Rice sl-MH-1 mutant induces cell death and confers blast resistance via the synergistic roles of signaling systems

Dgng Tin,Yn Lin,Shengping Li,Yiyng Co,Gng Li,Xinrui Guo,Ziqing Chen,Zijie Chen,Feng Wng,Zonghu Wng

a Fujian Key Laboratory of Genetic Engineering for Agriculture,Biotechnology Research Institute,Fujian Academy of Agricultural Sciences,Fuzhou 350003,Fujian,China

b State Key Laboratory of Ecological Control of Fujian-Taiwan Crop Pests,Key Laboratory of Ministry of Education for Genetics,Breeding and Multiple Utilization of Crops,Plant Immunity Center,Fujian Agriculture and Forestry University,Fuzhou 350003,Fujian,China

Keywords:Sekiguchi lesion Cell death Phytohormone M.oryzae Rice

ABSTRACT Serotonin is ubiquitous across all forms of life and functions in responses to biotic and abiotic stresses.In rice,the conversion of tryptamine to serotonin is catalyzed by Sekiguchi lesion(SL).Previous studies have identified an sl mutation(a null mutation of SL)in several rice varieties and confirmed its increase of resistance and cell death.However,a systematic understanding of the reprogrammed cellular processes causing cell death and resistance is lacking.We performed a multi-omics analysis to clarify the fundamental mechanisms at the protein,gene transcript,and metabolite levels.We found that cell death and Magnaporthe oryzae(M.oryzae)infection of the sl-MH-1 mutant activated plant hormone signal transduction involving salicylic acid(SA),jasmonic acid(JA),and abscisic acid(ABA)in multiple regulatory layers.We characterized the dynamic changes of several key hormone levels during disease progression and under the cell death conditions and showed that SA and JA positively regulated rice cell death and disease resistance.SL-overexpressing lines confirmed that the sl-MH-1 mutant positively regulated rice resistance to M.oryzae.Our studies shed light on cell death and facilitate further mechanistic dissection of programmed cell death in rice.

1.Introduction

Lesion-mimic mutants display phenotypes similar to those of pathogen infection-induced cell death under natural growth conditions.Lesion formation is usually accompanied by changes in the levels of salicylic acid(SA),jasmonic acid(JA),abscisic acid(ABA),ethylene(ET),reactive oxygen species(ROS),and other substances.These phytohormones subsequently promote the production of antimicrobial metabolites such as phenols,phytoalexins,and antioxidants[1-6].Thus,the molecular mechanisms of lesion-mimic formation and increased resistance are likely to be associated with endogenous signaling pathways.

Serotonin(5-hydroxytryptamine)is a well-known neurotransmitter in mammals and plants[7].Recently,serotonin has been reported[6,8,9]to activate intracellular defense mechanisms against Magnaporthe oryzae(the causal agent of rice blast)but to suppress resistance to brown planthopper and striped stem borer.We have identified an sl mutation(a null mutation of SL)and confirmed its increase of resistance and cell death.In the sl mutant,the inactivated SL gene induced SA accumulation at the expense of serotonin biosynthesis,and SA may be involved in resistance to brown planthopper(BPH)and striped stem borer(SSB)[9].Treatments with exogenous serotonin,vitamin C supplementation,or light-induced accumulation of tryptamine increase resistance to rice blast or reduce lesion formation in sl-MH-1 mutants,showing the effectiveness of ROS in controlling BPH,SSB,and M.oryzae[6,9,10-13].The sl-MH-1 mutant accumulated higher levels of JA and JA-isoleucine(JA-Ile)than WT plants[6].Thus,SA,ROS,and JAs(JA and JA-Ile)have been confirmed to be involved in sl-mediated responses to abiotic and biotic stresses,functioning as toxic byproducts to promote cell death and as protective signal transduction molecules to increase resistance to BPH,SSB,and M.oryzae.However,the synergistic action of these phytohormones in mitigating the effects of stress on the whole plant remains largely elusive.

Recently,rice proteomics research has progressed,generating functional information for proteins responding to various abiotic and biotic stresses[14-17].Transcriptomics analyses using RNA-Seq(RNA-Sequencing)have shed light on the gene expression profiling of lesion-mimic production[18-21],mediated by the resistance gene Pi9 and its coexpression network with the resistance genes Pi54 and Pi1[22,23].Metabolomic analyses[24]have contributed to provide direct evidence about events that occurred in the metabolic differences and explain biochemical phenotypes.Targeted analyses of phytohormones[25,26]have been used to characterize signaling and metabolic networks.Up to date,multiomics analysis of proteomic,transcriptomic,metabolomic,and phytohormone data[27,28]have been widely applied under cold,dehydration,and drought-stress conditions in rice.

The present study performed an integrated analysis of the levels of proteins,gene transcripts,metabolites,and characterized these events underwent the cell death and upon M.oryzae infection in the sl-MH-1 mutant compared with WT plants(Fig.S1).Hormone measurements were performed to characterize the changes in levels of SA,JA,ABA,indole-3-acetic acid(IAA)and ET during disease progression and cell death.We identified and characterized the involvement of SA,JA,and the sl gene in cell death and resistance to blast,which enabled us to identify crosstalk between hormones or signaling molecules,tryptophan,and secondary metabolism involved in the sl-MH-1 mutant in response to biotic and abiotic stress.These results provide evidence for molecular events at multiple levels and could serve as a valuable resource for further investigation of cell death in rice.

