Xurun Yu,Yufei Jing,Huihui Yo,Liping Rn,Yong Zng,Fei Xiong,*
a Jiangsu Key Laboratory of Crop Genetics and Physiology/Jiangsu Key Laboratory of Crop Cultivation and Physiology,Agricultural College of Yangzhou University;Co-Innovation Center for Modern Production Technology of Grain Crops,Yangzhou University,Yangzhou 225009,Jiangsu,China
b Yangzhou University Guangling College,Yangzhou 225000,Jiangsu,China
Keywords:Development Low temperature Spike Transcriptome Wheat
ABSTRACT Low temperature in early spring impairs wheat growth and grain yield.However,little is known about the cytological and molecular mechanisms underlying low temperature regulation of wheat spike development.Microstructure observation and transcriptome sequencing of wheat spikes under low temperature were conducted.Low temperature slowed spike development,reduced the yield-component parameters of wheat spikes at the harvest stage,delayed the formation of lateral spikelets and tissue development,and induced the early differentiation of terminal spikelets.Low temperature increased the content of abscisic acid and caused the upregulation of genes in the abscisic acid signaling pathway,including those encoding PP2Cs,SnRK2s,and bZIP transcription factors.Low temperature also induced the upregulation of 33 cold-responsive genes involved in wheat response to low-temperature stress and regulation of abscisic acid biosynthesis and metabolism of other substances.The wheat spike adapted to cold conditions by changing the expression levels of genes involved in spike morphogenesis,including the transcription-factor genes MADS6,ERF4,ERF78,WOX6,and NAC48.These findings suggest that low temperature in early spring delays wheat spike development by increasing abscisic acid content or affecting the expression of genes involved in morphogenesis.
Climate change has brought many low-temperature events,causing risk to global wheat production [1].These events often occur in China,the United States,Canada,Australia,Russia,Ukraine,Japan,and other countries,but are most widespread in China [2–4].In the Yangtze-Huai and Huang-Huai river basins,low temperature-induced freezing injury is one of the main meteorological dangers to wheat production [5].
Freezing injury caused by temperatures below 0 °C or severe temperature changes during overwintering.It can be divided into four categories:pre-winter freeze injury,overwintering freeze injury,early spring freeze injury,and late frost damage [6].Of these,early spring freeze injury incurs the greatest damage to wheat growth,because the wheat is in the jointing stage and early reproductive phase at this time [7,8].When wheat enters the reproductive growth stage,low-temperature acclimation is suspended and the plant enters the low-temperature tolerance stage[9,10].Wheat is more vulnerable to low temperature in the reproductive than in the vegetative growth period [11].Low temperature can change the chloroplast ultrastructure and carbohydrate metabolism of wheat,but an increase in antioxidant enzyme activity and osmotic adjustment substance content can alleviate the resulting oxidative damage to chloroplast structure[12].The yield reduction caused by low temperature in spring varies from 30% to 90% depending on the low-temperature conditions and reproductive growth period [13–16].
Wheat yield is determined by three factors:spike number per plant,kernel number per spike,and 1000-kernel weight[17].These factors influence spike development,which can be divided into four periods:pre-double ridge stage,spikelet differentiation stage,and early and late floret differentiation stages[18].Spike development is a morphogenetic process in which cell proliferation and differentiation are coordinated [19,20].When wheat was exposed to frost between stem elongation and grain filling stage,the number of aborted kernels increased and the number of kernels per spike decreased[21].Frost before and after anthesis exerted a similar effect on number of kernels per spike [16].In several studies[22–24],low temperature reduced the number of kernels per spike mainly by increasing the sterility of pollen grains on spikelets.When the low temperature occurred at grain filling stage,the kernels in the spikelet were small,shriveled,or shrunken or showed a blistered appearance [25].
Beyond the effect of low temperature on wheat spike development,few studies have investigated its effect on wheat spike tissues,such as lemma,glume,anther,and pistil.The development of these tissues strongly influences the final number of kernels per spike and ultimately determines wheat yield.Some genes regulating spike development have been identified.Mutation of the wheat FRIZZY PANICLE gene,which encodes a member of the APETALA2/ethylene response transcription factor family,generated supernumerary spikelets [26].The TaMOC1 gene,whose protein was proposed to function as a transcription factor,is involved in wheat spikelet development [27].Wheat FRIZZY PANICLE (WFZP)directly activated VERNALIZATION1(VRN1)and wheat HOMEOBOX4(TaHOX4) to regulate the initiation and development of spikelets[28].
