Qioling Luo,Wn Teng,Shung Fng,Hongw ei Li,Bin Li,Jinfng Chu,Zhensheng Li,Qi Zheng,*

a State Key Laboratory of Plant Cell and Chromosome Engineering,Institute of Genetics and Developmental Biology,Chinese Academy of Sciences,Beijing 100101,China

b University of Chinese Academy of Sciences,Beijing 100049,China

c National Centre for Plant Gene Research(Beijing),Institute of Genetics and Developmental Biology,Chinese Academy of Sciences,Beijing 100101,China

Keywords:Jasm onic acid Photosynthesis RNA-Seq Salt tolerance Triticum aestivum

A B S T R A C T Xiaoyan 60(XY60)is a new wheat variety bred in the laboratory of Zhensheng Li.After salt treatm ent,seedlings of XY60 m aintain green leaves and produce longer roots than the high yielding cultivar Zhongm ai 175(ZM175).To explain these different phenotypes w e carried out an RNA-Seq analysis using 12 samples from three tissues of both varieties subjected to salt and control treatm ents.By com paring data from the salt treated plants with the control,703,979,and 1197 differentially expressed genes(DEGs)w ere detected in new leaves,old leaves,and roots of XY60,respectively.The corresponding num bers for ZM175 w ere 613,1401,and 1301.The m ost significantly enriched Gene Ontology(GO)term s and KEGGpathw ays w ere associated w ith polyunsaturated fatty acid(PUFA)m etabolism in both new and old leaves from XY60.They were associated w ith photosynthesis and energy m etabolism in ZM175.The most significantly enriched KEGG pathway in roots of both varieties w as“glucosinolate biosynthesis”.In addition,jasm onic acid(JA)concentration in XY60 w as higher than in ZM175,although it increased significantly in both varieties follow ing salt treatm ent.Trends in relative expression levels of AOS,MYC2,and JAZ revealed by q RT-PCR w ere concordant w ith those from RNA-Seq.Our results suggest that PUFAs m ay contribute to salt tolerance in com m on w heat by enhancing the photosynthetic system and JA-related pathw ays.

1.Introduction

Salinity is a major abiotic stress that reduces the grain output w orldw ide.More than 800 m illion hectares of land are dam aged by salt globally[1],the area of salt-affected arable soil increases yearly due to the trace am ounts of sodium chloride(NaCl)from irrigation water,and rising w ater tables due to non-sustainable irrigation practices in som e regions and dryland salinity in other places[2-5].Bread w heat(Triticum aestivum L.)is one of the three m ost im portant crops in the w orld,and its grain yield is significantly influenced by salinity stress.

Plants respond to salt stress in tw o distinct phases.The first is the osm otic phase,w hich starts im m ediately w hen the roots sense excessive salt concentrations;and the second is the ion-toxic phase that begins w hen sodium accum ulates to toxic concentrations in older leaves[4,5].The osmotic effect leads to decreased size of stom atal apertures to balance the w ater pressure,eventually affecting photosynthesis[4].Excess reactive oxygen species(ROS)are in turn generated due to the reduced rate of photosynthesis,m eaning that plants m ust synthesize substantial oxidoreductases such as superoxide dism utase(SOD),peroxidase(POD),ascorbate peroxidase(APX),and catalase(CAT)[6]to resist the adverse environment[6,7].Recent proteomic comparisons revealed that a superior photosynthetic capacity could contribute to salt tolerance in w heat[8].As tim e passes,Na+finally accumulates m ainly in leaf blades,rather than in root tissue[9].Therefore,reducing root Na+intake and translocation,especially to above-ground parts,is extrem ely im portant.Although SOS1 m ay be the Na+sensor in plants[10],sensors of hyperosmosis and Na+have remained elusive[4,5,11,12].Fortunately,som e interm ediary signaling com ponents have been identified[4,5,10,11],and these include the SOSpathw ay[10],high-affinity potassium(K+)uptake transporter HKT[13,14],Na+/H+exchanger NHX [15],Arabidopsis thaliana vacuolar H+-translocating pyrophosphatase AVP[16,17],Ca2+signals[18-21],and m any m em bers of core transcription factor fam ilies,such as basic leucine zipper(b ZIP)[22],WRKY[23],MYB[24],basis helix-loop-helix(bHLH)[25],and NAC[26].Polyunsaturated fatty acid(PUFA)contents affect the fluidity of cytoplasm ic and thylakoid m em branes and the function of proteins such as the Na+/H+antiporter attached to them[27].Moreover,derivatives of PUFAs such as jasm onic acid(JA)[28-35]function as signal m olecules under abiotic stress.Several recent studies have im plicated JA signaling in salinity stress in plants[29-35].

RNA-Seq uses deep-sequencing technology that provides a far m ore precise m easurem ent of the levels of transcripts and their isoform s than other m ethods[36].With these advantages and low cost,RNA-Seq has been applied in numerous studies on response to abiotic stress[37-40].In this study,w e used RNA-Seq to explore reasons w hy tw o w heat varieties perform differently after being subjected to salinity stress.The information obtained revealed useful clues to explain differences in salinity stress response in w heat.

