The brittle stalk-5 ^214S2L (bk-5 ^214S2L ) mutant, characterized by its fragile stems and overall plant frailty, is a weak allelic variant of the bk-5 maize mutant (Li et al., 2022). All mutant materials were homozygous variants in the ethyl methanesulfonate (EMS)-induced mutant library of the inbred line RP125 (considered the wild type in this study). RP125 pollen grains were treated with a 0.1% EMS solution (Sigma-Aldrich, M0880) dissolved in mineral oil (Sigma-Aldrich, M8410) for 30–40 min. Subsequently, 60–70 ears were subjected to pollination to induce the mutation. Using this method, our group established an EMS mutant library in the RP125 background (Nie et al., 2021). Within this library, systematic screening for mutants exhibiting brittle culm traits led to the discovery of a stably inherited brittle culm mutant, named bk-5 ^214S2L , by successive selfing. Throughout the growth phases—from germination and seedling to nodulation, silking, and maturity—bk-5 ^214S2L plants underwent field observations for phenotypic evaluation. At the nodulation stage, a stem-breaking experiment was performed. Upon reaching maturity, various agronomic traits, including compressive strength, plant height, spike height, spike length, spike thickness, grain dimensions, spike weight, and 100-grain weight, were measured in both the mutant and RP125. Furthermore, bk-5 ^214S2L was crossed with Mo17 to generate F1 and F2 populations for subsequent genetic analyses and gene mapping.

Rind Penetrometer Resistance: Ten plants from each of RP125 and bk-5 ^214S2L exhibiting comparable growth potential were selected at the nodulation stage. Using a penetrometer (YYD-1, ToPu YunNong) with a pressure plate, the rind penetrometer resistance was assessed in the middle of the third internode of the stem. The measurements were performed until the stems ruptured, capturing the peaks of resistance of plants.

Stem Bending Strength (Three-Point Bending Test, TPBT): Likewise, ten plants from each of RP125 and bk-5 ^214S2L were selected at the nodulation stage. All stems from node 1 to node 6 of the stalk were excised with a knife. Subsequently, the internodal length of the third internode of the stem was measured, along with its diameter. The stem peak strength was measured using a YYD-1 stem strength tester. The stem bending strength was calculated as follows: M _max = PL/2, where M _max represents the bending strength, P denotes the stem peak strength, and L signifies the internodal length (Hou et al., 2022). During the measurement, we set up six biological replicates.

To visualize cellulose and hemicellulose within the cell wall, leaf veins from the third internode of the aboveground stem of RP125 and bk-5 ^214S2L plants were selected at the jointing (elongation) stage. In addition, leaf blades at the spike were included in the study. Using an ultrathin tissue slicer, internodes and leaf veins were precisely sectioned into 60-μm transverse sections. Subsequently, staining was conducted using cellulose staining solution (zinc chloride iodide; Leagene, DP0406) for 2 min. The sections were then observed under a light microscope (Leica, Wetzlar, Germany). For lignin observation, the third internode was similarly sectioned and stained using resorcinol (Ookawa et al., 2014). The sections were fixed in a 1% resorcinol/alcohol solution (v/v) for 2 min (Sigma, 79330), followed by color development by the addition of 18% HCl solution (Hossain et al., 2012). The stained sections were covered with a coverslip and observed under a light microscope (Leica) (Li et al., 2022).

Stems, along with leaves, from RP125 and bk-5 ^214S2L plants at the jointing stage were used for cellulose content determination. A cellulose measurement kit (Solarbio, BC4280) was used, and the steps included cell wall material (CWM) extraction, CWM drying, and cellulose extraction. Absorbance analysis, coupled with a chromogenic reaction of concentrated sulfuric acid with anthrone, facilitated the quantification. For lignin measurement, a lignin measurement kit (Solarbio, BC4200) was used. The internodes were dried and ground, followed by extraction and determination of lignin content using perchloric acid and glacial acetic acid. Air-dried or oven-dried samples were ground into powder and sieved for hemicellulose content determination. Hemicellulose hydrolysis included detecting reducing sugars by a characteristic absorption peak at 540 nm, which enabled the quantitative assessment using the hemicellulose content determination kit (Solarbio, BC4440). During the measurement, we set up three biological replicates and six experimental replicates.

At the jointing stage, stems (from the third internode above ground) of RP125 and bk-5 ^214S2L plants were collected on ice.