Transcriptome sequencing reveals global gene expression profiles.When winter wheat was exposed to a temperature of 0 °C,genes involved in metabolism in the wheat spike were differentially expressed,including those that encode Ca2+binding proteins,protein kinases,and inorganic pyrophosphatase [29].Proteomic analysis [30] showed that vernalization-responsive protein and cold shock proteins were accumulated in shoot apical meristem at low temperatures,enhancing the cold adaptation of wheat.Functional assignments using GO annotations showed [31] that genes involved in transport,redox,and stress response were differentially expressed at the three-leaf stage in wheat seedlings under low temperature.Physiological and biochemical changes in coldhardy and cold-sensitive wheat cultivars caused by low temperature were associated largely with the expression of genes encoding effector molecules that participate directly to alleviate stress,signal transduction-associated proteins,and transcription factors[32].But no studies have investigated the relationship between gene expression and spike morphological development in wheat under low temperature in early spring using transcriptome sequencing and microstructure observation.
The objective of the present study was to investigate gene expression in young wheat spikes under simulated early-spring low-temperature treatment using transcriptome sequencing and identify temperature-responsive genetic pathways affecting spike development.
The high-yield wheat cultivar Yangmai 15,which is widely planted in the east China plain,was provided by Lixiahe Agricultural Science Institute in Jiangsu province.It was grown in pots(30 × 30 cm) in the greenhouse at Yangzhou University (32°39′N,119°42′E),the Key Laboratory of Crop Genetics and Physiology of Jiangsu Province,from November 2018 to May 2019.The potted soil was sandy loam containing 24.5 g kg-1organic materials and 106,33.8,and 66.4 mg kg-1of available N,P,and K,respectively.
A total of 20 potted plants were divided equally into the control group and a low temperature treatment (LTT) group (n=10).LTT was applied at early spring (before vernal equinox) from March 6,2019 to March 8,2019.At this time,wheat seedling was in the regreening stage,specifically the differentiation stage of the pistil and stamen primordium of the main stem,on the basis of stereomicroscopic images.In a cryogenic room humidity was maintained at 75% RH,and CO2concentration at 500 μmol mol-1,and day and night temperatures at 0 °C (08:30–20:30) and -3 °C(20:30–08:30),respectively[17,33].This LTT lasted for 3 days,during which time the normal temperature treatment (NTT) group,which was defined as the control group,was grown in another cryogenic room,where the day and night temperatures were 10 °C and 5 °C,respectively.After LTT,the potted plants were grown in the greenhouse together with the control group.To each pot was added 1.8 g urea (46.67% N) before seeding and at the booting stage.Normal watering was supplied throughout the growth period.
In both NTT and LTT groups,spikes were collected at 0 and 6 days after treatment(DAT),and spikelets at 12 and 21 DAT.These samples were immediately soaked in a fixative containing 2.5%glutaraldehyde in phosphate buffer (pH 7.2) at 4 °C for 24 h.The plant tissue was then rinsed,dehydrated,substituted with 1,2-propylene oxide,embedded,and polymerized to generate resin-embedded samples,which were then cut longitudinally or transversely into 1-mm slices with an ultramicrotome(Leica Ultracut R,Weztlar,Germany).Slices were spread on glass slides,stained with 0.5% methyl violet for 10 min,and photographed under a light microscope (Leica DMLS) equipped with a digital camera (Nikon D3200,Ayutthaya,Thailand).
Spike length,spikelet number per spike,kernel number per spike,grain weight per spike,and 1000-kernel weight were recorded at maturity.Each parameter was recorded for 8 independent replicates,except for 1000-kernel weight.
A 200-mg sample of wheat spike tissues was ground in liquid nitrogen with 10 mL 80% (v/v) methanol extraction solution.The extract was incubated at 4 °C for 4 h and then centrifuged at 5000×g for 10 min.The supernatant was further purified on a C18-SPE column,dried under nitrogen,and dissolved in 2 mL phosphate-buffered saline solution containing 0.1% Tween 20 and gelatin (pH 7.5).The ABA content was determined on an enzyme-linked immunosorbent assay instrument(TENCAN SPARK,M?nnedorf,Switzerland)using an ABA ELISA Kit(Cominbio ABA-3-T,Suzhou,Jiangsu,China).
