Kexin Wng,Ynrong Liu,c,Fengkui Teng,Huifng Cen,Jinpin Yn,Shiwen Lin,Dyong Li,Wnjun Zhng,b,*

a College of Grassland Science and Technology,China Agricultural University,Beijing 100193,China

b Key Lab of Grassland Science in Beijing,China Agricultural University,Beijing 100193,China

c College of Biological Science,China Agricultural University,Beijing 100193,China

d College of Life Sciences,Shandong Normal University,Jinan 250014,Shandong,China

Keywords:Alfalfa miR156 Biomass yield Forage quality Abiotic stress

ABSTRACT Alfalfa (Medicago sativa L.) is the most widely cultivated perennial leguminous forage crop woldwide.MicroRNA156(miR156)precursor genes from dicotyledonous species are reportedly useful for improving alfalfa plant architecture and abiotic stress resistance.However,there has been no report on whether a miR156 precursor gene from a monocotyledonous species functions in alfalfa.We introduced two tandem precursor genes of miR156,rice Osa-MIR156b and Osa-MIR156c (Osa-MIR156bc),into alfalfa.The expression of miR156 in the transgenic (TG) alfalfa was significantly elevated.Compared to wild-type plants,the TG plants overexpressing miR156 had more branches and leaves and showed improved salt and drought tolerance.Overexpression of miR156 slightly reduced plant height,but the biomass yield of TG plants grown in flowerpots was still increased.Forage quality of TG plants was markedly improved by reduction of acid detergent lignin(ADL)content and increase in crude protein content.The expression of the putative miR156 target genes MsSPL6,MsSPL12,and MsSPL13 in TG plants was repressed by miR156 overexpression,and that of all tested MsSPLs would be sharply increased under drought or salt stress.RNA sequencing revealed that overexpression of miR156 affected the expression of genes associated with abiotic stress resistance and plant development in multiple pathways.This first report of overexpression of monocot miR156 precursors in alfalfa sheds light on the function of miRNA156 precursors from the monocot species rice that could be used for genetic improvement of the dicot forage crop alfalfa.

1.Introduction

Alfalfa(Medicago sativaL.)is the most widely cultivated perennial leguminous forage crops worldwide for its high yield,nutritive value and stress resistance ability.In recent decades,breeders have worked on improving alfalfa biomass yield,forage quality,and/or abiotic stress tolerance by genetic transformation,with some successes[1,2].These traits are complex and often controlled by many interacting genes[3],and side effects such as reduced plant height and changed leaf size are also observed[4].MicroRNAs,one kind of small non-coding RNA,are emerging as a powerful tool for improving crop plants by regulating the expression of targeted genes [5].Manipulating miRNA precursor gene expression promises to be a useful method for improving plant architecture and abiotic stress resistance simultaneously [6–8].

MicroRNA156 (miR156) is one of the most conserved small RNAs in plants.It plays a vital role in regulating plant architecture and abiotic stress resistance by post-transcriptionally repressing the expression of genes encoding SQUAMOSA PROMOTER BINDING PROTEIN LIKE (SPL) transcription factor [2].The MiR156-SPLregulatory pathway (miR156-MsSPLmodule) has been reported as a versatile toolbox for regulating root architecture [9],stem elongation[1,2],leaf germination[10],flowering time[1,2],panicle morphology [11],and branch number [12,13],also improving drought[14,15],salt[16]and heat stress[17]mainly by repressingMsSPL13expression.Overexpression of miR156 increased tillering in switchgrass and branching in alfalfa but reduced plant height,in a dose-dependent manner [1,2,4].Fourfold overexpression ofMsmiR156din alfalfa also increased forage quality [1],whereas a moderate increase in the miR156 level showed no effect on forage quality[2].The expression level of miR156 may strongly influence alfalfa stress resistance,biomass yield,and forage quality.

