LIU Kai,LI MinjuanZHANG Bin,YIN XumingXIA XinjieWANG ManlingCUI Yanchun
(1Key Laboratory of Agro-Ecological Processes in Subtropical Region,Institute of Subtropical Agriculture,Chinese Academy of Sciences,Changsha 410125,China;2University of the Chinese Academy of Sciences,Beijing 100049,China;3College of Agriculture and Biotechnology,Hunan University of Humanities,Science and Technology,Loudi 417000,China)
Abstract:Grain yield is a polygenic trait that can be influenced by environmental factors and genetic compositions at all plant growth stages.Currently,the molecular mechanisms behind the coordination of the interaction between grain yield-related traits remain unknown.In this study,we characterized the function of four STRESS_tolerance and GRAIN_ LENGTH (OsSGL) Poaceae ortholog genes that are transcribed into DUF1645 domain-containing proteins in relation to the grain length,grain weight,and drought stress-tolerance of rice.The transgenic plants with overexpressing or heterologous high levels of Poaceae OsSGL ortholog genes exhibited longer grain size than the wild type plants.Larger cells were seen in panicles of the four transgenic lines with paraffin sectioning and scanning electron microscopy analyses.In addition,four Poaceae OsSGL ortholog genes positively affected the drought tolerance of rice.Four transgenic plants displayed higher resistance to drought stress at the seedling and vegetative stages.RNA-sequencing and qRT-PCR results indicated that over-or heterologous-expression of four Poaceae OsSGL ortholog genes also affected the transcriptome of rice plants.These genes may play a role in auxin and cytokinin biosynthesis and their transduction pathways.Taken together,these results suggested that the four OsSGL orthologs have a conserved function in the regulation of stress-tolerance and cell growth by modulating hormonal biosynthesis and signaling.
Key words:graminaceous cereal crop;OsSGL gene;drought resistance;grain length;grain weight
More than 60% of the total worldwide agricultural production is from domesticated graminaceous cereal crops including rice (Oryza sativa),wheat (Triticum aestivum),maize (Zea mays),barley (Hordeum vulgare),sorghum (Sorghum biocolor),and small millets (Setaria italica) (Zhou et al,2016).Crop yield is controlled by multiple genes simultaneously and heavily influenced by the surrounding environments (Wing et al,2018;Panda et al,2021).Numerous genes have been identified as being important in regulating plant stress-tolerance at various developmental stages through diverse mechanisms (Pardo,2010;Zhou et al,2021).These genes are involved in different transduction pathways,and therefore,it is necessary to understand the crosstalk between these genes and their regulation of grain yield and crop stress response (Cao et al,2021).The genes,which may enhance crop productivity and improve plant stress-tolerance,can potentially be used in breeding of new crop varieties that can withstand unforeseen environmental stresses and changes (Zuo and Li,2014;Scheres and van der Putten,2017).
The Pfam database (http://pfam.xfam.org/) contains a collection of protein families which present with multiple sequence alignments and hidden Markov models.According to this database,there are some highly conserved plant-specific domain of unknown function (DUF) proteins that play important roles in plant growth and development,defense against diseases and insect pests,and adaption responses to abiotic stress.For example,DUF26 (PF01657),a duplicated domain in the rice root meander curling protein (OsRMC) that has also been found in ginkbilobin-2 fromGinkgo biloba,is annotated to be in the salt stress response/antifungal families (Sawano et al,2007;Zhang et al,2009).DUF640-containing genesG1,TRIANGULAR HULL 1(TH1)/BEAK-SHAPED GRAIN 1(BSG1) andABNROMAL FLOWER AND DWARF 1(AFD1) are expressed in young rice inflorescences and spikelet lemmas and paleas,and affect plant height,floral development,and grain yield through the regulation of cell division and expansion-related genes (Yoshida et al,2009;Li et al,2012;Ren et al,2016).The DUF1644 genesOsSIDP366andOsSIDP409may function as regulators of processing bodies/stress granules and can positively regulate salt and drought resistance in rice (Guo et al,2016).Overexpression of a novel ABA-responsiveTRITICUM AESTIVUMSALT-RELATEDHYPOTHETICAL PROTEIN(TaSRHP) gene that encodes a conserved DUF581 domain has been shown to enhance resistance to salt stress inArabidopsis thaliana(Hou et al,2013).