Wheat spikes on the main stem in the NTT and LTT groups were removed by hand from the enveloping leaf sheaths at 0 DAT.Spikes were immediately frozen in liquid nitrogen and stored at–78 °C,and total RNA was extracted from the tissues with Trizol.mRNA was enriched using magnetic beads containing oligo(dT)and then fragmented using fragmentation reagent.Single-strand cDNA was synthesized using template (fragmented mRNA) and six-base random primers.Double-strand cDNA was synthesized from the single-strand cDNA with RNase H,DNA polymerase I,and dNTPs.After purification,end repair,and poly A addition,the doublestrand cDNA was connected to the adapter.PCR amplification was performed after fragment-size selection.After the constructed cDNA library passed quality inspection with a bioanalyzer(Agilent 2100,Palo Alto,USA),transcriptome sequencing was performed on a sequencing platform(Illumina HiSeq 2500,San Diego,USA).After quality control,clean reads were obtained by filtering and then mapped to a reference sequence in European Molecular Biology Laboratory’s European Bioinformatics Institute,Cambridge,United Kingdom (http://plants.ensembl.org/Triticum_aestivum/ Info/Index) with Tophat and Bowtie 2.After mapping,the distribution and coverage of reads on the reference sequence were examined to judge whether they passed the second quality control.After this second quality control,the gene expression levels were measured.Differentially expressed genes (DEGs) between the NTT and LTT groups were extracted from the raw data.Genes with expression fold changes of ≥2 or ≤0.5 were assigned as DEGs.
GO functional enrichment analysis of target genes was performed with Goatools (https://github.com/tanghaibao/goatools).Corrected P-values <0.05 were considered to indicate a significant enrichment of function.Gene Set Enrichment Analysis(GSEA) was then performed to identify genes with no significant difference in expression but with biological implications.
Eight DEGs were selected from the transcriptome sequencing to verify relative expression levels by quantitative real-time polymerase chain reaction (qRT-PCR).The primer sequences used are listed in Table S1.The unigene cluster Ta54825,one of the most stable genes under low temperature,was used as an internal reference gene.The PCR reaction was performed on a fluorescence quantitative PCR system (Analytikjena qTOWER3,Jena,Germany)using the following conditions:95 °C for 30 s;40 cycles of 95 °C for 10 s and 54°C for 30 s,and 72°C for 15 s.Relative gene expression levels were calculated by the ΔΔCTmethod[34].Each expression level was the mean of three biological replicates.
SPSS (SPSS Inc.,Chicago,IL,USA) was used to identify differences between the NTT and LTT groups by Fisher’s protected least significant difference test at P <0.05.
The morphology of the potted seedlings after LTT is shown in Fig.1A.The wheat seedlings grew less after LTT than after NTT.Plant height gradually increased from 0 to 27 DAT and was significantly reduced by LTT,especially from 12 to 21 DAT (Fig.1B,D).Unlike plant height,which showed a gradually increasing trend,spike length presented a typical S curve:that is,a slow–quick–slow trend (Fig.1C1,C2 and D).LTT shortened spike length during all stages (Fig.1C1,C2,and D) and reduced the spikelet number,as seen from 18 to 27 DAT (Fig.1C2).
The agronomic trait values at harvest are presented in Table 1.LTT significantly reduced spike length and spikelet number,kernel number and grain weight per spike,and kernel number per spikelet.However,no dramatic differences in 1000-kernel weight were found between the NTT and LTT groups.
Table 1 Agronomic traits of wheat spike under low temperature in early spring.
Thus,low temperature in early spring delayed plant growth and spike development and reduced spikelet and kernel numbers per spike.
The microstructure of the longitudinal section of wheat spikes at 0 and 6 DAT is shown in Fig.2.At 0 DAT,the lateral spikelets differentiated but showed different degrees of differentiation between NTT and LTT groups.At this time,spikelet primordia were observed at the apex position,indicating that lateral spikelets were continuing to be differentiated (Fig.2A,C,D,and E).In NTT group,the meristem of a single spikelet differentiated into one distinct floret,which was composed of glume,lemma,stamen,and pistil primordia(Fig.2B).However,the lateral spikelets in the LTT group showed slower differentiation than those in NTT because the floret organ was incompletely differentiated,presenting only glume and lemma primordia (Fig.2F).As indicated by asterisks in Fig.2C,D,the numbers of lateral spikelets in the NTT and LTT groups were 13 and 11,respectively.
At 6 DAT,the floret primordium differentiated into glume,lemma,stamen,and pistil primordia in the NTT and LTT groups(Fig.2H,I,J,and L).At this time,a new tissue,the terminal spikelet,emerged in the LTT group,and showed a complex structure consisting of lower glume,upper glume,degenerate spikelet,floret glume,second floret lemma,floret primordium,and second floret primordium (Fig.2J,K).In the NTT group,terminal spikelets were not differentiated,and only the spikelet primordium could be observed at the top region of the spike (Fig.2I,G).Compared with 0 DAT,the number of lateral spikelets in the NTT and LTT groups increased to respectively 15 and 14 (Fig.2I,J).