2.Materials and m ethods

2.1.Materials

The parents of a recom binant inbred line(RIL)population,nam ely,Xiaoyan 60(XY60)and Zhongm ai 175(ZM175),w ere used in this study.XY60,bred at the Institute of Genetics and Developm ental Biology(IGDB),Chinese Academ y of Sciences(CAS)has shown consistent salt tolerance and can be cultivated in soils w ith salt content(≤0.3%m/m).ZM175 is a high-yield variety cultivated in the Huanghuai and Northern w inter w heat region of China.We grew seedlings as described below,and then harvested and sam pled three tissues of each seedling:roots,old leaves(the first and second leaves),and new leaves(all leaves except the old leaves).Three biological replicates of each tissue w ere pooled as one sam ple for RNASeq.Finally,12 sam ples from the three tissues of the tw o varieties under tw o treatm ents w ere prepared and nam ed CKXY60-new,CKXY60-old,CKXY60-root,SXY60-new,SXY60-old,SXY60-root,CKZM175-new,CKZM175-old,CKZM175-root,SZM175-new,SZM175-old,and SZM175-root.CK refers to the untreated control,Srefers to salt treatm ent,new refers to new leaf tissue,old refers to old leaf tissue,and root refers to root tissue.

2.2.Plant grow th and salt treatment

Sixty plum p seeds of each variety w ere neatly placed on a double layer of m oistened filter paper in tw o Petri dishes and germinated for 3 days.Thirty uniform seedlings of each variety w ere planted in w et soil in Ray Leach Cone-tainers(Stuew e&Sons,Inc.,USA),w hich w ere soaked in nutrient solution containing 1 m m ol L-1of Ca(NO3)2and 0.2 m m ol L-1of KH2PO4.On the second day,the plants of each variety w ere divided into tw o groups and processed w ith 0 or 3 g L-1sea salt-nutrient solution.After 3 days,the sea salt solution w as changed to a concentration of 6 g L-1,and 5 days later the concentration was increased to 9 g L-1.The nutrient solution of the zero-control group w as also replaced at the sam e tim es.The solution w as subsequently renew ed every 5 days until the plants w ere harvested.The total treatm ent lasted 21-22 days.The plants w ere kept in a greenhouse maintained under a 16/8-h light/darkness cycle at 22°C/18°Cduring their grow th.

2.3.RNA sequencing and data analysis

2.3.1.Total RNA isolation and qualification

Total RNA w as isolated from the above sam ples using a RNeasy Plant Mini Kit(Cat.No.74904;Qiagen)and treated w ith DNase I during RNA purification.RNA degradation and contam ination w ere m onitored on 1.5%agarose gels.Optical densities at OD260/280and OD260/230were examined using a Nano Drop 2000(Therm oFisher Scientific,CA,USA).RNA concentration and integrity w ere determ ined using an RNA Nano 6000 Assay Kit for the Agilent Bioanalyzer 2100 system(Agilent Technologies,CA,USA).

2.3.2.Library preparation for transcriptome sequencing

About 5μg of RNA per sam ple w ere used for transcriptom e sequencing.Sequencing libraries w ere obtained follow ing the handbook for the NEBNext Ultra RNA Library Prep Kit for Illum ina(NEB,USA).Index codes w ere added to attribute sequences to each sam ple.Poly(A)m RNA w as isolated from total RNA using oligo(d T)m agnetic beads,and then fragm ented in fragm entation buffer.First-strand c DNA w as synthesized w ith random hexam er prim ers using M-Mu LV Reverse Transcriptase(RNase H-).Second-strand c DNA w as then synthesized using DNA polym erase I,RNase H,d NTPs,and buffer.Rem aining overhangs w ere changed into blunt ends using exonuclease/polymerase.NEBNext adaptor(hairpin loop structure)w as ligated to the DNA fragm ents(adenylation of 3′ends)to prepare for hybridization.The library fragm ents w ere purified to select the c DNA fragm ents of 250-300 bp by using the AMPure XP system(Beckm an Coulter,Beverly,CA,USA).ThreeμL of USER Enzym e(NEB)w ere used w ith size-selected,adaptor-ligated c DNA at 37°C for 15 m in.Next,the universal PCRprim ers,Q5 Hot Start HiFi DNA polymerase,and index(X)primer,were added for PCR.Finally,w e purified the PCRproducts(AMPure XPsystem)and assessed the library quality on an Agilent Bioanalyzer 2100 system.

2.3.3.Sequencing and read mapping

The clustering of index-coded sam ples proceeded on a c Bot Cluster Generation System using a Tru Seq PE Cluster Kit v4-c Bot-HS(Illumina),following the m anufacturer's recommendations.After cluster generation,library preparations w ere sequenced on an Illum ina Hiseq 4000 platform and pairedend reads w ere generated.To analyze RNA-Seq data,the reference genom e and related gene m odel annotation files w ere dow nloaded directly from the w heat genom e w ebsite.Using Bow tie v2.2.3,w e built an index of the reference genome,and blasted clean reads to the wheat reference genom e (https://urgi.versailles.inra.fr/dow nload/iw gsc/Gene_m odels/)using Top Hat v2.0.12 (http://ccb.jhu.edu/softw are/tophat/index.shtm l).

2.3.4.Quantification of gene expression level

The expression level of each gene w as m easured as the num ber of reads per kilobase of transcript sequence per m illion base pairs(RPKM),which was the expected num ber of reads per kilobase of the exon region in a gene per m illion m apped reads.RPKM is currently the m ost w idely used m ethod for calculating gene expression levels,because it considers not only the effect of sequencing depth but also that of gene length for the read count[41].

2.3.5.Differential expression analysis

Prior to analysis of differential gene expression,the read counts w ere adjusted for each sequenced library using the edgeR program package through one scaling norm alized factor.Differential expression analysis between two samples w as perform ed using the DEGseq Rpackage.Using the edgeR function of Bioconductor, w e screened differentially expressed genes(DEGs)using the follow ing thresholds:FDR≤0.05 and|log2ratio|≥1[FDRis the false discovery rate,log2ratio m eans log2(fold change)up/dow nregulation(B/A)].

2.3.6.GO and KEGG enrichment analyses of differentially expressed genes

GO enrichm ent analysis of DEGs w as carried out using top GO softw are.The significant threshold of GO term s w as P≤0.05.KEGG pathw ay enrichm ent of DEGs w as perform ed using R softw are.