Using the RNA extraction kit (TIANGEN, DP419), total RNA was carefully extracted from distinct tissues of both RP125 and bk-5 ^214S2L at the nodulation stage. Subsequently, cDNA was synthesized by reverse transcription with the Evo M-MLV Reverse Transcription Kit II (AG, AG11711) and Evo M-MLV Plus cDNA Synthesis Kit (AG, AG11615). For gene expression analysis, qRT-PCR was conducted in an Applied Biosystems thermal cycler. The SYBR Premix Ex TaqII (Tli RNase H Plus) Kit (TaKaRa, RR820A) was used for qRT-PCR assays, ensuring accurate and reliable quantification of gene expression levels. RNA Denaturation is 65°C for 5 min. Reverse Transcription is 45°C for 30–60 min. Enzyme Inactivation is 70°C for 10 min. Pre-denaturation is 95°C for 3 min. Denaturation: 95°C for 10–30 s. Annealing: 55°C for 30 s. Extension is 72°C for 30 s. Cycles number are 40 (Nie et al., 2021). Primer sequences are provided in Supplementary Table S3. During the measurement, we set up six experimental replicates.

A genetically stable maize brittle stalk mutant was identified from the EMS mutant library (Nie et al., 2021). This mutant was named bk-5 ^214S2L (brittle stalk-5 ^214S2L ) based on its “brittle and breakable” phenotype. Phenotypic characterization of bk-5 ^214S2L showed that mutant plants were brittle from the three-leaf stage, with brittle stems till even the nodulation stage and maturity. Field phenotyping of the mutant showed no change in plant shape compared with RP125 at either seedling or nodulation stage (Figure 1A). In stem-breaking experiments in which the same portion of the stem at the jointing stage was bent with the same force, bk-5 ^214S2L stems broke completely, with neat and smooth breakage, whereas RP125 stems remained attached and did not undergo complete breakage (Figures 1B, C). Therefore, the stem lodging resistance and stem compressive strength were measured (Figures 1D, E), and a significant decrease was observed in both, with a 39.2% decrease in stem lodging resistance and a 23.3% decrease in stem compressive strength. To determine whether the mutation of the candidate gene affected the yield traits of plants, the spike length, spike weight, 100-grain weight, grain length, grain width, and grain thickness were measured at maturity. All yield traits were significantly lower in mutant plants than in RP125 plants (Supplementary Figure 1). Thus, the mutation of the candidate gene affected not only the plant’s stem bending resistance but also its yield.

The reduced stem mechanical strength suggests that the cell wall composition of mutant plants was altered. Therefore, we performed histochemical staining and determined the contents of cellulose, hemicellulose, and lignin in the stems and leaf veins (aboveground third node) and leaf blades at the spike of bk-5 ^214S2L and RP125 plants at the nodulation stage (Figure 2; Supplementary Figure 2). In the staining results, the color change of cellulose and hemicellulose staining was lesser in bk-5 ^214S2L stem sections than in RP125 stem sections, indicating that the difference in cellulose and lignin content of the stem was small, whereas the staining colors of cellulose and hemicellulose in leaf vein sections were significantly lighter, indicating that the cellulose and hemicellulose contents of the leaf veins were significantly reduced (Figures 2A-D). The results of the cellulose content assay showed that the stem cellulose content was 223.2 mg/g in RP125 and 180.3 mg/g in bk-5 ^214S2L , which was significantly reduced by 19.22% compared with RP125 (Figure 2E). The leaf cellulose content was 357.3 mg/g in RP125 and 251.3 mg/g in bk-5 ^214S2L , which was highly significantly reduced by 29.67% compared with RP125 (Supplementary Figure 2E). Lignin staining of stem sections of bk-5 ^214S2L was significantly lighter, indicating that the lignin content was significantly reduced in the mutant, whereas lignin staining of leaf vein sections was less variable, indicating no significant difference in the lignin content of leaf veins. The lignin content assay showed no significant change in the leaf lignin content of bk-5 ^214S2L compared with RP125, but the stem lignin content was highly significantly reduced by 23.56% (Supplementary Figure 2F; Figure 2F). The results of the hemicellulose content assay showed that the stem hemicellulose content was 102.7 mg/g in RP125 and 76.2 mg/g in bk-5 ^214S2L , which was highly significantly reduced by 25.8% compared with RP125 (Figure 2G). The leaf hemicellulose content was 192.6 mg/g in RP125 and 174 mg/g in bk-5 ^214S2L , which was significantly reduced by 9.66% compared with RP125 (Supplementary Figure 2G). These results suggest that the mutant gene plays an important role in the biosynthesis of cell wall components. The results also indicate that the main reason for the low compressive strength of bk-5 ^214S2L was the reduced contents of cell wall components.

