APP下载

Characterization and map-based cloning of miniature2-m1,a gene controlling kernel size in maize

2020-10-02GUANHaiyingDONGYongbinLUShoupingLIUTieshanHEChunmeiLIUChunxiaoLIUQiangDONGRuiWANGJuanLIYulingQIShijunWANGLiming

Journal of Integrative Agriculture 2020年8期

GUAN Hai-ying,DONG Yong-bin,LU Shou-ping,LIU Tie-shan,HE Chun-mei,LIU Chun-xiao,LIU Qiang,DONG Rui,WANG Juan,LI Yu-ling,QI Shi-jun,WANG Li-ming

1 Maize Research Institute,Shandong Academy of Agricultural Sciences/National Engineering Laboratory of Wheat and Maize/Key Laboratory of Biology and Genetic Improvement of Maize in Northern Yellow-Huai River Plain,Ministry of Agriculture,Jinan 250100,P.R.China

2 College of Agronomy,Henan Agricultural University/Collaborative Innovation Center of Henan Grain Crops/National Key Laboratory of Wheat and Maize Crop Science,Zhengzhou 450046,P.R.China

Abstract Kernel development plays an important role in determining kernel size in maize.Here we present the cloning and characterization of a maize gene,nitrate transporter1.5 (NRT1.5),which controls small kernel phenotype by playing an important role in kernel development.A novel recessive small kernel mutant miniature2-m1 (mn2-m1) was isolated from self-pollinated progenies of breeding materials.The mutant spontaneously showed small kernel character arresting both embryo and endosperm development at an early stage after pollination.Utilizing 21 polymorphic SSR markers,the mn2-m1 locus was limited to a 209.9-kb interval using 9 176 recessive individuals of a BC1 segregating population from mn2-m1/B73.Only one annotated gene was located in this 209.9 kb region,Zm00001d019294,which was predicted to encode nitrate transporter1.5 (NRT1.5).Allelism tests confirmed that mn2-m1 was allelic to miniature2-m2 (mn2-m2) and miniature2-710B(mn2-710B).The mn2-m1 and mn2-m2 alleles both had nucleotide deletions in the coding region resulting in premature termination,and the mn2-710B allele had some missence mutations.Subcellular localization showed that Miniature 2 (MN2)is localized in the plasma membrane.Quantitative real-time PCR (qRT-PCR) analysis revealed that the expression of MN2 and some genes involved in the basal endosperm transfer layer (BETL) and embryo surrounding region (ESR) development were affected in mn2-m1 seeds.These results suggested that MN2 plays an important role in maize seed development.

Keywords:miniature2-m1,nitrate transporter1.5,frame shift mutation,allelism tests,subcellular localization

1.Introduction

Maize is cultivated worldwide and its grains are one of the most important renewable resources for staple food,animal feed,and industrial raw materials.Maize grain yield is a complex quantitative genetic trait,and can be divided into smaller components (Austin and Lee 1996).Kernel size(including kernel length,width and thickness) is one of the most important factors affecting grain yield,and appears to be highly heritable (Doebleyet al.2006;Guptaet al.2006;Vega-Rojaset al.2016;Zhang Xet al.2016;Zhang Z Het al.2016).In addition,kernel size is related to maize end-use quality (Revillaet al.1999;Guptaet al.2006).Thus,understanding the genetic basis of kernel size is extremely important for the breeding of high-yield and high quality of maize.

The maize grain consists of three main tissues:embryo,endosperm and pericarp,and the endosperm comprises of four major regions:the embryo surrounding region (ESR),the basal endosperm transfer layer (BETL),the starchy endosperm (SE),and the aleurone layer (AL) (Sabelli and Larkins 2009;Dollet al.2017).There are three types of maize kernel mutants according to the defective regions,which affects embryo,endosperm and both embryo and endosperm development,respectively.

The characterization of numerous genes has yielded major advancements in understanding genetic mechanisms of grain size in model plants in recent years,such asArabidopsisand rice (Liet al.2011;Xiaet al.2013;Eckertet al.2014;Huanget al.2017).In maize,several genes that regulate grain size/shape also have been characterized,includingSh1andSu1that affect starch and protein accumulation in endosperm (Jameset al.1995;Guanet al.2017);PPR8522andEMB12that promote embryogenesis(Sossoet al.2012;Shenet al.2013);BSL1,Glol-1,Mn1,MPPR6,andRGF1that maintain the structure of the basal endosperm transfer layer (BETL) (Chenget al.1996;Maitzet al.2000;Costaet al.2003;Gutiérrez-Marcoset al.2006;Kanget al.2009;Manavskiet al.2012);Dek1,Sal1andThk1that regulate the development of aleurone cell layers(Lidet al.2002;Shenet al.2003;Yiet al.2011);EMP4,EMP5andEMP11that regulate kernel development by affecting endosperm development (Gutiérrez-Marcoset al.2007;Liuet al.2013;Renet al.2017);andDek10,Dek35,Dek36,Dek37,Dek44,Rgh3,andUBL1that regulate kernel development by affecting both endosperm and embryo development (Fouquetet al.2011;Chenet al.2017;Liet al.2017;Qiet al.2017,2019;Wanget al.2017;Daiet al.2018).