2.Materials and methods

2.1.Plant materials and conditions

The sl-MH-1 mutant was obtained as a tissue-culture variant of the indica rice cultivar Minghui 86[6,12].Minghui 86 and the sl-MH-1 mutant were grown in the greenhouse and field by conventional culture during the summer season in Fuzhou,Fujian province,China.A few small lesions appeared on the leaves at the seedling stage,and numerous lesions covered the leaves at the tilling stage in the sl-MH-1 mutant(Fig.S1).As the whole process of cell death was spontaneously occur,we named it as natural cell death(NCD).The plants in the growth chambers were used for proteomic,transcriptomic,and metabolomic studies when the sl-MH-1 mutant underwent cell death,and the plants in the greenhouse were used for transcriptomic and metabolomic measurements post-inoculation with M.oryzae at the seedling stage.

2.2.Blast fungus inoculation

The M.oryzae isolate FJ-1(virulent to Minghui 86)was used.For blast fungus inoculation,plants were grown in a greenhouse.M.oryzae inoculations were performed as previously described[17].The FJ-1 isolate was cultured on oatmeal agar medium for two weeks and spray-inoculated at a concentration of 5×105spores mL-1on 3-4-week-old plant leaves.The inoculated seedlings were maintained under high humidity until disease was visible.Mockinoculated(control)plants were treated identically except that the pathogen suspension was replaced by water.The pathogen inoculation experiments were repeated three times.

2.3.Metabolic analysis

NMR nontargeted metabolomics was used to determine the effects in the sl-MH-1 mutant showing cell death on its leaves.Freeze-dried plant material of 3 g was weighed and suspended in 1000μL of methanol:water(v/v,1:1).A 4-s on/off cycling program was used for 8 cycles for in-solution ultrasonic extraction(Sonics VX-130,Newtown,CT,USA).Samples were centrifuged at 13,000 r min-1for 15 min,and the supernatant was lyophilized before being redissolved in 450μL of water.An aqueous layer with a volume of 450μL was transferred to a clean 2-mL centrifuge tube followed by the addition of 50μL of DSS(4,4-dimethyl-4-silapen tane-1-sulfonate)standard solution.Samples were mixed well before their transfer to a 5 mm NMR tube(Norell,Morganton,NC,USA).Spectra were collected with a Bruker AV III 600-MHz spectrometer equipped with an inverse cryoprobe.The first increment of a 2D-1H,1H-NOESY pulse sequence was used for the acquisition of 1H NMR data and for suppressing the solvent signal.Experiments employed a 100 ms mixing time with 990 ms of presaturation(～80 Hz gamma B1).Spectra were collected at 25 °C with 64 scans over a period of 7 min.

The collected free induction decay signal was automatically zero filled and Fourier transformed in the processing module in Chenomx NMR Suite 8.1(Chenomx Inc.,Edmonton,Canada).The data were then phased,and the baseline was corrected by a technician experienced with the Chenomx processor.All of the spectra were referenced to the internal standard,DSS,and analyzed by experienced analysts against the Chenomx Compound Library.Among the 24 spectra,35 metabolites were identified and quantified.All of the concentration information of metabolites was exported to Microsoft Excel(Microsoft Corp.,Redmond,WA,USA)and normalized by weight across all samples before use in later multivariate analysis.

PCA(principal component analysis)and PLS-DA(partial leastsquares discrimination analysis)were performed using the PCA Methods Bioconductor Package[29]and PLS package[30]respectively.Plots were drawn with ggplot2[31].

Levels of tryptophan metabolism in leaf parts of the sl-MH-1 mutant exhibiting cell death and WT plants(100 mg)were quantified as previously described[6,32].Reproducibility was assessed with six biological replicates in each experiment.This experiment was performed in collaboration with Shanghai Biotree Biotech Co.,Ltd.,Shanghai,China(https://biotreeaq.bioon.com.cn/).

2.4.Isobaric tags for relative and absolute quantification(iTRAQ)-based proteomic assay and RNA-Seq

Protein extraction,digestion,and iTRAQ labeling were performed as previously described[33].Total protein was extracted from the leaves of the sl-MH-1 mutant exhibiting cell death and from wild-type(WT)plants.The proteins were reduced with dithiothreitol and alkylated with iodoacetamide prior to trypsin digestion.The digests were concentrated to 30μL and then labeled with iTRAQ 4-plex labels(AB Sciex,Foster City,CA,USA)according to the manufacturer’s instructions.The labeled samples were combined and separated by high-pH reverse-phase LC(RPLC)followed by second-dimension RPLC as previously described[34].The eluate was introduced to an in-line QTrap 4000 through a nanoSpray II source(AB SCIEX)using an uncoated fused-silica Pico tip(New Objective,Woburn,MA,USA).MS/MS data were acquired as previously described[34].The experiment was performed by Shanghai Applied Protein Technology Co.,Ltd.(https://www.aptbiotech.com/).

For RNA-Seq,leaves of WT plants and the sl-MH-1 mutant exhibiting cell death at 48 hpi(hour post-inoculation)were used for transcriptome analysis as described in Tian et al.[6].The RNA-Seq was conducted by Novogene,Beijing,China(https://www.novogene.com/).