The diverse and redundant roles of miRNA precursor genes in plant morphology and development have been well documented,although the mature miRNA sequences are the same or similar[18].In rice,precursor genes of miR156 were classified into two subfamilies,and showed different functions in regulating seed dormancy and plant architecture [19].Similarly,overexpression ofAtMIR156bin tomato resulted in abnormal fruit structure [20],but the phenomenon was not observed in plants overexpressingSlymiR156a[21].However,overexpression ofAtmiR156bandSlymiR156aresulted in delayed flowering and reduced fruit number.In alfalfa,ectopic expression ofLjmiR156increased soluble sugar contents of transgenic plants,in contrast to the effect observed in theMsmiR156doverexpression lines [1,2].Obviously,different precursor genes of miR156 may also exert different effects,especially when expressed in heterologous plant species.For this reason,the functions of miR156 precursor genes should be individually verified before their use as molecular tools in plant trait modification.

In this report,to investigate whether rice miR156 precursor genes can be used as molecular tools to improve the traits of alfalfa,a dicotyledonous plant species,we characterize the function of two naturally tandem rice miR156 precursor genes,Osa-MIR156bandOsa-MIR156cin alfalfa by heterologously coexpressing.The phenotype of alfalfa transgenic plants including branching number,biomass yield,forage quality,and tolerance to drought and salt were tested in comparison to wild type plants.

2.Materials and methods

2.1.Production of Osa-MIR156bc transgenic alfalfa

A rice cDNA fragment (AK110797.1) containing two naturally tandem precursor genes,Osa-MIR156bandOsa-MIR156c(Osa-MIR156bc) (Figs.1A,S2),was cloned from ajaponicarice (Oryza sativaL.) cv.Nipponbare with the primer pair 156-XbaⅠ-F and 156-SalⅠ-R (Table S1).The secondary structure ofOsa-MIR156bcpremicroRNA was characterized at the RNAfold WebServer(http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAfold.cgi).The miRBase database (http://www.mirbase.org/) was used to predict the mature miR156 sequence produced byOsa-MIR156bandOsa-MIR156c.Osa-MIR156bandOsa-MIR156cgenes,inserted between the restriction enzyme sitesXbaI andSalI of the binary vector pZH01,were driven by the CaMV 35S promoter and linked to the hygromycin resistance gene (hpt) as a selectable marker (Fig.1B)[22].The vector was then transformed intoAgrobacterium tumefaciensstrain EHA105 and used for transformation of alfalfa cultivar Zhongmu No.1[23].Transgenic plants were cultured in flowerpots with nutrient soil in a greenhouse with natural lighting.

2.2.Identification of Osa-MIR156bc transgenic plants

Genomic DNA was extracted from leaves of wild type(WT)and putative transgenic (TG) plants using the CTAB method for polymerase chain reaction (PCR) test as previously described [24].TG plants were tested by amplifying a 1097-bp fragment that spans the CaMV 35S promoter and thehptgene using primer pair 35SF and hpt-R (Table S1).Total RNA was isolated from young leaves with Trizol reagent (Thermo Fisher Scientific,San Jose,CA,USA).One microgram of RNA was used in a reverse-transcription reaction to synthesize first-strand cDNA for RT-PCR and quantitative real-time PCR (qRT-PCR) following the kit procedure (Takara,Dalian,China).To quantify miR156 accumulation level by stemloop qRT-PCR,a miR156 stem-loop RT primer was used in the first-strand cDNA synthesis reaction (Table S1).The TB GreenPremix Ex Taq(Takara RR420) was used for qRT-PCR analysis.The reaction was performed on the EcoTM Real-Time PCR System EC-100-1001 (Illumina,Westlake Village,CA,USA) according to the manufacturer’s instructions.The relative expression level was calculated by the 2-ΔΔCTmethod [25].The primers used for qRT-PCR testing of putative target genes are listed in Table S2.The β-actingene (JQ028730.1) was used as an internal control to normalize gene expression[26].At least three biological replicates were used for statistical analyses in the experiments.

