Dong Luo,Xi Zhng,Jie Liu,Yuguo Wu,Qing Zhou,Longf Fng,Zhipeng Liu,*

a State Key Laboratory of Grassland Agro-ecosystems,Key Laboratory of Grassland Livestock Industry Innovation,Ministry of Agriculture and Rural Affairs,College of Pastoral Agriculture Science and Technology,Lanzhou University,Lanzhou 730000,Gansu,China

b Key Laboratory of Western China’s Environmental Systems(Ministry of Education),College of Earth and Environmental Sciences,Lanzhou University,Lanzhou 730000,Gansu,China

Keywords:Medicago sativa L.MsDIUP1 Drought stress Antioxidant defense Osmotic adjustment

ABSTRACT Alfalfa is the most widely cultivated perennial legume forage crop worldwide.Drought is one of the major environmental factors influencing alfalfa productivity.However,the molecular mechanisms underlying alfalfa responses to drought stress are still largely unknown.This study identified a drought-inducible gene of unknown function,designated as Medicago sativa DROUGHT-INDUCED UNKNOWN PROTEIN 1(MsDIUP1).MsDIUP1 was localized to the nucleus,chloroplast,and plasma membranes.Overexpression of MsDIUP1 in Arabidopsis resulted in increased tolerance to drought,with higher seed germination,root length,fresh weight,and survival rate than in wild-type(WT)plants.Consistently,analysis of MsDIUP1 over-expression(OE)alfalfa plants revealed that MsDIUP1 also increased tolerance to drought stress,accompanied by physiological changes including reduced malondialdehyde(MDA)content and increased osmoprotectants accumulation(free proline and soluble sugar),relative to the WT.In contrast,disruption of MsDIUP1 expression by RNA interference(RNAi)in alfalfa resulted in a droughthypersensitive phenotype,with a lower chlorophyll content,higher MDA content,and less osmoprotectants accumulation than that of the WT.Transcript profiling of alfalfa WT,OE,and RNAi plants during drought stress showed differential responses for genes involved in stress signaling,antioxidant defense,and osmotic adjustment.Taken together,these results reveal a positive role for MsDIUP1 in regulating drought tolerance.

1.Introduction

Plant growth and productivity are severely impaired by various environmental factors during their life cycles,with drought stress the main factor preventing plants from reaching their full genetic potential and limiting crop yields worldwide[1].Drought stress profoundly impairs leaf gas exchange properties,nutrient uptake,photosynthesis,cell membrane stability,and osmotic homeostasis,owing mainly to osmotic stress–imposed constraints on plant processes[2,3].Upon stress-signal perception,plants deploy complex defensive systems at physiological and biochemical levels,including maintaining the integrity of the cell membrane,regulating water balance,scavenging reactive oxygen species(ROS),accumulating osmoprotectants(e.g.,free proline and soluble sugar),and growth regulation[4].The physiological responses of plants acclimating to adverse environments are all initiated at a molecular level upon the activation of a wide range of stressinducible genes,encoding receptors,signaling molecules,transporters,transcription factors,enzymes,molecular chaperones,and osmoprotectants,as well as proteins of unknown function[5].These genes form cascades of defense networks within signaling pathways,such as abscisic acid(ABA)-independent and ABAdependent pathways,for tolerating or combating drought stress[6].

Recently[7],increasing evidence has shown that many genes of unknown function operate in abiotic stress-response signaling,or general acclimation mechanisms.In Arabidopsis,41 unknown genes that respond to oxidative stress were constitutively expressed in transgenic plants,with more than 70% of the expressed unknown proteins conferring tolerance to oxidative stress and about 50% of the expressed unknown genes rendering plants more susceptible to osmotic or salinity stress[8].When 1007 unknown genes in Arabidopsis were chosen to test the tolerance of their corresponding homozygous T-DNA insertional mutants to multiple abiotic stresses,more than 80%of the mutants showed tolerance or sensitivity to more than one abiotic stress treatment and links were found between the acclimation to osmotic,salinity,oxidative,hypoxia,and ABA stresses[9].In view of the extreme diversity in their sequences and structures,a great variety of novel defense mechanisms can be expected in this uncharacterized group of genes.Once the functions of unknown genes are revealed,it is likely that our current view of the defense networks of plants’responses to drought stress will undergo a dramatic change.

Alfalfa(Medicago sativa L.)is the most extensively cultivated perennial forage legume crop,with more than 40 million hectares worldwide[10,11].With its high nutritional value and wide adaptation,this species is often called‘‘the queen of forages”and is used as hay,silage,and pasture for ruminants and dairy production[12].In China,alfalfa plantation areas are distributed mainly in arid and semi-arid areas,including the north,northwest,and northeast regions[13].The increasing desertification of the soil in these areas is severely reducing the quality and yield of the alfalfa,causing high production losses[11,14].Therefore,understanding the molecular mechanisms of alfalfa in responses to drought stress is critical for breeding of drought-tolerant cultivars.Recently,several potential drought-responsive alfalfa genes have been studied in alfalfa itself using genetic techniques.Overexpression of the genes encoding microRNA MsmiR156[15],methyltransferase MsTMT[16],squamosa promoter binding protein MsSPL9[17],or transcription factor MsWRKY11[18]have been shown to strongly increase the drought tolerance of transgenic alfalfa,illustrating the key roles of these genes in drought-stress regulatory networks.However,most of these studies have focused on known genes that their corresponding homologous genes have been characterized or annotated in other model plant species,while the mechanisms of action of the genes of unknown function(about 24% of the alfalfa genome lacking functional assignments[19])remain largely unknown.

In this study,we identified a Medicago sativa DROUGHTINDUCED UNKNOWN PROTEIN 1(MsDIUP1).To elucidate its biological role,the MsDIUP1 gene was overexpressed in Arabidopsis and its drought-tolerance characteristics were investigated.MsDIUP1 was then transformed into alfalfa by overexpressing and RNA interfering and the drought-tolerant phenotype,physiological change,and transcriptional response of transgenic alfalfa plants under drought stress were systematically studied.The objective of this study was to characterize the function of unknown genes and generate a superior drought-tolerant alfalfa cultivar.