2.5.Bioinformatics and statistical anal ysis

The proteins and RNA-Seq data were conducted on the Majorbio I-Sanger Cloud Platform(https://www.i-sanger.com).For protein data,the MS/MS data were processed and expression analysis was performed using software created by Majorbio.Proteins were considered significantly differentially abundant if they had P-values≤0.05(t-test)and mean fold changes between sl-MH-1 mutant and WT≥1.2 or≤0.83.Proteome Discoverer software(Thermo Fisher)was used to convert the O-Exactive-generated raw data into mgf files.The Mascot search engine(Matrixscience.com)was used to identify the proteins using a database containing the nonredundant protein database(NCBInr),UniProt,and other databases.All raw MS files are available for download as UniProt_rice_144684_20151027.fasta(https://www.uniprot.org)with project number 144,684.The proteins identified were functionally annotated by GO database.To classify and group the identified proteins,the Kyoto Encyclopedia of Genes and Genomes(KEGG)(https://www.genome.jp/kegg/)(KOBAS 2.1.1)was used.GO functional enrichment and KEGG pathway analyses were performed with Goatools(https://github.com/tanghaibao/Goatools)and KOBAS(https://kobas.cbi.pku.edu.cn/home.do)[35].

Raw paired-end reads were trimmed and quality-controlled with SeqPrep(https://github.com/jstjohn/SeqPrep)and Sickle(https://github.com/najoshi/sickle)with default parameters.The clean reads were aligned to a reference genome,the O.sativa japonica cultivar Nipponbare sequence according to IRGSP-1.0 with orientation mode using TopHat software(https://tophat.cbcb.umd.edu/,version 2.0.0)[36].The mapping criteria of bowtie were as follows:sequencing reads must be uniquely matched to the genome,allowing up to 2 mismatches without insertions or deletions.Then,the region of the gene was expanded following the depths of the sites,and the operon was obtained.The whole genome was split into multiple 15-kb windows overlapping by 5 kb.Newly transcribed regions were defined as more than 2 consecutive windows without overlapping regions of genes,where at least 2 reads mapped per window in the same orientation.

To identify DEGs(differential expression genes)in different samples,the expression level of each transcript was calculated according to the fragments per kilobase of exon per million mapped reads(FRKM)method.RSEM(https://deweylab.biostat.wisc.edu/rsem/)was used to quantify gene abundances[37].The R statistical package software EdgeR(Empirical analysis of Digital Gene Expression in R(https://www.bioconductor.org/packages/2.12/bioc/html/edgeR.html))was used for differential expression analysis[38].Functional enrichment analysis,including GO and KEGG analyses,was performed to identify DEGs that were significantly enriched in GO terms and metabolic pathways at a Bonferroni-corrected P-value≤0.05 compared with the wholetranscriptome background.

2.6.RNA extraction and qPCR assay

Total RNA from leaf blades exhibiting cell death was isolated using TRIzol reagent(Invitrogen,Carlsbad,CA,USA).For realtime quantitative PCR(qPCR)assays,three technical and three biological replicates were used for each gene.qPCR was used to validate the DEGs obtained from RNA-Seq.Primers were designed using Primer 5(PREMIER Biosoft)and are summarized in Table S1.A ProtoScript M-MuLV First Strand cDNA Synthesis Kit(NEB)was used to synthesize cDNA.The reaction was performed in a LightCycler R 480 II PCR system with a volume of 10μL containing 5μL of SYBR Green I Master,0.2μL of forward primer,0.2μL reverse primer,2μL of 1:5 diluted cDNA template,and RNase-free water.Actin was used as internal control.The amplification efficiency of qPCR was calculated in both control and target samples,and the FC(Fold Change)was calculated using the CT method.

2.7.Measurements of SA,JA,JA-Ile,ET,ABA,ET,and IAA contents and chemical treatment

Hormone contents were quantified with MetWare(https://www.metware.cn/)based on the AB Sciex QTRAP 6500 LC-MS/MS platform.Three replicates of each assay were performed.A 500-mg sample of powder from the sl-MH-1 mutant and WT plants was extracted with 5 mL of 90%aqueous methanol and 2 ng of each D-labeled hormone was added to the extraction solvents as internal standards for content measurement.After the crude extracts were passed through tandem MAX and MCX cartridges(6 mL,500 mg)(Milford,MA,USA),the eluate was reconstituted with 200μL of 80% methanol and subjected to UPLC-MS/MS.All procedures followed the manufacturer’s instructions.Hormones were detected in negative multiple reaction monitoring(MRM)mode.Each hormone was quantified with another hormone using an MRM transition.The source parameters were set as follows:IS voltage-4500 V,TEM 600 °C,GS1 45,GS2 55,and curtain gas 28.These hormones were measured at the Institute of Genetics and Developmental Biology of the Chinese Academy of Sciences,Beijing,China(https://www.genetics.ac.cn/jspt/zwjs/).