2.3.Agronomic traits and forage quality of Osa-MIR156bc transgenic alfalfa

WT and TG plants were propagated by the stem-cutting method.After growth in a greenhouse for four months,four plants of each line were selected for phenotype analyses.The primary branch number,secondary branch number,and internode length(the third internode from the top) from five tillers in each of the four plants were recorded.Leaves of the third internode of branches of WT and TG plants were photographed with a Nikon P2-DBL camera and the leaf area was estimated with NISElement BR 4.60.00 software (Nikon,Shizuoka,Japan).Aboveground materials of each plant were harvested,dried at 65 °C for 48 h,and weighed.Stems and leaves were weighed separately to calculate stem/leaf ratio.The contents of neutral detergent fiber(NDF),acid detergent fiber (ADF) and acid detergent lignin (ADL)in dry matter were measured by Van Soest herbage analysis using ANKOM A2000i semi-automatic fiber analyzer(ANKOM,Macedon,NY,USA) [6].Using the Kjeldahl method,total nitrogen content was measured with KJ2300 Kjeldahl analyzer(FOSS,Hiller?d,Denmark) and the content of crude protein was calculated based on total nitrogen content.Soluble sugar content was determined by anthrone colorimetry[27]with four biological replicates and three technical repeats.

2.4.Salt and drought treatments

WT and TG plants were cultured in polyethylene tubes(diameter=5 cm,height=25 cm) filled with sand for four weeks under a 16 h light/8 h dark photoperiod [23].Each line had four replicates.Plants were watered with 0.1× Hoagland solution containing 250 mmol L-1NaCl every other day for salt treatment.Physiological parameters including Na+concentration,K+/Na+,H2O2,leaf electrolyte leakage (leaf EL),and malondialdehyde(MDA) content were measured two weeks after salt treatment[28].Na+and K+concentrations were measured with a flame photometer[7].H2O2content was tested following the instructions for a reagent test kit A064-1-1 (Nanjing Jiancheng Bioengineering Institute,Nanjing,China).

For drought treatments,the WT plants and three TG lines were well watered to the maximum soil water capacity(100%)and then grown in the greenhouse with water withheld.When the weight of tubes had decreased to 40% (after two weeks) of the weight at maximum water capacity,physiological parameters including relative water content (RWC) and leaf EL were tested to evaluate drought tolerance.Four biological replicates were used in these tests.

2.5.Identification of putative miR156-targeted genes

Putative miR156-targeted genes of alfalfa were searched against the alfalfa transcript database (https://www.alfalfatoolbox.org) usingAtSPLsandMtSPLscDNA sequences,and miR156-targeted site was predicted with psRNATarget (http://plantgrn.noble.org/psRNATarget) [29].Sequences alignments ofMsSPLs,MtSPLs,AtSPLsandOsSPLswas performed with Clustal W[30],and a phylogenetic tree was constructed with MEGA 5.0(https://www.megasoftware.net/) [31].The amino acid sequences ofSPLswere multiply aligned with DNAMAN 8.0 software (Lynnon-Biosoft,Ontario,Canada).Expression patterns ofMsSPLswere detected by qRT-PCR,using primers listed in Table S2.

2.6.RNA sequencing and data analysis

For RNA sequencing,total RNA of WT and TG alfalfa leaves,collected from the third internodes of branches,was extracted using Trizol reagent,and quantified and tested with a NanoPhotometer spectrophotometer (IMPLEN GmbH,Munich,Germany) and RNA Nano 6000 Assay Kit (Agilent,B?blingen,Germany).Six samples collected from three WT plants and three TG1 plants (TG1,TG15 and TG4)were used to characterize the global gene expression profiles.A 1.5-μg aliquot of total RNA from each sample was used to generate sequencing libraries using NEBNext Ultra RNA Library Prep Kit for Illumina(NEB,Ipswich,MA,USA)following the manufacturer’s recommendations.The libraries were sequenced on an Illumina Hiseq 2000 platform(Shenzhen).High-quality clean reads were assembled using Trinity software version v2.0.6 package(http://trinityrnaseq.sourceforge.net).Gene functions were annotated using NCBI (www.ncbi.nlm.nih.gov) non-redundant protein sequences (NR),NCBI nucleotide sequences (NT),Gene Ontology(GO,http://www.geneontology.org),and Clusters of Orthologous Groups of proteins (COG,http://www.ncbi.nlm.nih.gov/cog/),Kyoto Encyclopaedia of Genes and Genomes (KEGG,http://www.genome.ad.jp/kegg/) and a manually annotated and reviewed protein sequence database (Swiss-Prot,https://www.expasy.org/)databases.Gene expression levels were estimated as described[32].Ten genes were selected randomly for measurement of their expression by qRT-PCR to verify the RNA sequencing results.Primers are listed in Table S3.

