Wide grain 4,encoding an alpha-tubulin,regulates grain size by affecting cell expansion in rice

Yi Liu, Lianan Guo, Guoli Qu, Yang Xiang, Xu Zhao, Hua Yuan, Ting Li, Liangzhu Kang, Shiwen Tang,Bin Tu, Bingtian Ma, Yuping Wang, Shigui Li, Weilan Chen*, Peng Qin*

State Key Laboratory of Crop Gene Exploration and Utilization in Southwest China, Rice Research Institute, Sichuan Agricultural University, Chengdu 611130, Sichuan, China

Keywords:Cell size Microtubules Oryza sativa WG4

ABSTRACT Rice is one of the three most important food crops in the world.Increasing rice yield is an effective way to ensure food security.Grain size is a key factor affecting rice yield; however, the genetic and molecular mechanisms regulating grain size have not been fully investigated.In this study, we identified a rice mutant, wide grain 4-D (wg4-D), that exhibited a significant increase in grain width and a decrease in grain length.Histological analysis demonstrated that WG4 affects cell expansion thereby regulating grain size.MutMap-based gene mapping and complementary transgenic experiments revealed that WG4 encodes an alpha-tubulin,OsTubA1.A SNP mutation in WG4 affected the arrangement of cortical microtubules and caused a wide-grain phenotype.WG4 is located in nuclei and cytoplasm and expressed in various tissues.Our results provide insights into the function of tubulin in rice and identifies novel targets the regulation of grain size in crop breeding.

1.Introduction

Rice(Oryza sativa L.)is one of the three world’s most important food crops, providing 35%–60% of humanity’s need for energy and protein.Rice yields have a direct impact on global food security,but rice production has been stagnant for an extended period during which the global population has increased,and the area of arable land has fallen, exacerbating a potential global food crisis.The yield of rice comprises three main factors:effective number of panicles,number of grains per panicle,and grain weight,with the last being largely determined by grain size[1].The spikelet hull determines the storage capacity of the grain and limits its growth, and therefore plays a dominant role in determining grain size.The development of spikelet hulls is regulated by cell division and cell size.In the early stages of development, cells in the spikelet hull undergo numerous divisions after which cell division slows and cells expand in size.

Numerous genes related to grain size have been identified.These genes are spread across different regulatory pathways,including GS3,RGA1,GGC2,and DEP1 in the G protein signal transduction pathway [2–4], OsMKKK10, OsMKK4, OsMAPK6, and OsMKKK70 in the mitogen-activated protein kinase(MAPK)signaling pathway [5–7], CLG1, GW2, WG1, WTG1, and OsUBP15 in the ubiquitin–proteasome degradation pathway [8–11], phytohormones signal transduction-related GSE5, SG1, GL3.1, GSK2, DLT,WG3, and GS5 [12–18], transcriptional regulatory factors GW8,GLW7, and SPL18 [19–21], as well as additional grain size-related genes, such as SRS5 and SRS3 [22–24].Some of these genes, such as GS5, GSK2, and DLT, regulate both cell division and cell expansion,some,including GW2,GS3,GSE5,control grain size by regulating cell division, and others, such as WTG1, GLW7, OsBSK2, SRS3,and SRS5, control grain size by regulating cell expansion [25].

Microtubules are the main components of the cytoskeleton and are found in all eukaryotic cells.Tubulin is highly conserved throughout the biological world and plays an important role in the maintenance of cell morphology, intracellular material transport, cell activity, signal transduction, and cell division during plant growth and development [26].During cell division, tubulin plays a role in determining the location of cell division and promoting nuclear separation.In the process of cell expansion, it can affect the deposition of cellulose by changing the direction of microtubule arrangement, thus inhibiting cell elongation [27–30].Treatment with colchicine causes damage to microtubules and cells become rounded, suggesting that microtubules are also important for maintaining cellular asymmetry [31].Microtubules in plants, are mainly composed of alpha-tubulin (TUA) and betatubulin (TUB).There are 6 identified TUA and 9 TUB proteins in Arabidopsis thaliana [32], and 4 TUA proteins and 8 TUB proteins in rice[30,33].One of them,OsTubA1,is mainly expressed in highly fragmented tissues,and its preferred expression is regulated by its first intron [33].SRS5/TID1, which is highly homologous to OsTubA1, regulates cell elongation independently of the BR signaling pathway, and mutation results in significantly shorter cell length in the glumes and small,round seeds.It also affects the orientation of microtubule arrangement and microtubule dynamics,thus affecting growth tropism [22,34].Mutations in SRS5 gene homologs TUA6 and TUA4 in Arabidopsis thaliana affect the structure and stability of microtubule dimers, resulting in clockwise left-hand spiral growth and elongation [35].OsTUB8, encoding a beta-tubulin, is specifically expressed in rice anthers [36].