Chemical treatments of the sl-MH-1 plants with the JA-Ile and JA biosynthesis inhibitor DIECA(200μmol L-1;Sigma-Aldrich),and SA biosynthesis inhibitor PAC(300μmol L-1;Seebio)were applied as described previously[39].All inhibitors were prepared in water containing 0.02%(v/v)Tween 20 and sprayed onto 3-week-old rice seedlings two times per day(at 9:00 and 15:00)for three days.Water spraying was used as a control.A conidial suspension of M.oryzae isolate FJ-1 at 5×105spores mL-1was then sprayed on the leaves of the plants.After inoculation,the seedlings were grown as described above.Samples were harvested at 24 and 48 hpi for qPCR.qPCR was performed as described above for SA and JA biosynthesis and signaling pathways.The primer sequences are listed in Table S1.

2.8.Construction of SL overexpression and CRISPR-CAS9 allene oxide synthase 2(AOS2)mutant

The SL overexpression and CRISPR-Cas9 AOS2 were obtained from the Wuhan Institute of Biotechnology,Wuhan,China.A fulllength cDNA fragment(1572 bp)of SL(LOC_Os12g16720)was PCR-amplified from cDNA prepared from Minghui 86 using KOD DNA polymerase(TOYOBO(Shanghai)Biotech Co.,Ltd.)and confirmed by sequencing.SL CDS was cloned in the modified plant transformation vector pCAMBIA1300.The binary construct was confirmed by colony PCR,restriction digestion,and DNA sequencing.The confirmed clone was transformed into Agrobacterium strain EHA105 and used for genetic transformation of rice.

A guide RNA was designed to target exons of the AOS2 gene using the website(https://chopchop.cbu.uib.no/)and is described in Table S1.The overexpression and CRISPR vectors were transformed into,respectively,Minghui 86 and the sl-MH-1 mutant by Agrobacterium-mediated transformation.Individual T1 plants of SL over-expression and CRISPR-CAS9 AOS2 were validated,respectively,by quantitative reverse-transcription PCR(qRT-PCR)and sequencing the DNA products,using primers listed in Table S1.

3.Results

3.1.Loss of function of sl altered the metabolite profile of the tryptophan pathway and led to sugar accumulation in the sl-MH-1 mutant

The levels of shikimic acid,tryptamine,L-tryptophan,and 4-hydroxybenzoic acid were higher in the sl-MH-1 mutant showing cell death than in WT plants,with the exception of 5-hydroxytryptamine(Fig.1A).Thirty-five different metabolites were identified according to the concentration of dextran sodium sulfate and peak area(Table S2).In comparison with WT plants,the levels of respectively 26 and 9 metabolites were upregulated and downregulated in the sl-MH-1 mutant.PCA(Fig.1B-a)indicated that the metabolic profiles of the sl-MH-1 mutant and WT plants were different.The first principal component(PC1)values were positive for WT plants and negative for the sl-MH-1 mutant,reflecting the similarities in these 35 metabolites among the six biological repeats in both the sl-MH-1 mutant and WT plants.The cumulative contribution of the PCA was 96.9%up to the second principal component(PC2),and the PC2 values reflected differences among biological replicates in the sl-MH-1 mutant and WT plants.

The corresponding PCA loading map reflected the contribution of each metabolite to PC1 and PC2.The levels of sucrose metabolites were highest in both loading 1 and loading 2 compared with other metabolites.The levels of fructose and glucose were also higher under loading 2 than under other metabolites.The further the metabolite was from the center(0,0)in the direction of loading 1 and loading 2,the greater the contribution of the metabolite to PC1 and PC2.Thus,sucrose,fructose,and glucose contributed most to PC1 and PC2(Fig.1B-b).

We further selected significant differentially abundant metabolites using PLS-DA.The vertical coordinates showed that sucrose,fructose,glucose,serine,glutamine,1,3-dimethylurate,glutamate,alanine,proline,valine,4-aminobutyrate,trehalose,tryptophan,threonine,and galacturate were the 15 most affected metabolites,whose concentration differences between the sl-MH-1 mutant and WT plants are shown by the color code on the right(Fig.1B-c).The variable importance in projection(VIP)plot indicated that most of those secondary metabolites were up-regulated in the sl-MH-1 mutant and are core metabolites in response to cell death.

3.2.Mutation of SL stimulated genes and proteins associated with signal transduction and defense response in a similar manner

The total spectrum,identified spectrum,peptide number,protein number,and protein group number are shown in Fig.S2.A total of 2987 differential expressed proteins(DEPs)were identified in the sl-MH-1 mutant and WT plants with a false discovery rate of 1%,including 419 proteins that were upregulated and 237 that were down-regulated with a threshold of 1.2-fold change value(a ratio of 1.2 or 0.83)and P-value＜0.05(Fig.S3A;Table S3).

Following RNA-Seq,38,673,939 and 83,263,641 sequences were mapped in the sl-MH-1 mutant and WT plants in each biological replicate(Fig.S2;Table S4).These mapped loci showed＞94%alignment with the rice genome(version 7).All biological replicates(both the sl-MH-1 mutant and WT plants)were repeatable(Pearson’s r＞0.98)(Table S5).The number of significant DEGs between the sl-MH-1 mutant and WT plants with a fold change≥2 was 734(429 up-regulated and 305 down-regulated)(Fig.S3B;Table S6).qPCR of eight randomly selected genes yielded reproducible expression patterns for all genes,supporting the reliability of the DEGs(Fig.S3C;Table S1).