2.7.Statistical analyses

In the experiments,means from at least three biological replicates with at least three technical repeats were compared by one-way ANOVA with SPSS 20.0 (P<0.05).

3.Results

3.1.MiR156 is highly expressed in young leaves of alfalfa

MiR156 showed significantly higher expression in leaves,especially young leaves,than in other tissues(Fig.S1).This finding suggested that miR156 is involved in leaf development in alfalfa.

3.2.Generation and molecular verification of transgenic alfalfa plants overexpressing rice Osa-MIR156bc

Nineteen PCR-positive lines ofOsa-MIR156bc-containing alfalfa were generated (Fig.1C),of which 16 were positive by RT-PCR assay (Fig.1D).Transgenic lines 6,23,and 24 showed no expression ofOsa-MIR156bc,possibly owing to gene silencing.QRT-PCR tests showed that miR156 was significantly higher in 11 TG lines than in WT (Fig.1E).Two lines (TG1 and TG14) with moderate expression of miR156 (about four times that of WT),and two(TG15 and TG4) with higher expression of miR156 (6–8 times higher than WT) were selected for further study.

3.3.Overexpression of miR156 altered alfalfa phenotype and improved biomass yield

As shown in Fig.2A,the height of four-month-old TG plants was less than that of WT,possibly owing to the shortened internode length(Fig.2B,C,and G).However,the TG plants showed luxuriant aboveground growth (Fig.2A) and more primary branches (threeto sixfold) and secondary branches (two fold) than WT controls(Fig.2D–G).Thus,the TG plants had a significantly higher leafto-stem ratio than WT (Fig.2H),although the leaves of TG plants were smaller than those of WT (Fig.2I and J).Overall,overexpression of miR156 (OE-miR156) resulted in a significantly higher aboveground biomass yield (two-to fourfold) in alfalfa TG plants than in WT controls (Fig.2K).

3.4.OE-miR156 improved alfalfa forage quality

As shown in Table 1,there was no significant difference between WT and TG plants in soluble sugar content except for TG4.The NDF and ADF of the tested plants also showed no significant difference (Table 1),but the ADL content of TG plants (6.2%–7.7%)was significantly lower than that of WT(9.7%)(Table 1).The crude protein content of TG plants was significantly higher at 1.5(TG1 and TG14)to 3.5(TG15 and TG4)percentage points than that of WT(Table 1).These results suggested that OE-miR156 improved alfalfa forage quality improvement,especially protein content.

3.5.OE-miR156 increased alfalfa salt tolerance

Four-week-old WT and TG plants (Fig.3A) were subjected to salt stress treatment with 250 mmol L-1NaCl.After two weeks,the leaves of WT showed clear curling and chlorosis (Fig.3B),whereas the leaves of TG plants showed barely any damage symptoms (Fig.3B).The leaves of TG plants showed significantly lower Na+content than those of WT (Fig.3C),resulting in a significantly higher K+/Na+ratio in TG than in WT plants(Fig.3D).The MDA content of TG plants was also significantly lower than that in WT plants (Fig.3E).EL of leaves in TG plants was about 40% less than that in WT plants (Fig.3F).Lower hydrogen peroxide (H2O2)content was observed in TG than in WT plants (Fig.3G).Thus,OE-miR156 in alfalfa increased its salt tolerance via multiple physiological pathways.

3.6.OE-miR156 increased alfalfa drought tolerance

Fig.2.Phenotypic comparison of WT and Osa-MIR156bc transgenic alfalfa plants.(A)Four-month-old WT and transgenic lines grown in pots.(B,C)Statistical analysis of plant height(B)and internode length(C)of WT and TG plants.(D)Photograph of a primary branch containing top three internodes of WT and TG plants.(E,F)Comparison of the primary branch(E)and secondary branch(F)numbers of WT and TG plants.(G)A typical photograph of top three internodes and secondary branches of a primary branch.(H)Leaf to stem ratio of the fresh weight of WT and TG plants.(I,J)Comparison of leaf areas(I)and leaf phenotypes(J)of WT and TG plants.Bar represents 1 cm.(K)Comparison of dry matter of WT and TG plants.Values are means ± SD (n=4).Different letters represent significant difference (P <0.05).