2.4.Quantitative real-time PCR

Tw elve RNA sam ples w ere obtained from three tissues of both varieties w ith or w ithout salt treatm ent.c DNA w as synthesized using a Rever Tra Ace q PCR RT Master Mix w ith gDNA rem over(Toyobo,Japan;Code No.FSQ-301).Quantitative real-tim e PCRw as perform ed on a Step OnePlus Real-tim e PCR System(Applied Biosystem s,USA)using SYBR Green Realtime PCR Master Mix(Toyobo;Code No.QPK-201).The reference sequence w as the w heat actin gene,because it is expressed m ost stably in different tissues and developm ental stages w ith or w ithout stress.Three biological replicates and three technical replicates w ere analyzed for each sample.The prim ers used for q RT-PCR in this study are detailed in Table S4.

2.5.JA quantification

Fresh plant tissues(～200 m g)w ere hom ogenized and extracted in 80%m ethanol(v/v)for 24 h w ith2H5JA as an internal standard.After centrifugation,purification w as perform ed using an Oasis Max solid-phase extract cartridge based on a previously described m ethod[42].LC-MS/MS analysis w as performed on a UPLC system(Waters,USA)coupled to the 5500 Q-Trap System(AB SCIEX,USA).The sam ple w as injected onto a BEH C18 colum n(1.7 m m,2.1 m m×150.0 m m;Waters)filled w ith m obile phase 0.05%acetic acid(A)and 0.05%acetic acid in acetonitrile(B).Quantification w as perform ed using the m ultiple reaction m onitoring(MRM)m ode and the transitions w ere 214.0＞59.0 for2H5JA,and 209.0＞59.0 for JA.Three biological replicates w ere analyzed for each sample in this study.

3.Results

3.1.Different performances of XY60 and ZM175 after salt treatment

We adopted an experimental method to sim ulate the actual conditions experienced by plants in coastal regions.Under control conditions,shoots and roots of XY60 and ZM175 exhibited sim ilar grow th(Fig.1-A).When subjected to salt stress,the growth of roots of both XY60 and ZM175 was clearly inhibited,but ZM175's roots w ere m ore seriously im pacted(Fig.1-B).Prolonged salt treatm ent leads to the accum ulation of Na+to a toxic level in plant cells,and older leaves turn yellow or even die[9].In this study,after about 3 weeks of salt treatm ent,the first or even the second leaf of ZM175 turned yellow,w hile the leaves of XY60 w ere still green(Fig.1-B).The fresh w eights of stem s,new leaves,and total shoots show ed no significant difference betw een XY60 and ZM175 under the control treatm ent,but they w ere significantly higher in XY60 than in ZM175 under the salt treatm ent(Fig.1-C).The grow th of both XY60 and ZM175 w ere restrained by salt stress,but based on the dry w eights of different tissues,the biom asses of all tissues and the w hole plant of XY60 w ere also significantly greater than those of ZM175 after salt treatm ent(Fig.1-D).

3.2.RNA-Seq analysis

RNA sequencing by Illum ina Hiseq 4000 w as applied to three tissues(new leaves,old leaves,and roots)of XY60 and ZM175 w ith or w ithout salt treatm ent;1,387,578,054 raw reads w ere obtained from 12 sam ples,and average read num ber w as 115,631,504.Average GC content w as 57.28%,and an average 94.70%of raw reads had quality scores at Q20(an error probability for base calling of 1%).After rem oving the adapter,poly-N,and low-quality reads,clean data represented＞97%of the raw data(Fig.S1).Regarding the clean data,the total num ber of reads w as 933,971,458,the m axim um num ber of reads in 12 sam ples w as 79,427,316(CKXY60-root),the m inim um w as 76,116,790(CKXY60-old),and the average w as 77,830,955.The average percentage of GC w as 57.57%and an average of 95.28%of the clean data had quality scores of Q20(Table S1).

Fig.1-Grow th of XY60 and ZM175 seed lings w ith or w ithout salt treatment.(A)Phenotypes under control conditions.(B)Phenotypes under salt treatm ent.(C)Fresh w eights of different tissues.(D)Dry w eights of different tissues.All data in Cand D are m eans±stand ard d erivations(SD)d erived from three indep endent ex perim ents.*,significant at P＜0.05,**,significant at P＜0.01 determ ined by the Stud ent's t-test.

Before further analysis,a series of assessm ents w ere perform ed to confirm the sequencing quality of the 12 sam ples(Fig.S2).We m apped the clean data from each sam ple to the w heat reference genom e sequence w ith the softw are Top Hat and found that the proportion of total mapped reads ranged from 63.67%to 70.57%(49,770,278 to 54,386,796 reads)(Table S2).The m ean percentage of m apped paired reads w as 55.49%.We also aligned the clean data w ith the reference gene using the softw are SOAPaligner/SOAP2[43].The percentage of total reads m apped to a gene ranged from 60.70%to 66.91%,and the percentage of reads that perfectly m atched a gene w as 34.88%-39.19%in all 12 sam ples.The unique m atched reads constituted about 20%of the total and m ulti-position-m atched reads m ade up about 40%(Table S3).

To obtain an overview of the relationship am ong these sam ples,w e conducted PCA w ith the clean data of 12 sam ples.The tw o-and three-dim ensional diagram s show ed that samples from the sam e tissue clustered together(Fig.2).The sam ple SZM175-old w as clearly exceptional am ong old tissues,agreeing w ith its m arkedly different phenotype.SZM175-new w as a relative outlier in new tissue.