Although numerous factors regulating grain size have been identified, the role of tubulin remains largely unknown.In this study, we isolated a gain-of-function mutant, wide grain 4-D(wg4-D),which showed increased grain width and decreased grain length.MutMap-based cloning and transgenic experiments demonstrated that WG4 corresponds to OsTubA1, which encodes alpha-tubulin and has not been functionally reported.Our results suggest that WG4 plays an important role in regulating seed development and provide insights into the genetic regulatory network of grain size in rice.

2.Materials and methods

2.1.Plant materials and growth conditions

The wg4-D mutant was identified in an ethyl methanesulfonate mutant library of the indica restorer line Shuhui 498 (R498).All plants were grown under natural conditions in experimental fields of the Rice Research Institute, Sichuan Agricultural University.

2.2.Agronomic traits and phenotypic analysis

To analyze differences among the wild-type (WT) R498, wg4-D mutant and the transgenic lines (complementation & knock-out)phenotypes, 18 mature plant samples were randomly selected for agronomic trait analysis.We used a Mini1600 automatic analysis system(Jie Lai Mei Technology Co.,Ltd.,Chengdu,Sichuan)to measure 1000-grain weight and grain size.Microsoft Excel 2021(Microsoft, Redmond, WA, USA) was used to calculate the mean value and standard deviation (SD), and to conduct Student’s ttests.Duncan’s multiple comparisons were conducted using Data Processing System software [37].

2.3.Scanning electron microscopy

Pre-anthesis spikelet hulls of R498 and wg4-D mutant were collected and fixed in 2.5% glutaraldehyde for 48 h at 4 °C before dehydration in a graded ethanol series (from 30% to 100%) and observation under a scanning electron microscope(SEM)[38].Cell length and cell number within fields of view were measured using Image Pro Plus 6.0 software (Media Cybernetics, Rockville, MD,USA).

2.4.Genetic analysis and gene mapping

The WG4 gene was cloned by a MutMap strategy [39].We first crossed the wg4-D mutant with wild-type R498 to generate an F2population.Grain sizes from mature plants were measured, and segregation ratios were assessed by chi-squared tests.Twentyfive F2plants with wide grain phenotypes were selected and their leaves were pooled to extract DNA that was used in whole-genome resequencing and alignment of sequence reads to the R498 genome [40].Finally, we calculated single nucleotide polymorphism(SNP) indices and performed linkage analysis.

2.5.PCR and quantitative real-time PCR (qRT-PCR) analysis

Genomic DNA was extracted from leaves of R498 and wg4-D mutant plants using the cetyltrimethylammonium bromide(CTAB)method.Polymerase chain reaction (PCR) was performed using 2× PCR HERO MIX (dye) (FOREGENE, Chengdu, Sichuan).), Total RNA was extracted from various tissues of R498 and wg4-D mutant using a total plant RNA extraction kit(FOREGENE)according to the product manual.A PrimeScript RT kit and gDNA eraser (TaKaRa,Dalian, Liaoning, China) were used to reverse transcribe total RNA (500 ng) into cDNA and qRT-PCR was performed on a qTOWER3G real-time PCR thermal cycler(Analytik Jena, Jena,Germany) using real-time PCR Easy-SYBR Green I (FOREGENE).UFC1 and FhaB were used as internal controls [41].Each analysis was repeated three times.The primers for qRT-PCR are listed in Table S1.