2.Materials and methods

2.1.Plant materials and stress treatments

The alfalfa cv.Zhongmu 1 was used for gene expression analysis.To study the expression pattern of MsDIUP1 in multiple tissues,the roots,stems,leaves,petiole,flowers,and seeds of alfalfa grown under normal conditions in a greenhouse were collected.Three biological replicates were included for each tissue.To study the expression pattern of MsDIUP1 under stresses,the alfalfa seedlings were grown in 1/2 MS solution(pH=5.8),under the following conditions:22 °C,16 h light/8 h dark,a flux density of 180 μmol m-2s-1,and a relative humidity of 80%.During seedling growth,the nutrient solution was changed every-two days.Twelve-day-old alfalfa seedlings were separated into two groups:a multipleconcentrations group and a multiple-time points group.For the first group,the seedlings were exposed for 24 h to 1/2 MS nutrient solution containing various concentrations of mannitol(0,100,200,300,400,and 500 mmol L-1),NaCl(0,50,100,150,200,and 250 mmol L-1),or ABA(0,20,40,60,80,and 100 μmol L-1).For the second group,the seedlings were then treated with 1/2 MS solution containing 500 mmol L-1mannitol,150 mmol L-1NaCl,or 80 μmol L-1ABA for various exposure times(0,1,3,6,12,and 24 h).Three biological replicates were included for each treatment.All samples were immediately frozen in liquid nitrogen for RNA preparation.

2.2.Quantitative reverse transcription PCR(qRT-PCR)analysis

Total RNA was extracted from the samples with a TRIzol Kit(Sangon Biotech,Shanghai,China)following the manufacturer’s instructions.First-strand cDNA was synthesized using a FastQuant RT Kit(with gDNase)(Tiangen Biotech,Beijing,China).Genespecific primers for qRT-PCR were designed with DNAMAN software(Lynnon BioSoft,Vandreuil,Quebec,Canada)and are listed in Table S1.qRT-PCR was performed using 2xSG Fast qPCR Master Mix(Sangon Biotech)on a 7500 Fast Real-Time PCR system(Applied Biosystems,Foster City,CA,USA),and the procedure was as follows:95 °C for 30 s,40 cycles of 95 °C for 5 s,and 60 °C for 30 s.Three technical replicates were performed for each sample.The expression levels of each gene were normalized to those of ACTIN,and relative gene expression levels were calculated by the 2-ΔΔCTmethod[20].

2.3.Sequence analysis of MsDIUP1

The conserved residue analysis was performed using SMART database(https://smart.embl-heidelberg.de/)and Pfam database(https://pfam.xfam.org/).The homologous protein sequences of MsDIUP1 were retrieved in the publicly available database:National Center for Biotechnology Information(NCBI)(https://www.ncbi.nlm.nih.gov/).The sequence alignment was performed using ClustalX 2.1 software(https://www.clustal.org/clustal2/).The phylogenetic tree was generated by the neighbor-joining method,using the MEGA 6.06 software(https://www.megasoftware.net/).

The subcellular localization of MsDIUP1 was predicted using BaCelLo database(https://gpcr2.biocomp.unibo.it/bacello/pred.htm)and Plant-mPLoc database(https://www.csbio.sjtu.edu.cn/bioinf/plant/).

2.4.Vector construction and plant transformation

For a promoter-β-glucuronidase(GUS)construct,the recombinant pBGWSF7.0-ProMsDIUP1:GUS was generated by amplifying the 3219 bp 5′-flanking DNA of the MsDIUP1 coding region,and cloned into the pBGWSF7.0 vector for sequence confirmation.The Agrobacterium tumefaciens EHA105 strain carrying the construct pBGWSF7.0-ProMsDIUP1:GUS was transformed into the Arabidopsis ecotype Columbia-0 via floral dip transformation[21].

For green fluorescent protein(GFP)and MsDIUP1 co-expression construct,the coding region of MsDIUP1 was amplified by PCR and introduced into the XhoI and SalI restriction sites of the pBI121 vector to yield a pBI121-Pro35S:MsDIUP1:GFP construct.The pBI121 empty vector was used as control.The A.tumefaciens EHA105 strain carrying the fused construct pBI121-Pro35S:MsDIUP1:GFP and control vector pBI121-Pro35S:GFP were co-injected with 0,50,150,or 300 mmol L-1mannitol into tobacco(Nicotiana benthamiana)leaves for 48 h.

For over-expression(OE)and RNA interference(RNAi)constructs, recombinant pEarleyGate100-Pro35S:MsDIUP1 and

pANDA35HK-Pro35S:MsDIUP1 plasmid constructs were generated using respectively the full-length coding sequence of MsDIUP1 and a partial region of coding sequence of MsDIUP1 amplified from alfalfa Regen SY-4D(a high-efficiency genetic transformation genotype).The A.tumefaciens EHA105 strain carrying the construct

pEarleyGate100-Pro35S:MsDIUP1 was used to transform Arabidopsis

as described above.The A.tumefaciens EHA105 strain carrying the construct pEarleyGate100-Pro35S:MsDIUP1 or pANDA35HK-Pro35S:MsDIUP1 was separately used to generate stably transformed seedlings of alfalfa Regen SY-4D as previously described[22,23]with some modifications.

2.5.Promoter activity and subcellular localization analysis

For GUS assay of MsDIUP1 promoter in tissues,nine tissues from transgenic Arabidopsis expressing pBGWSF7.0-ProMsDIUP1:GUS at multiple developmental stages were collected.These tissues included root,seedling,rosette leaf,cauline leaf,petiole,stem,flower,silique,and seed.For the GUS assay of MsDIUP1 promoter under stress conditions,transgenic Arabidopsis seeds were germinated and grown on MS medium under control conditions(16 h light and 8 h dark at 22 °C)for three weeks.The seedlings were then transferred to vertically oriented plates containing MS medium with 400 mmol L-1mannitol for 0,6,24,and 48 h.The samples were submerged in X-Gluc staining solution(0.5 mmol L-1X-Gluc,0.5 mmol L-1K3Fe(CN)6,0.5 mmol L-1K4Fe(CN)6,0.2%Triton X-100,20%(v/v)methanol,50 mmol L-1phosphate buffer,pH=7.0)under vacuum for 1 h and then incubated at 37 °C for 24 h.After removal of X-Gluc staining solution,the stained samples were incubated in 75%ethanol until chlorophyll removal and then incubated in ethanol:acetic acid(1:1)solution for 4 h to improve clearing.The images were captured using a Zeiss SteREO Discovery V20 stereo microscope(Carl Zeiss AG,Oberkochen,Germany).