3.3.Screening of DEGs between pairs of samples

The numbers of the DEGs betw een pairs of sam ples w ere determ ined(Fig.3).When com paring sam ples from salt treatm ent w ith those of the control treatm ent,there w ere 703,979,and 1197 DEGs in new leaves,old leaves,and roots of XY60,respectively.For ZM175,the corresponding num bers w ere 613,1401,and 1301,respectively.More genes w ere generally upregulated by salt treatm ent than w ere dow nregulated(S/CK).The CKZM175-old-vs.-SZM175-old group w as particularly pronounced in this regard because the upregulated genes outnum bered the dow nregulated ones by 3.5-fold.However,the opposite situation w as found in the CKZM175-root-vs.-SZM175-root group,in w hich 251 m ore genes w ere dow nregulated by salt treatm ent than w ere upregulated by it.

Fig.2-Principal com p onents analysis(PCA)of the 12 sam ples.(A)Tw o-dim ensional chart p roduced w ith the clean d ata from 12 sam ples.(B)Three-dim ensional chart prod uced w ith the clean data from 12 sam ples.

When comparing XY60 and ZM175 under the control treatm ent,w e found that root tissue sam ples contained the m ost DEGs,w hich could indicate large inherent differences in root tissue betw een the tw o varieties.

Fig.3-Com parison of DEGs betw een p airs of sam ples.DEGs of A-vs.-B are those w ith gene exp ression fold change(B/A)w ith FDRthreshold s of≤0.05 and|log2ratio|≥1.

Venn diagram s w ere created to analyze the sim ilarities and differences of XY60 and ZM175 in the sam e tissue(Fig.4).There w ere m any com m on DEGs in the sam e tissue sam ples,but more DEGs w ere exclusive to either XY60 or ZM175.The DEG num bers of the new leaf tissue and root tissue of the tw o varieties w ere relatively sim ilar(Fig.4-A,C);how ever,the num bers w ere quite different in old leaf tissue(Fig.4-B).This accorded w ith the pronounced phenotypic difference in old leaves between XY60 and ZM175.The new leaf tissue of both varieties had the least DEGs and com m on genes,w hereas the root tissue had the m ost DEGs and com m on genes.This m ay be related to the different degrees of salt toxicity in different tissues.

Fig.4-Com p arisons of DEGs betw een sam ples of com m on tissues from XY60 and ZM 175.(A)New leaf tissue;(B)old leaf tissue;(C)root tissue.

Regarding the specific identities of genes show ing differential expression,w e first focused on classical ion transporter genes related to salt stress,but found that only one allele of HKT8(Traes_4BL_5A58CACB2)w as significantly upregulated in root tissues of XY60 and ZM175 after salt treatm ent.Moreover,there w as no significant difference at the transcriptional level in the NHX1,AVP,and other HKT genes in com parisons of salt treatm ent and the corresponding control in all 12 sam ples.Notably,no differentially expressed SOS gene w as detected in any sam ple.Next,w e analyzed the top 20 DEGs w ith m inim um FDR and m axim um|log2ratio|to compare the salt treatment w ith the control,and found many genes responding to abiotic stress.In new leaf tissue of XY60,som e DEGs w ith oxidoreductase(cytochrom e P450 and Lascorbate oxidase),WRKY transcription factor 55,and translation initiation factor 6-2 activities w ere identified.Dehydrin,allene oxide synthase,and putative lipoxygenase 3 w ere upregulated in old leaf tissue of XY60.Epoxide hydrolase 2,ribonuclease 1,putative LRR receptor-like serine/threonineprotein kinase,and putative laccase-9 w ere also am ong the top 20 DEGs of old leaves of XY60.Regarding ZM175,photosystem II 10 k Da polypeptide w as the m ost dow nregulated gene in both new and old leaves,and some genes w ith oxidoreductase activity w ere also detected,including peroxidase 2,dehydrogenase 1,gibberellin 20 oxidase 2,and flavanone 3-hydroxylase.Moreover,som e genes involved in starch and sucrose metabolism,such as fructan 6-fructosyltransferase and 0/6(G)-fructosyltransferase,w ere strongly upregulated in new leaf tissue.An isocitrate lyase functioning in cellular carbohydrate metabolic processes was greatly enhanced in old leaves.Regarding root tissue,comm on DEGs in XY60 and ZM175 w ere found;for exam ple,dehydrins DHN2 and DHN3,solute carrier fam ily 22 m em ber 3,and sugar porter.How ever,defensin Tk-AMP-D1.1,prolinerich protein precursor,MADS-box transcription factor 27,and aspartic proteinase Asp1-like w ere found in XY60.Sim ultaneously,a putative flavin-containing m onooxygenase 1,tonoplast dicarboxylate transporter-like,and som e predicted proteins w ere included in the top 20 DEGs in the root tissue of ZM175.

3.4.GO enrichment analysis of the DEGs

Based on the DEG list,w e conducted GO enrichm ent analysis using the top GO function of Bioconductor.DEGs(S/CK)of six groups (CKXY60-new-vs.-SXY60-new, CKZM175-new-vs.-SZM175-new,CKXY60-old-vs.-SXY60-old,CKZM175-old-vs.-SZM175-old,CKXY60-root-vs.-SXY60-root and CKZM175-root-vs.-SZM175-root)and all w ere classified in three categories:‘biological processes',‘cellular components'and‘molecular functions'(Fig.5,Figs.S3,and S4).More than 70%of the DEGs in the category‘biological processes'in each of the six groups w ere concentrated into four function term s,nam ely,‘m etabolic processes',‘single-organism processes',‘cellular processes'and‘response to stim ulus'.Seventy-80%of DEGs in the category‘cellular com ponents'in each of the six groups w ere classified into three m ajor term s,viz.‘cell',‘cell parts'and‘organelle'.The top three function terms involving most DEGs in the‘m olecular functions'category w ere‘catalytic activity',‘binding'and‘transporter activity'in all six groups;and the first tw o term s m ade up about 90%of the DEGs in each group.