2.6.Vector construction and transformation

Functional annotations of candidate genes were obtained from the current reference genome of indica cultivar R498 (https://ricerc.sicau.edu.cn/).One sgRNA target in exon 2 was selected to construct a CRISPR/Cas9 vector using a CRISPR/Cas kit (BIOGLE,Hangzhou, Zhejiang).The final CRISPR/Cas9 construct was introduced into R498 by Agrobacterium tumefaciens-mediated transformation.Two k promoter sequences upstream of ATG and the full-length coding sequence of WG4 without stop codon were amplified from R498 or wg4-D mutant and cloned into a pCAMBIA1300 vector and introduced into wg4-D mutant or R498.In addition, the full-length coding sequence of WG4 without a stop codon was amplified from R498 and cloned into the pCAMBIA2300-eGFP vector to generate a 35S::WG4-eGFP construct.The 2 k promoter sequence upstream of ATG in WG4 was amplified from ZH11 and cloned into a DX2181 vector.The primers for vector construction are listed in Table S1.

2.7.Subcellular localization

35S::eGFP and 35S::WG4-eGFP were transiently expressed in rice protoplasts to observe the subcellular localization of WG4.The GFP signal image was captured using a Leica laser confocal microscope.We used Agrobacterium tumefacienss-mediated transformation to transfer 35S::eGFP and 35S::WG4-eGFP into Nicotiana benthamiana leaves.Fluorescence was observed and photographed with a Leica laser confocal microscope after culturing for two days.

2.8.Immunofluorescence

Microtubule immunofluorescence followed the methods of Baluska [42] and Lloyd [43].Root tips from 3-day-old R498 and wg4-D seedlings were fixed at room temperature in a PEMT(50 mmol L-1PIPES, 1 mmol L-1EGTA, 1 mmol L-1MgCl2, 0.05%TritonX-100) solution containing 4% paraformaldehyde for 1 h,washed with PEMT buffer for 10 min,then placed in the enzymatic hydrolysate containing 2% cellulase and 1% pectinase and hydrolyzed at 37 °C for 1 h.The softened root tips were gently rinsed 3 times with the PEM buffer,each time for 10 min,and then rinsed with the PBS buffer for another 10 min prior to incubation overnight at 4 °C in a buffer containing 1500 diluted rabbit antialpha-tubulin antibody (Proteintech, 11224–1-AP).The root tips were washed five times with PBST.AlexaFluor488 sheep antirabbit antibody diluted in PBST was used for hybridization at room temperature for 5 h.The microtubules were washed with PBST 4 times and their structures were observed using a Leica confocal microscope.

3.Result

3.1.Characterization of the wg4-D mutant

We generated an ethyl methanesulfonate mutant library of the indica variety R498 and isolated the wg4-D mutant(Fig.1A)used to study the molecular genetic mechanism of grain size.The grain width of the wg4-D mutant was increased by 15.89% relative to R498 (Fig.1B, C), the grain length was decreased by 22.04%(Fig.1D) and 1000-grain weight was reduced by 12.56% (Fig.1E).Among additional agronomic traits assessed for R498 and the wg4-D mutant the number of tillers in the wg4-D mutant was reduced and panicle length was decreased by 25.80%, whereas grain density and number of primary branches increased by 24.08%and 8.41%,respectively.There was no significant difference in the number of grains per panicle (Table S2).Since the wg4-D mutant exhibited a potential wide-grain phenotype before anthesis (Fig.1F), grain size was determined by cell division or cell expansion [44].Therefore, we observed cell size and cell number inside the lemma of the R498 and wg4-D mutant using SEM(Fig.1G, H).The epidermal cell length of the lemma of the wg4-D mutant was significantly shorter and wider than that of R498(Fig.1I, J), whereas the number of cells was unchanged (Fig.1K).These results suggested that the wg4-D mutation inhibited elongation of longitudinal cells but promoted expansion of transverse cells.