Cellulose is an important cell wall component, and changes in its content affect the formation and structure of the cell wall. Therefore, we investigated the anatomical characteristics of thick-walled cells in the leaf veins and stems of RP125 and bk-5 ^214S2L plants. Changes in the cell wall structure may be an important factor contributing to the reduced compressive strength of bk-5 ^214S2L plants. Transverse sections of leaf veins showed that the thickness of thick-walled tissues (number of cell layers) in the leaf veins of bk-5 ^214S2L was significantly reduced, becoming fragile and irregular (Supplementary Figure 3). SEM of thick-walled tissues showed a significant reduction in wall thickness of cells around the vascular bundles in the leaf veins of bk-5 ^214S2L compared with RP125, and a significant reduction in cell wall thickness of thick-walled tissues was observed in bk-5 ^214S2L stems (Figures 3A–D). The wall thickness of cells around the vascular bundles was quantified using ImageJ software, and bk-5 ^214S2L cell wall thickness was highly significantly reduced by 47.5% compared with RP125 (Figure 3E). Longitudinal sections of leaf veins showed that the thickness of thick-walled tissues (number of cell layers) in the leaf veins of bk-5 ^214S2L was significantly reduced and irregular (Supplementary Figure 4). Therefore, reduced cell wall thickness and the thinning of thick-walled tissues (reduction in the number of cell layers) may be responsible for the reduced mechanical strength.

The F1 generation was obtained by crossing bk-5 ^214S2L (male parent) with Mo17 (female parent), and the F2 segregation population was obtained by self-crossing. The phenotypes of the F2 generation were counted and were normal:mutant = 3:1, in accordance with Mendel’s law of inheritance, which suggested that the phenotypes were controlled by a single recessive mutation (Supplementary Table 1). Thereafter, 15–20 leaves of normal and brittle stalk plants from the F2 generation were taken, mixed in equal quantities, and DNA was extracted to make the dominant and recessive pools for bulk segregation analysis. More than 200 pairs of SSR markers covering 10 chromosomes of maize were used to screen RP125, Mo17, F1, and the dominant and recessive pools of RP125. SSR markers were used to screen RP125, Mo17, F1, and the dominant and recessive pools for polymorphism. As a result of the screening, markers ID-1 and ID-3 located on chromosome 3 were strongly associated with the target genes (Supplementary Figure 5). Based on the two SSR markers obtained, 160 F2 brittle culm monocotyledons were amplified, and the monocotyledon exchange rate was calculated, and it was initially determined that the localization interval was located upstream of marker ID-4 (Supplementary Figure 5). To determine the location of the bk-5 ^214S2L mutant gene, new InDel markers were developed and screened upstream of marker ID-4 using the maize Mo17 reference genome information as well as RP125 sequence data (Supplementary Table 2). All brittle culm monocotyledons in the F2 generation population were screened with these markers, the bands were counted, and the exchange rate was calculated. The target region was narrowed down step by step to 93 kb between SSR1 and SSR3, and the locus was mapped (Figure 4A).

According to the fine-mapping results, combined with the reference genome gene annotation information, three ORFs were found in the 93-kb region between SSR1 and SSR3 (Figure 4A). Thus, the corresponding specific primers were designed according to the three ORFs, and the three genes were amplified and sequenced using DNA of RP125 and bk-5 ^214S2L as templates to detect the mutation sites. The sequencing results showed that only Zm00001d043477 had a single-base mutation from C to T (Figure 4A), and the sequencing peak map showed that the SNP site was a single peak, indicating that the sequencing results were credible. For verification, 20 leaves of brittle culms were selected from the F2 generation plants of Mo17 × bk-5 ^214S2L , and DNA was extracted and mixed in equal amounts to make a cryptic mixed pool. The obtained SNP locus was amplified and sequenced using the mixed pool DNA as a template. The sequencing results showed that this locus in the recessive pool was also a pure mutation, and the sequencing peaks were single peaks (Figure 4B), which indicated that Zm00001d043477 was the mutant gene of bk-5 ^214S2L . Analysis of the Gramene database revealed that the Zm00001d043477 gene was 4,464 bp in length, with the longest transcript of 3,248 bp encoding 535 amino acids, which was functionally annotated as encoding the 11 subunits of cellulose synthase. A single-base mutation changed its 214th amino acid from serine (S) to leucine (L); thus, it was named ZmCesA11 ^S214>L .

