Guoting Liang·Jing Guo·Shuyong Zhang·Guangcan Zhang

Abstract Plant species produce different types of allelopathic chemicals in nature;however,little is known about the differential regulation of plant allelopathy.Because allelopathy caused by p-hydroxybenzoic acid (pHBA) is one of the main obstacles to continuous cropping of Populus × euramericana ‘Neva’,we examined gene expression dynamics in Neva leaves induced by pHBA.The expression of genes related to photosynthesis and respiration changed,while mRNA involved in regulating gene expression during allelopathy was degraded.Some miRNAs that are involved in plant allelopathy target crucial genes for regulating reactive oxygen species.Moreover,coexpression regulatory networks were constructed based on profiles of the identified miRNA-target interactions and the differentially expressed miRNA—target pairs.These findings provide a mechanistic framework for understanding allelopathy in plants.

Keywords Populus × euramericana ‘Neva’·Allelopathic response·miRNA·Transcriptome·Degradome

Introduction

Plant allelopathy mediates crucial interactions between the environment and plant life cycle during the growth of plants(Li and Xiao 2012).Various plant species use different types of allelochemicals to regulate the growth and development of plants and help seedlings survive allelopathic conditions(Csiszár 2009;Chaudhary et al.2015),so we need to understand the allelochemicals involved (Latif et al.2017).Plant allelochemicals primarily consist of phenols,terpenoids,alkaloids and other secondary metabolites (Kaczmarek 2009),so they are synthesized by secondary metabolism(Weston et al.2012).Allelochemical stress can alter the physiology,biochemistry and levels of componds in plants(Liang et al.2017).Due to the importance of allelopathy in the plant growth and development process,particularly for agriculture and forestry,many studies have focused on the regulatory mechanisms underlying plant allelopathy (Jabran et al.2015).

Genetic studies have revealed genes that regulate plant allelopathy,particularly genes involved in photosynthesis,respiration,secondary metabolism and signal transduction (Huang et al.2015;Barchenger and Bosland 2016).Additionally,exogenous application of allelochemicals can modify plant development (Liang et al.2017).Recently,emerging ‘-omics’ technologies have increased the current understanding of the regulation of plant allelopathy at the whole-genome level.Indeed,numerous genes with various functions are differentially expressed under allelochemical stress,indicating that a highly complex network regulates plant allelopathy (Dmitrović et al.2015;Li et al.2017).

Despite the distinct gene expression programs between allelochemical-treated and untreated plants (Song et al.2008),little is known about the regulation of gene expression during allelopathy.Notably,mRNA degradation is an important step in regulating gene expression in eukaryotic cells.The mRNA decapping mediated by decapping enzymes (DCPs) represents an essential step in mRNA degradation.Decapped mRNA intermediates are then degraded in a 3′—5′ orientation by an exosome complex or digested by the 5′—3′ exonuclease XRN1 (Chanfreau 2015).In addition,an endonuclease and small RNAs (miRNAs or siRNAs) can initiate the internal cleavage of mRNAs (Durand et al.2 017).The uncapped mRNAs,which are characterized by shortened poly(A) tail and 5′ monophosphate,are produced by the active decapping enzymes and internally cleaved by endonucleases and small RNAs (Kramer 2017).Alternatively,uncapped mRNAs could be produced by nonsense-mediated mRNA decay (NMD) (Schoenberg and Lynne 2012).Originally,parallel analysis of RNA ends (PARE) was developed to identify miRNA cleavage on target mRNAs (German et al.2008).PARE has been demonstrated to be a useful tool for prof iling uncapped mRNAs in eukaryotic cells (Addo-Quaye et al.2009;Zhang et al.2013).Taking advantage of a modified 5′ rapid amplif ication of cDNA ends (RACE)technique,PARE selectively captures the mRNA degradation intermediates,poly(A) RNAs with 5′ monophosphate(German et al.2009).Thus far,most studies of mRNA degradation have been based on yeast and mammalian models(Hu et al.2015);however,the process and roles of mRNA degradation remain unknown in plants,particularly during allelopathy.

