3.1 | Characterization of the wgls1 mutant
The wgsl1 mutant was discovered during the screening of an EMS-treated population of M2 mutant lines of RP125. The wgsl1 mutant was self-pollinated for over five generations before researched further. The phenotype of the wgsl1 mutant was first observed at the five-leaf stage (Fig. 1A; Supplementary Fig. 1), when it exhibited white and green striate leaves, a characteristic that remained visible throughout its life cycle, including in petiole and bracts (Fig. 1B_D). To investigate the impact of the mutant phenotype on photosynthesis, we analyzed photosynthesis pigment contents, including Chl a, Chl b, and carotenoid, in wgsl1 and the WT. The results showed that Chl a, Chl b, and carotenoid contents in the wgsl1 mutant were significantly lower than the WT (Fig. 1E). Additionally, given the critical role of chlorophyll in photosynthesis, we measured the photosynthesis rate of the wgsl1 mutant and the WT. As anticipated, there was no significant change in stomatal conductance and transpiration rate compared with the WT (Supplementary Fig. 2A, C), but the wgsl1 mutant had a significantly lower photosynthetic rate (Fig. 1F) and a significantly higher intercellular CO2 concentration (Supplementary Fig. 2B), consistent with the results of the chlorophyll analysis.
All of these phenotypes can affect yield traits in field growing conditions. Therefore, we evaluated plant, ear, and kernel phenotypes. The plant height of wgsl1 was 17.8% (P < 0.01) lower than that of the WT, and the ear height was 32.67% (P < 0.05) lower (Supplementary Fig. 3A), indicating that changes in leaf color impacted plant growth. Further investigation revealed that the ear length and ear diameter of wgsl1 were 19.4% and 11.8% smaller (P < 0.05), respectively, than those of the WT (Supplementary Fig. 3B). Kernel length and width were 5.8% and 3. 37% smaller (P < 0.05), respectively, than those of the WT (Supplementary Fig. 3C), and ear weight was 61.7% smaller (P < 0.01). However, there was no significant difference in 100-grain weight (Supplementary Fig. 3D). In conclusion, wgsl1 had white and green striate leaves throughout the plant from the five-leaf stage, and its development, ears, and grain yield were affected.
3.2 | Defective chloroplast development in wgsl1
The development of chloroplasts is closely linked to chlorophyll content and photosynthesis. To determine whether chloroplast development was impacted in the wgsl1 mutant, we conducted an analysis of leaf sections at the five-leaf stage using a microscope. The semi-thin section study results revealed that WT leaves had more chloroplasts (Fig. 2A, B), whereas in the mutant, almost no chloroplasts were observed (Fig. 2D, E). This finding suggested that the white-green stripes in the leaves were caused by the lack of chloroplasts in these regions. Cryo-electron microscopy results further supported this conclusion. Figure 2C, F clearly shows that there was minimal chlorophyll in the cross-section of the mutant leaves near the vascular bundle, whereas in the WT cross-section, there was a significant accumulation of chlorophyll.
Additionally, the ultrastructure of the chloroplasts was observed using TEM. The WT chloroplasts showed normal development, with dense and well-structured grana stacks (Fig. 3A_C). By contrast, wgsl1 had chloroplasts with disrupted architecture, lacking stromal thylakoids (Fig. 3D_F); Some mesophyll cells even lacked chloroplasts altogether. These findings indicate that wgsl1 has severely impaired chloroplast development.
3.3 | Genetic analysis and map-based cloning of ZmWGSL1
To uncover the molecular mechanism behind the phenotype of the wgsl1 mutant, a mapping population was generated by crossing the wgsl1 mutant with inbred B73. The crossing produced only F1 plants with the normal phenotype, and the phenotype segregation in the F2 population was consistent with a 3:1 ratio of WT to mutant, as tested by a chi-square test with χ2 < χ2 (0.05, 1) = 3.84 (Supplementary Table 1). These results suggest that the wgsl1 phenotype is controlled by a single recessive nuclear locus.
