2019).
CAS PubMed Google Scholar

Download references

Acknowledgements

We thank Prof. José Ramón Botella, Changqing Zhang, Zongwei Cai, Guiling Sun and Dr. Yue Rui for valuable suggestions. We would like to thank Ms. Lingyu Mi, Qi Zheng and Ruiting Sun at the electron microscope platform of State Key Laboratory of Crop Stress Adaptation and Improvement, Henan University. This work was supported by the National Natural Science Foundation of China (U21A20206 and 31970808) and the Program for Innovative Research Team (in Science and Technology) at University of Henan Province (21IRTSTHN019). D.W.G. was supported by grants from the USDA to the College of Agriculture and Life Sciences at the University of Arizona.

Author information

These authors contributed equally: Yusen Zhou, Tian Zhang, Xiaocui Wang, Wenqiang Wu.

Authors and Affiliations

State Key Laboratory of Crop Stress Adaptation and Improvement, School of Life Sciences, Henan University, Jinming avenue 1, Kaifeng, 475004, China
Yusen Zhou, Tian Zhang, Wenqiang Wu, Jingjing Xing, Zuliang Li, Xin Qiao, Xiaohang Wang, Guangshun Wang, Shenglong Bai, Zhi Li, Yanli Niu, Junli Zhang, Chen Lan, Zhubing Hu, Baozhu Li, Xuebin Zhang, Wei Wang, David W. Galbraith, Siyi Guo & Chun-Peng Song
Sanya Institute, Henan University, Sanya, 572025, China
Yusen Zhou, Wenhui Li, Zhubing Hu, Siyi Guo & Chun-Peng Song
State Key Laboratory of Molecular Developmental Biology, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing, China; Innovative Academy of Seed Design, Chinese Academy of Sciences, Beijing, 100101, China
Xiaocui Wang, Chunrui Zhang & Yuhang Chen
College of Advanced Agricultural Sciences, University of Chinese Academy of Sciences, Beijing, 100049, China
Xiaocui Wang, Chunrui Zhang & Yuhang Chen
Biomedical Pioneering Innovation Center, School of Life Sciences, Beijing Advanced Innovation Center for Genomics, Peking University, Beijing, 100871, China
Yuanzhen Suo
Joint National Laboratory for Antibody Drug Engineering, Henan University, Kaifeng, 475004, China
Jiajia Wang
School of Plant Sciences and Bio5 Institute, The University of Arizona, Tucson, AZ, 85721, USA
David W. Galbraith

Authors

Yusen Zhou
View author publications
You can also search for this author in PubMed Google Scholar
Tian Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Xiaocui Wang
View author publications
You can also search for this author in PubMed Google Scholar
Wenqiang Wu
View author publications
You can also search for this author in PubMed Google Scholar
Jingjing Xing
View author publications
You can also search for this author in PubMed Google Scholar
Zuliang Li
View author publications
You can also search for this author in PubMed Google Scholar
Xin Qiao
View author publications
You can also search for this author in PubMed Google Scholar
Chunrui Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Xiaohang Wang
View author publications
You can also search for this author in PubMed Google Scholar
Guangshun Wang
View author publications
You can also search for this author in PubMed Google Scholar
Wenhui Li
View author publications
You can also search for this author in PubMed Google Scholar
Shenglong Bai
View author publications
You can also search for this author in PubMed Google Scholar
Zhi Li
View author publications
You can also search for this author in PubMed Google Scholar
Yuanzhen Suo
View author publications
You can also search for this author in PubMed Google Scholar
Jiajia Wang
View author publications
You can also search for this author in PubMed Google Scholar
Yanli Niu
View author publications
You can also search for this author in PubMed Google Scholar
Junli Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Chen Lan
View author publications
You can also search for this author in PubMed Google Scholar
Zhubing Hu
View author publications
You can also search for this author in PubMed Google Scholar
Baozhu Li
View author publications
You can also search for this author in PubMed Google Scholar
Xuebin Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Wei Wang
View author publications
You can also search for this author in PubMed Google Scholar
David W. Galbraith
View author publications
You can also search for this author in PubMed Google Scholar
Yuhang Chen
View author publications
You can also search for this author in PubMed Google Scholar
Siyi Guo
View author publications
You can also search for this author in PubMed Google Scholar
Chun-Peng Song
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

C.-P.S. and S.G. designed and supervised this study. Y.Z., S.G., and C.-P.S. designed the experiment; Y.Z., J.X., X.W. (Xiaohang Wang), X.Q., G.W., W.L., Zu.Li., Y.S., J.W., Zh.Li., Y.N., J.Z., C.L., Z.H., X.Z., S.B., and B.L. performed the experiments and data analysis; Xi.Wa., W.Wu., C.Z., W.Wa., and Y.C. performed crystallization and structure analysis; T.Z., S.G., D.W.G., and C.-P.S. prepared the manuscript with the inputs from other authors. All authors have read, edited, and approved the content of the manuscript.

Corresponding authors

Correspondence to Siyi Guo or Chun-Peng Song.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Description of Additional Supplementary Files

Supplementary Data S1

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Zhou, Y., Zhang, T., Wang, X. et al. A maize epimerase modulates cell wall synthesis and glycosylation during stomatal morphogenesis. Nat Commun 14, 4384 (2023). https://doi.org/10.1038/s41467-023-40013-6

Download citation

Received: 10 January 2023
Accepted: 09 July 2023
Published: 20 July 2023
DOI: https://doi.org/10.1038/s41467-023-40013-6

This article is cited by

Cell wall anisotropy plays a key role in Zea mays stomatal complex movement: the possible role of the cell wall matrix
- K. Gkolemis
- E. Giannoutsou
- P. Apostolakos
Plant Molecular Biology (2023)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.