Fig.5-Functional classification of GO term s from DEGs for new leaf tissue.(A)CKXY60-new-vs.-SXY60-new.(B)CKZM 175-new-vs.-SZM 175-new.The y-ax is on the left rep resents the p ercentage of DEGs in the GO term;the y-axis on the right rep resents the num ber of DEGs.

The m ost significantly enriched GO term s w ere m arkedly different am ong the six groups.For XY60,‘fatty acid biosynthesis processes'w as the m ost significantly enriched GO term in the category of biological processes in new leaf tissue.Besides this com m on term,‘proline biosynthetic process'w as also significantly enriched in the old leaf tissue.Most of the DEGs associated w ith these tw o functional term s w ere upregulated.Proline is know n as a hallmark of response to abiotic stress.According to the GO analysis,proline accum ulated in XY60,indicating that the salt resistance system had been activated.The GO term of the cellular component‘Golgi apparatus part'w as the most significantly enriched in new and old leaf tissues,w hereas‘extracellular regions'w as significantly enriched in the new leaf tissue and‘cell w all'w as enriched in the old leaf tissue.In the m olecular function category,the three m ost significantly enriched term s w ere‘fructosyltransferase activity',‘dioxygenase activity'and‘oxidoreductase activity,acting on single electron donors w ith incorporation of tw o atom s of oxygen'in new leaves,and the latter of these and‘glucosidase activity'w ere the m ost significantly enriched in old leaves.In contrast to XY60,for the CKZM175-new-vs.-SZM175-new group,the GO term‘cellular response to starvation'w as the m ost significantly enriched in biological processes,and in old leaf tissue from ZM175,the other tw o term s w ere‘response to red or far red light'and‘sulfate transport'.Alm ost all of the DEGs in‘response to red or far red light'w ere dow nregulated corresponding to chlorosis in old leaves.In addition,all of the DEGs associated w ith the term s‘sulfate transport'and‘cellular response to starvation'w ere upregulated.These results proved that the photosystem of ZM175 w as dam aged after salt stress and that the energy supply w as insufficient to m aintain grow th.In the cellular com ponent category,‘external encapsulating structure'and‘cell periphery'w ere the m ost significantly enriched in new leaf tissue,w hereas in old leaf tissue,the m ost enriched categories w ere‘photosystem'and‘chloroplast strom a'.In the m olecular function category,‘a(chǎn)lpha-1,4-glucosidase activity'w as the m ost significantly enriched term in new leaves,and additionally,‘hydrolase activity,acting on glycosyl bonds'and‘fructosyltransferase activity'w ere also significantly enriched in old leaf tissue of ZM175.

In contrast to the results for shoot tissue,in the CKXY60-root-vs.-SXY60-root group,the GO terms‘response to stress',‘response to hypoxia'and‘m alate transport'w ere the m ost significantly enriched in the category of biological processes,w hereas the term s‘extracellular region',‘plant-type vacuole'and‘external encapsulating structure'w ere in the cellular com ponent category,and the term s‘tetrapyrrole binding',‘hydrolase activity, acting on glycosyl bonds' and‘m onooxygenase activity'w ere in the m olecular function category.In the CKZM175-root-vs.-SZM175-root group,the term s‘cellular response to reactive oxygen species',‘response to iron ion'and‘carbon fixation'w ere the m ost significantly enriched in the biological process category,w hile‘extracellular region'w as in the cellular com ponent category and‘tetrapyrrole binding'and‘oxidoreductase activity'w ere in the m olecular function category.The above analyses illustrate that responses to salt stress occurred in the roots of both XY60 and ZM175,but differed in detail,as reflected by the different enriched GO term s.

3.5.Significant enrichment analysis of Kyoto Encyclopedia of Genes and Genomes(KEGG)pathw ays

KEGG is a prom inent reference know ledge base for the integration and interpretation of large-scale molecular data sets generated by genom e sequencing and other highthroughput experim ental technologies[44-46].To obtain a deeper understanding of the functions of DEGs,w e thus conducted KEGG pathw ay enrichment analysis.Five com mon pathw ays,nam ely,‘m etabolic pathw ay',‘biosynthesis of secondary m etabolites',‘starch and sucrose metabolism',‘phenylpropanooid biosynthesis'and‘benzoxazinoid biosynthesis'were found in all six groups.Overall,the top 20 enriched pathw ays in the new leaf tissue of XY60 w ere very similar to those in ZM175,as w as the casefor root tissues.In contrast,only 10 of the top 20 pathways were shared between XY60 and ZM175 in old leaf tissue,agreeing w ith the m arked differences in phenotype betw een XY60 and ZM175 in old leaves.

The three pathw ays w ith the greatest enrichm ent in the CKXY60-old-vs.-SXY60-old group were‘linoleic acid m etabolism',‘a(chǎn)lpha-linolenic acid m etabolism'and‘m onoterpenoid biosynthesis'(Fig.6-A).In contrast,in the CKZM175-old-vs.-SZM175-old group,the pathw ay‘photosynthesis-antenna proteins'had greatest enrichment,far larger than those of the other pathw ays(Fig.6-B).These enriched KEGG pathw ays dem onstrated that the photosynthetic system of ZM175 w as significantly affected,w hereas PUFA m etabolism in XY60 w as markedly influenced.The CKZM175-old-vs.-SZM175-old group featured m ore DEGs than the CKXY60-old-vs.-SXY60-old group in all of the com m on pathw ays.The specific pathw ays of CKXY60-old-vs.-SXY60-old consisted of eight prim ary m etabolism pathw ays,such as ‘citrate cycle',‘phenylalanine m etabolism' and ‘a(chǎn)lpha-linolenic acid metabolism',and tw o secondary metabolism-related pathw ays,nam ely,‘carotenoid biosynthesis'and‘plant horm one signal transduction'.The specific pathw ays of CKZM175-oldvs.-SZM175-old tended to involve secondary m etabolism,for exam ple,‘flavonoid biosynthesis',‘diterpenoid biosynthesis'and ‘zeatin biosynthesis', w ith only three prim ary m etabolism-related pathw ays,nam ely,‘photosynthesisantenna proteins',‘a(chǎn)scorbate and aldarate m etabolism',and‘tryptophan metabolism'.