Fig.3.OE-miR156 improved alfalfa salt tolerance.(A,B)Phenotype of WT and TG plants before(A)salt treatment and after(B)250 mmol L-1 NaCl treatment for 14 days.(C–G) Leaf Na+ content (C),K+/Na+ ratio (D),MDA content (E),electrolytic leakage (EL) (F) and H2O2 content (G) of WT and TG plants after 250 mmol L-1 NaCl treatment for 14 days.Values are means ± SD (n=4).Different letters represent significant difference (P <0.05).

Drought tolerance of WT and TG plants was evaluated by waterwithholding treatment (Fig.4A).When the tube weight was 40%(water withholding for two weeks)of the weight of the maximum water capacity,WT plants showed marked wilting and leaf chlorosis.However,the TG plants were still green without visible wilting(Fig.4B).No significant differences in leaf RWC and EL were observed between WT and TG plants under well-watered conditions,whereas the RWC of TG plants was 20–30%higher than that of WT plants after water withholding for two weeks(Fig.4C).Leaf EL increased with the extension of drought stress time both in WT and TG plants,whereas the EL of TG plants was significantly lower than that of WT plants(Fig.4D).Obviously,the results showed that drought-elicited cell membrane damage in TG plants was lower than that in WT plants.OE-miR156 increased alfalfa drought tolerance.

3.7.MiR156-target genes MsSPL12/13 respond to salt and drought stress

Ten genes harboring an SBP domain and miR156 targeted site were selected(Table S2;Fig.S3A),and their evolutionary relationships with SPLs ofArabidopsis,rice,alfalfa andMedicago truncatulawere determined (Fig.S3B).According to these relationships,the 10 alfalfaSPLgenes were namedMsSPL6a/b,MsSPL9a/b,MsSPL12a/b,MsSPL13a/b,andMsSPL16a/b.The twoMsSPLhomologs genes(a and b)are highly similar in nucleotide sequence and were amplified by the same pair of primers in qRT-PCR tests.Moreover,MsSPL9,MsSPL12,andMsSPL13share the same miR156-targeted sequence (Fig.S4).Further study showed thatMsSPL12andMsSPL13were significantly down-regulated in TG plants compared to WT,andMsSPL6in OE-miR156 lines showed lower expression than WT but had no significant difference.Expression ofMsSPL9andMsSPL16showed no significant difference from that of WT(Fig.5).These results indicate thatMsSPL6/12/13were repressed by miR156.

Fig.4.OE-miR156 increased drought tolerance of alfalfa.(A,B)Phenotype of WT and TG plants before(A)and after(B)drought treatment by water withholding for 14 days.(C,D)Leaf relative water content(RWC)(C)and leaf EL(D)of WT and TG plants were measured before and after drought treatment.Values are means±SD(n=4).Different letters represent significant difference (P <0.05).

Fig.5.Relative expression levels of MsSPL6,MsSPL9,MsSPL12,MsSPL13,and MsSPL16 in WT and TG plants.Values are means±SD(n=3).Different letters represent significant difference (P <0.05).

To investigate the responses of miR156-mediated salt and drought tolerance,we evaluated the expression pattern of five putative miR156 target genes after salt and dehydration treatments.MsSPL6/12/13were significantly up-regulated after salt treatment for 6 h,while the expression ofMsSPL12showed sharp reduction after salt stress treatment for 24 h.The expression ofMsSPL9andMsSPL16showed no significant difference between before and 6 h after salt treatment,but dramatically increased 24 h after treatment (Fig.6A).The expression of bothMsSPL12andMsSPL13showed a trend of fast rising-declining-rising pattern under dehydration treatment (Fig.6B).These results suggested thatMsSPL12andMsSPL13play vital roles in miR156-mediated salt and drought tolerance.

Fig.6.Relative expression levels of miR156 and its putative target genes in WT and TG plants after salt (A) and dehydration treatment (B).

Values are means ± SD (n=3).The different letters represent significant difference (P<0.05).