For subcellular localization assay of MsDIUP1,fluorescence signals in the transgenic tobacco leaves were detected with a Leica SP8 confocal laser scanning microscope(Leica Microsystem,Heidelberg,Germany)with selective settings for GFP(excitation with a 488 nm laser and emission with a 505–540 nm band pass filter)and chloroplast autofluorescence(excitation with a 552 nm laser and emission with a 620–700 nm band pass filter).

2.6.Evaluation of drought stress tolerance of overexpressing Arabidopsis

To investigate the effects of drought stress on seed germination,seeds of Arabidopsis WT and OE lines were sterilized and germinated on MS plates supplemented with 0 or 300 mmol L-1of mannitol for 10 days.Germination was indicated by the emergence of a 1-mm radicle tip through the endosperm.

To investigate the effects of drought stress on root growth of young seedlings,seeds of Arabidopsis WT and OE lines were grown on MS medium under control conditions.After seven days,the seedlings were transferred to vertically oriented plates containing MS medium with 0 or 300 mmol L-1of mannitol.On day 12 after plant transfer,root length elongation and fresh weight were recorded.

For drought stress assays of the survivability of plants,seeds of Arabidopsis WT and OE lines were germinated on MS medium for seven days and then transferred into soil and grown with sufficient watering for two weeks.The plants were then subjected to drought treatment by withholding irrigation.When WT plants exhibited the lethal effects of dehydration,watering was resumed and the plants were allowed to grow for four additional days.All plants were grown under 16 h light and 8 h dark at 22°C.All assays were repeated at least three times.Photographs were taken before and after stress treatment and the survival rate was calculated.

2.7.Evaluation of drought tolerance of alfalfa in the greenhouse

T0transgenic alfalfa lines and WT plants were propagated from stem cuttings as described previously[24].Two-week-old alfalfa seedlings were transferred into plastic culture pots containing vermiculite under a 16-h photoperiod(with a light intensity of～800 μmol m-2s-1)at 26±2°C and 60%±5%of relative humidity(RH),and watered with 1/2 MS solution every-two days to field water capacity.For drought-stress experiment,two-month-old alfalfa WT,OE,and RNAi plants were grown in separate pots and watering was withheld until the WT plants showed wilting.For the comparative drought stress experiment,one-month-old alfalfa plants were divided into two sets:1)WT and OE40 lines;and 2)WT and RNAi1 lines,with two plants for each line and the plants of each set grown in the same pot.Watering was withheld from all plants until more than half of the plants showed wilting,after which all plants were re-watered to field water capacity for eight additional days.Three independent experiments were performed.Photographs were taken before and after stress treatment and the survival rate was calculated.The leaves of two-month-old alfalfa WT and transgenic lines were harvested before and after stress treatment and used for further physiological analysis.

2.8.Determination of physiological indicators

Leaf samples from before and after stress treatment were immediately assessed for four physiological indexes,with three biological replicates for each line.Total chlorophyll,malondialdehyde(MDA),proline,and soluble sugar were measured using Comin Biochemical Test Kits(CPL-2-G,MDA-2-Y,PRO-2-Y,and KT-2-Y,respectively;Cominbio,Suzhou,Jiangsu,China;https://www.cominbio.com/)in accordance with the manufacturer’s instructions.

2.9.RNA extraction,RNA-seq,and transcriptome data analysis

Two-month-old alfalfa plants were air-dried on paper towels in a hood for 2 h at 22°C under continuous light(180 μmol m-2s-1)as drought-stress treatment.Plants grown with sufficient watering under room temperature were treated as non-stressed controls.A total of 18 leaf samples receiving three WT(three biological replicates),three OE lines(OE6,OE31,and OE40,with each genotype as a biological replicate for each other),three RNAi lines(RNAi1,RNAi7,and RNAi33,with each genotype as a biological replicate for each other)were collected from the non-stressed control and drought stress treatment groups.RNA extractions,as well as the qualitative and quantitative measurements of all 18 alfalfa samples,were performed as previously described[25].For RNA-seq,the total RNA from all 18 alfalfa samples was used to prepare separate cDNA libraries for sequencing on an Illumina Hiseq 2000 platform(Illumina,San Diego,CA,USA)[26].

Clean reads were separated from the raw reads by removal of reads containing adapters,reads containing poly-N,and lowquality reads.These clean reads were then mapped to the alfalfa reference genome sequence(https://figshare.com/projects/whole_genome_sequencing_and_assembly_of_Medicago_sativa/66380)

with Hisat2[27].All the genes were then annotated based on the following databases:NCBI non-redundant protein sequences(nr)and Gene Ontology(GO).Gene expression levels were estimated by the fragments per kilobase per million fragments mapped(FPKM)method[28].Twelve candidate genes were selected to verify gene expression by qRT-PCR analysis(for primers,see Table S1).

Based on the mean FPKM values in each treatment,differential expression between the two conditions or groups(three biological replicates per condition)was assessed with DESeq 2[29].Both the absolute value of the fold change≥1.5 and the P-value≤0.01 were used as thresholds to identify differentially expressed genes(DEGs).Expression pattern assessment for the DEGs was performed with MEV 4.9(https://sourceforge.net/projects/mevtm4/files/mev-tm4/)using hierarchical clustering.For functional annotation analyses,GO enrichment analyses for DEGs were performed with Omicshare(https://www.omicshare.com/tools/Home/Soft/gogseasenior).GO terms were identified as overrepresented based on a Q-value＜0.05.

3.Results

3.1.Sequence analysis of MsDIUP1

In this study,large-scale genes involved in drought response were screened from model plant M.truncatula microarray data(https://mtgea.noble.org/v3/),and a gene of unknown function that was dramatically induced by drought stress was identified,Medtr5g020060(Fig.S1).This gene showed high sequence identity(97.9%)to its homolog from alfalfa,MSAD_213496.t1(https://www.alfalfatoolbox.org/),which was also highly induced by mannitolstimulated drought stress,and was thus designated as MsDIUP1.