3.2.WG4 was predicted to encode alpha-tubulin

To identify the gene involved in the wg4-D mutant phenotype,we generated an F2population by crossing the wg4-D mutant with R498.The wg4-D mutation was partially dominant:there were 157 plants with wide and short grains, 349 plants with intermediate grains, and 173 plants with normal (WT) grains conforming to a 1:2:1 segregation ratio (χ2c= 0.66 < χ20.05(2)= 5.99) (Fig.S1E, F).There was a significant negative correlation between grain width and length among F2plants (Fig.S1G).These results indicated that grain width and length in the wg4-D mutant was controlled by a single allele.We then identified the wg4-D mutation site using the Mut-Map method.A distinct SNP cluster of four mutation sites was found on chromosome 3 (Fig.S1H, I).However, alignment with the reference genome indica cultivar R498 identified only one mutation site co-segregating with the grain width phenotype and it was located in the coding region of LOC_Os03g51600, which encodes an alphatubulin.A C to A substitution in the third exon led to a nonsynonymous mutation from histidine to asparagine (Fig.S1I).This amino acid (His192) was located in the tubulin domain of LOC_Os03g51600 and was conserved among all orthologs identified in a wide range of plant species (Figs.S2, S3), suggesting that this amino acid is required for LOC_Os03g51600 function.Therefore, the mutation in LOC_Os03g51600 was likely responsible for the wide grain phenotype of the wg4-D mutant, and hereafter,LOC_Os03g51600 is named WG4.

To verify that the single-nucleotide substitution in WG4 was responsible for the grain phenotype of the wg4-D mutant, we introduced a WG4 genomic fragment into the wg4-D mutant and introduced a wg4-D mutant genomic fragment into WT R498 (Fig.S4B).As expected, in the background of the wg4-D mutant, the grain width and length of gWG4 transgenic lines were between those of R498 and the wg4-D mutant,and the performances of gwg4-D transgenic lines in R498 were consistent with those of the wg4-D mutant (Fig.S4C–F).These results indicated that the grain phenotype of the wg4-D mutant was caused by a gain-of-function mutation in WG4.We also generated two knockout (KO) mutants of WG4 in R498 background; both had premature stop codons and lost the two tubulin domains(Fig.S5C).Grain size in the two KO mutants was not changed relative to R498 (Fig.S5E, F).Taken together, our results supported the prediction that WG4 encoded alpha-tubulin and that a gainof-function mutation in WG4 was responsible for the wide grain phenotype of the wg4-D mutant.

3.3.WG4 influences the transverse orientation of cortical microtubules

Changes in tubulin can affect cell mitosis and the arrangement of microtubules [22].A comparison of mitosis and cortical microtubule arrays in the root elongation zones of R498 and wg4-D mutant cells using immunofluorescence staining found no significant difference in mitosis (Fig.1L–S).Whereas cortical microtubules in transverse arrays in R498 were well organized and highly aligned (Fig.1T) there was a lower percentage of cells with highly-ordered transverse arrays in the wg4-D mutant (Fig.1U)and a larger range in of the angle of distribution of microtubules.These results suggested that the wg4-D mutation affected the arrangement of cortical microtubule arrays.

Fig.2.Expression profile of WG4.(A) Expression pattern of WG4 in different tissues.R and LB, root and leaf blade at the seedling stage; C and LS, culm and leaf sheath at booting;YP1-YP9,young panicles with different lengths(cm);H9-H15,hulls from panicles with different lengths(cm);E5,E10 and E15,caryopses at 5,10 and 15 days after pollination;UFC1 and FhaB were used as internal controls.Data are means±SD(n=3 replicates).(B–F)Expression pattern detected in plants harboring the fusion construct PromoterWG4::GUS.GUS activity detected in a primary root(B),developing culm(C),leaf blade(D),leaf sheath(E),and spikelet hull(F).(G)Subcellular localization of WG4-eGFP in rice protoplasts.(H) Co-localization of WG4 with nuclear marker in Nicotiana benthamiana leaves.