Fig.1.Loss of function of SL induces accumulation of sugars and secondary metabolites.(A)The value in the box below each metabolite shows the fold change between the sl-MH-1 mutant underwent cell death and wild-type plants.Values are means±SE(n=6).P-values of the ANOVAs are indicated.*,P＜0.0.5;**,P＜0.01;***,P＜0.001.ND,not determined.(B)a,b,and c signify PCA,PLS-DA,and VIP of the differential expressed metabolites.B and C in panels a and c represent sl-MH-1 mutant and wild-type plants,respectively.

GO enrichment analysis and KEGG analyses of DEPs and DEGs were performed as shown in Fig.2A and B.Pathways enriched in these DEPs and DEGs included alpha-linolenic acid metabolism,glutathione metabolism,phenylalanine,tyrosine and tryptophan biosynthesis,plant hormone signal transduction,flavonoid biosynthesis,MAPK signaling pathway,phenylpropanoid biosynthesis,and diterpenoid biosynthesis.Among them,plant hormone signal transduction was induced in both samples,consistent with the above altered metabolites.Glutathione metabolism,phenylalanine,tyrosine and tryptophan biosynthesis were significantly enriched among changed metabolites and DEPs,but not DEGs,suggesting that some post-translational regulation mechanisms were involved in an association analysis of RNA-Seq and the iTRAQ data,most of these DEGs/DEPs were up-regulated at both RNA and protein levels in response to cell death(Table 1).GO enrichment analysis suggested close associations with carboxylyase activity,carbon-carbon lyase activity,oxidation-reduction process,and defense response,and KEGG enrichment analysis showed that the most significant pathways included plant hormone signal transduction,flavonoid biosynthesis,and MAPK signaling pathway(Fig.2C).

Fig.2.Classification of differential expressed genes and proteins in the sl-MH-1 mutant showing cell death compared with wild-type plants.(A)GO and KEGG enrichment analysis of differential expressed proteins in the sl-MH-1 mutant exhibiting cell death compared with wild-type plants.(B)GO and KEGG enrichment analysis of differential expressed genes in the sl-MH-1 mutant exhibiting cell death compared with wild-type plants.(C)GO and KEGG enrichment analysis of integrated proteome and transcriptome database in the sl-MH-1 mutant exhibiting cell death compared with wild-type plants.

Taken together,these results show that sl-MH-1 positively activated plant hormone signal transduction in a similar manner at the transcriptional and translational levels under the sl-MH-1 cell death condition.

Table 1Common differential expressed genes and proteins and their assigned functions in the sl-MH-1 mutant under the cell death conditions compared with wild-type plants.

3.3.Plant hormone signal transduction was induced by M.oryzae infection in the sl-MH-1 mutant

To investigate the molecular mechanism in the sl-MH-1 mutant in response to M.oryzae infection,we performed a comparative transcriptome analysis between the sl-MH-1 mutant and WT plants at 48 hpi.The sequence information for each sample subjected to RNA-Seq and of the mapped loci is presented in Table S3.The comparison of significant DEGs was set as fold change≥2 or≤0.50.Respectively 805,1772,and 1125 DEGs were identified from these comparisons between sl-MH-1_mock and WT_mock(designated sl-MH-1 mutant and WT mock-treated),WT_mock and WT_48 hpi(designated WT plants between mock-treated and at 48 hpi)and sl-MH-1_mock and sl-MH-1_48 hpi(designated sl-MH-1 mutant between mock-treated and at 48 hpi)(Tables S7,8,9).

GO enrichment analysis was performed for all the DEGs.The top 20 GO terms in each comparison were selected.Analysis of DEGs in the WT_mock_vs_sl-MH-1_mock comparison showed that the most significant categories were cell wall,external encapsulating structure,extracellular region,hydrolase activity,and cell wall organization and biogenesis(Fig.S4A).In the WT_mock_vs_WT_48 hpi comparison,the most significant functional terms were plastid and chloroplast thylakoid lumen,chloroplast and tetrapyrrole biosynthetic process,photosynthesis,secondary metabolic process,and thylakoid part(Fig.S4B).For the sl-MH-1_mock_vs_sl-MH-1_48 hpi comparison,the most abundant enriched genes were associated with oxidation-reduction process,oxidoreductase activity,and extracellular region(Fig.S4C).

Further KEGG enrichment analysis showed that the most significantly enriched(P-value＜0.01)pathways in the WT_mock_vs_sl-MH-1_mock comparison were associated with fatty acid elongation,phenylpropanoid biosynthesis,cutin,suberine,and wax biosynthesis(Fig.S4A).DEGs in the WT_-mock_vs_WT_48 hpi comparison were involved in phenylpropanoid biosynthesis,porphyrin and chlorophyll metabolism,diterpenoid biosynthesis,and carbon fixation in photosynthetic organisms(Fig.S4B).In the sl-MH-1_mock_vs_sl-MH-1_48 hpi comparison,the most highly enriched pathways were plant hormone signal transduction,diterpenoid biosynthesis,cortisol synthesis and secretion,and brassinosteroid biosynthesis(Fig.S4C).In terms of plant hormone signal transduction,25 DEGs were involved in tryptophan metabolism,zeatin biosynthesis,diterpenoid biosynthesis,carotenoid biosynthesis,cysteine and methionine metabolism,and phenylalanine metabolism(Table 2).All genes in tryptophan metabolism encoding AUX/IAA,GH,and SAUR were up-regulated,with the exception of Os06g0499550 and Os07g0592600,while those genes encoding PP2C,SAPK9,and bZIP transcription factor 12 synthesis associated with ABA in the carotenoid biosynthesis pathway were downregulated in the sl-MH-1 mutant.SA from phenylalanine metabolism was observed to be activated,accompanied by an increase of NPR1,TGA,and PR1,and three genes of the cysteine and methionine metabolic pathway were differentially regulated to mediate accumulation of ET.