The putative MsDIUP1 gene was 1274 bp long,consisting of three exons and two introns,with an open reading frame of 381 bp,and encoded a predicted protein of 126 amino acids(Fig.S2).Its predicted molecular mass and isoelectric point were 13.89 kDa and 7.39,respectively.Analyses of the predicted protein using Pfam and SMART programs suggested that it lacked any confidently conserved domains(E-value＜0.05).

Multiple alignment analysis revealed that the amino acid sequence of MsDIUP1 showed high similarity with predicted protein products of many model plants,including M.truncatula(97.62%)and Glycine max(68.50%),and a relative similarity with Arabidopsis thaliana(53.91%),Brassica napus(50.78%),and Oryza sativa(29.31%)(Fig.1A).All of the sequences harbored two conserved regions:I and II,which were near the 3′-terminal(Fig.1A).Consistently,determination of the evolutionary relationships among plant DIUP1 proteins revealed that MsDIUP1 has close relationships with legume plants,but relatively distant relationships with Theobroma cacao and O.sativa(Fig.1B).Because to our knowledge,all the homologous predicted proteins of MsDIUP1 have not been characterized,MsDIUP1 appeared to be a novel protein of unknown function.

Fig.1.Sequence alignment and phylogenetic study of MsDIUP1.(A)Multiple sequence alignment of amino acid sequences of MsDIUP1 with DIUP1s from other plants.Sequences were aligned with ClustalX 2.1 software.(B)A phylogenetic tree of plant stress-responsive DIUP1s,constructed with MEGA 6.06 software.

3.2.Expression pattern of MsDIUP1

The tissue expression pattern of MsDIUP1 was investigated by qRT-PCR.MsDIUP1 was expressed mainly in seeds,but was found in low amounts in roots,stems,leaves,petioles,and flowers(Fig.2A).Similarly,the histochemical staining of transgenic Arabidopsis expressing ProMsDIUP1:GUS showed that GUS activity was clearly detected in mature seeds and weakly detected in flowers,but absent in roots,seedlings,rosette leaves,cauline leaves,petioles,stems,and siliques(Fig.2B).Notably,mature seeds from the same silique showed inconsistent GUS expression patterns,with higher expression in some seeds and lower expression in other seeds(Fig.2B).The inconsistency may have been directly caused by different GUS staining efficiencies among seeds.Alternatively,seeds from a silique initiated their maturation autonomously and showed slight developmental differences during the maturation process[30],thus showing differing levels of GUS expression.

The expression patterns of MsDIUP1 under various stresses were further investigated using qRT-PCR.First,the expression level of MsDIUP1 in alfalfa under several levels of stress was measured(Fig.3A).Under mannitol treatment,the expression of MsDIUP1 increased significantly with concentration(from 100 to 500 mmol L-1)in both shoots and roots.Under NaCl treatment,the highest expression levels of MsDIUP1 transcription levels were observed at 150 mmol L-1in both shoots and roots.Under ABA treatment,the highest MsDIUP1 transcription levels were observed at 80 μmol L-1in both shoots and roots.Second,the expression level of MsDIUP1 in alfalfa at several treatment times was analyzed(Fig.3B).Under 500 mmol L-1mannitol treatment,the expression of MsDIUP1 showed a pronounced increase from 1 to 24 h in shoots,while the expression of MsDIUP1 increased remarkably,peaking at 12 h and declining slightly at 24 h in roots.Under 150 mmol L-1NaCl treatment,the highest expression level of MsDIUP1 was observed at 12 h in shoots,while the highest expression level of MsDIUP1 was observed in roots at 3 h.Under 80 μmol L-1ABA treatment,the highest expression level of MsDIUP1 was observed at 12 h in both shoots and roots.In agreement with the qRT-PCR results mentioned above,the expression of ProMsDIUP1:GUS in Arabidopsis seedlings increased markedly in both shoots and roots during the 48-h mannitol treatment(Fig.3C).

3.3.Subcellular localization of MsDIUP1

BaCelLo was used to predict the subcellular localization of the MsDIUP1 protein,suggesting that MsDIUP1 was localized to the chloroplast.Plant-mPLoc indicated that MsDIUP1 was localized to the nucleus.To determine the actual localization of the MsDIUP1 protein in cells,MsDIUP1 was fused to GFP in the vector pBI121-Pro35S:MsDIUP1:GFP and then was transiently expressed in leaf epidermal cells in tobacco leaves using agroinfiltration(Fig.S3).The MsDIUP1-GFP signal overlapped exclusively with the chloroplast autofluorescence signal(Fig.4)and was also detected in the nucleus and plasma membrane(Fig.4).These results were consistent with the localizations predicted by BaCelLo and Plant-mPLoc and suggest that MsDIUP1 functions in these locations.

To determine whether drought stress was able to translocate MsDIUP1 within cells,we examined the movement of the MsDIUP1-GFP signal in tobacco leaves upon treatment with 50,150,or 300 mmol L-1mannitol(Fig.S3).Consistent with the observation under normal conditions,the MsDIUP1-GFP signal could also be detected in the nucleus,chloroplast,and plasma membrane under various mannitol treatments(Fig.4).Interestingly,the MsDIUP1-GFP signal accumulated markedly at some places near the plasma membrane under higher concentrations of mannitol(150 and 300 mmol L-1mannitol)(Fig.4).At first view,this finding suggested that drought stress altered the distribution of MsDIUP1 in cells.However,similar fluorescence signal accumulation was also observed in the control GFP-alone vector upon treatment with 150 or 300 mmol L-1mannitol(Fig.S4).A possible explanation for this phenomenon is that the hyperosmotic stimulus changes cell geometry and causes severe cell shrinkage,triggering the aggregation of GFP signals in the plasmolyzed area[31,32].Thus,our results suggested that overexpression of the MsDIUP1 protein is not sufficient to initiate a specific movement under drought stress.

3.4.Overexpression of MsDIUP1 in Arabidopsis conferred tolerance to drought stress

Fig.2.Tissue expression of MsDIUP1.(A,B)Expression of MsDIUP1 in multiple tissues at the transcriptional level in alfalfa(A)and at the protein level in Arabidopsis(B).Scale bars,1 mm in(B).

Fig.3.Stress-induced assay of MsDIUP1.(A,B)Gene expression of MsDIUP1 in alfalfa under six degrees of stress(A)or at six treatment times(B).(C)Promoter-derived GUS expression of MsDIUP1 under 400 mmol L-1 mannitol treatment in Arabidopsis.Scale bar,2 mm in(C).