Rice has been recognized as a model cereal not only because of its vast cultivated area and importance as a staple food but also because of its small genome size and availability of high-resolution linkage maps (Valluru et al,2014;Mustafiz et al,2016).Grain size,one of the major components that determines grain yield and quality,is an important selective target during domestication and breeding.To date,several genes or QTLs related to grain size have been identified and extensively studied using homology-based cloning approaches.GS3,the first characterized QTL for grain size regulation,encodes a putative phosphatidylethanolamine-binding protein that negatively regulates grain size (Fan et al,2006;Mao et al,2010).Its maize ortholog,ZmGS3,contains domains in common with the GS3 protein that uses different functional polymorphisms and possibly different mechanisms to control maize kernel development (Li et al,2010b).GW2,encoding a RING-type E3 ubiquitin ligase,negatively regulates rice grain weight (Song et al,2007).Two of its maize co-orthologsZmGW2-CHR4/5control some of the phenotypic variation of maize kernel size and weight (Li et al,2010a).Extensively research on three wheat homologs,TaGW2-A,TaGW2-BandTaGW2-D,includes gene cloning,expression analysis,evolution experiment,functional marker development and elucidation of the genetic effects of each homolog (Su et al,2011;Simmonds et al,2016;Qin et al,2017).In addition,GS5encodes a putative serine carboxypeptidase that can function as a positive regulator of grain size.Higher expression ofGS5increases grain width and grain yield by accelerating cell division and cell expansion in the spikelet hull (Xu et al,2015).Maize orthologZmGS5affects kernel development,suggesting a conserved function of this ortholog among different plant species (Liu et al,2015).TaGS5homoeologues in wheat have been isolated and mapped on chromosomes 3A,3B and 3D.TaGS5-3Ais a positive regulator of grain size,and its favored allele,TaGS5-3A-T,can potentially be applied and used for high-yield wheat breeding (Wang et al,2015;Ma et al,2016).
A DUF1645 (Pfam PF07816) domain-encoding gene,OsSGL(Oryza sativaSTRESS_tolerance and GRAIN_ LENGTH),has been identified as a powerful pleiotropic gene that positively regulates rice grain length,grain weight,and drought resistance (Cui et al,2016;Wang et al,2016).TheOsSGLhomolog in sorghum,SbSGL,also shows a conserved function in seed size control by regulating cell division (Zhang et al,2018).DUF1645 proteins are widespread and can be found in both monocots and eudicots,and may be more or less conserved in plants closely related to theOryzaspecies.Thus,we hypothesized thatOsSGLis strongly conserved in rice genomes and thatPoaceaeorthologs ofOsSGLmay play a role in increasing grain length and water-deficit stress tolerance in cereal plants.To test this,four full DUF1645 domain-containingOsSGLorthologs fromsorghum (SbSGL),maize (ZmSGL),millet (SiSGL),and rice (OsSGL2) were cloned and functionally characterized by heterologous-or over-expression in anindicarice cultivar.Phenotypic observations of these transgenic rice plants included increased grain size and weight,enhanced water-deficit stress tolerance during the seedling and vegetative stages,enlarged root systems,and higher osmolyte content.Genomic observations showed that there were highly overlapping transcript levels of cell cycle and stress-responsive genes that were altered.
FourOsSGLorthologs containing entire DUF1645 domains were cloned fromsorghum (SbSGL),maize (ZmSGL),millet (SiSGL) and rice (OsSGL2) and then transferred intoindicarice cultivars (Fig.1-A and -F,Tables S1 and S2).ZmSGL(NCBI Accession KT626002) is located in maize chromosome 4 and encodes a putative proline-rich receptor-like protein kinase PERK7.The cDNA ofSiSGL(NCBI Accession XM_004951915.4) is 1 300 bp long and contains an 876-bp open reading frame (ORF) flanked by a 5'-UTR (232 bp) and a 3'-UTR (192 bp).SiSGLencodes a putative 291-aa protein with a molecular mass of 30.3 kDa.The cDNA ofOsSGL2(NCBI Accession XM_015770529.2) is 1 424 bp long and contains a 747-bp ORF flanked by a 5'-UTR (133 bp) and a 3'-UTR (544 bp).OsSGL2encodes a putative 248-aa protein with a molecular mass of 26.7 kDa.There were low DNA similarities (up to 31% with sorghum) revealed when aligning the nucleotide sequences and ORFs ofOsSGLand non-Oryzaspecies,whereas the OsSGL protein is relatively closely related (up to 53.3% with maize) to its homologs because of the highly conserved DUF1645 domain (Wang et al,2016).However,the relationship between the two rice orthologs (OsSGL and OsSGL2) and other rice DUF1645 proteins is more divergent,suggesting that OsSGL2 plays different roles in rice (Fig.S1).