3.4.Expression profile and subcellular localization of WG4

qRT-PCR analysis of the accumulation of WG4 mRNA transcripts during plant development showed that WG4 was expressed in all tissues.WG4 mRNA mainly accumulated in young panicles at different developmental stages (Fig.2A).This expression pattern was consistent with its biological function in regulating grain size.GUS staining revealed that WG4 was expressed in roots, culms, leaf sheaths, and spikelet hulls(Fig.2B–F),consistent with the qRT-PCR results.Transient transformation experiments in rice protoplasts showed WG4-eGFP protein signals in the cytoplasm and nuclei (Fig.2G).WG4 and a nuclear-localization protein (D53) [45] were co-localized in tobacco, confirming that WG4 was partially located in the nucleus(Fig.2H).These results suggest that WG4 is a constitutively expressed gene that localizes to the nucleus and cytoplasm.

4.Discussion

In this study we isolated a partially dominant rice wg4-D mutant with decreased grain length and increased grain width (Fig.1C, D).SEM of cell size and number in inner epidermis cells of the lemma of the R498 parent and wg4-D mutant plants showed that the mutation affected cell expansion (Fig.1I–K).The WG4 gene encoded alpha-tubulin (Fig.S1), which influences the transverse orientation of cortical microtubules (Fig.1T, U).There are four alpha-tubulins in rice, and WG4 was highly homologous with the reported SRS5.An srs5 mutant was also partially dominant and exhibited shorter and rounder seeds in shorter panicles, similar to the wg4-D mutant[22].Moreover, the short grain phenotype of the srs5 mutant was due to effects on cell expansion.The mutation in SRS5 also caused changes in the arrangement of microtubules.These results strongly indicate that the arrangement of microtubules plays an important role in rice growth and development.

Microtubules are ubiquitous cytoskeletal components that are highly similar in structure across most cell types and organisms.However,their different protein subtypes,rich post-translational modifications, and various microtubule-binding proteins make them able to adapt to specific functions in different cell types.Moreover, these three important factors do not work alone but complement each other, and a tiny change in one can cause a subtle change in the function of microtubule populations.‘‘General-purpose” subtypes of alpha-tubulin and beta-tubulin are highly conserved between evolutionarily distant species,whereas less common subtypes evolve as new microtubule functions emerge.For example, Qinghai spruce pollen-specific tubulin TUA1 induced a higher pollen germination rate and increased pollen tube growth rate when expressed in Arabidopsis thaliana, suggesting that the isoform changes were accompanied by new functions.In this study,we found that the cortical microtubules of WT R498 cells in the elongation region of root tips were arranged horizontally, but in wg4-D mutant plants the orientation was significantly changed (Fig.1T, U), suggesting that changed functions may be accompanied by changes in the microtubule array.Changes in microtubule arrays can lead to changes in many plant traits, such as small-grain phenotypes as in this study,and other phenotypes,such as reduced plant height,lower tiller number and lower seed setting rate,that have not been further studied.The results of this study will contribute to the further exploration of the biological significance of tubulin in plant growth and development.

CRediT authorship contribution statement

Yi Liu:Data curation, Methodology, Formal analysis, Investigation,Writing–original draft,Writing–review&editing,Visualization.Lianan Guo:Data curation, Methodology, Validation,Investigation.Guoli Qu:Investigation.Yang Xiang:Validation.Xu Zhao:Validation.Hua Yuan:Methodology, Formal analysis.Ting Li:Methodology.Liangzhu Kang:Resources, Supervision.Shiwen Tang:Resources, Supervision.Bin Tu:Formal analysis.Bingtian Ma:Formal analysis.Yuping Wang:Formal analysis.Shigui Li:Conceptualization, Methodology.Weilan Chen:Resources,Supervision,Funding acquisition,Writing–review&editing.Peng Qin:Conceptualization, Methodology, Supervision, Project administration, Funding acquisition, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was supported by grants from the National Natural Science Foundation of China (32121003), the Sichuan Science and Technology Program(2022ZDZX0012,2022YFQ0026),and the Natural Science Foundation of Sichuan province (2022NSFSC1667).This work was supported by High-performance Computing Platform of Sichuan Agricultural University.

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2023.05.013.