To identify the functional roles of MsDIUP1 in drought response,we overexpressed MsDIUP1 in the model plant Arabidopsis.PCR and qRT-PCR analysis revealed that 18 independent OE positive plants expressed the transgene with various mRNA levels(Fig.S5).A preliminary drought-tolerance test in soil involving five OE lines with high(OE11,OE13,and OE19),moderate(OE4),and low(OE6)expression levels of MsDIUP1 allowed us to identify the first four lines with clear increases in drought tolerance(Fig.S6).The three typical OE lines(OE11,OE13,and OE19)with high expression levels of MsDIUP1 were then selected for further stress tolerance evaluation.

Because the MsDIUP1 protein localized to the chloroplast,the growth status of transgenic Arabidopsis seedlings under continuous dark conditions was evaluated at the germination stage.Both Arabidopsis WT and OE lines showed a long-hypocotyl phenotype when grown in dark conditions;however,there was no marked difference in fresh weight,hypocotyl length,and root length between dark-treated Arabidopsis WT and OE lines(Fig.S7),suggesting that MsDIUP1 might not be responsible for the plant growth under dark conditions.

Mannitol was used for simulating drought stress at the germination stage.Arabidopsis seed germination assays were performed on MS medium supplemented with 0 or 300 mmol L-1of mannitol.The seed germination rates in Arabidopsis WT and OE lines did not differ markedly on regular MS medium without mannitol(Fig.5A).Germination of both WT and OE seeds was significantly inhibited when mannitol was added to the medium,but the inhibition of WT seeds was greater than that of OE seeds(Fig.5A).Only 84%of the WT seeds germinated with 300 mmol L-1of mannitol by day 10,in contrast to the 94%–100% germination of the OE seeds(Fig.5A).

The growth of transgenic Arabidopsis under mannitol treatment was evaluated at the seedling stage.Seven-day-old seedlings of Arabidopsis WT and OE lines were transferred from MS medium onto MS medium supplemented with or without mannitol(Fig.5B).There was no significant difference in fresh weight and root growth among seedlings cultured on MS medium for 12 days.However,the fresh weight and root length of the three OE lines of Arabidopsis at 300 mmol L-1of mannitol were greater than those of the WT(Fig.5C).

The responses of transgenic Arabidopsis plants to drought stress were further tested in soil.Seven-day-old seedlings of Arabidopsis WT and OE lines were transferred from MS medium into pots containing soil for two weeks.Under normal growth conditions,no OE lines showed major changes in plant architecture or growth habit(Fig.5D).When the leaves of the WT plants began to dry up after 20 days of drought treatment,the leaves of the three OE lines showed a relatively mild phenotype,with some wilting and turning purple(Fig.5D).After rehydration,the leaves of the three OE lines regained green coloration and grew better than the WT plants(Fig.5D).As a statistical result,the survival rate(87%–94%)of the three OE lines was significantly higher(P＜0.05)than that of the WT plants(33%)under drought stress(Fig.5E).

Fig.4.Subcellular localization assay of MsDIUP1 upon treatment with 0,50,150,or 300 mmol L-1 mannitol.The rows show respectively MsDIUP1-GFP fusion protein,chloroplast autofluorescence,bright field,merged images,and enlarged images.The area shown in the white dotted box is enlarged fourfold on the rightmost side.The white arrowhead indicates the chloroplast;the white arrow indicates the nucleus;the black arrowhead indicates the plasma membrane;and the black arrow indicates accumulation of MsDIUP1-GFP fusion protein.Scale bars,50 μm.

3.5.MsDIUP1 is a positive regulator of drought response in alfalfa

Given that MsDIUP1 confers tolerance to drought stress in Arabidopsis(Figs.5,S6),the regulation of MsDIUP1 in drought response was investigated in alfalfa using OE and RNAi plants.In total,47 OE and 54 RNAi independent transgenic alfalfa plants were produced.PCR and qRT-PCR screening revealed that 20 OE and 15 RNAi positive lines expressed the MsDIUP1 gene with various mRNA levels(Fig.S8).A preliminary water-loss assay for all the alfalfa WT,OE,and RNAi lines using dissected leaves showed that 65% of the lines of OE leaves retained more water than the WT leaves,whereas 80% of the leaves of RNAi lines lost more water than WT leaves during the 6-h dehydration process(Fig.S9A).Among them,three MsDIUP1-highly expressing OE lines(OE6,OE31,and OE40)showed markedly slower water loss,whereas three MsDIUP1-lowly expressing RNAi lines(RNAi1,RNAi7,and RNAi33)showed faster water loss than the WT(Fig.S9B).These six lines together with one WT were chosen for further drought evaluation via T0generation stem cuttings.

First,we evaluated the growth status of two-month-old alfalfa WT,OE,and RNAi lines in separate pots without or with drought stress.Under normal conditions,there was no significant difference between the WT,OE,and RNAi lines of alfalfa(Fig.6A).After 11 days of drought stress,three OE lines grew normally with the leaves slightly yellowing,whereas WT plants showed a rolled-leaf phenotype;in contrast,the leaves of the three RNAi lines showed severe wilting and chlorosis(Fig.6A).Second,we performed a comparison test of one-month-old alfalfa WT and OE40 lines,or WT and RNAi1 lines in the same pot under drought stress.After 39 days without watering,OE40 plants showed fewer wilting symptoms than WT plants.Eight days after the resumption of watering,more than 80% of the OE40 plants recovered growth,compared with only about 40% for WT(Fig.6B).In contrast,the RNAi1 plants showed more wilting symptoms than the WT plants without watering for 38 days,and RNAi1 plants(33%)showed a lower survival rate than WT plants(44%)on the eighth day following re-watering(Fig.6B).Although the primary root length of the RNAi1 and WT after drought stress showed no pronounced difference,the OE40 exhibited a dramatically greater primary root length than that of the WT plants under drought stress(Fig.6C).Together,these phenotypic observations strongly suggested the positive affect of MsDIUP1 on drought tolerance.