Nucleotide sequence diversity and haplotype studies of theOsSGLlocus (LOC_Os02g04130) conducted in rice cultivars and wild rice accessions were performed using RiceVarMap v2.0.Only four single nucleotide polymorphisms in the ORF region corresponded to synonymous mutations,suggesting a conserved and important role ofOsSGLin rice growth (Table S3).Next,the expression profiles ofOsSGLandOsSGL2were analyzed from the young roots and leaves of 7-day-old seedlings,the flag leaves and young stems at the elongation stage,and the panicles and mature stems at the booting stage in rice.Quantitative real-time PCR (qRT-PCR) results showed thatOsSGLandOsSGL2were expressed in almost all tissues examined and that the expression was the greatest in young roots and panicles.This suggested thatOsSGLandOsSGL2may play important roles in regulating early germination and reproductive development in rice (Fig.S2).
To elucidate the possible functions of the four complete DUF1645 domain-containingOsSGLorthologs,we investigated the phenotypes of the four transgenic plants with over-or heterologous-expression ofPoaceae OsSGLortholog genes (Fig.1-A to -F).TheOsSGLortholog transcripts were abundant in the transgenic lines and there were large differences among them (Fig.S3).Transgenic plants expressingZmSGL(OE-ZmSGL) were slightly shorter in stature,but the grains were on average 7.97% longer,6.41% narrower,and 8.75% heavier than those of the wild type plants,leading to an average 13.84% increase in grain yield (Fig.1-H to -L and Table S4).The OE-SiSGLlines showed the smallest changes in grain size,with 5.43% longer in grain length,3.74% narrower in grain width,and 7.21% heavier in grain weight than the wild type (Fig.1-H to -L and Table S4).Overall,the average grain lengths,widths,and 1000-grain weights were dramatically increased in transgenic lines,while the plant heights of the transgenic plants were decreased relative to the wild type plants (Fig.1-J to -L).These results indicated that the expression ofOsSGLorthologs promote the formation of slightly dwarfed plants with slenderer and heavier grains.
Given that the over-or heterologous-expression transgenic lines displayed larger grain sizes,we analyzed the cell number and the size of the palea/lemma from the transgenic plants.The spikelet hulls of transgenic plants were more slender than those of the wild type plants before flowering (Fig.2-A).Transverse sections of the central parts of the palea/lemma of plant florets were microscopically inspected before flowering to investigate the increased organ size on a cellular basis (Fig.2-B to -K).It revealed that the inner parenchyma cell layer of the palea/lemma in the OE-ZmSGLplants contained 23.7%-32.2% more cells than those in the wild type and the cells were 12.6%-34.6% larger (Fig.2-D,-I and -L).Furthermore,we compared the center of the maturated spikelet hulls (lemma) in transgenic lines to the corresponding controls with a scanning electron microscopy (SEM).The spikelet hull cells of OE-ZmSGLwere larger in size (~21.7%) and significantly longer in longitudinal orientation (~24.9%) than those of the control,though the cell widths were decreased (~3.2%) and the estimated cell numbers were slightly lowered (Fig.2-D,-I,-M and -N).The longer and narrower spikelet hulls of OE-ZmSGLplants resulted in combination with increased cell length and size in the longitudinal direction and decreased cell widths and increased cell division in the transverse direction of epidermal cells from the inner and outer glumes.Broadly similar phenotypic changes were also observed among the OE-SiSGLand OE-OsSGL2transgenic plants,which is consistent with our previously reported results (Wang et al,2016;Zhang et al,2018).