Previous studies on the regulation of plant allelopathy have focused on herbaceous species (e.g.,Arabidopsis,wheat,and barnyard grass;Dmitrović et al.2015;Fang et al.2016);in contrast,information on trees remains limited.Populus×euramericana‘Neva’ is a common broadleaf tree in Europe,America and Asia (Liang et al.2017).It is a commercially important because it produces a large amount of soft and straight-textured wood.Therefore,much eff ort has focused on exploring the regulatory mechanisms underlying wood development,and high-throughput sequencing technology has been used to identify functional genes in poplars (Ren 2012).

To increase our understanding of the regulatory mechanisms underlying allelopathy in trees,we profiled global mRNA expression inPopulus×euramericana‘Neva’ leaves treated with various allelochemicals to explore the regulatory network of allelopathy.

Furthermore,PARE was used to investigate the role of mRNA degradation in gene expression in ‘Neva’ leaves.The long-term objective of the present study is to provide reference maps for the regulatory mechanisms of plant allelopathy,with an emphasis on the regulation of orchestrated gene expression during allelopathy.

Materials and methods

Plant material and pHBA treatments

One-year-old healthy branches were collected from 20Populus×euramericana‘Neva’ trees at a forestry ecostation of Shandong Agricultural University (36°16′ N,117°15′ E,134 m a.s.l.),Shadong Province,China on March 15,2016.Tree height averaged approximately 10 m,and mean diameter at breast height was 10.5 cm.Branches approximately 20 cm long with a diameter of approximately 2 cm were cut from the middle and lower parts of the crown.The collected cuttings were water-cultured in controlled conditions (25 °C,12 h,day;0 °C,12 h,night) in a light incubator for 90 days before initiation of photosynthesis tests.Three replicates each of 10 seedlings were treated with different pHBA concentrations solutions (0,2,4,and 8 mM L−1) for 3 days.

RNA isolation and RNA-sequencing analysis

Total RNA was isolated separately from L0,L2,L4 and L8 seedlings using RNAiso Plus and RNAiso-Mate for Plant Tissue (Takara,Dalian,China) using the protocol of the manufacturer.BioAnalyzer (Agilent Technologies,Santa Clara,CA,USA) was used to assess the integrity and quantity of total RNA.Four cDNA libraries (L0,L2,L4 and L8) were prepared and sent to LC Sciences for sequencing with the Illumina HiSeq 2500 platform (Hangzhou,China).Before transcriptome assembly,ambiguous residues (N’s),adapter contamination and low-quality reads were filtered from the raw reads.De novo assemblies of the global transcriptome for the four libraries were pooled to obtain clean reads using the program Trinity.The NCBI non-redundant protein database (NR),Swiss-Prot,Pfam,KOG,and KEGG were used for BLAST analysis of the assembled unigene sequences in the transcriptome for gene annotation.Subsequently,clean reads of the L0,L2,L4 and L8 libraries were mapped to the transcriptome with Bowtie 2 (Langmead and Salzberg 2012).The number of mapped clean reads per unigene was calculated and normalized to RPKM (the number of reads per kilobase per million clean reads) (Mortazavi et al.2008).The agriGO analysis tools were used to perform GO enrichment (http://bioin fo.cau.edu.cn/agriG O/) (Zhou et al.2010).The REVIGO and Cytoscape applications were used to visualize the enriched GO terms (Supek et al.2 011).

PARE analysis of uncapped transcripts

Total RNA was extracted from L0,L2,L4 and L8 seedlings as done for transcriptome sequencing.A PARE library from the mixture of equal amounts of RNA samples (L0,L2,L4 and L8) was constructed (German et al.2008).In brief,a modified RLM 5′-RACE technique was applied to isolate RNAs totally with a 3′ poly(A) tail and a 5′ monophosphate.An RNA adapter was ligated to the 5′ ends of single-stranded degraded RNAs with a 5′ monophosphate,then first strand cDNA was generated using oligo(dT) with a 3′ adapter.The cDNA was amplif ied by a short PCR and then digested with the restriction enzymeMmeI to cleave 20 bp from the 3′ end of the recognition site.The purif ied products were subjected to high-throughput sequencing using an Illumina HiSeq 2000 instrument (Hangzhou,China).The Uncapped mRNAs were analyzed as in previous studies (Zhang et al.2013;Cao et al.2016) by removing the reads corresponding to known rRNAs,tRNAs,small nuclear RNAs,small nucleolar RNAs,repeats,and transposons,then filtering out raw sequencing reads.The remaining clean reads were then mapped to the assembled transcriptome by Bowtie 2.When a PARE signature was mapped to multiple genes,reads were divided among genes.The total number of reads for specific transcripts was determined via custom Perl scripts and normalized to the number of tags per billion reads (TPB).