In the CKXY60-new-vs.-SXY60-new group,the four pathw ays w ith the highest richness factors w ere‘linoleic acid m etabolism',‘sesquiterpenoid and triterpenoid biosynthesis',‘m onoterpenoid biosynthesis'and‘a(chǎn)lphalinolenic acid m etabolism',w hereas in the CKZM175-new-vs.-SZM175-new group,they w ere‘biotin m etabolism',‘m onoterpenoid biosynthesis',‘linoleic acid m etabolism'and‘galactose m etabolism'(Fig.S5).The common pathw ay ‘linoleic acid m etabolism'in the CKXY60-new-vs.-SXY60-new group involved m ore DEGs than the CKZM175-new-vs.-SZM175-new group.Five pathw ays,namely,‘phenylalanine metabolism',‘valine,leucine and isoleucine biosynthesis',‘photosynthesis antenna proteins',‘glucosinolate biosynthesis',and‘biotin m etabolism',in CKZM175-new-vs.-SZM175-new differed from those in CKXY60-new-vs.-SXY60-new.Although the new leaves of SZM175 rem ained green,genes encoding photosynthetic antenna proteins w ere dow n regulated.

The enriched KEGG pathw ays in CKXY60-root-vs.-SXY60-root w ere similar to those in the CKZM175-root-vs.-SZM175-root group,w ith only three differing pathw ays.In XY60,‘zeatin biosynthesis',‘a(chǎn)m ino sugar and nucleotide sugar m etabolism',and‘tryptophan m etabolism'w ere enriched,whereas in ZM175,the different enriched pathw ays w ere‘glyoxylate and dicarboxylate m etabolism',‘galactose m etabolism',and‘carotenoid biosynthesis'.In contrast to new and old leaf tissues the highest richness factor for root tissue w as for the‘glucosinolate biosynthesis'pathw ay in both varieties,although the pathw ay‘diterpenoid biosynthesis'had the sam e richness factor in ZM175(Fig.S6).

3.6.Verification of changes related to the JA pathw ay in both varieties

JA concentrations w ere significantly increased after salt treatm ent in new leaf tissues of both varieties,but the background level of JA in XY60 w as significantly higher than that in ZM175(Fig.7-A).The relative exp ression levels of six alleles of allene oxide synthase(AOS),located on chrom osom es 4A,4B,and 4D,w ere higher in old leaf tissue from XY60 than from ZM175 w hen com paring the salt treatm ent w ith the control.Sim ilar results w ere obtained for new leaf tissue,except for chrom osom e 4D alleles w hose expression levels w ere sim ilar in both varieties(Fig.7-B-G).We also determ ined the expression levels of four MYC2 transcription factors(Traes_2AL_411B944D6,Traes_2BL_8FED05903,Traes_2DL_DE3909A32,Traes_5AL_331CE753D)and tw o jasm onate ZIM d om ain (JAZ) proteins (Traes_5AL_BB55F989A and Traes_5BL_7A6C3831E)in the new leaf tissues;they w ere all upregulated in XY60,but w ere dow nregulated or had low er relative expression levels in ZM175 after salt treatm ent(Fig.7-H).

Fig.6-The top 20 significantly enriched KEGG pathw ays for old leaf tissue.(A)CKXY60-old-vs.-SXY60-old.(B)CKZM 175-oldvs.-SZM 175-old.The y-axis represents KEGG p athw ays;the x-ax is represents enrichm ent factors;larger values indicate a higher percentage of the DEGs in the p athw ay.

Fig.7-JA concentrations and relative exp ression levels of genes in the JA-related pathw ay in tw o varieties.(A)JA concentration in new leaf tissues.The d ata are m eans±stand ard d erivation(SD).The SDs are d erived from three biological repetitions.(*,P＜0.05)d eterm ined by the Student's t-test.(B-G)Relative ex pression levels of six loci for allene oxide synthase(AOS)in old and new leaf tissues.(H)The relative exp ression levels of MYC2 transcription factors and jasmonate ZIM domain JAZ proteins in new leaf tissues.

3.7.Validation of DEGs by quantitative real-time PCR(qRTPCR)

The reliability of RNA-Seq data w as confirm ed in m any studies[37-39,47,48].Although there w as no biological replication in RNA-Seq in our study,＞80% of the differentially expressed genes w ere sim ilar betw een new and old leaves,and＞70%w ere sim ilarly exp ressed betw een leaves and stems(data not show n).In addition,the enriched GO term s and KEGG pathw ays of different tissues w ere also very sim ilar.These results can be regarded as additional support for the validity of the RNA-Seq data.To further confirm the accuracy of the DEGs found by RNA-Seq,＞30 genes w ere selected and analyzed by q RT-PCR.The upand dow nregulated trends of the q RT-PCR results w ere all consistent w ith the RNA-Seq d ata.Tw elve genes w ith the highest consistency are show n in Fig.8,including peroxid ase 2,photosystem II 10 k Da polyp eptide,lipoxygenase 2.2,chlorophyll a-b binding protein 1,MADS-box transcription factor 27,and ethylene-responsive transcription factor ERF012.