Kernel mutants affecting both endosperm and embryo developmentare mostly caused by mutations in pentatricopeptide repeat (PPR) proteins and non-PPR proteins.Dek10,Dek35,Dek36,andDek37encode mitochondria-targeted PPR proteins (Chenet al.2017;Qiet al.2017;Wanget al.2017;Daiet al.2018).Dek10is responsible for multiple editing sites innad3andcox2,and is essential for mitochondrial functions,embryo,endosperm,and seedling development (Qiet al.2017).Thedek10mutation reduces the function of complex I and complex IV in electron transport chain (ETC) and has significantly delayed seed development.Dek35affects thecis-splicing of mitochondrialnad4intron 1 and is required for mitochondrial function and seed development (Chenet al.2017).Thedek35mutation reduces the accumulation of DEK35 proteins and the splicing efficiency of mitochondrialnad4intron 1,leading to a lethal-seed phenotype.Dek36is required for RNA editing ofatp4-59,nad7-383 andccmFN-302 and is essential for embryo and endosperm development (Wanget al.2017).Thedek36mutation reduces the activities of mitochondrial complex I,complex III and complex IV,causing small and collapsed kernels,leading to embryos and/or seedlings lethality.Dek37is involved incis-splicing of mitochondrialnad2intron 1 and is required for complex I assembly,mitochondrial function and seed development in maize (Daiet al.2018).Thedek37mutation causes no detectable DEK37 protein in mutant seeds,decreases splicing efficiency of mitochondrialnad2intron 1,leading to reduced assembly and NADH dehydrogenase activity of complex I,and causes delayed endosperm and embryo development.On the other hand,Dek44,Rgh3andUBL1encode non-PPR proteins (Fouquetet al.2011;Liet al.2017;Qiet al.2019).Dek44encodes a mitochondrial ribosomal protein L9 which only accumulates in kernels(Qiet al.2019).The loss of function ofDek44affects the expression of mitochondrial respiration chain subunit genes,reduces the function of complexes in the ETC,inhibits cell proliferation,arrests mitochondrial oxidative phosphorylation,and further embryo and endosperm development.Rgh3encodes the maize U2AF35related protein (URP),an RNA splicing factor involved in both U2 and U12 splicing,which is required in the endosperm for embryos to develop and has an autonomous role in embryo and seedling development (Fouquetet al.2011).Thergh3mutation alters a fraction of noncanonical splicing events,causes defective endosperm cell differentiation,leading to a rough endosperm phenotype.UBL1encodes a putative RNA exonuclease belonging to the 2H phosphodiesterase superfamily in mazie,which plays an important role in kernel and seedling development by influencing pre-mRNA splicing.The loss of function ofUBL1decreases the level and constitution of small nuclear RNA (snRNA) U6,leading to an mRNA splicing defect.

Nitrate is one of the most abundant sources in the soil of nitrogen,an essential element affecting plant growth and development (Masselet al.2016;O’Brienet al.2016).Nitrate transporters are essential components responsible for uptake,source to sink distribution and signaling of nitrate(O’Brienet al.2016;Guan 2017).Two families of nitrate transporters,NRT1andNRT2,have been identified in higher plants (Crawford 1995;Daniel-Vedeleet al.1998;Stitt 1999;Tsayet al.2007).InArabidopsis,onlyAtNRT1.1toAtNRT1.8ofNRT1familyhad been better characterized (Liet al.2007;Plettet al.2010).Functional analyses indicate that these proteins do indeed transport NO3–.NRT1.1andNRT1.5are bidirectional transporters involved in root-toshoot nitrate translocation (Linet al.2008;Chenet al.2012;Léranset al.2013).None of the above genes regulating kernel size/shape especially these regulating kernel development by affecting both endosperm and embryo development in maize encodes nitrate transporter.In this study,a gene controlling small kernel phenotype,mn2-m1,predicted to encode a nitrate transporter NRT1.5 ofNRT1family,was cloned.Further physiological and molecular analysis revealed thatMN2plays an essential role in maize kernel development.

2.Materials and methods

2.1.Parent lines and populations

Mutantsmn2-m1andmn2-m2were isolated from two individual ears of breeding materials that were eliminated in 2011 (Zhengzhou,Henan Province) and 2016 (Jinan,Shandong Province),respectively.These two small kernel mutants were both preserved by selfing the small kernel seeds from the initial segregating ears.mn2-710B(Stock 710B) was requested from the Maize Genetics COOP Stock Center (Urbana,Illinois,USA),and the stock was in a polymorphic background.The small kernel ofmn2-m1was self-pollinated for four generations and used for crossing with maize inbred lines B73,Chang7-2,Z58,and Qi319 to generate F1plants,respectively.F2and BC1segregating populations were obtained from F1self-pollination or crosses ofmn2-m1with the F1,respectively.Recessivemn2-m1individuals of the B73/mn2-m1BC1segregating population were used for initial and fine mapping.All plants were grown in the light incubator or greenhouse,Hainan (Huangliu,Hainan) or Shandong (Zhangqiu,Shandong) experimental stations.