Thus,low temperature in early spring affected the differentiation of wheat spikes.Specifically,the differentiation of the tissues of lateral spikelets was delayed,and the differentiation of terminal spikelets was advanced,resulting in the early fixation of spikelet number per spike.
The microstructure of anthers in florets at the middle of the spike at 12 and 21 DAT is shown in Fig.3M.At 12 DAT,anthers in florets were at the microsporocyte formation stage in the NTT group (Fig.3A).Four layers of cells can be seen clearly outside the microspore mother cells:epidermis,fibrous layer,middle cell,and tapetum (Fig.3A,B).At this time,anthers in the LTT group were at the stage of sporogonium differentiation,and secondary sporogenous cells were observed in the middles of anthers(Fig.3C,D).The four layers were not completely differentiated,indicating that LTT slowed anther development.
At 21 DAT,the spikelet in the middle of the wheat spike differentiated into at least two florets,which were defined as the first and second florets (Fig.3E–H).For the first floret in NTT,anther development was at the meiosis stage,as the chromosomes of microspore mother cells can be clearly observed(Fig.3E,F).At this time,tapetal cells were degenerating,as indicated by the incomplete and degraded nuclei (Fig.3E).However,the microspore mother cells in the LTT group were at the microsporocyte formation stage,and four layers outside were well differentiated(Fig.3G,H).For the second floret,nucleoli of the microspore mother cells within the anther were clearly observed in NTT group,indicating that the microspore mother cells were in meiosis interphase (Fig.3I,J).At this time,LTT slowed the development of the anther,given that the microspore mother cells were at the microsporocyte formation stage(Fig.3K,L).Tapetal cells in the second floret showed larger and more complete nuclei than those in the first floret (Fig.3E,H,I,L).
Fig.1.Plant growth and spike development of wheat under low temperature in early spring.(A)Potted wheat seeding after low temperature treatment.(B)Morphology of wheat plant.(C1,C2) Wheat spike.(D) Time course of plant height and spike length.Numbers in (B),(C1),and (C2) indicate days after treatment.
Thus,low temperature in early spring delayed the development of the first and second anthers at the base of spikelets,and the development of microsporocytes in the NTT group was faster than that in the LTT group.
LTT induced the generation of 2787 DEGs,of which 1596 were upregulated and 1191 were downregulated(Fig.4A).Among these DEGs,33 that encoded cold-responsive proteins,cold-induced proteins,cold acclimation proteins,or cold shock proteins were examined.All DEGs were upregulated after LTT (Fig.4B).They were most abundant on chromosomes 2A,2B,and 2D and absent from chromosome 4 (Fig.4C).The number of cold-response proteins annotated by response to stimulus in the GO molecular function category was greatest (Fig.4D).Their functions included response to cold (GO:0009409),ABA (GO:0009737),response to alcohol(GO:0097305),and water deprivation (GO:0 009414;Fig.4E).GSEA showed that 15 cold-responsive protein-encoding genes were enriched in six GO terms:response to ABA (GO:0009737),cold acclimation (GO:0009631),response to alcohol (GO:0097305),response to water(GO:0009415),and response to water deprivation (GO:0009414).The enrichment degree of response to ABA (GO:0009737) was the highest among all gene sets,and the cold-responsive protein-encoding genes were highly expressed under LTT (Fig.4F).
Fig.2.Microstructure of longitudinal section of wheat spike under low temperature in early spring.(A,B,C) Normal temperature treatment at 0 DAT.A and C are magnifications of the red box labeled in C.(D,E,F) Low temperature in early spring at 0 DAT.E and F are magnifications of the red box labeled in D.(G,H,I) Normal temperature treatment at 6 DAT.G and H are magnifications of the red box labeled in I.(J,K,L)Low temperature in early spring at 6 DAT.K and L are magnifications of the red box labeled in J.Asterisks indicate lateral spikelets;DS,degenerate spikelet;FG,floret glume;FP,floret primordium;GP,glume primordium;LG,lower glume;LP,lemma primordium;LS,lateral spikelet;Nu,nucleus;PP,pistil primordium;SFL,second floret lemma;SFP,second floret primordium;SPr,spikelet primordium;SP,stamen primordium;TS,terminal spikelet;UG,upper glume;Va,vacuole.