4.Discussion

Fig.8-Valid ation of transcrip tional changes of 12 genes selected from the DEGs using q RT-PCR.DEGs in(A-D)new leaf tissue,(E-H)old leaf tissue,and(I-L)root tissue.(B and G)DEGs w ere d etected only in ZM 175,or only in XY60(K).

Plants exhibit tw o different phases of response to salinity[9,11].Many transcriptom ic analyses have focused on the first stage,namely,early response to salt exposure[38,47,49-51].Few er studies have concentrated on the second stage.In the present study,an attem pt w as m ade to sim ulate longer term exposure in typical coastal regions w hereby salt content gradually increased over tim e.The seedlings w ere cultivated for about 3 w eeks,and the old leaves of ZM175 becam e yellow,reflecting a salt-specific phenotype.Ow ing to the im portance of root tissue for salt stress,m any studies have focused on roots[47,50-52],but less research has concentrated on shoots[37,53].In this study,w e exam ined three tissues:new leaf tissue,old leaf tissue,and root tissue.Each tissue am ong the 12 sam ples clustered together in PCA(Fig.2),dem onstrating that the differences among the three tissues w ere larger than the differences not only betw een the tw o varieties but also betw een the salt treatm ent and the control.It is crucial that genes are expressed correctly in specific cell types or specific tissues,as a particular gene m ay have opposite effects in shoots and roots[4].This indicates the im portance of analyzing and com paring the DEGs in different tissues from entire plants.

Phytohormones and their crosstalk are key components of responses to abiotic stress.In our study,except for ethylene(ETH),pathw ays related to the synthesis of other phytohormones(including auxin(IAA),cytokinin(CTK),abscisic acid(ABA),gibberellin(GA),jasm onic acid(JA),brassinosteroids(BR),and salicylic acid(SA))w ere found in the top 20 KEGG pathw ays in six groups.For exam ple,the GA synthesis-related pathw ay‘diterpenoid biosynthesis'w as identified in every group except CKXY60-old-vs.-SXY60-old;due to the im portance of chloroplasts,the JA synthesis-related pathw ay‘a(chǎn)lpha-linolenic acid m etabolism'w as discovered only in the green leaf tissue groups,namely,CKXY60-new-vs.-SXY60-new,CKZM175-new-vs.-SZM175-new,and CKXY60-old-vs.-SXY60-old,but not in the yellow leaf tissue group CKZM175-old-vs.-SZM175-old;‘brassinosteroid biosynthesis'w as found only in the root tissues of the tw o varieties;ABA is the best know n phytohormone involved in abiotic stress,and w e identified its synthesis-related pathw ay‘carotenoid biosynthesis'in the CKXY60-old-vs.-SXY60-old and CKZM175-rootvs.-SZM175-root groups.We also detected m any transcription factor fam ilies including WRKY,MYB,MYC,NAC,and MADSbox,am ong the DEGs.Therefore,phytohorm ones had significant roles in salt response in both varieties.

The integration of the results of GO analysis and enriched KEGG pathw ays provided a deeper understanding of different responses to salt stress in the different cultivars and tissues.Photosynthesis converts light energy to chemical energy in chloroplasts,the m ost im portant energy process on earth.How ever,salt stress seriously affects photosynthetic efficiency ow ing to the associated stom atal closure and subsequent ROSproduction,w hich results in thylakoid mem brane dam age [4,5,12,54].In the CKZM175-old-vs.-SZM175-old group,alm ost all DEGs w ere dow nregulated for the GO term‘response to red or far-red light'and the KEGG pathw ay‘photosynthesis antenna proteins';in accordance w ith these processes the three m ost significantly enriched GO term s in the category of cell com ponents w ere‘photosystem',‘chloroplast strom a',and‘photosynthetic m em brane'.In addition,all DEGs associated w ith the GO term‘response to starvation'w ere upregulated,as w ere the ten DEGs predicted to be associated w ith sulfate transporters in the GO term‘sulfate transport'.Sulfate has an essential role in photosynthesis,and participates in the dark reaction after binding w ith iron in ferredoxin.All of these results indicate that the photosynthetic system w as disrupted and that nutrition and energy w ere lacking in the old leaves of SZM175.A sim ilar but less severe situation w as evident in new leaf tissue of ZM175.In contrast,old leaves of SXY60 w ere still green after salt treatm ent.The enriched GO term s and the top 20 KEGG pathw ays w ere not related to the photosynthetic system in the CKXY60-old-vs.-SXY60-old group and far few er DEGs w ere related to photosynthesis,indicating that the intact chloroplasts of XY60 w ere sufficiently robust to maintain norm al photosynthesis after salt treatm ent.Chloroplasts are not only the sites w here energy is produced by photosynthesis,but also the sites of synthesis of abiotic stress-related phytohorm ones,such as ABA and JA,and other m etabolites including enzymes for rem oving ROS[55].The salt tolerance of the w heat cultivar Shanrong 3[a salinity-tolerant derivative of an asym m etric som atic hybrid betw een cultivar Jinan 177 and tall w heatgrass(Thinopyrum ponticum)][50,56]w as partially attributed to its superior photosynthetic capacity[8].Therefore,the m ore robust photosynthesis system of XY60 relative to ZM175 could be a key reason that seedlings of XY60 are m ore salt-resistant.