2.2.Light microscopy of cytological sections

For scanning electron microscopy analysis,the separated endosperms from mature seeds ofmn2-m1and B73/mn2-m1from the same plants ofmn2-m1crossed with B73/mn2-m1were fixed in a solution of 2.5% (v/v)glutaraldehyde (pH 7.2).Samples were dehydrated through an ethanol series and critical point dried in liquid CO2.Then,they were mounted on specimen stubs using double-sided tape and coated with gold particles in an SCD040 sputter coater (Balzers AG,Balzers,Liechtenstein).Finally,the samples were examined using an S-3400N (Hitachi,Tokyo,Japan).

2.3.Allelism tests

For allelism tests betweenmn2-m1andmn2-m2,the pollen frommn2-m2plant crossed with five B73/mn2-m1plants,respectively.At the same time,themn2-m2plant was selfpollinated.The number of normal and small kernels from these five cross-pollinated ears was observed,respectively.For allelism tests betweenmn2-m1andmn2-710B,seven plants from normal kernels of the stock 710B were selfpollinated,at the same time their pollen crossed with sevenmn2-m1plants.The number of normal and small kernels from these self-pollinated and the corresponding crosspollinated ears was observed,respectively.

2.4.DNA extraction,marker development and gene prediction

Genomic DNA was extracted from leaves of 10-day-old plants by the CTAB method (Saghai-Maroofet al.1984).At the same time,20 normal andmn2-m1plants each were randomly selected from the BC1population ofmn2-m1/B73 to construct DNA pools (a wild type (WT) pool and a mutant pool).

For genetic mapping of theMN2locus,bulked segregant analysis (BSA) and simple sequence repeat (SSR)molecular markers were used.A total of 188 SSR markers distributed evenly on 10 chromosomes were selected from the Maize Genetics and Genomics Database (http://www.maizegdb.org) for screening polymorphisms both between the two parents (B73 andmn2-m1) and the two DNA pools(Appendix A).The polymorphic markers were used to genotype 218 BC1recessive individuals ofmn2-m1/B73,and to construct the initial genetic map.To enrich the primary map of the target locus,an additional 19 SSR markers between P1 and P2 were selected from published studies(Xuet al.2013),screened for polymorphism (Appendix B),and the polymorphic markers were used to genotype the recombinants identified by P1 and P2 in the 218 BC1recessive individuals.To fine map the target region,4 347 recessive plants of BC1segregating populations were screened to obtain more recombinant individuals by P5 and P6 and another 62 nearby SSR markers flanked by P5 and P6 were selected from published studies (Xuet al.2013),detected for polymorphism,and they were used to genotype the crossovers identified by P5 and P6 in the 4 565 (218+4 347) BC1recessive individuals (Appendix B).An additional 4 611 recessive plants from BC1segregating populations were screened by P15 and P17 and new SSR markers were developed by screening nearby sequences(between P15 and P17) of the B73 genome (http://ensembl.gramene.org/Zea_mays/Info/Index),using SSRHunter1.3 Software (Li and Wan 2005) and BLASTing against a genomic sequence database of high throughput genomic sequences (HTGS) to pick unique sequences,and primers flanking microsatellite repeats were designed by Primer 5.0 (Appendix C).The newly developed SSR markers were screened for polymorphism and they were used to genotype the total recombinants indentified by P15 and P16 in the 9 176 (218+4 347+4 611) BC1recessive individuals.SSR markers used for map-based cloning ofmn2-m1were listed in Appendix D.

Gene prediction and annotation within the target region flanked by P19 and P20 was conducted by Softberry (http://linux1.softberry.com) and according to the Maize Genetics and Genomics Database.

2.5.Gene cloning,sequencing analysis,multiple sequence alignment,and phylogenetic analysis

The DNA/cDNA sequence of candidate geneZm00001d019294was amplified from mutantsmn2-m1,mn2-m2andmn2-710Band their WT parent B73,and recombinants using primers G1-2F-G1-3R (F:5´-CACTACCGCCCTAGCAAA-3´;R:5´-TGCAAAAG TCAAAAACATACAAC-3´) and GR1-6F-GR1-6R1 (F:5´-ATGGCTGAGGGTAGCTG-3´;R:5´-GCCTTAGAAT GGATTGGTAT-3´),respectively.The target DNA/cDNA fragments were cut and recovered from the 1% agarose gel and purified by the DNA Gel Extraction Kit (TransGen,Beijing),then ligated to the pEASY-Blunt cloning vector and sequenced (Thermo Fisher Scientific,Beijing).The mutational sites of the candidate gene were ascertained by comparing and analyzing the sequencing results using DNAStar and BioXM2.6 Software (http://www.bio-soft.net/format/bioxm.htm).

The full-length protein sequences ofMN2(Zm00001d019294) inmn2-m1,mn2-m2,mn2-710B,and WT line B73 were inferred by translation of cDNA to protein sequences using BioXM 2.6 Software.Multiple cDNA/amino acid sequence alignment was conducted using DNAMAN version 5.2.2 Software.