Amonng DEGs associated with ABA,genes encoding serine/threonine protein kinase (TraesCS2A02G493800,TraesCS2B02G521800,TraesCS2D02G493700) and ABA 8′-hydroxylase (TraesCS5A02G238000,TraesCS5B02G236500,TraesCS5D02G244900)were upregulated,whereas those encoding ABA-intensive 5-like protein 3 (TraesCS4D02G177700) were downregulated after LTT (Fig.5A).In the ABA signal transduction pathway,ABA receptor PYR/PYL,type 2C protein phosphatase(PP2C),SNF1-related protein kinase 2 (SnRK2),and transcription factor AREB binding factor (ABF)/AREB,belonging to class A bZIP transcription factors are important members.Genes encoding the PP2C,SnRK2,and bZIP transcription factors were differentially expressed after LTT,but that encoding PYR/PYL was not (Fig.5).As seen in the heat map of DEGs,most of the genes encoding PP2C (TraesCS2A02G087100,TraesCS2B02G023600,TraesCS2B02 G101800,etc.;Fig.5B) and bZIP transcription factor (TraesCS 4D02G177700,TraesCS5A02G057500,TraesCS7B02G391800,etc.;Fig.5D) and all genes encoding SnRK2 (TraesCS2A02G493800,TraesCS6A02G132700,TraesCS6B02G160900,etc.;Fig.5C) were upregulated after LTT.Some GO items were associated with ABA,including response to ABA (GO:0009737),ABA catabolic process(GO:0046345),ABA-activated signaling pathway (GO:0009738),and ABA metabolic process (GO:0 009687;Fig.5E).Among these GO items,response to ABA pathway enriched the largest number of DEGs(Fig.5E).ABA content at 0,6,and 12 DAT was significantly increased by LTT (Fig.5F).
Analysis of GO functional enrichment of the DEGs showed 829 GO entries,three of which were associated with cold response:cold acclimation (GO:0009631),response to cold (GO:0009409),and cellular response to cold (GO:0 070417;Fig.6A);respectively 27,44,and four DEGs were enriched in these three GO entries.Cluster analysis indicated that almost all were upregulated after LTT (Fig.6B).Four DEGs:NAC48 (TraesCS3B02G439600),NAC48(TraesCS3A02G406000),MPK5 (TraesCS4D02G198600),and MPK5(TraesCS4A02G106400),were enriched in GO entries associated with the morphological development of the wheat spike,including plant organ development (GO:0099402),inflorescence development (GO:0010229),plant ovule development (GO:0048481),reproductive structure development (GO:0048608),reproductive shoot system development (GO:0090567),and pollen development (GO:0 009555;Fig.6C).The four DEGs encoded NAC transcription factors (TraesCS3B02G439600 and TraesCS3A02G4060 00) and mitogen-activated protein kinase (TraesCS4D02G198600 and TraesCS4A02G106400),and their expression levels were upregulated after LTT.
Fig.3.Microstructure of transverse section of an anther within a wheat spikelet under low temperature in early spring.(A,B)Normal temperature treatment at 12 DAT.(C,D)Low temperature in early spring at 12 DAT.(E–H)First floret at the base of spikelet.(E,F)Normal temperature treatment at 21 DAT.(G,H)Low temperature in early spring at 21 DAT.(I–L)Second floret at the base of spikelet.(I,J)Normal temperature treatment at 21 DAT.(K,L)Low temperature in early spring at 21 DAT.(M)Profile of a spike and a spikelet of wheat;figure shows the position of transverse section and the first and second floret in a spikelet.An,anther;CVB,connective vascular bundle;Ep,epidermis;FL,fibrous layer;FF,first floret;ML,middle cell;MMC,microspore mother cell;SF,second floret;SSC,secondary sporogenous cell;Ta,tapetum.
Six GO items possibly associated with wheat spike development were selected from the DEGs (Fig.6D).These GO items were enriched for 19 DEGs including GLO15 (TraesCS2A02G003200),ERF4 (TraesCS3A02G328000),ADO3 (TraesCS4A02G164000),GEML5(TraesCS5A02G488500),and WOX6(TraesCS2D02G100200,TraesCS2B02G117900,and TraesCS2A02G100700),which encodes very-long-chain aldehyde decarbonylase GL1-5,ethyleneresponsive transcription factor 4,adagio-like protein 3,GEM-like protein 5,and WUSCHEL-related homeobox 6,respectively.Six DEGs encoded transcription factors,including ERF4,WOX6,ERF78(TraesCS3B02G357500) and MADS6 (TraesCS6D02G240200),which encode ethylene-responsive transcription factor 4,WUSCHEL-related homeobox 6,ethylene-responsive transcription factor 4,and MADS-box transcription factor 6,respectively.The anatomical structure formation involved in morphogenesis (GO:0048646) was enriched with nine DEGs,while stamen formation(GO:0048455) and floral whorl structural organization (GO:0048459) was enriched with only one DEG(TraesCS6D02G240200).These genes participate in biological processes such as anatomical structure development (GO:0048856)and anatomical structure morphogenesis (GO:0 009653;Fig.S1),in wheat spike development.