To further uncover the cytological basis underlying the regulation of grain size byOsSGL,the phenotypes ofOsSGLmutant (Cas-OsSGL) seeds were investigated.In our previous studies,there were no obvious phenotypic differences between the wild type andOsSGL-RNAi transgenic plants during growth and development,and the average grain sizes and lengths of the transgenic lines with downregulatedOsSGLexpression were comparable to the wild type plants (Wang et al,2016).Consistent with previous studies,the seeds of knockout lineossgldisplayed no significant difference compared to the wild type seeds (Fig.S4).As expected,the cell size showed no obvious difference between theossgland wild type spikelet hulls based on the SEM results (Fig.S4).
The upregulation ofOsSGLorthologs resulted in a slight increase in the grain-filling rate.There were no significant differences observed in either endosperm fresh weight (FW) or dry weight (DW) between the transgenic and wild type plants at 6 d after flowering (DAF).Starting at 6 DAF,both FW and DW of transgenic lines increased significantly faster and at 12 DAF,FW of transgenic plants were slightly heavier than that of wild type plants.The endosperm FW and DW of transgenic plants were heavier than those of wild type plants starting at 24 DAF and reached their maximums at 36 DAF,which was consistent with their longer ovaries,slenderer grains and increased rice endosperm weights (Fig.S5).The RiceXPro expression data onOsSGL(https://ricexpro.dna.affrc.go.jp/field-development.php?featurenum=16985) showed that its expression is especially high in the reproductive organs (inflorescence,anther,pistill,lemma and palea) during the development stage,as well as in the embryo and endosperm at the ripening stage.Thus,OsSGLorthologs may also positively impact dry matter accumulation during grain filling and may improve higher in the transgenic plants (144.9-187.3 and 10.96-13.20 mg/g,respectively) than in the wild type plants (86.4 and 7.44 mg/g,respectively).Moreover,the MDA contents of the transgenic plants (13.4-16.7 nmol/g) were significantly lower than those of the wild type plants (24.6 nmol/g) (Fig.S6).
To further assess drought tolerance at the vegetative stage,we withheld irrigation for 12 d before returning to a normal watering regimen.After this simulated drought treatment,the leaves of wild type plants were curled up and wilted,while the transgenic lines displayed less changes.At 5 d after resuming normal irrigation,the transgenic plants recovered more quickly and grew faster than the wild type plants (Fig.3-A).After three weeks of recovery,only 19.6% of the wild type plants still displayed green tissues,while up to 58.2% of the OE-OsSGL2plant leaves and 67.1% of the OE-ZmSGLplant leaves had survived (Fig.3-B to -F).The over-expression plants forOsSGLand itsPoaceaeorthologs conferred significantly improved drought tolerances in rice.rice endosperm growth,thereby regulating grain weight.
Consistented with the results obtained from theOsSGLheterologous-or over-expressing transgenic rice andArabidopsisplant experiments (Cui et al,2016),similar osmotic-tolerance phenotypes were observed in otherOsSGLorthologs over-or heterologous-transgenic rice lines.In comparison to the wild type plants,the average shoot length of transgenic plants displayed no distinguishable differences under normal conditions.However,the transgenic lines exhibited longer shoots under osmotic stress at the post-germination stage.The average root length in transgenic lines was longer than that of the wild type plants under both normal and osmotic conditions (Fig.S6).We investigated three drought stress-relevant parameters,namely proline,soluble sugar and malondialdehyde (MDA) at the seedling stage.Under normal conditions,no content differences were observed between the wild type and transgenic plants.However,under osmotic stress,the proline and soluble sugar contents were significantly elevated in all genotypes,though the increases were significantly
To further explore the possible roles of four DUF1645 contained genes in rice,we performed RNA-sequencing and qRT-PCR with the young inflorescence buds (2-4 cm) of four transgenic and wild type plants at the early booting stage (Fig.4-A to -D).We observed a stack of non-additively expressed genes (20 916-23 232) that differed significantly from the wild type plants,including more than 3 000 differentially expressed genes (DEGs) from 25 372 expressed genes shared between the two lines based on the RNA-sequencing results (Table S5).Gene Ontology (GO) analysis showed that the genes affected by the over-or heterologous-expression of the four genes were involved in cellular macromolecule biosynthetic processes,gene expression,or cell division and elongation (Fig.4-E).Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of the non-additively expressed genes revealed that these genes were involved in a variety of biological pathways.Most of the enriched pathways were related to the ribosome,RNA transport,RNA polymerase,or DNA replication (Fig.4-F).Interestingly,as shown in volcano plots,the upregulated (red) and downregulated (blue) genes were distributed in almost the same areas (Fig.4-A).Furthermore,Venn diagram analysis showed that only 8.3% of the DEGs were specific to a construct and that most of the DEGs overlapped (73.8%) (Fig.4-B and -C).Thus,the expression of the four transgeneOsSGLorthologs in rice had a similar effect on the transcriptome composition,suggesting that these genes have similar functions.