Taken together, the comparisons between sl-MH-1_mock_vs_sl-MH-1_48 hpi,WT_mock_vs_sl-MH-1_mock,and WT_mock_vs_WT_48 hpi suggest that rice responded to M.oryzae infection to regulate the expression of many defenseassociated genes in a similar manner.Furthermore,some genes with respect to secondary metabolite and hormone in plant hormone signal transduction were obviously affected in the sl-MH-1 mutant.

Table 2Differential expressed transcripts involved in hormone response pathways in the sl-MH-1 mutant at 48 hpi with M.oryzae compared to mock-treated.

3.4.Synergy among major phytohormones led to cell death and increased resistance to rice blast disease in the sl-MH-1 mutant

Because our results,along with some reported earlier[4,6,9,11,40],indicated that SA,JAs,ABA,IAA,and ET were involved in cell death and resistance to pathogens in the sl mutants,we sought to compare their free status levels between the sl-MH-1 mutant and WT plants at 0,24,48,and 72 hpi,and under the cell death conditions.

Fig.3.Contrasting expression of phytohormones SA,JA,JA-Ile,ACC,ABA,and IAA between the sl mutant and wild-type plants at 0,24,48,and 72 hpi with M.oryzae and under the cell death conditions.SA,salicylic acid;JA,jasmonic acid;JA-Ile,jasmonoyl-L-isoleucine;ABA,abscisic acid;ACC,1-aminocyclopropanecarboxylic acid;IAA,indole-3-acetic acid;NCD,natural cell death;hpi,hour post-inoculation.Values are means±s.e.m.of three biologically independent experiments.*,P＜0.05；**,P＜0.01(Student’s ttest).

For SA,the free status levels in the sl-MH-1 mutant were up,down,and up-regulated at 24,48,and 72 hpi,respectively,and reached the highest under the cell death conditions.Further investigation showed that the free status levels of SA were always higher in the sl-MH-1 mutant than in WT plants,especially at 72 hpi or under the cell death conditions(Fig.3;Table S10).In contrast,the free status levels of JA and JA-Ile were suppressed at 24 hpi,but gradually up-regulated at 48 and 72 hpi,and reached their highest level in the sl-MH-1 mutant under the cell death conditions.The free status levels of JA and JA-Ile in the sl-MH-1 mutant were lower at 24 and 48 hpi but higher at 72 hpi and under the cell death conditions than in WT plants(Fig.3;Table S10).

Although the free status levels of ABA and ACC were not visibly different between the sl-MH-1 mutant and WT plants at 0,24 hpi,and 48 hpi,there was an obviously increase in the sl-MH-1 mutant under the cell death conditions,suggesting that the two substances may be coordinated during disease progression or cell death(Fig.3;Table S10).

Although IAA and serotonin have the same precursor in the tryptophan metabolic pathway,mutation of SL did not cause a change in the free status levels of IAA in the sl-MH-1 mutant compared with WT plants(Fig.3).The above study in 3.3 demonstrated that GH3.2 and SAUR were up-regulated in the sl-MH-1 mutant compared with WT plants at 48 hpi.GH3 genes encode IAA-amid synthetases that are involved in the regulation of auxin homeostasis by conjugating excess IAA to amino acids for the activation or inactivation of IAA.Thus,overexpression of the GH3-family gene OsGH3-2 increased basal disease resistance by suppressing pathogen-induced IAA accumulation[41],implying that the level of IAA was not markedly changed in the sl-MH-1 mutant compared with WT plants.

Fig.4.Increased susceptibility to M.oryzae or cell death symptom,and suppression of expression of gene in the SA and JA biosynthesis and signaling pathways in the sl-MH-1 mutant by chemical treatment with DIECA(JA and JA-Ile biosynthesis inhibitor)and PAC(SA biosynthesis inhibitor).(A)Increase of disease symptoms of M.oryzae or cell death symptoms in DIECA-and PAC-treated line.(B,C).Relative expression levels of genes involved in SA and JA synthesis and signaling in mock buffer and PAC,Buffer and DIECA treatment sl-MH-1 mutant.(D)Mutations in three knockout mutants of AOS2 in the sl-MH-1 background.(E)The aos2 knockout line showed more lesions than the sl-MH-1 mutant.Values are means±s.e.m.(n=3).*,P＜0.05;**,P＜0.01(Significant difference between Buffer and inhibitor-treated sl-MH-1 mutant by Student’s t-test).Fold change is the ratio of sl-MH-1 to WT.