Thus,LTT induced the differential expression of genes in wheat spikes,including genes responding to cold adaptation,genes encoding transcription factors,and genes involved in biosynthesis,metabolism,and signal transduction of abscisic acid that were directly or indirectly involved in plant organ morphogenesis.
The eight DEGs PP2C,CRPM1,MPK5,NAC48,WOX6,GEML5,KING1,and MADS6) were selected for transcription validation by qRT-PCR (Fig.7).Their expression abundances were consistent with those from transcriptome sequencing,which showed seven upregulated (Fig.7A–N) and one downregulated (Fig.7O,P) DEG.
Fig.4.Gene expression profile and functional analysis of cold acclimation proteins.(A)Number of differentially expressed genes after low temperature treatment.(B)Cluster analysis of 33 genes encoding cold acclimation and cold shock proteins.(C)Chromosomal locations of 33 genes.(D)GO annotation of 33 genes.(E)GO functional enrichment of 33 genes.(F)Profileofgenesetenrichmentanalysis ofresponsetoabscisicacid(GO:0009737).The curve intheupperpartofthefigurerepresents thedynamicvalueofenrichmentscore,andthehighestpointrepresentstheenrichmentscorevalueofthisgeneset.Theblackverticallineinthemiddleofthefigurerepresentsthegenenumberinthegeneset(GO:0009737).The lower curve of the figure shows the sorting value,where the negative value indicates that the genes are associated with the low temperature treatment.
Fig.5.Clustering of genes involved in ABA signal transduction.(A)Gene encoding ABA-inducible protein kinase,ABA 8′-hydroxylase 3 and ABA-insensitive 5-like protein.(B)Gene encoding protein phosphatase 2C.(C) Gene encoding the SNF1-related protein kinase.(D) Gene encoding bZIP transcription factor.(E) Chordal diagram of some GO terms involved in synthesis,metabolism,and signal transduction of ABA.(F) Time course of ABA content after low temperature treatment.
A well-differentiated wheat spike consists of lateral spikelets,rachis,and terminal spikelets (Fig.S2).A single wheat spike contains only one terminal spikelet but has multiple lateral spikelets,on which multiple caryopses develop,determining grain yield(Fig.S2).Previous studies [2,5,11,17,35] have confirmed that low temperature reduces the number of kernels per spike and the grain yield of wheat,and our findings are consistent with these.In a previous study[36],low temperature during pistil and stamen differentiation stage caused a high proportion of empty glumes and finally reduced the number of kernels per spike.With respect to the mechanism whereby low temperature decreases the yield of wheat,researchers [3,37,38] have proposed that low temperature,especially temperatures below 0 °C,causes extensive changes to cellular carbohydrates and lipids,cell viability,and uptake ability of ions or nutrients;impairment of photosynthesis;morphological and anatomical changes such as to epidermis and cell wall thickness;and changes in gene expression;and ultimately affects the growth of the wheat spike.In comparison with the NTT,the rate of spike growth was significantly reduced by LTT,as reflected by plant height,spike length,and differentiation of lateral and terminal spikelets.Each lateral spikelet is composed of several florets,which grow on the spikelet axis in an interactive manner.At 0 DAT,the first floret in the NTT group differentiated the whorls of the flower,including the glume,lemma,stamen,and pistil.However,the LTT group differentiated only the glume and lemma.This result indicated that low temperature in early spring can slow the development of the floret at the base of the spikelet and eventually reduce spike length and plant height.Similarly,Valluru et al.[39]reported that low temperature reduced the size and biomass of wheat spikes.The numbers of kernels and spikelets per spike are associated with the development of the terminal spikelet.The terminal spikelet is located at the top of the rachis,with late differentiation and few fertile florets,and usually contains one or two fertile florets (Fig.S2).In a previous study [40],the floret primordium at the middle lateral spikelet was found to appear with the terminal spikelets.During wheat spike differentiation,once the terminal spikelet appears,the number of lateral spikelets no longer increases [18].Ledent and Moss [41] pointed out that the appearance of terminal spikelets is affected by multiple factors including temperature,light,and nutrition.In the present study,low temperature in early spring triggered the early appearance of terminal spikelets,so that the number of lateral spikelets ceased to increase,possibly accounting for the reduction by low temperature of the number of lateral spikelets and kernel number per spike.