To confirm the RNA-sequencing data and obtain further insight into the function ofOsSGLorthologs,we selected 63 key DEGs that participate in hormone response,biosynthesis and signaling pathways for qRT-PCR expression analyses from four over-expression transgenic and wild type plants (Table S6 and Fig.S7).Compared with the wild type plants,the transcript levels ofOsPIN1andOsPIN2,two auxin transport genes,were obviously decreased while the transcript levels ofYUCCA4,YUCCA6andYUCCA9,three auxin synthesis genes,were greatly increased in the young panicles of the over-expression lines.The expression levels of five cytokinin (CK) signaling pathway-related genes,OsCKX4,RR1,RR5,RR7andRR10,were considerably different in the transgenic plants.In summary,the reliability of our RNA-sequencing data of the transcriptome was validated with qRT-PCR.The altered transcript levels of the analyzed genes in the over-or heterologous-expression lines along with our previous data suggested that the fourOsSGLorthologs may play a conserved dual role in regulating stress response and meristematic activity through a plant hormone-mediated pathway.
Various abiotic and biotic stresses can result in 30%-60% crop yield loss globally every year (Seo et al,2011).Ensuring a constant future food supply necessitates the development of crop varieties that can respond better to stress or that have improved yields.Therefore,understanding the molecular mechanisms that regulate grain size and stress resistance can enable the development of new strategies to improve the yield of cereal crops.In this study,the over-or heterologous-expression of four DUF1645-containingOsSGLorthologs positively increased cell numbers and sizes by promoting mitotic division and cell expansion in the inflorescence meristem,which ultimately resulted in an increase in grain length and yield.Moreover,it also significantly enhanced drought tolerance of the rice plant,which might due to the promotion of more extensive root systems,an increased accumulation of osmolytes,and altered transcript levels of stress-responsive genes.This study provided an important basis for the functional analysis of DUF1645-containingOsSGLortholog genes in improving rice stress-tolerance and increasing grain length as a potential means to improve crop yields.
Phytohormones such as auxins and cytokinins regulate numerous biological processes in plants including cell proliferation and expansion,reproductive development,grain size,grain yield,abiotic-stress responses,root development and growth,and root architecture via a complex signaling network (Liu et al,2017;Korver et al,2018).For example,Big Grain1(BG1) encodes a novel membrane-localized protein involved in auxin transport that can increase cell proliferation and elongation of spikelet hulls (Liu et al,2015).TGW6negatively regulates the endosperm growth or grain filling by modulating the endogenous auxin levels in rice (Ishimaru et al,2013).Moreover,auxin-cytokinin interaction plays a key role in root development,which is important since having appropriate root architecture is essential for plants to maintain a steady supply of nutrients and water (Ishimaru et al,2013;Korver et al,2018).Our data showed that some auxins and cytokinin genes related to cell response/ biosynthesis/signaling/interaction are either activated or inhibited in rice over-expression lines.Based on these current results and previous data (Cui et al,2016;Wang et al,2016;Zhang et al,2018),we speculated that DUF1645-containingOsSGLorthologs share the same functions in thePoaceaefamily and may either act as positive modulators upstream of auxin and/or cytokinin signaling or may indirectly affect the plant hormone pathway.Future studies should be performed to investigate the possibility that DUF1645-containing proteins may interact with similar protein or gene targets that affect inflorescence meristem grain elongation or that affect abiotic-stress tolerance of root systems by regulating cell proliferation and cell size.