3.5.Suppression of the biosynthesis of SA and JA increased the sl-MH-1 mutant susceptibility to M.oryzae or cell death

To further determine whether SA and JA are required for defense against M.oryzae,we treated the sl-MH-1 mutant exogenously with an inhibitor of SA biosynthesis,paclobutrazol(PAC)[42],and a lipoxygenase(LOX)inhibitor,diethyldithiocarbamic acid(DIECA),and then inoculated them with M.oryzae.PAC-and DIECA-treated plants showed higher disease symptoms than mock-treated plants(Fig.4A).qRT-PCR analysis revealed that the PAC-treated sl-MH-1 mutant showed lower expression of the SA synthesis and signaling molecule genes OsPAD4,OsPAL,OsICS1 and OsEDS1 at 24 and 48 hpi(Fig.4B).Blocking JA and JA-Ile biosynthesis in the sl-MH-1 mutant also led to decreases in the expression levels of PLDa1,LOX2,LOX5,AOS2,AOC,COI1b,COI2,JAZ5,and JAZ8 at 24 hpi and 48 hpi,but not those of PLDa1,LOX2,and JAZ5,which showed higher expression in WT plants(Fig.4C).These results indicate that SA and JA confer resistance to M.oryzae infection in the sl-MH-1 mutant.

To further investigate whether JA is involved in cell death in the sl-MH-1 mutant,AOS2,catalyzing(13S)-hydroperoxyoctadecatrie noic acid to an allene oxide in the first step of JA biosynthesis,was edited in the sl-MH-1 background.We obtained three types:aos2-I,aos2-II,and aos2-III mutants(Fig.4D;Table S11).Because the aos2-I and aos2-III mutant could not survive until the mature stage in the T0 generation,we harvested only aos2-II for phenotype identification.The aos2-II null mutant plants showed fewer lesions than the sl-MH-1 mutant at the seedling stage(Fig.4E),suggesting that AOS2 might be associated with cell death in the sl-MH-1 mutant.

3.6.SL negatively regulates rice resistance to M.oryzae

The multi-omics analysis suggested that SL negatively regulates rice defense against M.oryzae.To confirm this hypothesis,we overexpressed this gene in the Minghui 86 background by Agrobacterium-mediated genetic transformation.We obtained 20 overexpressing transgenic plants and selected the SL transgenic overexpression lines for inoculation testing.The transgenic lines were more susceptible than Minghui 86 to the virulent blast strain FJ-1(Fig.5A),and produced significantly higher biomass than Minghui 86(Fig.5B).

Three experiments of mock-treated,before-inoculation,and 7 days post-inoculation(referred to as MT,BI,and PI)of Minghui 86 and transgenic overexpression lines were harvested for qPCR analysis.We designed two pairs of primers located in respectively the 5′and 3′-terminal coding regions of SL(Table S1).The results showed significant upregulation of SL in transgenic overexpression lines as compared with Minghui 86 at the three experiments,and the two pairs of primers also showed consistent results(Fig.5C).Taken together,these findings again suggest that SL reduces regulates resistance to rice blast disease.

4.Discussion

4.1 Synergistic roles among phytohormones and metabolites contribute to cell death or increased resistance to blast disease in the sl-MH-1 mutant

Fig.5.Disease resistance of rice SL overexpressing lines.Blast resistance of SL overexpressing line compared with wild-type Minghui 86.Three representative leaves were collected at 7 days after spray inoculation with the blast isolate FJ-1.(B)Fungal biomass of the inoculated leaves is indicated.WT,Minghui 86.(C,D)Three deals of MT,BI,and PI of the overexpressing transgenic lines were qPCR using two pairs of primers targeted 5′and 3′to the SL gene.MT,mock-treated,BI,before-inoculation,and PI,7 days postinoculation.**,P＜0.01(Student’s t-test).

In the study,our metabolic analysis indicated that the sl mutant directly causes the accumulation of upstream metabolites and induces the metabolism of other compounds.Some metabolics such as phenylalanine,tyrosine,tryptophan,and tryptamine etc were upregulated in the sl-MH-1 mutant,all of which have the common precursor of shikimic acid,facilitating the production of H2O2[6,9,10].As sugars affect gene expression through sugarspecific signaling cascades,which regulate the expression of stress-related genes such as superoxide oxidase,they are also newly emerging ROS scavengers that protect chloroplasts from oxidative damage[48].Some metabolites such as sugars(sucrose,fructose,glucose,and trehalose),prolin,and tryptophan,are involved in the regulation of ROS levels[8,43-46].Thus,signal transduction was associated with glucose and fructose accumulation in the sl-MH-1 mutant exhibiting cell death.

Signaling by phytohormones such as SA,JA,ABA,and ET is involved in leaf senescence and in the hypersensitive response to lesion browning after M.oryzae infection[47-49],and SA and JA are recognized as major defense hormones[50].Given that light,M.oryzae,and BPH induced tryptamine and SA accumulation in the sl-MH-1 mutant,the level of SA was up-regulated at the early stage following infection by M.oryzae[9,10].The present study indicated that Os01g0808100,which encodes transcription factor TGAL1,was down-regulated in the sl-MH-1 mutant at 48 hpi,and increasing evidences indicates[51]that TGAL1 positively regulates the NPR1-dependent SA signaling pathway,which is also required for SA-mediated repression of the JA response.Thus,suppression of TGA may reduce the repression of SA in the JA response.