Fig.6.Expression and GO functional enrichment of genes differentially expressed under low temperature.(A) GO terms associated with cold.(B) Cluster analysis of genes enriched in cold-associated GO terms.(C) Chordal diagram of GO terms from DEGs in Fig.6B.(D) Chordal diagram of GO terms associated with plant organ development among all DEGs.
In addition to spike development,flower organ development was affected by low temperature.Low temperature in early spring delayed anther development,including that of the four layers:epidermis,fibrous layer,middle cell,and tapetum,and the microspore mother cell.Demotes-Mainard et al.[36]found that low temperature (–2 °C for 1 h for 3 consecutive days) exerted no significant effect on meiosis on microspore mother cells,a finding inconsistent with ours.However,Bennett et al.[42] showed that meiotic duration and pollen maturation time increase with decreasing temperature.We also found that the degradation of the tapetum became slower after LTT.In rice,the degree of tapetal degeneration was correlated with cold-temperature tolerance and was important for the accurate development of pollen[43].In several studies[44–47],low temperature during spike development increased spikelet sterility.Our finding that low temperature in early spring reduced kernel number per spikelet and that this decline was closely associated with floret sterility is in accord with these reports.
DEGs involved in wheat spike development under low temperature in early spring were identified.These DEGs included mainly genes encoding cold-responsive proteins,proteins involved in ABA signal transduction,and transcription factors.A process(Fig.8) by which low temperature delays wheat spike development by affecting the differential expression of genes is proposed on the basis of the results.
Fig.7.Verification of relative expression level of DEGs.(A,B) PP2C,encoding protein phosphatase 2C 37.(C,D) CRPM1,encoding cold acclimation protein WCOR413.(E,F)MPK5,encoding mitogen-activated protein kinase 5.(G,H)NAC48,encoding NAC transcription factor 6B.(H,I)WOX6,encoding WUSCHEL-related homeobox 6.(J,K)GEML5,encoding GEM-like protein 5.(L,M) KING1,encoding SNF1-related protein kinase regulatory subunit gamma-1.(O,P) MADS6,encoding MADS-box transcription factor 6.Relative gene expression level was validated from qRT-PCR and transcripts per million (TPM) value was obtained from transcriptome sequencing.Asterisks indicate significant TPM differences between the LTT and NTT groups.Each value is the mean of three biological replicates.LTT,low temperature treatment;NTT,normal temperature treatment.
LTT induced the upregulation of genes encoding cold-responsive proteins,which included cold-response protein,cold-induced protein,cold shock protein,and late embryogenesis-abundant (LEA) protein,which could increase plant cold tolerance [48,49].In the present study,these coldresponsive proteins not only participated in response to ABA but also regulated the biosynthesis and metabolism of ABA and ultimately increased ABA content,suggesting an interactive relationship between them.This conjecture is supported by previous studies.Tsuda et al.[50] and Sasaki et al.[51] reported that some genes encoding cold-responsive proteins,such as LEA,were induced by low temperature and ABA.In the ABA signaling pathway,ABA was captured by ABA receptors PYR/PYL/RCAR,inhibiting the activity of PP2Cs,which inactivated SnRK2s by direct dephosphorylation [52].In the present study,the genes encoding PP2Cs and SnRK2s were all upregulated under low temperature as in a previous study[53].Most ABA response genes contain a conserved ABA response element (AREB) in the promoter region,and the proteins interacting with this element are AREB binding or ABFs,which belong to type A bZIP transcription factors [54,55].In the present study,of the 15 differentially expressed bZIP transcription factor genes detected,most were upregulated after LTT.According to Finkelstein et al.[56],the gene ABA INSENSITIVE 5 (ABI5) that encodes ABA-regulated transcription factor acts in seedling development and seed maturation.In the present study,the expression of ABI5 decreased under the condition of delayed spike development caused by low temperature,a finding possibly accounted for by the difference between wheat and Arabidopsis in its mechanism of function.The ABA content in wheat spike increased under different low-temperature conditions [57],in agreement with the present study.We speculate that low temperature in early spring delays wheat spike development by inducing the differential expression of genes involved in synthesis,metabolism,and signal transduction of ABA and increasing ABA content.
Fig.8.Proposed process by which low temperature in early spring affects wheat spike development.Low temperature induces the upregulation of genes encoding cold response protein,which are involved in response to cold,ABA,stress and temperature stimuli and in regulation of ABA biosynthesis,isoprenoid metabolic process,lipid biosynthetic process,and other processes.In the ABA signal transduction pathway,genes including PP2Cs,SnRK2s,and bZIP are up-regulated after low-temperature treatment.Under low temperature,some genes are upregulated in response to cold acclimation.MPK5 and NAC48,which encode respectively mitogen-activated protein kinase and NAC transcription factor,are involved in plant organ,inflorescence,and ovule development.Low temperature also induces the differential expression of transcription factor genes such as MADS6, ERF4, ERF78,and WOX6,affecting plant organ formation,stamen formation,floral whorl structural organization,and other processes.