In conclusion,this study provided fundamental information concerning grain length,yield,and abiotic-stress responses ofPoaceaefamily plants,and enhanced our understanding of the molecular mechanisms used by DUF1645 inPoaceaeplants.Our findings indicated thatOsSGLPoaceaeorthologs have a highly conserved function that regulates both drought stress-tolerance and yield of rice plants through plant hormone responses,biosynthesis,signaling and crosstalk pathways.We hypothesized that this highly conserved DUF1645 family in cereal crops has not been subjected to natural or artificial selection during domestication or breeding.These findings bridged the gap in the understanding of the plant hormone crosstalk between environmental stressors and genetic interactions that influence grain yield.
The rice varieties included KH2 (O.sativaL.ssp.indica),lines over-expressingOsSGL(OE-OsSGL) andOsSGL2(OE-OsSGL2),lines heterologously-expressingSbSGL(OE-SbSGL),ZmSGL(OE-ZmSGL) andSiSGL(OE-SiSGL),and a line whereOsSGLwas knocked out (Cas9-OsSGL).The seeds of KH2,Nipponbare (O.sativaL.ssp.japanica),sorghum Xingxiangliang 2 (S.bicolor),maize Zhongnuo 1 (Z.mays),millet Yugu 1 (S.italica),OsSGLmutant lines,and transgenic plants were provided by the Institute of Subtropical Agriculture,Chinese Academy of Sciences (CAS) in Changsha,Hunan Province,China.Seeds were surface sterilized with 70% ethanol for 1 min and 2% sodium hypochlorite for 15 min,washed 3 times with sterile distilled water,and soaked in water for 3 d at 25 °C (with daily water changes).Field-grown plants were cultivated under normal field conditions during the standard season in the experimental field of the Institute of Subtropical Agriculture,CAS,Changsha,China.Plants were cultivated with one plant per hill and spaces of 20 cm × 20 cm.Field management,including irrigation,fertilizer application,and pest control,followed the normal agricultural practices.For rice root measurements,sterilized seeds were germinated on 1/2 Murashige and Skoog (MS) medium for 7 d at 28 °C under a 12 h light/12 h darkness photoperiod.
To identify genes homologous toOsSGLinPoaceae,we first performed a BLAST search with theOsSGLsequence using the NCBI database (http://www.ncib.nlm.nih.gov/BLAST/).Next,the conserved DUF1645 domain of OsSGL-homologous proteins fromPoaceaewere inputted and analyzed using the NCBI Batch Web CD-Search Tool (https://www.ncbi.nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi).The cDNA fragments containing the whole ORFs and entire DUF1645 domains ofOsSGL,SbSGL,ZmSGL,SiSGL,andOsSGL2were amplified from total RNA isolations with specific forward and reverse primers (Table S1),using an RT-PCR kit (Promega,Madison,USA) according to the manufacturer’s instructions.The details of all 24 OsSGLPoaceaehomologous proteins are listed in Table S2.The nucleotide sequence diversity ofOsSGL(LOC_Os02g04130)was analyzed with RiceVarMap v2.0 (Table S3).The nucleotide and amino acid sequences ofOsSGLwere aligned with their homologs of other cereal species with CLUSTALW2 (http:// www.ebi.ac.uk/tools/clustalw2).The phylogenetic tree was constructed by multiple sequence alignments via MEGA6 using the Neighbor-Joining method with 1 000 replicates bootstrap analysis.