For phylogenetic analysis,the full-length protein sequence of MN2 was obtained from GenBank (http://www.ncbi.nlm.nih.gov),and used as a query to search against the nonredundant protein sequences (nr) database by BLASTP,and its homologs from other species (SbNRT1.5X1,SbNRT1.5X2,SiNRT1.5,and AtNRT1.5) were obtained.Additionally,some of its homologs (ZmNRT1.5A (Chr.5),ZmNRT1.5B (Chr.4),SbNRT1.5A,SbNRT1.5B,OsNRT1.5A,and OsNRT1.5B) were obtained from previous study (Plettet al.2010).Multiple amino acid sequence alignment and phylogenetic analysis was conducted using MEGA7.0.14 with the ClustalW and the rooted neighbor-joining tree,respectively.

2.6.Subcellular localization of MN2

To generate a translational fusion between the MN2 and green fluorescent protein (GFP),the cDNA sequence of full-length MN2 protein was amplified by RT-PCR using primers GFP-MN2-1F (5´-CAGTCTAGACAT GGCTGAGGGTAGCTGGACC-3´) and GFP-MN2-1R(5´-CAGGGATCCTCAAATGCATTCCTCCGCATC-3´),and introduced into the binary vector pCAMBIA1301S by restriction enzymes (XbaI andBamHI) and ligases.The GFP-MN2 fusion protein was placed under the CaMV35S promoter for driving constitutive expression.The construct was then introduced into tobacco (Nicotiana benthamiana)leaf protoplasts by the polyethylene glycol (PEG) mediated transformation method.GFP green signals were detected using Zeiss LSM 510 META Confocal Microscopy(Oberkochen,Germany).CD3-mCherry (CD3-1007) was used as the plasma membrane marker (Nelsonet al.2007).

2.7.Materials preparation,RNA extraction and quantitative real-time PCR

For expression profile analysis,different tissues of root,leaf,stem,ear,tassel,and developing seeds were collected from the maize B73 inbred line.Root and leaf were collected from three 15-day-old seedlings,stem was collected from the 7th to the 8th joints of above the ground from 3 adult plants,ear was collected from 3 plants with 5-cm-long ear,tassel was collected from 3 plants at tasseling stage,and each 12 developing seeds were collected from 3 ears at 15 and 21 days after pollination (DAP),respectively.Each sample was collected with 3 repeats,respectively.Each 8 kernels of normal andmn2-m1from the same ear of BC1(mn2-m1/B73//mn2-m1mn2-m1) at 15 and 20 DAP were collected,respectively,with 3 repeats.These samples were ground to powder in liquid nitrogen with a mortar and a pestle,respectively.Total RNA was isolated using about 100 mg of the powder with an RNA Isolation Kit (Tiangen,Beijing).Approximately 4 μg of total RNA was used for firststrand cDNA generation using oligo (dT) and Prime Script II(TaKaRa,Dalian).Specific primers ofMN2and other genes used for quantitative real-time PCR (qRT-PCR) were designed using Primer 5.0 or in previous studies (Fouquetet al.2011;Garnettet al.2013;Liet al.2017;Wenet al.2017) (Appendix E).Theactingene (Zm00001d010159)was amplified as a control.qRT-PCR was performed on ABI 7500 Real-time PCR System (ABI,USA) using SYBR Green PCR Master Mix (TaKaRa).Relative expression levels of genes were calculated using the 2−ΔΔCTmethod (Livak and Schmittgen 2001).The experiment was performed with 3 biological and experimental replicates.

3.Results

3.1.Phenotypic and genetic characterization of mn2-m1

A spontaneous small kernel mutant,designatedmn2-m1,was isolated from an ear of breeding materials that were eliminated in Zhengzhou in 2011.All F1seeds of themn2-m1mutant crossed with WT inbred lines B73,Chang7-2,Z58,and Qi319 displayed normal phenotypes,and their F2populations all showed segregation ratios (normal:small)of 3:1 (χ23:1<3.84,χ20.05=3.84,Table 1;Appendix F).These results indicated that the small kernel phenotype in themn2-m1mutant was controlled by a single recessive nuclear gene.

mn2-m1kernels were clearly distinguishable from WT at 15 DAP,and became small at mature stages (Fig.1-A and B;Appendix G).The embryo length,width and thickness ofmn2-m1were also affected (Appendices G and H,P≤0.01,Student’st-test).Grain length,width,thickness,and weight(100 kernels) ofmn2-m1kernels were significantly affected in the B73,Chang7-2,Z58,and Qi319 genetic backgrounds except grain length in Chang7-2 (Fig.1-C and D;Appendix I,P≤0.01,Student’st-test).Cytological analysis showed that starch granules in WT maize endosperm were surrounded with many protein bodies and mostly were in long column or irregular in shape,non-uniform (Fig.1-E).mn2-m1starch granules were surrounded with fewer protein bodies,mostly spherical in shape and loosely packed (Fig.1-F).

3.2.Fine mapping of mn2-m1

Preliminary mapping ofmn2-m1was conducted using 218 BC1recessive individuals obtained from the cross ofmn2-m1/B73 withmn2-m1.We found only two SSR markers of these 188 markers,P1 and P2 on the short armof chromosome 7 to exhibit polymorphism between the two parents and the DNA pools (Appendices A and D).They were used to screen the 218 BC1recessive individuals.P1 and P2 were positioned on two sides ofmn2-m1,toward the telomeric end and centromeric region of chromosome 7,with 7 and 3 recombinants,respectively (Fig.2-A).Among 19 SSR markers between P1 and P2 obtained from previous studies (Xuet al.2013) (Appendix B),seven showed polymorphisms and were used to screen the 10 recombinant individuals detected by markers P1 and P2.Totals of 7,7,2,2,3,3,and 3 plants with crossovers were found with P3,P4,P5,P6,P7,P8,and P9,respectively (Fig.2-A;Appendices B and D).Thus,mn2-m1was limited to an interval of 11.78 Mb flanked by P5 and P6.