Fig.5.Overexpression of MsDIUP1 in Arabidopsis increased drought tolerance.(A)The seed germination rate of Arabidopsis WT and OE lines grown on MS medium containing 0 or 300 mmol L-1 of mannitol.(B)The phenotypes of Arabidopsis OE and WT seedlings grown on MS medium containing 0 or 300 mmol L-1 of mannitol.(C)The fresh weight and primary root length of OE and WT seedlings.(D)The phenotypes of Arabidopsis OE and WT plants during drought stress.(E)The survival rate of Arabidopsis OE and WT plants recovered from drought stress.Scale bars,1 cm in(B).*,P＜0.05;**,P＜0.01.

Severe drought stress usually results in unavoidable consequences for the structure and functioning of chloroplasts,which are reflected in changes in chlorophyll content,thus this key indicator was monitored.The total chlorophyll content in the WT,OE,and RNAi lines was dramatically reduced under drought stress.The degree of reduction of chlorophyll content in the three OE lines was compatible with that of WT,but in the three RNAi lines was more markedly reduced than in the WT under drought stress(Fig.6D).The MDA content,which is an important indicator representing the degree of cell membrane damage of droughtstressed plants,was examined.After drought stress,MDA content increased less in OE lines but more in RNAi lines than that of the WT(Fig.6E).Proline and soluble sugar can be used as osmoprotectants enabling plants to tolerate drought stress.Therefore,the contents of these two osmoprotectants were tested.After drought stress,the OE lines accumulated more free proline and soluble sugar than the WT,whereas these two osmoprotectants in the RNAi lines were accumulated less than in the WT(Fig.6F,G).These results suggest that the function of MsDIUP1 in response to drought stress helps to maintain alfalfa’s physiological balance.

3.6.Transcriptome changes in alfalfa WT,OE,and RNAi with or without drought stress

All the above findings indicate that MsDIUP1 is necessary for conferring plant tolerance to drought stress.To further reveal the potential molecular network of MsDIUP1-associated drought tolerance,the transcriptome profiles of alfalfa WT,OE,and RNAi plants under normal and drought conditions were acquired by RNA-seq.Two-month-old alfalfa plants were exposed to drought stress for 2 h.Many(about 180,000)genes,including splice alternatives,were identified in this experiment.qRT-PCR validation of 12 selected genes revealed a consistent expression pattern(Fig.S10A)and a positive correlation coefficient(R2=0.909)(Fig.S10B)between the qRT-PCR and RNA-Seq data,confirming the reliability of the transcriptome data.

Fig.6.Investigation of drought tolerance of MsDIUP1-transformed alfalfa plants.(A)Phenotypes of two-month-old alfalfa WT,OE,and RNAi plants during drought stress.(B)Phenotypes of one-month-old alfalfa WT and OE40,or WT and RNAi1 under drought stress.(C)Growth of roots of alfalfa WT and OE40,or WT and RNAi1 after drought stress.(D–G)Comparison of four physiological indexes including contents of chlorophyll(D),MDA(E),proline(F),and soluble sugar(G)among alfalfa WT,OE,and RNAi plants under drought stress.Scale bars,5 cm in(C).*,P＜0.05;**,P＜0.01.

To investigate transcriptome changes in transgenic alfalfa plants with or without drought stress,seven-two-way comparisons of gene expression profiles were made to show that gene regulation differed in each pair of RNAi/WT(RNAi relative to WT),OE/WT(OE relative to WT),RNAi-D/WT-D(drought-treated RNAi relative to drought-treated WT),OE-D/WT-D(drought-treated OE relative to drought-treated WT),RNAi-D/RNAi(drought-treated RNAi relative to RNAi),WT-D/WT(drought-treated WT relative to WT),and OE-D/OE(drought-treated OE relative to OE).Respectively 222,456,233,127,1844,2443,and 2230 DEGs were found in comparisons of RNAi/WT,OE/WT,RNAi-D/WT-D,OE-D/WT-D,RNAi-D/RNAi,WT-D/WT,and OE-D/OE(Fig.7A).Venn diagrams were constructed to display both the genotypic variation of MsDIUP1(Fig.7B)and the effects of drought treatment(Fig.7C).

To investigate the characteristics of the different responses,we performed a comparative analysis of GO category enrichment between 863 MsDIUP1-regulated and 3441 drought-regulated DEGs.Respectively 84 and 254 GO terms were identified as overrepresented based on a Q-value＜0.05(Tables S2,S3).The MsDIUP1-regulated DEGs were involved mainly in binding activity,biosynthetic process,and catabolic process,whereas the droughtregulated DEGs were associated mainly with stress signaling,kinase activity,and transporter activity(Tables S2,S3).Both DEG classes were enriched in common in 37 GO categories,which were related mainly to transferase activity,catalytic activity,oxidoreductase activity,and carbohydrate metabolism(Fig.7D).This finding suggests substantial crosstalk between MsDIUP1-and droughtinduced responses.

The expression patterns of all DEGs enriched in the significantly overrepresented functional categories were further analyzed.Among the DEGs modulated by MsDIUP1,several showed differential expression among alfalfa WT,OE,and RNAi plants without or with drought stress.These included the stress signalingassociated gene PP2C53(Protein phosphatase 2C 53),the antioxidant defense-associated gene AER(2-alkenal reductase),the osmotic adjustment-associated genes HSP81-2(Heat shock protein 81–2)and ERD14(Early response to dehydration 14,dehydrin b)(Fig.8A).Similarly,many drought-regulated DEGs showed differential induction among alfalfa WT,OE,and RNAi plants during drought stress.These included the stress signaling-associated genes including two RLKs(Receptor-like kinases),one STPK(Serine/threonineprotein kinase),one CDPK(Calcium-dependent protein kinase),and two PP2Cs(PP2C2 and PP2C63),the antioxidant defense-associated genes ZEP(Zeaxanthin epoxidase),NUDX20(Nudix hydrolase 20),PPR(Pentatricopeptide repeat-containing protein),and AER,the osmotic adjustment-associated genes P5CR(Pyrroline-5-carboxylate reductase),two PGlcTs(Plastidic glucose transporters),HSP81-2,and ERD14(Fig.8B).The majority of MsDIUP1-regulated DEGs were also changed by drought stress,and both DEGs were involved in common in the processes of stress signaling,antioxidant defense,and osmotic adjustment(Fig.8).These results suggested that MsDIUP1 functions in reprogramming the transcriptional networks of these three processes to confer the drought tolerance of alfalfa.