3.8.Global gene expression analyses of WT and OE-miR156 plants

Among 1078 significantly differentially expressed genes(DEGs,with an adjustedP-value<0.05)in TG and WT plants,511 were upregulated and 567 were down-regulated (Fig.S5).Ten genes were randomly selected from the DEGs to verify the RNA sequencing data by qRT-PCR (Fig.S6).Their expression by qRT-PCR showed no significant difference (P<0.05) from their expression by RNA sequencing.GO classification indicated that most of the DEGs were enriched in membrane,cell membrane part,and membraneenclosed lumen(Fig.S7).Genes associated with plant abiotic stresses were isolated from DEGs and are listed in Table S4.These genes include drought-responsive genes[33],an Na+/H+antiporter 1gene[34],and abiotic stress-associated transcription-factor genes encodingApetala 2/Ethylene Response Factors(AP2/ERF) [35] andWRKY[36] (Fig.7A–C).Several DEGs were associated with plant development including nuclear transcription factor Y subunit C3(NF-YC3) and lignin biosynthesis-associated genes,which may contribute to improved crude protein content and reduced ADL content in TGs (Table S3;Fig.7D and E) [37].Thus,miR156-mediated regulation of alfalfa development and abiotic stress response employs multiple and complex pathways.

4.Discussion

Fig.7.Relative expression levels of genes associated with plant development and abiotic stress.Relative expression of ERD15(A), Na+/H+ antiporter 1(B),NFYC3(C), ERF008(D) and HCT (E).Values are means ± SD (n=3).Different letters represent significant difference (P <0.05).

MiR156 has been recognized as a key genetic resource for improving forage traits [1,2].In this study,we up-regulated miR156 expression in alfalfa by heterogeneous expression of two naturally tandem genes from the monocot species rice,Osa-MIR156bandOsa-MIR156c,resulting in an ideal plant architecture type with dramatically increased biomass yield and protein content,reduced ADL content,and improved salt and drought tolerance.Despite the effect of OE-miR156 on reducing plant height,OE-miR156 lines with miR156 expression level similar to that previously reported were only about 5%shorter than WT.The different effects of miR156 on alfalfa plant height could be due to the difference in miR156 precursors and/or the specific genetic background of the alfalfa cultivar.Functional differences of pre-miRNA genes are present in different plant species even though they produce the same or similar miRNA sequences.For example,LjmiR156increased alfalfa soluble sugar content[1],which was not the case whenMsmiR156dwas overexpressed [2].Overexpression ofAtmiR156bin tomato resulted in abnormal fruit structure [20],but this effect was not observed in tomato plants overexpressingSlymiR156a[21].But the same precursor microRNA gene could also show different functions when expressed in different plant species.LjmiR156represses root elongation inLotus japonicusbut not in alfalfa [1,2].In rice,a loss-of-function mutation of a miR156 subfamily (MIR156a–c,k,andl) showed no significant effect on shoot architecture [19].In the present study,heterogeneous expression ofOsa-MIR156bc,generating the same miR156 sequence,increased transgenic alfalfa branching number and biomass yield and showed little effect on plant height.

Protein content is a key parameter for assessing forage nutritive value.Forage with high protein content can reduce or eliminate protein supplementation by costly off-farm feeds[2].Plant protein content is affected by multiple metabolic pathways [38–40],and also by leaf biomass and leaf-to-stem ratio [2].OE-miR156 or reducing the level of its targetedSPLstends to increase leaf number[10,41,42].We confirmed that miR156 accumulated in leaves,especially young leaves,in agreement with observations inArabidopsisand rice[13,43].OE-miR156 increased alfalfa leaf number and secondary branching.OE-miR156 alfalfa plants have a higher leaf biomass yield and leaf-to-stem ratio,which may be attributed to the increased crude protein content and reduced ADL level [41].InArabidopsis,overexpression ofAtNF-YC4,a gene homolog ofNF-YC3,increased protein content and reduced starch levels [37].Here,we noticed that the expression ofNF-YC3was higher in TG than in WT plants.However,overexpression of miR156 had no effect on starch content in alfalfa [1,2].Thus,miR156-mediated increases in protein content may involve complex regulatory pathways,which merit further study.