Because mutations in the bk-5 ^214S2L gene affected plant brittleness, we investigated its expression in different tissue types of plants at the jointing stage. qRT-PCR results showed that bk-5 ^214S2L was most highly expressed in male spikes at the V14 stage (Figure 5A). To understand the expression of CesA related to cell wall formation, we performed differential expression analysis of CesA involved in the cell wall biosynthesis pathway using male spikes at stage V14 with the highest expression of bk-5 ^214S2L as samples (Supplementary Table 3). The expression of all CesAs involved in maize cell wall biosynthesis was significantly downregulated (Figure 5B). This indicates that bk-5 ^214S2L is essential to primary cell wall and secondary wall biosynthesis in maize.

The phenotypic properties of mutant fibers provide theoretical guidance for improving the palatability of maize silage; therefore, we determined the relevant indices for RP125, bk-5 ^214S2L , and bk-5 silage evaluation, including ADF, NDF, tannin, starch, soluble sugar, and crude protein contents. The ADF percentages in the stems of RP125, bk-5 ^214S2L , and bk-5 were 24.23%, 21.15%, and 22.51%, respectively, and the ADF content of bk-5 ^214S2L and bk-5 stems was significantly lower than that of RP125 stems, and the ADF content of bk-5 ^214S2L stems was 12.71% lower than that of RP125 stems (Figure 6A). However, the ADF content of bk-5 leaves was 12.71% lower than that of RP125 leaves. bk-5 leaves had significantly lower ADF content than RP125 leaves, and ADF content of bk-5 ^214S2L leaves was not significantly different from RP125 leaves (Supplementary Figure 6A). The NDF percentage of bk-5 ^214S2L stems was 12.98%, which was significantly lower than that of RP125 stems by 44.43%, and the NDF percentage of bk-5 stems was 12.18%, which was significantly lower than that of RP125 stems by 47.86% (Figure 6B). NDF content was similarly significantly reduced in the leaves of bk-5 ^214S2L and bk-5 compared with RP125 (Supplementary Figure 6B). The detergent fiber contents of both bk-5 ^214S2L and bk-5 complied with the regulations for primary silage maize. The tannin content of bk-5 ^214S2L stems was significantly reduced by 17.46% and that of bk-5 stems was significantly reduced by 14.29% compared with RP125 (Figure 6C). In leaves, no significant difference was found between the tannin contents of bk-5 ^214S2L , bk-5, and RP125 (Supplementary Figure 6C). The starch content of bk-5 ^214S2L stems was significantly higher than that of RP125 stems by 22.6%, and the starch content of bk-5 stems was significantly lower than that of RP125 stems by 45.1% (Figure 6D). The starch content of bk-5 ^214S2L leaves was not significantly different from that of RP125 leaves, and the starch content of bk-5 leaves was significantly reduced by 14.7% compared with RP125 and by 10.2% compared with bk-5 ^214S2L (Supplementary Figure 6D). For soluble sugars, the soluble sugar content was not significantly different between RP125 and bk-5 ^214S2L stems, whereas the soluble sugar content of bk-5 stems was significantly lower than that of both RP125 and bk-5 ^214S2L stems (Figure 6E); the soluble sugar content of both bk-5 ^214S2L and bk-5 leaves was significantly higher than that of RP125 leaves (Supplementary Figure 6E). The crude protein content of both bk-5 ^214S2L and bk-5 stems and leaves was slightly elevated compared with RP125 (Figure 6F; Supplementary Figure 6F). Thus, based on phenotypic properties and silage-related indices, the weak allelic variant of bk-5, bk-5 ^214S2L , can be used as a high-quality roughage resource to provide sufficient effective fibers for livestock to reduce the risk of rumen acidosis and ensure rumen health of livestock. It can also provide easily digestible NDF and ADF, increase roughage energy, reduce concentrate feed usage, and lower feed cost.