PUFAs,m ajor com ponents of biom em branes,m odify ROS[55]as w ell as enzym es such as SOD,POD,APX,and CAT.If extensive PUFAs were peroxidated,the fluidity of the membranes w ould decrease,increasing leakiness and causing secondary dam age to m em brane proteins.Increases in PUFAs significantly enhance tolerance to salt stress by decreasing the salt-induced dam age to photosynthetic m achinery and increase ability to restore photosynthetic and Na+/H+antiporter system s[31,57-59].Arabidopsis m utant fad2,w hich lacks functional FAD2,has low er plasm a m em brane PUFAs and reduced Na+/H+antiporter activity that results in cytoplasm ic Na+accum ulation and hence salt sensitivity com pared w ith the w ild type[60].Other studies also suggest that the activity of the Na+/H+antiporter protein depends on m em brane fluidity[27,57,58,61].GO analysis indicated that the term‘fatty acid biosynthetic processes'instead of term s related to the photosynthetic system w as significantly enriched,and all of the DEGs involved w ere upregulated in the CKXY60-old-vs.-SXY60-old comparison.Consistent w ith this,the PUFA-related pathw ays‘a(chǎn)lpha-linolenic acid m etabolism'and‘linoleic acid m etabolism'w ere highly ranked in the top 20 enriched KEGG pathw ays.A similar situation w as evident for the CKXY60-new-vs.-SXY60-new group.Finally,relative expression levels of AOS w ere higher in XY60 than in ZM175 after salt treatm ent.Therefore,enhanced PUFA synthesis could m ake a major contribution to a robust chloroplast photosynthetic system,and ultim ately lead to stronger salt resistance.

As signal m olecules,PUFAs and their derivatives have major regulatory roles in response to abiotic stress[50,62].An exam ple of this is alpha-linolenic acid,a precursor of JA and MeJA.Many studies have dem onstrated that JA,MeJA,JAleucine,JA-isoleucine,and other jasm onates are vital to crucial processes such as direct and indirect defense,secondary m etabolism,reproduction,senescence,and fruit developm ent[30,63-65].Previous studies indicated that pretreatm ent with JA before salinization diminished the inhibitory effect of a high salt concentration on grow th and photosynthesis[29,66,67].Ding et al.[32]also show ed that jasm onate could complement the function of Arabidopsis lipoxygenase 3 in salinity stress response.Moreover,Valenzuela et al.found that salt stress response triggered activation of the jasm onate signaling pathw ay,leading to the inhibition of cell elongation in prim ary roots of Arabidopsis[33].After salt treatm ent,PUFA synthesis w as enhanced in the salt-tolerant w heat cultivar Shanrong 3 com pared w ith that in its parent Jinan 177,and the gene coding allene oxide synthase(jasm onate synthesis)w as also upregulated[50].Other studies revealed the role of JA in response to salt stress in other species[68-72].Ow ing to the significant roles of PUFAs and JAs in responses to abiotic stress,w e deduced that the genes associated w ith the GO term‘fatty acid biosynthetic process'and the KEGG pathw ays‘a(chǎn)lpha-linolenic acid m etabolism'and‘linoleic acid m etabolism'contributed to the salt resistance of XY60.

JAs are conserved elicitors of plant secondary m etabolism,and induce synthesis of alm ost all secondary metabolites[73-75].The top 20 significantly enriched KEGG pathw ays in this study involved m any secondary m etabolites such as flavonoid,terpenoids,glucosinolate,and phenylpropanoid in all three tissues of the tw o varieties.The pathw ays w ith the greatest enrichm ent factors in the new and old leaf tissues betw een XY60 and ZM175 w ere quite different(Fig.S5,Fig.6).However,in root tissue samples,the highest enrichment factor belonged to the pathw ay‘glucosinolate biosynthesis'(Fig.S6).Glucosinolates(GSLs),induced by JAs[76,77],are indispensable in plants and act as direct m ediators of biotic and abiotic stress response[77].Although DEGs involved in the JA synthesis pathw ay w ere upregulated in new leaves of ZM175,w e found that they w ere dow nregulated in SZM175-new w hen com pared w ith the level in SXY60-new(Fig.S7-AC).Similarly,DEGs involved in dow nstream regulation of JA w ere dow nregulated in SZM175-new w hen com pared w ith the level in SXY60-new(Fig.S7-D-F).It is w orth noting that the JA concentrations w ere increased in the new leaf tissues of both varieties after salt treatm ent,and that the trends of relative expression levels of AOS,MYC2,and JAZ w ere concordant w ith RNA-Seq.Therefore,JA-related pathw ays m ay regulate the salt tolerance in both XY60 and ZM175,but w ith effects in XY60 being stronger.

We analyzed transcriptom ic profiles of salt-response in three tissues of tw o varieties,and found that PUFAs m ay contribute to salt tolerance in wheat by enhancing the photosynthetic system and JA-related pathw ays.We detected salt-tolerance QTL in the RIL population developed from a cross betw een XY60 and ZM175.The DEGs provide im portant clues for future gene cloning.

Conflict of interest

Authors declare that there are no conflicts of interest.

Acknow ledgm ents

This research w as supported by the Strategic Priority Research Program of the Chinese Academ y of Sciences(XDA08030105),Key Programs of the Chinese Academy of Sciences(KFZD-SW-112),and STS Project of Chinese Academy of Sciences(KFJSTS-ZDTP-024).We particularly thank Professor Robert Mc Intosh from The University of Sydney for editing the English text of this m anuscript.

Author contributions

ZSL and QZ supervised the research.QZ,QLL,and WT designed the experim ent.QLL perform ed phenotyping and gene expression analyses,and w rote the m anuscript.SF and JFC determined JA concentrations.QZ also put forw ard many constructive suggestions and edited the m anuscript.BL and HWLprovided substantial help in preparing the m aterials and perform ing the experim ents.All authors read and approved the final m anuscript.

App endix A.Supp lem entary d ata

Supplem entary data for this article can be found online at https://doi.org/10.1016/j.cj.2018.11.009.