The ZmWGSL1 gene was apprpximately mapped using SSR markers and BSA. The mapping process initially identified multiple markers (umc2165, bnlg1759a, umc1779, and umc1653) on chromosome 6 that co-segregated with the ZmWGSL1 gene through the screening of 206 SSR polymorphic markers between B73 and wgsl1. Additionally, insertion-deletion polymorphism markers were developed for further mapping (Supplementary Table 2). The ZmWGSL1 gene was ultimately mapped to the region between markers Indel-1 (five recombinants) and Indel-13 (four recombinants) through the analysis of 573 individuals from the B73/wgsl1 F2 population. The markers Indel-1 and Phi123 on the long arm were linked to the ZmWGSL1 gene in a 91.4 kb region. Six gene models were predicted in this region based on the maize B73 genome sequence (version 4.0) (Fig. 4A). The predicted region was then sequenced to detect mutations, revealing a ‘G’ to ‘A’ substitution in the protein-coding region of the target gene ZmWGSL1 (Zm00001d039036) (Fig. 4B). This gene has nine exons and eight introns, and the mutation occurs in the second exon, resulting in a change of the amino acid Gly to Arg (Fig. 4B). As a result, ZmWGSL1 is the most likely candidate for the wgsl1 locus. To confirm that the maize ZmWGSL1 mutation causes a white-green stripe phenotype, we verified the knockout of maize ZmWGSL1. Twenty-four T0 generation maize plants (unique plant numbers 195-1-195-24) were obtained after genetic transformation. DNA from these single plants was then extracted separately and PCR was designed to amplify fragments containing CRISPR loci using the genome as a template. The amplification products were 375 bp in size and were sequenced. The results showed that seven strains, including monoculture 195 − 10, 195 − 11, 195 − 14, 195 − 17, 195 − 19, 195 − 21 and 195 − 24, had large deletions in the target gene sequence region (Fig. 4D). Phenotypic observation of T0 generation seedlings revealed that ZmWGSL1 mutation showed an albino and lethal phenotype (Fig. 4E); we determined from the phenotypes and sequencing results of the maize transgenic knockout lines that mutation of the ZmWGSL1 candidate gene results in a change in leaf color, with a strong mutation causing an albino and lethal phenotype and a weak mutation, wgsl1, resulting in a white-green stripe phenotype.
3.4 | ZmWGSL1 encodes a 16S rRNA processing protein
Gene prediction indicated that ZmWGSL1 encodes 16S rRNA processing protein belonging to the RimM family with 661 amino acids and a calculated molecular mass of 73.2 kDa. Domain analysis using the NCBI conserved Domain database (https://www.ncbi.nlm.nih.gov/Structure/cdd/-wrpsb.cgi) revealed that ZmWGSL1 contain rimM motif, and is a member of the RimM superfamily. The mutation site was located in the rimM superfamily motif, resulting in amino acid change from Gly to Arg (Fig. 5A). As the wgsl1 mutant plants exhibit a white-green stripe, this mutation would appear to play an important role in the functional integrity of ZmWGSL1 protein. BLASTp analysis of the NCBI database showed that there are several homologous proteins in Arabidopsis., Sorghum bicolor, Brachypodium distachyon, and Medicago truncatula. Among these species, ZmWGSL1 showed the highest degree of identity with S. bicolor (71.19%), and moderate similarity with B. distachyon (21.52%). It showed a low degree of identity with Arabidopsis (8.64%) and M. truncatula (11.21%). The mutation site in the RimM motif is highly conserved in plants, suggesting that it plays a crucial role in the function of the protein (Fig. 5B). A phylogenetic analysis of the WGSL1 protein and related proteins was performed to study their evolutionary relationships. The results showed a clear divergence between monocots and dicots (Fig. 5C). However, the functions of these homologous proteins are not yet well understood.
3.5 | Candidate gene expression profiling, ribosome profile and subcellular location of ZmWGSL1
As the ZmWGSL1 mutation affects plant leaf color, we examined ZmWGSL1 expression in different tissues of the whole plant at the plucking stage. RT-qPCR showed that ZmWGSL1 is highly expressed in stems and roots at the plucking stage (Figure. 6A). To examine the impact of the wgsl1 mutation on ribosomal subunit biogenesis and the formation of monosome and polysome complexes, polysome profiles were analyzed with five-leaf stage leaf extracts using 5–50% sucrose gradient centrifugation. In the leaf mutant, the 40S and 80S ribosomal subunits were reduced, compared with 60S ribosomal subunits (Fig. 6B). The levels of monosome and polysome complexes were calculated by measuring the peak areas of absorbance at 254 nm (A254). The wgsl1 mutant had a higher 60S/40S ratio and a lower 80S/40S ratio than the WT. The calculation of the peak area of A254 revealed that 40.2% of the ribosomes in the WT leaf extracts were in polysomes, and, the level of polysome complexes was reduced in the wgsl1 mutant.