4.Discussion

Although genes of unknown or poorly characterized functions are not usually selected by researchers as a subject for deep analysis owing to their complexity and challenges[33],increased evidence suggests that the function of many of these proteins is important for plant tolerance to abiotic stresses and that these genes should be included in future functional studies in this field[8,9,34].In this study,we identified a putative stress-responsive gene,MsDIUP1,with an unknown functional annotation(Fig.1),which inspired an interest in identifying its role in the tolerance of abiotic stress.

Fig.8.Two-way comparisons of gene expression profiles show DEGs involved in stress signaling,antioxidant response,and osmotic adjustment processes in alfalfa.(A)Heat map of the expression levels of the MsDIUP1-regulated DEGs differing in each pair of RNAi/WT,OE/WT,RNAi-D/WT-D,and OE-D/WT-D groups.(B)Heat map of the expression levels of drought-regulated DEGs differing in each pair of RNAi-D/RNAi,WT-D/WT,and OE-D/OE groups.DEGs modulated by both MsDIUP1 and drought stress are marked in red.The red and blue colors indicate high and low expression levels,respectively.

That seed germination begins with the uptake of water by dehydrated mature seeds[35],In our study,regarding the high expression of MsDIUP1 in mature seeds(Fig.2),a commonly asked question is whether MsDIUP1 is involved in regulating seed germination.Although the seed germination rate between Arabidopsis WT and OE lines was not obviously different on regular MS medium,this indicator was less inhibited in Arabidopsis OE lines than in the WT in the presence of mannitol(Fig.5A),implying that the Arabidopsis OE seeds showed higher adaptation to the adverse environment during the germination stage.Plant roots function in plant growth by providing anchorage,uptake,storage,and transportation of minerals and water[36].Abiotic stresses impair rootsystem architecture traits such as root positioning,length,angle,branching,surface area(including root hairs),coverage,and diameter,thus impeding root growth and development[37,38].The finding that the induced expression of MsDIUP1 occurred in roots under various stress conditions at both the transcriptional and protein level(Fig.3),suggesting that MsDIUP1 may be involved in plant root growth and development under stress conditions.As expected,overexpression of MsDIUP1 in Arabidopsis dramatically increased the drought stress tolerance of seedlings,with higher root length and fresh weight than in WT seedlings in the early developmental stages(Fig.5B,C).The effects of drought are evident at all phenological stages of plant growth,whether the water deficit occurs at early or late developmental stages[39].Indeed,the Arabidopsis OE plants also displayed more drought-stress tolerance than WT at the late development stage,showing a higher survival rate after drought stress(Fig.5D,E).Similar to the observation of the morphological or agronomic traits found in Arabidopsis,overexpression of MsDIUP1 increased,while knockdown of MsDIUP1 reduced,the drought tolerance of alfalfa plants,whether at the early or late developmental stages(Fig.6A,B).Therefore,our phenotypic evidence supports the notion that MsDIUP1 functions as a positive modulator in plant response to drought stress.It should be noted that the leaf water loss of five alfalfa OE lines(OE27,OE47,OE3,OE45,and OE44,accounting for 25% of the OE lines)and three alfalfa RNAi lines(RNAi46,RNAi54,and RNAi32,accounting for 20%of the RNAi lines)showed non-significant difference with that of WT,suggesting that their drought tolerance might not have changed(Fig.S9A).This observation could possibly be explained by a slight alteration in the MsDIUP1 gene expression levels in transgenic plants[40].However,the notion is contrary to the observation that the leaf water loss of two alfalfa OE lines(OE30 and OE41,accounting for 10% of the OE lines)with highly expressed MsDIUP1 was even faster than that of WT(Fig.S9A).One possible reason is that leaf water loss is not the only factor determining MsDIUP1-mediated drought tolerance,which may also be affected by other factors such as antioxidant defense and osmotic adjustment[41].Another reason could be the differential influence of the T-DNA integration and expression of MsDIUP1 on the expression of endogenous genes among the transgenic lines[42].

It is well known[41]that water deprivation accelerates the degradation of chlorophyll pigments(chlorophylls a and b)owing to the hydrolysis of chloroplast proteins.As observed in other drought-treated species,the chlorophyll content was reduced in all alfalfa plants by water-deficit treatment.Although there was no marked difference in chlorophyll content between droughtstressed alfalfa WT and OE plants,it was lower in droughtstressed RNAi plants than in WT plants(Fig.6D),suggesting that impaired metabolism processes in chloroplast are at least partly responsible for MsDIUP1-mediated drought tolerance.Plants suffering from abiotic stresses exhibit profound adverse effects on cell membranes,meaning that lipid peroxidation is aggravated[43].MDA content has been used as a criterion for the degree of plant membrane lipid peroxidation.Our results showed that a lower MDA content was present in OE plants and a higher content in RNAi plants than in WT plants under drought stress(Fig.6E),indicating that MsDIUP1 effectively prevents lipid peroxidation in alfalfa under drought stress.To alleviate these adverse impacts,plants accumulate more compatible osmolytes,such as free proline[44]and soluble sugar[45],both of which function as osmoprotectants so that plants can tolerate stress.Under drought stress,alfalfa that overexpressed MsDIUP1 accumulated much more proline and soluble sugar than WT plants(Fig.6F,G),while alfalfa with reduced MsDIUP1 accumulated less proline and soluble sugar than WT plants,suggesting the involvement of MsDIUP1 in regulating the synthesis of osmolytes to reduce osmotic stress in alfalfa.

The physiological responses of plants acclimating to unfavorable environments are all initiated upon the activation of cascades of molecular networks involved in stress perception,signal transduction,and expression of stress-responsive genes and metabolites[11].In early stress-signaling events,RLKs,STPKs,CDPKs,and PP2Cs are central regulators of signal perception and transduction,connecting the perception of external stimuli to cellular responses[6,11].Many studies have suggested that these signal sensors and transducers act in optimizing plant responses to drought stress,including GsRLCK in soybean[46],TaSnRK2.4 in Arabidopsis[47],ZmCPK4 and ZmPP2C55 in maize[48,49].In this study,a subset of genes involved in the positive regulation of stress signaling were activated by MsDIUP1 or drought stress,including two RLKs,one STPK,one CDPK,and three PP2Cs(Fig.8).The higher expression of these genes in OE than in WT or RNAi plants without or with drought stress explains the increased drought tolerance of the alfalfa OE plants.