Generally,SA modulates defense responses primarily against biotrophic pathogens,while JA is involved in plant defenses against primarily necrotrophs.The interaction of rice with M.oryzae comprises an initial biotrophic phase that allows an explosive surge of SA and a subsequent necrotrophic phase induction of JA biosynthesis[52,53],drivering the switch between SA and JA.

Fig.6.A proposed model summarizing the action of phytohormones and cellular pathways by sunlight and M.oryzae.Red arrow represents activation or up-regulation,dashed arrow indicates repression,and dotted arrow indicates undetermined pathway.Fold change in the sl-MH-1 mutant compared with wild-type plant at 0,24,48,72 hpi and under NCD condition were calculated based on a-log scale and are displayed in those five-box circle with different colors to the left of each hormone name.SA,salicylic acid;JA,jasmonates;JA-Ile,jasmonic acid-isoleucine;JAR1,jasmonate resistant1;ABA,abscisic acid;ET,ethylene;ROS,reactive oxygen species;NPR1,non-expressor of pathogenesis related gene 1;TGA,transcription factor TGAL;MAPK6,mitogen activated protein kinase 6;PP2C,protein phosphatase 2C;IAA,indole-3-acetic acid;PAMPs,pathogen-associated molecular patterns.

There is an intricate relationship between ABA production and ROS levels in plants.The ABA-dependent antioxidant defense system in serotonin biosynthesis operates in the hypersensitive reaction that results in lesion browning and leaf senescence[54,55].However,the increase of ABA increases susceptibility and pathogen attacks[56].For example,abscisic acid2-1(aba2-1)and aba3-1 ABA-deficient mutants in Arabidopsis displayed increased resistance to Golovinomyces and Erwinia chrysanthemi,respectively[57,58].Consistent with those previous studies,we found that the level of ABA did not change markedly at 24 and 48 hpi but increased dramatically under the cell death conditions in the sl-MH-1,it may be,we speculate coupled with accumulated glucose and fructose to induce cell death[59].

4.2 SL combines with SA and JA to regulate rice resistance to M.oryzae

Serotonin is found in many species and acts in slowing leaf senescence and responding to many biotic stresses[6,9-13].SL encodes the CYP71A1 protein that has tryptamine 5-hydroxylase activity to catalyze the conversion of tryptamine to serotonin and is highly conserved in the genomes of all cereal crops[8].Previous studies[6,9,11]found that loss of function of SL increased resistance to much biotic stress by reducing in ROS scavenging and the content level of serotonin,and regulating the levels of hormones via the regulation of signaling molecules.In the present study,SL-overexpressing lines were more susceptible to M.oryzae than the wild-type Minghui 86.But we also found that several hormones were involved in cell death and resistance to M.oryzae in sl-MH-1 plants.Inhibition of expression of SA and JA in the sl-MH-1 mutant increased susceptibility to the blast fungus in comparison with sl-MH-1 plants,suggesting that the SA and JA signaling pathways positively participated in cell death and blast resistance of sl-MH-1 plants,in a manner contrary to the SL.Taken together,We propose a working model of how phytohormones affect cell death and increase resistance in the sl-MH-1 mutant involving many functionally divergent downstream phytohormones and including dynamic interactions among SA,JAs,ABA,and ET(Fig.6).These crosstalks among metabolites and hormones provides the plant with a powerful capacity to finely regulate its immune response and utilize its resources in a cost-effective manner[60].Thus,the form and timing of crosstalk determine disease trends[61].

5.Conclusions

The results of the present study led to speculation that SA plays an antagonistic role with respect to other phytohormones such as JAs and ABA at the earlier stage against M.oryzae infection,but with the disease progresses,or underwent cell death,these phytohormones act in a synergistic manner to confer resistance or cell death.This comprehensive study sheds light on the role of SL and suggests the prominent role of suppression of serotonin biosynthesis in mediating broad-spectrum disease resistance in rice.

Data availability

RNA-seq data were submitted to the international repository GEO(Gene Expression Omnibus,https://www.ncbi.nlm.nih.gov/bioproject/634690)with BioProjectID:PRJNA634690.All other data generated and analyses performed during this study are included in this published article and its supplemental data files.

CRediT authorship contribution statement

Dagang Tian:Conceptualization,Funding acquisition,Project administration,Visualization,Writing-original draft,Writingreview & editing.Yan Lin:Investigation.Shengping Li:Investigation.Yiyang Cao:Investigation.Gang Li:Validation.Xinrui Guo:Resources.Ziqiang Chen:Data curation.Zaijie Chen:Formal analysis.Feng Wang:Funding acquisition.Zonghua Wang:Writingreview & editing,Funding acquisition.

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 Collaborative Innovation Engineering‘‘5511”(XTCXGC2021002),the National Natural Science Foundation of China(U1805232),the Youth Program of National Natural Science Foundation of China(31301654),and the Youth Program of Fujian Academy of Agricultural Sciences(YC2019004).We thank Shanghai Biotree Biotech Co.,Ltd.for assistance in metabolomics analysis,and Jinfang Chu at Institute of Genetics and Developmental Biology,Chinese Academy of Sciences,Beijing,China for helping us measure phytohormones.

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2022.03.005.