Identification of miRNAs and their target genes

Previously identified miRNAs from L0,L2,L4 and L8 seedlings were used in the present study.The assembled transcriptome obtained in the present study was used as transcriptome input.With the Cleave L and 3.0 pipeline,the degradome sequencing reads with 20/21 nucleotides were used to identify potential cleaved targets.The degradome reads for ‘Neva’ were mapped to the transcriptome data.The identified targets were categorized as 0,1,2,3,or 4 in the present study (Yang et al.2013).

QRT-PCR analysis

The qRT-PCR was employed to validate the expression levels of the transcriptome sequences.Total RNA was isolated from L0,L2,L4 and L8 samples as described above.Gene-specific primers for qRT-PCR are listed in Table S1.The ABI 7300 Real-Time PCR System (Bio-Rad,Hercules,CA,USA) was used for PCR reactions.The 2−ΔΔCTmethod was used to calculate the relative gene expression levels(Livak and Thomas 2001),and the actin gene was used as an internal reference for mRNA expression.The RNA-seq data generated in the present study have been submitted to the Sequence Read Archive database in NCBI (accession GSE114411).

Construction of the miRNA-target interaction network

CytoScape (version 3.4.1) software (Shannon et al.2003)was used to generate networks of the predicted miRNA-target interactions.

Results

De novo assembly and gene annotation of the global transcriptome

De novo transcriptome assembly was conducted using the Trinity pipeline and all clean reads from the four libraries(L0,L2,L4 and L8 leaves).After rigorous data cleanup and quality checks,50,489,630,52,572,060,52,289,212,and 40,646,242 clean reads from the L0,L2,L4,and L8 leaf libraries,respectively,were obtained.The assembly generated 63,356 transcripts with a total length of 84,568,251 bp and an average length of 1334.91 bp.Homologous transcripts with >95% similarity were clustered,producing 28,946 unigenes.The length of the unigenes ranged from 297 to 7930 bp,with an average of 1268.82 bp (Table S2).

All unigenes were annotated with BLASTX searches against the Swiss-Prot,NCBI non-redundant protein sequences (NR) and Pfam databases.The identified 10,604 unigenes matching known genes in the Swiss-Prot database with E-value <10−5accounted for 36.63% of the total unigenes (Table S3).Similarly,28,225 (97.51%) unigenes were identified in the NR database and 6371 (22.01%) in the Pfam databases.GO (Gene Ontology),COG (Eukaryotic Ortholog Groups) and KEGG (Kyoto Encyclopedia of Genes and Genomes) assignments were used to complete the annotation of unigenes in our global transcriptome.

Based on sequence similarity,10,189 unigenes (35.20%of total unigenes) were annotated in the GO database (http://geneo ntolo gy.org/).A total of 20,559 unigenes (71.03% of total unigenes) were categorized into 25 COG functional groups;“function unknown”represented the largest group,followed by“posttranslational modification,protein turnover,chaperons”,“translation,ribosomal structure and biogenesis”,“amino acid transport and metabolism”,“carbohydrate transport and metabolism”,“transcription”and“signal transduction”(Fig.1).There were 22,778 (78.69%)unigenes assigned to 129 KEGG pathways;“ribosome”,“carbon metabolism”,“protein processing in endoplasmic reticulum”,“biosynthesis of amino acids”,and“spliceosome”were the most highly represented pathways.