ChloroP (http://www.cbs.dtu.dk/services/ChloroP/) and TargetP (http://www.cbs.dtu.dk/services/TargetP/) were used to predict the subcellular localization of ZmWGSL1: a chloroplast localization was forecasted. To confirm this, two constructs were generated: p35S:WGSL1-GFP, containing the full length WGSL CDS (coding sequence) fused to green fluorescent protein (GFP) without the stop codon, and a control empty vector p35S: GFP. The two plasmids were introduced into maize protoplasts using a polyethylene glycol-mediated method (Harthill et al 2006). Confocal microscopy was used to visualize the fluorescent signals, and the results showed that the WGSL1-GFP fusion protein co-localized with the autofluorescence signals of the chloroplasts in the maize protoplasts. By contrast, the GFP signals were observed in the plasma membrane, cytoplasm, and nucleus in the protoplasts transformed with the control vector (Fig. 6C). These findings demonstrate that ZmWGSL1 is targeted to the chloroplasts.
3.6 | RNA-seq analysis
The expression of plastid genes is often linked to defective chloroplast development. To examine the impact of the ZmWGSL1 mutation on gene expression, we performed RNA-seq analysis of WT and wgsl1 five-leaf stage leaves. Results showed that 1098 genes displayed differential expression (843 up-regulated and 255 down-regulated) between the wgsl1 mutant and the WT, with a threshold fold change greater than 1 and a q-value less than 0.05. To verify the accuracy of the transcriptome data, we selected differential genes related to ribosome synthesis pathway genes and, porphyrin and chlorophyll II metabolic pathways for quantitative validation; differential expression analysis of these genes proved the accuracy of the transcriptome data (Supplementary Fig. 5A, B). To further understand the functions of these DEGs, we conducted gene ontology (GO) annotation and enrichment analysis. A total of 36 GO terms were significantly enriched (corrected P < 0.05), including biological process (13), cellular component (13), and molecular function (10) (Supplementary Fig. 4A). The eight most relevant GO terms were GO: 0005840 (ribosome, corrected P = 2.5987E-16), GO: 0003735 (structural constituent of ribosome, corrected P = 1.19E-15), GO: 00005198 (structural molecule activity, corrected P = 6.4724E-18), GO: 0030529 (intracellular ribonucleoprotein complex, corrected P = 4.33E-14), GO: 1990904 (ribonucleoprotein complex, corrected P = 4.33E-14), GO: 0006518 (peptide metabolic process, corrected P = 1.70E-11), GO: 0043603 (cellular amide metabolic process, corrected P = 1.70E-11), and GO: 0006412(translation, corrected P = 2.17E-11). Sixty-eight DEGs classified under GO: 0005840 (ribosome) were divided into four categories: 40S ribosomal protein, 60S ribosomal protein, plastid-specific 30S ribosomal protein, and translation elongation factor EF1B. All of these were up-regulated in the wgsl1 mutant compared with the WT. In the biological process category, 71 DEGs classified under GO: 0006412 (translation) were divided into categories. A Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis was performed (Supplementary Fig. 4B), which indicated that the ribosome was the most prominent pathway (corrected P = 8.29E-16), consistent with the GO analysis (Fig. 7A). Furthermore, porphyrin and chlorophyll metabolism were also enriched, with all genes up-regulated in both pathways (Fig. 7B). Collectively, the transcriptomic data provide additional evidence that ZmWGSL1 is responsible for the formation of associated ribosomes in chloroplasts.
3.7 | Gene knockout to verify gene function
Thirty T0 generation rice plants (unique plant numbers M1-M31) were obtained after genetic transformation (some seedlings died). DNA from these monocultures was then extracted separately and PCR was designed to amplify the fragments containing the CRISPR loci using the genome as a template. The amplification products were 500 bp in length and were sequenced. The results showed that the peak curves of monocultures M17, M18, M20, and M24 had sets of peaks from the CRISPR loci, indicating that the target genes of these monocultures were mutated; monocultures M19, M27, and M30 had single-base insertions (Fig. 8B_D). The PCR results indicated that M20, M27, M29, and M30 are pure mutations, and M17, M18, M24, and M31 are heterozygous mutations. M20, M27, M29, and M30 T0 generation plants exhibited the same white-green streak trait as wgsl1 maize (Supplementary Fig. 6); phenotype and sequencing results of transgenic rice knockout lines indicated that the mutation in the ZmWGSL1 candidate gene causes the appearance of the white-green streak trait in leaves.