Many other genes differentially expressed in the wheat spike to adapt to low-temperature conditions,including those encoding transcription factors,were directly involved in the morphological development of plant organs.The differentially expressed transcription factor genes included five upregulated genes:ERF4,ERF78,WRK71,NAC48,and WOX6,and one downregulated gene,MADS6,and encoded ethylene-responsive transcription factor 4,WRKY transcription factor WRKY71,NAC transcription factor 6B,MADS-box transcription factor 6,and WUSCHEL-related homeobox 6,respectively.These transcription factors are involved in plant organ development,inflorescence development,plant ovule development,stamen formation,and floral whorl structural organization.These processes probably participate in the development of spike tissues such as terminal spikelet,lateral spikelet,anther,and pistil,and ultimately affect wheat spike development.Talanova et al.[58] reported that the expression of two WRKY transcription factor genes,Wrab17 and Wrab19,was promoted by cold exposure and increased the cold resistance of plants.Geraldine et al.[59] reported that low-temperature stress during the early vegetative stage in rice reduced the dry weight,length,and diameter of roots and induced the high expression of many genes,including ethylene-responsive transcription factor.These results suggested that the upregulated expression of ethylene-responsive transcription factor was negatively correlated with plant size and biomass,which was similar to the results of the present study.A single amino-acid substitution in the DNA-binding domain of COM2 that encodes ethylene-responsive transcription factor generates ‘‘miracle wheat”,whose spikelets are replaced by lateral branch-like structures [60].Derbyshire et al.[61] reported that MOS1,which disrupts the expression of an AP2 transcription factor,is involved in the formation of terminal spikelets in wheat.WUSCHEL transcription factor could be involved in the auxinmediated differentiation of the apical meristem in plant [62].In accord with this finding,the high expression of WUSCHEL transcription factor gene WOX6 may be involved in the differentiation of the terminal spikelet at the top of the wheat spike.MADS6,which encodes a type of MADS-box transcription factor,is probably involved in spike development,given that the MADS-box gene plays an important role in determining meristem identity in barley[63].In a previous study[64],MADS-box genes functioned in organ differentiation and flower development in wheat,and their development could be upregulated or downregulated.Thus,low temperature-induced downregulation of MADS6 participates in the delay of spike development.
In addition to transcription factor genes,some nontranscription factor genes including MPK5,GEML5,and GLO15,which encode mitogen-activated protein kinase (MAPK) 5,GEMlike protein 5,and very-long-chain aldehyde decarbonylase GL1-5,respectively,were markedly downregulated by low temperature.GO analysis showed that these genes were involved in plant organ formation,stamen formation,and floral whorl structural organization during wheat spike development(Fig.6).MAPK genes in other plants,such as OSMAP1 in rice[65]and ZmMPK17 in maize[66],were reportedly involved in the low-temperature response or signal pathway.These reports can support the conclusion of the present study that the upregulation of MPK5 is involved in spike development under low temperature.Aohara et al.[67] reported that the germin-like protein gene PtXSP24 was associated with various stresses,had oxalate oxidase and/or superoxide dismutase activities in the cell wall,and was involved in tolerance to nonfreezing low temperature.Thus,these differentially expressed transcription factor genes and non-transcription factor genes are involved in wheat organ morphogenesis and ultimately change the development process of the wheat spike.
Low temperature in early spring led to differential expression of genes in the wheat spike,including cold-response protein genes,ABA signaling pathway-associated genes,and morphological development-associated genes.The differential expression of these genes caused the response of spike development to low temperature stress,and changed the some metabolic processes;led to delayed development of the wheat spike expressed as a decrease in spike length,delay of flower organ development,and early emergence of the terminal spikelet;and finally reduced the kernel number per spike of wheat.
CRediT authorship contribution statement
Xurun Yu:Conceptualization,Writing– original draft,Funding acquisition.Yufei Jiang:Software.Huihui Yao:Validation.Liping Ran:Data curation.Yong Zang:Investigation,Methodology.Fei Xiong:Supervision.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This study was supported by the Natural Science Foundation of China (31801269,31971810),the Natural Science Foundation of Jiangsu Province (BK20180275),the China Postdoctoral Science Foundation (2018M642332),and A Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD).
Appendix A.Supplementary data
Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2021.08.008.