Vegetative-to-floral phase transition is determined largely by miR156-SPLmodule [1,2,44].Overexpression of miR156 in alfalfa resulted in extended vegetative phase,andLjmiR156awas more influential thanMsmiR156din prevention of precocious flowering[1,2].Overexpression ofOsa-MIR156bcdelayed flowering time by about 2 weeks,possibly contributing to increased biomass yield and forage quality inOsa-MIR156bctransgenic plants [45].We observed similar effects of OE-miR156 on flowering,suggesting that different miR156 precursors play a conserved role in delaying flowering time.

Leaf area and number are not only important to plant photosynthesis and respiration,but also closely associated with plant abiotic stress [46–48].In this study,miR156 reduced alfalfa leaf area.Reduced leaf area is considered a positive trait in drought tolerance,owing to the reduced transpiration area [49].However,leaf number was greatly increased in OE-miR156 plants.Overexpression of miR156 reportedly improved drought and salt tolerance[14–17].TG plants showed greater water retention and cell membrane integrity than WT plants under salt and drought stress.DEGs in TG and WT encoded mostly proteins enriched in membrane,cell membrane part,and membrane-enclosed lumen (Fig.S7).Several drought-or dehydration-induced protein-encoding genes showed up-regulated expression in TG plants,such as the dehydrationinduced geneERD15[50].AndWD40-1,a gene encoding a positive regulator of drought tolerance,was regulated by miR156-SPLmodule.It showed significantly up-regulated expression in TG plants[51].OE-miR156 in alfalfa also reduced Na+accumulation in leaves subjected to severe salt stress,possibly owing to the up-regulatedNa+/H+ antiporter 1gene expression in TG plants [34].The expression of several transcription factor genes encodingApetala 2/Ethylene Response Factors(AP2/ERF),WRKYandNAC,which have been reported to be modulators of plant abiotic stress resistance[35,36],was changed.For example,MsERF8increased the salt tolerance of tobacco andArabidopsis[52] and was upregulated in our OE-miR156 plants.

MiR156 regulates plant development and abiotic stress via post-transcriptional regulation ofSPLsexpression [1,2,12–15].The expression levels of sevenSPL genes(MsSPL2/3/4/6/9/12/13)were down-regulated inMsmiR156dtransgenic plants [53].In the present study,onlyMsSPL12andMsSPL13showed reduced expression in OE-miR156 alfalfa plants,possibly owing to the different repression effects on different target genes exerted by various miR156 precursors [4,44].SPL13was involved in plant abiotic stress resistance and inflorescence morphogenesis [51,54].In the present study,MsSPL12was highly induced in Zhongmu No.1 alfalfa leaf after salt and drought stress treatment in comparison with other target genes of miR156.These results suggest thatMsSPL12could act downstream of miR156 to regulate alfalfa development and abiotic stress tolerance.

In summary,overexpression of miR156 by heterogeneous expression ofOsa-MIR156bcincreased salt and drought tolerance of alfalfa,increased biomass yield by increasing leaf and branching number,and improved forage quality by increasing protein content and reducing ADL content.Rice miRNA156 precursor genes could be used for alfalfa genetic improvement.

5.Conclusions

Overexpression of monocot miR156 precursors in alfalfa improved forage quality by increasing crude protein and biomass yield and reducing plant height and ADL content.Transgenic plants showed increased salt and drought tolerance compared to WT.This is the first report of overexpression of monocot miR156 precursors in alfalfa.These results shed light on the function of miRNA156 precursors from a monocotyledonous species,rice.These precursors could be used for genetic improvement of alfalfa,a dicot species.More specifically,miR156 could be used as a molecular tool for forage quality improvement.

CRediT authorship contribution statement

Kexin Wang,Yanrong Liu,and Wanjun Zhang:conceived and designed the research.Kexin Wang,Fengkui Teng,and Huifang Cen:conducted experiments.Huifang Cen:contributed to transgenic plant generation.Kexin Wang,Jianpin Yan,Shiwen Lin,and Dayong Li:analyzed data.Kexin Wang,Yanrong Liu,and Wanjun Zhang:wrote the manuscript.All authors read and approved the final manuscript.

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 work was supported by the National Natural Science Foundation of China (31971755) and the Beijing Municiple Natural Science Foundation (6192011) of China.We thank Dr.Rongda Qu of North Carolina State University,Dr.Hong Luo of Clemson University,and Dr.James C.Nelson of Kansas State University for reading the manuscript.

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.11.009.