After sequence verification,the various amplified DNA fragments were inserted into the multi-cloning sites of the binary expression vector pCAMBIA1300-pJITl63.The resulting constructs overexpressed pCaMV35S::X::NOS(XrepresentedOsSGL,SbSGL,ZmSGL,SiSGLorOsSGL2) and containedhptII (hygromycin resistance gene) as a selection marker.They were introduced intoAgrobacterium tumefaciensstrain EHA105 by electroporation.All the final constructs were sequenced to ensure the correctness of the introduced segments before being used to transform immature embryogenic calli induced from rice KH2.Hygromycin resistance was used to select for positive transgenic plants.In total,we obtained 67 independent transgenic plants forOsSGL,29 forSbSGL,24 forZmSGL,14 forSiSGL,and 21 forOsSGL2.Using primers specific for the hygromycin phosphotransferase gene,we performed PCR to confirm the presence of T-DNA in the transformants.The expression levels ofOsSGL,SbSGL,ZmSGL,SiSGL,andOsSGL2in corresponding T0positive transgenic plants were assessed by analyzing the middle part of the 7th young leaf from different independent transgenic lines with qRT-PCR whereOsActinwas used as an internal control.The transgene transcripts were abundant in all transgenic lines but displayed large differences among the lines.Homozygous T3plants were used for subsequent experiments.The Cas9-OsSGLplants were obtained from KH2 using the CRISPR-Cas9 technology and characterized two independent mutant alleles (Fig.S8).The T2generation of Cas9-OsSGLplants were used.
Total RNA was extracted with TRIzol reagent (Invitrogen,Burlington,Canada) according to the manufacturer’s instructions.DNAase-treated RNA (1 µg) was reverse transcribed with a PrimeScriptTM1st Strand cDNA synthesis kit (TaKaRa,Japan) according to the manufacturer’s protocol.qRT-PCR analysis was conducted with a Fast Start Universal SYBR Green Master Mix (Roche,Swiss) and reactions were performed in an ABI 7900HT thermocycler (Applied Biosystems,USA) at 95 °C for 10 min,followed by 40 cycles of 95 °C for 15 s and 58 °C for 30 s.The rice housekeeping geneOsActinwas used as an internal control.Individual threshold cycles for each of the target genes and controls were optimized manually.Real-time PCR efficiencies were calculated from the standard curve slopes from each gene.Samples from the unstressed group were used for calibration.After normalization to theOsActintranscript levels,the relative transcript levels were averaged from three independent replicates of each sample.Relative amounts of mRNA were calculated using the comparative threshold cycle method (Maier et al,2016).
Harvested rice grains were air-dried and stored at room temperature for at least two months before testing.Fully filled seeds (including the hulls) were used to measure grain length,width,and weight.Sixty full seeds were randomly selected from each cultivar/line and divided into three equal groups.All seeds were aligned lengthwise along a Vernier caliper to measure seed length and then arranged crosswise to measure seed width.Length and width measurements were calculated from the average of three measurements.Seed thickness was measured with a Vernier caliper.More than 600 seeds from each plant were used to determine the 1000-grain weight.To measure the panicles,three medium-sized main panicles were obtained from each transgenic and corresponding control plants.We measured the panicle length and counted the number of panicles and the number of grains per panicle.To measure the flag leaves,sixty healthy plants were randomly selected from each cultivar/line and divided into three groups equally.The flag leaf length (FLL,cm) and flag leaf width (FLW,cm) were measured,and the flag leaf area (FLA,cm2) was calculated using the formula:FLA=FLL×FLW× 0.75.Duncan or Dunnett tests were performed to compare the means of all the measurements from different lines or cultivars/lines using SPSS 19.0 (SPSS Inc,IBM Company).
Young spikelet hulls were fixed in FAA (50% ethanol,5% glacial acetic acid,and 5% formaldehyde) for 48 h and sent to Servicebio Company (http://www.servicebio.cn/) for paraffin sectioning.The sections were observed under a microscope (Leica DMR,Germany) and scanned by a Pannoramic MIDI digital slide scanner (3D HISTECH,Hungary).Area measurements of vascular elements were taken using both the Pannoramic Scanner and Caseviewer (C.V 2.3) software.For glume cell observation,samples were fixed in 2.5% glutaraldehyde (40.5 mL of 0.2 mol/L Na2HPO4,9.5 mL of 0.2 mol/L NaH2PO4,10 mL of 25% glutaraldehyde and 40 mL H2O) for 48 h,cut at the longitudinal middle of the spikelet hulls,covered with gold nanoparticles by vacuum sputtering,and observed with SEM (HITACHI,S-3000N,Japan) at an accelerating voltage of 10 kV.