Fig.1 Eukaryotic Ortholog Groups (COG) categories of the transcriptome

Transcriptomic alterations in leaves during allelopathy induction

To assess alterations in gene expression during the induction of plant allelopathy inPopulus×euramericana“Neva”seedlings,we globally profiled mRNA expression in L0,L2,L4 and L8 leaves.We detected 27,252,27,548,21,098,and 24,595 unigenes (>1 RPKM) in L0,L2,L4 and L8 leaves,respectively (Table S4).By further analyzing differential gene expression between L0 and treatment (L2,L4 and L8) seedlings,we found 511 differentially (|log2 foldchange| >2,p<0.05) expressed genes (DEGs) between L0 and L2 seedlings,2505 DEGs between L0 and L4 seedlings,and 1990 DEGs between L0 and L8 seedlings (Table S5).In the hierarchical cluster analysis,these DEGs grouped into six clusters (Fig.2).Genes in Clusters 1,3 and 4 showed increased expression when the pHBA concentration was lower (L0 and L2) and decreased expression when pHBA concentrations were higher (L4 and L8).In contrast,the expression pattern of genes in Cluster 2 was the opposite.All the results suggested a tight linkage of these genes with plant allelopathy inPopulus×euramericana“Neva”seedlings.

Fig.2 Hierarchical cluster analysis of the DEGs and expression patterns of the gene clusters.Gene expression data were log2-transformed and color-coded.Blue:low expression levels;red:high expression levels

In the GO enrichment analysis to investigate biological functions of these crucial DEGs,Cluster 1 genes were enriched in GO terms associated with energy metabolism,TCA,amide metabolism,ROS metabolism,transport and proteolysis.Cluster 2 genes were enriched in GO terms related to amino acid metabolism,lipid and acid metabolism,oomycetes response,secondary metabolism,Flavonoid metabolism,ethylene metabolism and ROS metabolism.Cluster 3 genes were enriched in GO terms related to cell wall modification,amino acid metabolism,secondary metabolism,one carbon metabolism,f lavonoid metabolism,development and defense response.Cluster 4 genes were enriched in GO terms related to stimulus response,development,abscisic acid metabolism,rhythmic process and galactose metabolism.Cluster 5 genes were enriched in GO terms related to secondary metabolism and transport(Fig.3).Quantitative real-time PCR (qRT-PCR) were used to validate the DEGs identified by high throughput sequencing,similar expression patterns were found,supporting the reliability of the sequencing data (Fig.4).RNA-seq and qRT-PCR data were closely correlated (Pearson’s correlation analysis,r=0.903,p<0.0001).

Fig.3 Enriched GO (BP) terms for the five gene clusters

Genes regulating respiration play a role in plant response to pHBA stress

Since GO enrichment analysis showed that genes in Cluster 1 were enriched in GO terms related to respiration,we paid special attention to genes involved in respiration.We identified all 57 unigenes involved in glycolysis,25 unigenes involved in the respiratory electron transport chain,69 unigenes participating in oxidative phosphorylation and 35 unigenes participating in the tricarboxylic acid cycle (TCA)in the four libraries for the ‘Neva’ leaf samples (Table S6).Notably,all DEGs that encoded critical TCA enzymes,including citrate synthase,malate dehydrogenase,pyruvate dehydrogenase,succinate dehydrogenase,succinyl-CoA synthetase,aconitate hydratase,were more highly expressed in L0 and L2 leaves.Interestingly,DEGs encoding critical glycolytic enzymes,including glyceraldehyde-3-phosphate dehydrogenase,alcohol dehydrogenase,enolase showed similar expression patterns.Moreover,most of the DEGs related to respiratory electron transport chain and oxidative phosphorylation were downregulated in L4 and L8 leaves(Fig.5).