Table 1 Segregation of F2 progenies from four genetic backgrounds

Fig.1 Phenotypic characterization of wild type (WT) and mn2-m1 mutant kernels.A and B,phenotypic observation of WT and mn2-m1 mutant kernels (arrow) from F2 segregating population of B73 background at 15 days after pollination (DAP) and mature stage.C and D,the length and width of mature kernels which were randomly selected from F2 ears of B73×mn2-m1 (WT,up;mn2-m1 mutant,down).E and F,scanning electron microscopy analysis of starch granules in endosperm from mature kernels of WT and mn2-m1.SG,starch granule;PB,protein body.Bars in A,B,C and D,1 cm;bars in E and F,20 μm.

To fine mapmn2-m1,an additional 62 SSR markers flanked by P5 and P6 from previous studies were analyzed,and nine were polymorphic (Appendix B).Meanwhile,4 347 recessive plants of BC1segregating populations were screened,identifying 31 plants with crossovers,including eight and 23 detected by P5 and P6,respectively.All 35(31+2+2) recombinant individuals were genotyped using the 9 new polymorphic SSR markers P10,P11,P12,P13,P14,P15,P16,P17,and P18,respectively (Fig.2-B;Appendices B and D).Totals of 9,9,9,9,7,3,0,4 and 5 recombinants were identified,respectively.P10–P15 markers were on the side of P5,P17 and P18 were on the opposite side.Thus,mn2-m1was narrowed down to a genomic region of about 1.34 Mb flanked by markers P15 and P17(Fig.2-B).Subsequently,4 611 recessive plants from the BC1segregating population were screened by P15 and P17,identifying 5 and 12 recombinants,respectively.Additionally,the B73 genome sequence in the target region between P15 and P17 was used to develop new SSR markers.Among 135 SSR fragments generated using SSRHunter1.3,33 were identified to be single copy (Appendix C).A total of 11 pairs of SSR markers were developed based on their physical position and 3 were polymorphic (P19,P20 and P21) (Appendices C and D).All 24 (17+3+4) crossovers were screened by P16,P19,P20 and P21,2,2,2 and 2 recombinants were identified,respectively (Fig.2-C).Finally,mn2-m1was mapped to an interval of 209.9 kb flanked by markers P19 and P20 (Fig.2-C).

Fig.2 Map-based cloning of mn2-m1.The mn2-m1 gene was initially mapped to an interval of 11.78 Mb flanked by markers P5 and P6 on the short arm of chromosome 7 (A),more precisely mapped to an interval of 1.34 Mb flanked by markers P15 and P17(B),and finally mapped to an interval of 209.9 kb between markers P19 and P20 (C).D,schematic diagram of MN2 coding region.Black boxes and lines indicate exons and introns,and numbers above them indicate their length,respectively.Vertical lines indicate SNPs (red and green) or InDels (blue and orange) in the coding region that cause missense mutations compared to the wild type,respectively.The vertical lines with asterisk indicate two SNPs nearby.In mn2-m1,the red and blue vertical lines indicate nucleotide changes located in the exons that resulted in missense mutations in the encoded product,from left to right were 50 (6 bp of the first exon +44 bp at the first part of the second exon) (A/G)–17 (Asp/Gly),101 (A/T)–34 (His/Leu),111 (–––/AGT) 112-37 (–/Ser) 38,128 (C/G)–43 (Ala/Gly),131 (T/C)–44 (Met/Thr),187 (C/A)–63 (Arg/Ser),844 (A/C)–282 (IIe/Leu),937,938 (AC/CA)–313 (Thr/Gln),1 078 (G/C)–360 (Ala/Pro),1 258 (A/G)–420 (Thr/Ala),and 1 379 (A/G)–460 (Lys/Arg),respectively,and the last one 1 395 (G/–) (blue box) which resulted in a frameshift mutation and a premature stop codon.In mn2-m2,the red vertical lines were the same with those in mn2-m1,and the blue vertical line indicates a 46-bp (1 455–1 500 bp)deletion (blue box) that caused a frameshift mutation and a premature stop codon.In mn2-710B,the red and blue vertical lines were the same with those in mn2-m1,and the green orange vertical lines indicate missense mutations specific for mn2-710B,respectively.From left to right,the 6 specific missense mutations were 118 (G/A)–40 (Ala/Thr),433 (–––/TGC) 434–144 (–/Leu) 145,985 (G/A)–329 (Gly/Ser),1 072 (T/C)–358 (Phe/Leu),1 085 (––––––/CGGCGG),1 086–362 (––/AlaAla) 363,and 1 897 (G/A)–633 (Ala/Thr),respectively.The numbers of 50 (6 bp of the first exon +44 bp of the second exon) and so on indicate the positions of nucleotides in the open reading frame (ORFs) starting from the initial codon of ATG,respectively.ATG and TGA represent the initial and stop codons,respectively.