To determine drought tolerance in transgenic rice,T3homozygous seeds were used.To test osmotic stress at the post-germination stage,sterilized seeds were sown on 1/2 MS medium for 5 d in a growth room while healthy germinated seeds were transferred to 1/2 MS medium containing 0 or 400 mmol/L mannitol for 6 d in a growth room.Three replicates were performed.Lengths of shoot and primary,adventitious,and lateral roots were measured at the end of the treatments.To measure the adventitious root,the five longest adventitious roots on each seedling were counted.Similarly,lateral root length was measured by taking the 15 longest lateral roots on each primary root.
Drought assays were performed in a controlled growth chamber PGC15.5 (Percival,Perry,USA).Germinated seeds were transferred to an incubator with a photoperiod of 12 h light (30 °C)/12 h dark (25 °C) for 5 d.After 5 d,20 seedlings from each transgenic line were planted in three rows (one plot) alongside the wild type plants at a randomized complete block design with three replicates.All plants were grown in polyvinyl chloride pots (diameter 30 cm and height 45 cm) under natural conditions prior to stress treatment.At the 4-leaf-seedling stage,watering was withheld from all plants for about 14 d until 90% of the leaves coiled.After one week recovery with normal watering,survival rates were determined.
After 6 d of osmotic treatment,shoots of wild type and transgenic plants were prepared for biochemical analysis.Proline and soluble sugar levels of harvested tissue samples were measured according to the sulphosalicylic acid (Troll and Lindsley,1955) and the anthrone (Morris,1948) methods,respectively.The levels of MDA were determined with thiobarbituric acid (Kramer et al,1991).
Total RNA was isolated with TRIzol reagent from 2-4 cm long samples of young panicles of over-or heterologous-expressing transgenic and wild type KH2 plants at the early booting stage as described earlier.Materials from 10 plants of each genotype were pooled for RNA extraction.RNA quantification,qualification,library preparation for strand-specific transcriptome sequencing,clustering and sequencing,data analysis and quality control were conducted by Novogene (http://www.novogene.com/) according to their protocols.Differential expression analysis of the genes in the mutant,transgenic,and wild type samples were performed using the DESeq R package (version 1.18.0).The resultingP-values were adjusted using the Benjamini and Hochberg method for controlling false discovery rate.Genes with an adjustedP-value < 0.05 found by DESeq were assigned as DEGs.The common DEGs among four transgenic plants were characterized by Venn diagram analysis and were used to carry out GO and KEGG analyses.
ACKNOWLEDGEMENTS
This study was supported by the National Natural Science Foundation of China (Grant Nos.31501393,31671671 and 31671612) and the Open Research Fund of State Key Laboratory of Hybrid Rice (Wuhan University,China) (Grant No.KF201803).We thank the Public Service Technology Center,Institute of Subtropical Agriculture,Chinese Academy of Science for technical support.
SUPPLEMENTAL DATA
The following materials are available in the online version of this article at http://www.sciencedirect.com/journal/rice-science;http://www.ricescience.org.
Fig.S1.Sequence homology and conserved DUF1645 domains inOsSGLPoaceaeorthologs.
Fig.S2.Expression pattern ofOsSGLandOsSGL2in rice.
Fig.S3.Expression levels ofOsSGLorthologs in over-expression transgenic rice.
Fig.S4.OsSGLregulatesgrain length.
Fig.S5.OsSGLorthologs regulate endosperm size and grain milk filling in KH2 and transgenic lines at indicated days after fertilization (DAF).
Fig.S6.Osmotic treatment in transgenic and wild-type rice at seedling stage.
Fig.S7.Validation of the RNA-sequencing data by qRT-PCR.
Fig.S8.Characterization of theOsSGLmutant created via CRISPR-Cas9 system.
Table S1.Primers used for cloning ofOsSGLPoaceaeorthologs.
Table S2.OsSGLPoaceae orthologs and their corresponding proteins in the DUF1645 super-family.
Table S3.Sequence and haplotype analysis of theOsSGLlocus in cultivated rice varieties and wild rice accessions.
Table S4.Phenotypical measurements of wild type rice KH2 and corresponding over-or heterologous-expression transgenic rice plants.
Table S5.Up-and down-regulated differentially expressed genes (DEGs) in any two lines.
Table S6.Primer pairs used for qRT-PCR analysis.