Photosynthesis in leaves during pHBA stress

The key genes in three parts of photosynthesis were examined and are discussed here:regulation of photosynthetic electron transport chain,photophosphorylation and Calvin cycle.9 genes involved in photophosphorylation were differential expressed during the different pHBA treatments in ‘Neva’ leaves (Table S7).Furthermore,24 genes involved in photosynthetic electron transport chain exhibited differential expression between pHBA-treated(L2,L4 and L8) seedlings and untreated (L0) seedlings(Fig.6).Intriguingly,the genes regulating photosynthetic electron transport chain and photophosphorylation had distinct expression patterns.The photosystem II core complex proteins gene (PSBY),ferredoxin gene (FER) and most genes related to photosystem I were more highly expressed in L2 leaves.Interestingly,the expression pattern of genes involved in photophosphorylation was similar to that of the gene encoding cytochrome b6-f complex iron-sulfur subunit (UCRIA);both were higher in L4 and L8 leaves.All 27 DEGs participated in the Calvin cycle.Notably,the genes encoding critical enzymes in the Calvin cycle,(ribulose bisphosphate carboxylase,fructose-bisphosphate aldolase,glyceraldehyde-3-phosphate dehydrogenase and phosphoglycerate kinase) were more highly expressed in L0 and L2 leaves,whereas transketolase,phosphoribulokinase and sedoheptulose-1,7-bisphosphatase were more highly expressed in L4 and L8 leaves.

Fig.6 Expression profiles of differentially expressed genes that regulate photosynthesis in L0,L2,L4 and L8 seedlings.Expression levels of genes (RPKM) based on RNA-seq were log2-transformed and color-coded.Blue:low expression levels;orange:high expression levels

Osmoregulatory substance in leaves during pHBA stress

To investigate the osmoregulation ofPopulus×euramericana“Neva”under pHBA stress,we separately analyzed the expression of genes involved in starch and sucrose metabolism,amino acid metabolism and purine metabolism.We detected 72 genes (>1 RPKM) regulating starch and sucrose metabolism that were differentially expressed in leaves in the different treatments (Table S8).Thirtyfour DEGs between pHBA treated (L2,L4 and L8) seedlings and untreated (L0) seedlings were related to amino acid metabolism (Fig.7),and 46 unigenes participated in purine metabolism.Most of the osmoregulatory DEGs were downregulated as the pHBA stress increased.

Fig.7 Expression profiles of differentially expressed genes that regulate starch and sucrose,amino acid and purine metabolism in L0,L2,L4 and L8 seedlings.Expression levels of genes (RPKM) based on RNA-seq were log2-transformed and color-coded.Blue:low expression levels;orange:high expression levels

Reactive oxygen species regulation in leaves during pHBA stress

Reactive oxygen species (ROS) regulation is an important index during plant allelopathy.Here 16 unigenes related to ROS metabolism were significantly differentially expressed(Table S9),and most of these were more highly expressed in L0 and L2 leaves (Fig.8).

Fig.8 Expression profiles of differentially expressed genes that regulate reactive oxygen species (ROS) metabolism in L0,L2,L4 and L8 seedlings.Expression levels of genes(RPKM) based on RNA-seq were log2-transformed and color-coded.Blue:low expression levels;orange:high expression levels

Transcriptional and translational regulation of plant allelopathy

The significant alteration in gene expression programs between pHBA concentration treated (L2,L4 and L8) seedlings and untreated (L0) seedlings prompted us to investigate the regulation of gene expression inPopulus×euramericana‘Neva’ during allelopathy.We measured the expression levels of genes encoding transcription initiation factors,translation initiation factors,transcription elongation factors and translation elongation factors in leaves,and the expression of 10 translation initiation factor genes,31 translation initiation factor genes,4 transcription elongation factor genes and 6 translation elongation factor genes differed by more than fivefold (p<0.05) between pHBA-treated (L2,L4 or L8) seedlings and the untreated (L0) seedlings (Fig.9).Moreover,8 of 10 transcription initiation factor genes,30 of 31 translation initiation factor genes,all 4 transcription elongation factor genes and 6 translation elongation factor genes were more highly expressed in untreated (L0) seedlings and lower than in the pHBA-treated (L2) seedlings than those treated with higher concentrations of pHBA (L4 and L8) seedlings (Fig.9).Expression of these genes varied greatly between pHBA-treated and untreated seedlings,with an average fold-change of 0.987 (1.911 for L2,0.358 for L4,0.693 for L8 separately) and a maximum of 12.98 (for translation initiation factor genes).