3.3.mn2-m1 was allelic to mn2-m2 and mn2-710B

mn2-m1,mn2-m2andmn2-710Bshared similar phenotypes of small kernels,additionallymn2-710Bwas also mapped to the short arm of chromosome 7 (http://www.maizegdb.org).To validate whether themn2-m1mutation is or not allelic tomn2-m2andmn2-710B,genetic allelism tests were conducted (Appendices J and K).The self-pollination ofmn2-m2ear showed small kernel,and all 5 F1ears from the crosses of B73/mn2-m1withmn2-m2mn2-m2showed segregation ratios (normal:small) of 1:1 (χ21:1<3.84,χ20.05=3.84;Appendices J and K),noting that whenmn2-m2was crossed to WT line B73,all F1had WT phenotypes (data not shown).The self-pollination ofmn2-m1ear showed small kernel (Appendix J),3 of 7 ears of +/mn2-710Bselfpollination showed segregation ratios (normal:small) of 3:1(χ23:1<3.84,χ20.05=3.84;Appendix J,data not completely shown),and the corresponding 3 F1ears ofmn2-m1mn2-m1crossed with +/mn2-710Bshowed segregation ratios(normal:small) of 1:1 (χ21:1<3.84,χ20.05=3.84;Appendices J and K).These results confirmed that mutationsmn2-m1,mn2-m2andmn2-710Boccurred at the same locus.

3.4.Candidate gene annotation,sequence analysis and subcellular localization

Gene prediction in the 209.9 kb region of the maize genome containingmn2-m1revealed only 1 gene,Zm00001d019294,predicted to encode nitrate transporter 1.5(NRT1.5) comprised of 640 amino acids harboring 2 major facilitator superfamily (MFS) conserved regions (Fig.2-C;Appendices L and M).We sequenced theZm00001d019294gene from amplified genomic DNA and cDNA of mutant linesmn2-m1,mn2-m2andmn2-710Band WT inbred line B73,respectively.A single nucleotide was deleted at position 1 395 bp (G/–) from the ATG start codon in the fifth exon ofmn2-m1,and a 46-bp deletion occurred from 1 455 to 1 500 bp in the same exon ofmn2-m2,and a 12 (3+3+6)-bp insertion occurred in the 2nd,3rd and 5th exon ofmn2-710B,respectively (Fig.2-D;Appendix N).Both of these deletions inmn2-m1andmn2-m2resulted in premature termination codon mutations.The resulting 1 431 bp truncated translational region encodes a predicted protein of 476 amino acids inmn2-m1,and a 1 506-bp truncated translational region encodes a predicted protein of 501 amino acids inmn2-m2compared to a 1 923-bp open reading frame (ORF) encoding predicted proteins of 640 amino acids in WT inbred line B73,respectively.The deletions inmn2-m1andmn2-m2both changed the conserved region of the 2nd MFS and some of it disappeared compared to the WT inbred line B73 (Appendix L).Of the 11 missense mutations inmn2-m1,only five before the deletions shared bymn2-m1,mn2-m2andmn2-710Bwere located in the second MFS conserved region,others being outside of the MFS conserved region (Fig.2-D;Appendix L).Of the 6 missense mutations specific formn2-710B,the 1st and last were outside of the MFS conserved region,the 2nd(one amino acid insertion) was located in the 1st MFS and other three (amino acids changed and insertions) being in the 2nd MFS.Sequencing of the key recombinants revealed that only the last missense mutation shared bymn2-m1,mn2-m2,andmn2-710Band that specific formn2-710B,and themn2-m1andmn2-m2deletions fell in the target region (Appendices L and O).Additionally,co-segregation analysis with CAPS1 marker developed based on the SNP(G/-) utilizing 492 normal and 484mn2-m1individuals of BC1population,respectively,revealed that there was no recombinant (Appendix P).

We investigated the possible phylogenetic relationships between MN2 and its homologs of ZmNRT1.5A and ZmNRT1.5B (Plettet al.2010) in maize and other homologs from other species.The results showed that MN2 and its homologs SbNRT1.5X1 and SbNRT1.5X2 fromSorghum bicolorand SiNRT1.5 fromSetaria italicaare divided clearly into 1 subgroup,the proteins ZmNRT1.5A and ZmNRT1.5B and other proteins SbNRT1.5A and SbNRT1.5B fromS.bicolor,OsNRT1.5A and OsNRT1.5B fromOryza sativaform another subgroup,while the protein AtNRT1.5 from dicotyledonous plantArabidopsis thalianaform a single subgroup (Fig.3).These data indicate that the MN2 is conserved in plants.

To examine the subcellular localization of MN2 protein,we generated a construct,p35S::GFP-MN2,which harbored the full-length MN2 coding sequence fused to the C terminus of GFP.This construct was then transiently expressed in the tobacco (N.tabacum) leaf protoplasts.Confocal laser scanning microscopy showed that the GFP signal produced from p35S::GFP-MN2 overlapped well with CD3-mCherry red,demonstrating that MN2 is targeted to the plasma membrane (Fig.4).