In plant cells,chloroplasts constitute a source of ROS,which includes singlet oxygen(1O2)at photosystem(PS)II and superoxide anion(O2-)at PS I[50].To avoid the oxidative damage of ROS and to allow their signaling activity,chloroplasts harbor antioxidant systems,both enzymatic and non-enzymatic[50].ZEP is an enzyme that controls the direction of the chloroplast photoprotective xanthophyll cycle,which is one of the most efficient mechanisms protecting plants under oxidative stress conditions[51].Overexpression of enzymes with general ZEP activity strengthened the antioxidant defense system and increased abiotic stress tolerance in various plant models[52,53].PPR proteins are a large family of RNA-binding proteins that usually carry out specific RNA processing in chloroplasts,where RNAs are processed to become functional rRNAs and mRNAs[54].The inactivation of one PPR gene,WSL,led to reduced translation efficiency in chloroplast,and the wsl mutant showed increased sensitivity to ABA,salinity,and sugars and accumulated more H2O2than WT plants[55].NUDXs are a protein family that possesses pyrophosphohydrolase activity toward a wide variety of nucleoside diphosphate derivatives including the cofactors NAD(P)H and FAD in chloroplasts[56].It has been shown[57]that a genetically defined nudix hydrolase AtNUDX19 in Arabidopsis acts as an NADPH pyrophosphohydrolase to modulate the cellular levels and redox states of pyridine nucleotides and fine-tunes photooxidative stress response.In our transcriptome study,these genes,including one ZEP,one PPR,and one NUDX20,were highly activated by MsDIUP1 or drought stress;they were more induced in alfalfa OE plants but less induced in RNAi plants than in WT plants during drought stress(Fig.8).Given that one common feature of these genes is that they are involved in oxidative stress tolerance in chloroplasts[51–57],it is reasonable to speculate that MsDIUP1 is required for regulating cellular redox homeostasis in chloroplasts,ensuring various metabolic activities in chloroplasts in response to stress.In plants,the antioxidant defense system functions not only in chloroplasts but also in most of the other subcellular compartments.AER has an NADPH-dependent oxidoreductase activity and functions in the detoxification of lipid peroxidation-derived reactive carbonyl species(RCS)and α,β-unsaturated aldehydes in cytoplasm and nucleus[58,59].Studies[58,60]have shown that overexpression of AER in tobacco and Arabidopsis can effectively reduce the cellular RCS content,increasing tolerance to oxidative damage under drought and salt stress conditions.As expected,our transcriptome analysis revealed that one AER was highly activated by MsDIUP1 or drought stress(Fig.8).The higher expression or induction of this gene in OE than in WT plants under drought stress further confirms that MsDIUP1-regulated antioxidants are widely distributed in various subcellular compartments to alleviate oxidative damage.

In addition to beneficial antioxidants,plants usually respond to drought stress by accumulating osmoprotectants such as proline and soluble sugar.P5CR functions in proline biosynthesis pathways.In higher plants,proline can be synthesized via two main pathways,one starting from glutamate and the other from ornithine;both routes converge at the generation of P5C,which in the final step is reduced to L-proline by P5CR[61].PGlcTs are required for the export of glucose from the stroma as a product of amylolytic starch degradation,and may contribute to sugar accumulation under abiotic stresses[62,63].The higher induction of respectively one P5CR and two PGlcTs in alfalfa OE plants than in WT or RNAi plants during drought stress explains the increased proline and soluble sugar accumulation of the alfalfa OE plants at the physiological level(Figs.6F,G,and 8).These results suggested that MsDIUP1 could offer an increased osmotic adjustment ability for alfalfa plants under drought stress.Protective protein accumulation is also involved in abiotic stress tolerance mechanisms such as HSPs or dehydrins that play protective roles during stress events[64].Transgenic plants overexpressing genes encoding HSPs or dehydrins showed a higher content of osmoprotectants and a marked tolerance to abiotic stresses,suggesting that these molecules improve the protection ability of plants from stress damage via osmotic adjustment[65,66].In the present study,two genes modulated by MsDIUP1 or drought stress,including one HSP81-2 and one EDR14,were more highly expressed or induced in OE than in WT plants under drought stress(Fig.8),confirming that MsDIUP1-regulated protective proteins contribute to osmotic adjustment during stress conditions.Taken together,our physiological and molecular discoveries show the positive roles of MsDIUP1 in regulating the processes of lipid peroxidation prevention,stress signaling,antioxidant defense,and osmotic adjustment,as well as,ultimately,plant drought tolerance(Fig.9).

Fig.9.Model describing how MsDIUP1 confers alfalfa drought tolerance.Red arrows represent positive regulation.Blue barred lines represent negative regulation.

5.Conclusions

We identified MsDIUP1,a novel gene of unknown function,in alfalfa and provided evidence of its positive role in drought tolerance improvement.MsDIUP1 was localized to the nucleus,chloroplast,and plasma membranes,and its expression was highly induced in response to drought,salt,and ABA treatment.Overexpression of MsDIUP1 in Arabidopsis resulted in increased tolerance to drought,with higher seed germination,root elongation,and fresh weight of seedlings,as well as a higher survival rate for plants relative to the WT.In alfalfa OE and RNAi plants,upregulation of MsDIUP1 increased,while downregulation of MsDIUP1 reduced,the contents of proline and soluble sugar and,ultimately,alfalfa tolerance to drought stress.Key genes were altered in the processes of stress signaling,antioxidant defense,and osmotic adjustment,further validating the positive role of MsDIUP1 in response to drought stress.

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.

CRediT authorship contribution statement

Dong Luo:Investigation,Data curation,Writing–original draft.Xi Zhang:Data curation,Resources.Jie Liu:Data curation,Formal analysis.Yuguo Wu:Data curation,Formal analysis.Qiang Zhou:Data curation,Formal analysis.Longfa Fang:Methodology.Zhipeng Liu:Conceptualization,Resources,Writing–review & editing,Funding acquisition,Supervision.

Acknowledgments

This research was supported by the Strategic Pilot Projects of Chinese Academy of Sciences(XDA26030103),the National Natural Science Foundation of China(31722055 and 31672476),and the Key Science and Technology Foundation of Gansu Province(19ZD2NA002).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2022.05.013.