3.5.Expression profile analysis of MN2,some genes related to endosperm development and some ZmNRT genes

Fig.3 Phylogenetic anaysis of the MN2 protein and its related proteins.Accession numbers of the protein sequence were as follows:Zea mays (MN2,GenBank:ONM52643;ZmNRT1.5A,GenBank: AQK73708;ZmNRT1.5B,GenBank accession:AQK54750);Sorghum bicolor (SbNRT1.5X1,NCBI Reference Sequence:XP_002459620.2;SbNRT1.5X2,GenBank:KXG34784.1;SbNRT1.5A,GenBank:KXG30977.1;SbNRT1.5B,GenBank:EES05624.1);Setaria italica (SiNRT1.5,GenBank:RCV47059.1);Oryza sativa (OsNRT1.5A,MSU ID:LOC_Os02g46460.1;OsNRT1.5B,MSU ID:LOC_Os02g48570.1);Arabidopsis thaliana (AtNRT1.5,NCBI Reference Sequence:NP_174523.2).

The results of expression profiles revealed thatMN2was mainly expressed in developing seeds,ZmNRT1.5Awas expressed in all the tested tissues with the highest level in stem,andZmNRT1.5Bwas mainly expressed in ear and developing seeds with the highest level (Fig.5).To investigate whethermn2-m1mutations affected the transcript levels of its own mRNA and some marker genes relative to BETL and ESR development,such asMaternally Expressed Gene-1(MEG-1),Basal Endosperm Transfer Layer-2(BETL-2),Basal Endosperm Transfer Layer-10(BETL-10),Transfer Cell Response Regulator-1(TCRR-1),andEmbryo Surrounding Region-6(ESR-6) (Hueroset al.1999;Gutiérrez-Marcoset al.2004;Balandínet al.2005;Muñizet al.2006;Gómezet al.2009;Fouquetet al.2011;Lopatoet al.2014;Liet al.2017),qRT-PCR analysis was performed using seeds ofmn2-m1and WT at 15 and 20 DAP.The expression levels ofMN2and all the tested marker genes,MEG-1,BETL-2,BETL-10,TCRR-1,andESR-6were decreased inmn2-m1kernels (Fig.6-A).Expression analysis of someZmNRTSwas also performed in developing seeds ofmn2-m1and WT (Fig.6-B).The expression ofZmNRT1.5A,ZmNRT1.1AandZmNRT1.1Bwere increased inmn2-m1at 15 DAP,both 15 and 20 DAP,and 20 DAP,respectively,andZmNRT1.5Bshowed no significant difference between WT andmn2-m1.

4.Discussion

4.1.MN2 plays an important role during maize kernel development

Themn2-m1mutants showed pleiotropic phenotypes.Embryo-specific(emb) andempty pericarp(emp) mutants show strong effects predominantly in embryo development,and have strong defects in endosperm formation,respectively,defectivekernel (dek) andminiature(mn)mutants are characterized by kernel phenotypes involving profound effects on both endosperm and embryo developpment,and a drastic reduction in kernel size,respectively (Neuffer and Sheridan 1980;Scanlonet al.1994;Dollet al.2017).mn2-m1owns dual attributes of bothdefectivekernel (dek) andminiature(mn) mutants and presents an obvious small kernel phenotype with development defects in both embryo and endosperm.The physical characteristics ofmn2-m1kernels such as length,width,thickness,and weight were also affected in multiple F2segregating populations of different genetic backgrounds.The kernel length ofmn2-m1was not significantly affected in anmn2-m1/Chang7-2 F2population,perhaps owing to its own genetic background.

Fig.4 Subcellular localization of MN2.GFP-MN2 fusion protein was transiently expressed in tobacco leaf protoplasts.CD3-mCherry was used to lable the plasma membrane.The bright-field image,GFP fluorescence (green),CD3-mCherry fluorescence (red),and a merge (yellow) of the bright-field,GFP fluorescence and CD3-mCherry fluorescence images are shown in the respective panels.Bars=20 μm.

Fig.5 Expression profiles of MN2 and its homologs ZmNRT1.5A and ZmNRT1.5B.R,root;S,stem;L,leaf;E,ear;T,tassel.DAP,day after pollination.These tissues were collected from maize inbred line B73.Tissues of root and leaf were collected from 15-day-old seedlings,stem was collected from the 7th to the 8th joints of above the ground from the adult plants.The results represent three biological and experimental replicates.Bars mean SD.

4.2.NRT1.5 confers the mn2-m1 mutation

In our study,only one candidate gene was identified in the 209.9 kb target region ofmn2-m1,Zm00001d019294,predicted to encode NRT1.5.No other genes related to kernel phenotypes have been previously cloned in this target region,so we assert thatmn2-m1is a novel mutation conferring small kernel size in maize.Our sequencing results ofmn2-m1,key recombinants and another two alleles(mn2-m2andmn2-710B),and co-segregated analysis with CAPS1 marker,and further genetic allelism tests provided evidences thatZm00001d019294(NRT1.5) was the gene responsible formn2-m1in maize.Further,the expression levels of its own mRNA and other genes relative to BETL and ESR development were significantly suppressed inmn2-m1kernels.Themn2-m1mutant showed significantly reduced kernel length,width,thickness,and weight relative to WT plants,and retarded development of endosperm and embryo from early stages.We infer that the defect inmn2-m1led to the kernel development deficiency in themn2-m1mutant,affecting development of embryo and endosperm and causing small kernel size.Several small kernel mutants displaying both embryo and endosperm defects had been cloned and characterized in maize (Chenet al.2017;Liet al.2017;Qiet al.2017,2019;Wanget al.2017;Daiet al.2018).Dek10,Dek35,Dek36,andDek37encode PPR proteins,which are essential for mitochondrial functions and seed development (Chenet al.2017;Qiet al.2017;Wanget al.2017;Daiet al.2018).Dek44andUBL1encode a putative 50S ribosomal protein L9 and an RNA exonuclease,respectively,loss of function of them both arrest embryo and endosperm development (Liet al.2017;Qiet al.2019).To our knowledge,no small kernel mutants caused by mutations in nitrate transporters was reported in maize.Consequently,NRT1.5 is the first identified nitrate transporter responsible for small kernel phenotype arresting embryo and endosperm development in maize.

NRT1.1andNRT1.5have been characterized as essential transporters ofNRT1transporter families,involved in root-to-shoot nitrate translocation inArabidopsis(Linet al.2008;Chenet al.2012;Léranset al.2013).Knockout mutation ofNRT1.1(chl1-5)inArabidopsisreduced the amount of15NO3–enrichment in the leaves (Léranset al.2013).Knockdown or knockout mutations ofNRT1.5inArabidopsisreduced the amount of nitrate transported from the root to the shoot (Linet al.2008).Our qRT-PCR results revealed thatZmNRT1.5A,ZmNRT1.1AandZmNRT1.1Bwere expressed at higher levels andMN2(ZmNRT1.5) was expressed at a lower level inmn2-m1developing seeds than WT,respectively.The increased expression ofZmNRT1.5A,ZmNRT1.1AandZmNRT1.1Binmn2-m1seed might be caused by the feedback mechanism in response to lower expression ofMN2.Thus,some compensatory effects might exist between them.Different expression profiles ofMN2and its homologsZmNRT1.5AandZmNRT1.5Bin maize indicate a diversity of functions may exist between them andMN2may play the major role in developing seeds.NRT1.5mRNA was mainly expressed inArabidopsisroots and was located in the plasma membrane based on expressing the NRT1.5-GFP fusion protein inArabidopsisprotoplasts (Linet al.2008;Liet al.2010).Our results revealedNRT1.5to be mainly expressed in developing seeds and MN2 is also targeted to the plasma membrane by transient transforming the GFP-MN2 fusion protein into tobacco leaf protoplasts.We postulated that some mutation in maizeNRT1.5will result in low levels of nitrate transported from the root to the upper organs,or a novel functional deficiency which conferred low efficiency of kernel development.However,whethermn2-m1mutation reduced levels of nitrate transported from the root to the upper organs still remains to be investigated.

Fig.6 Quantitative real time PCR analysis.A,expression analysis of MN2 and other genes relative to BETL (MEG-1,BETL-2,BETL-10,and TCRR-1) and ESR (ESR6) development in seeds at 15 and 20 days after pollination (DAP) from wild type and mn2-m1.B,expression analysis of ZmNRT1.5A,ZmNRT1.5B, ZmNRT1.1A,and ZmNRT1.1B genes in seeds of wild type and mn2-m1 at 15 and 20 DAP.* indicates P≤0.05 (Student’s t-test).The results represent 3 biological and experimental replicates.Bars mean SD (n=9).

ArabidopsisNRT1.5 is also involved in nitrate allocation to roots,and the regulation of tolerance to salt,drought and cadmium stresses,however,it remains to be evaluated in maize (Chenet al.2012).The upstream and downstream components that interact with NRT1.5 remain to be identified and how the pathway is controlled still remains to be discovered.We suggest that the nitrate transporter NRT1.5 plays an important role in the kernel development of maize.In the future,NRT1.5might be used to achieve higher maize yields and nitrogen use efficiency through overexpression of genetic transformation using constitutive promoters,or generating mutants using CRISPR/Cas9 gene editing technology,or mining elite alleles by association mapping,or molecular marker-assisted breeding.

5.Conclusion

In this study,a small kernel mutantmn2-m1was isolated in maize.TheMN2locus was cloned by map-based cloning,further genetic allelism tests and qRT-PCR confirmed thatZm00001d019294,which was predicted to encode nitrate transporter1.5 (NRT1.5),was the candidate gene.Our results provided essential information thatMN2plays an important role during kernel development in maize.

Acknowledgements

We thank Marty Sachs in the Maize Genetics COOP Stock Center (Urbana,Illinois,USA) for kindly furnishing seeds ofmn2-710B(Stock 710B).This work was supported by the National Key Research and Development Program of China (2017YFD0101204),the National Natural Science Foundation of China (31701443),the Agricultural Science and Technology Innovation Project of the Shandong Academy of Agricultural Sciences,China (CXGC2017B01),and the Natural Science Foundation of Shandong Province,China (ZR2016CB52).

Appendiesassociated with this paper can be available on http://www.ChinaAgriSci.com/V2/En/appendix.htm


登录APP查看全文