Fig.9 Diff erential expression of genes regulating transcription and translation.(A) Transcription initiation factors.(B) Translation initiation factors.(C) Transcription elongation factors.(D) Translation elongation factors.The numbers on the x-axis represent unigenes

mRNA degradation in ‘Neva’ leaves

To globally analyze mRNA degradation inPopulus×euramericana‘Neva’ leaves,we performed high-throughput sequencing of PARE libraries from the mixture of L0,L2,L4 and L8 leaves.In total,we identified 26,444 uncapped mRNA species in leaves (Fig.10).A total of 26,370 unigenes were detected in both capped(transcriptome libraries) and uncapped forms (PARE libraries),accounting for 42.84% of transcriptome unigenes (Fig.10).Correlation analysis revealed a positive correlation (Pearson correlation,r=0.403,p<0.0001)between uncapped mRNA abundance and transcriptomic mRNA abundance in L0 leaves (Fig.11 B).Analysis of the L2,L4 and L8 libraries conf irmed these positive correlations (Fig.11 B).Transcriptomic mRNA abundance was also positively correlated to uncapped mRNA length in mixture leaves samples (Pearson correlation,r=0.123,p<0.0001,Fig.11 B).

Fig.10 Number of uncapped mRNA species in the mixture of L0,L2,L4 and L8 leaves

Fig.11 Characterization of uncapped mRNA expression in Populus × euramericana‘Neva’ leaves.(A) Correlations between uncapped mRNA abundance and transcriptomic mRNA abundance in pHBAtreated (L2,L4 and L8) and untreated (L0) leaves.(B)Correlations between uncapped mRNA abundance and transcriptomic mRNA abundance in the mixture of leaf samples (L0,L2,L4 and L8).Data for gene expression levels and mRNA length were log2-transformed

We also detected 24 genes encoding mRNA decay factors (includingDDX6,RCK,DHH1,groEL,HSPD1,CNOT,NOT1,PATL1,MOT2,EDC4,RRP,EXOSC10,ENO,DHX36,RHAU,RCD1,CAF40,CCR4-NOT,MTR4,SKIV2L2andPABPC).The expression levels of all but 8 of these genes were similar among L0,L2,L4 and L8 leaves;8 genes were significantly more highly expressed (|log2 foldchange| >2,p<0.05) in L2 leaves than in other treatments(Table S10).

Roles of miRNAs in plant allelopathy in ‘Neva’

PARE analysis revealed 386 miRNA-target mRNA pairs in the four libraries (L0,L2,L4 and L8),and 247 of these miRNAs and 234 targets were identif eid (Table S11),and we analyzed the relationship among them.Eight-six miRNA-target mRNA pairs were revealed among the four different libraries from ‘Neva’ leaves (Table S12).For example,targets of 8 miRNAs (aly-miR397a-5p_R+1_1ss21GTmiR397,athmiR397a_R+1,ptc-miR397a,ptc-MIR1444c-p3_1ss3CT,ptc-miR1444a_R+1,ptc-miR7833,PC-3p-79202_59,ptcmiR6424) act in oxidoreductase activity;cme-miR399d and ptc-miR399e regulate one target,SUC2 (ducrose transport protein),which acts in carbohydrate transport.Targets of 7 miRNAs (ath-miR164c-5p,mdm-miR164b_R+3,mtrmiR164a_1ss21AT,gma-miR167c_R+2_1ss21GT,mdmmiR167h_R+1,ptc-miR164a,ptc-MIR6474-p3) act in the regulation of transcription (Table S12).In addition,GSTX4_TOBAC (probable glutathioneS-transferase),which is involved in the auxin-activated signaling pathway,was identified as a target of ptc-MIR7828-p3_1ss2TA.

Construction of the miRNA-target interaction network

To visualize the miRNA-targets interactions directly,we used CytoScape to generate networks between them.Two networks were generated.One network was used to identify miRNA/target interactions,and the other network was used to identify interactions between the differentially expressed miRNAs and differentially expressed targets (Fig.12).

Fig.12 Networks of miRNA-targets interactions.a Identified miRNA—target interactions.miRNA,purple circle;target gene,green square. b Interactions between the differentially expressed miRNAs and the differentially expressed targets.miRNA,blue circle;target gene,green square

Discussion

Many plants have mechanisms against allelopathy to improve their fitness in ecosystems.Allelochemicals are produced in plants under various environmental stress,and they can affect plant growth and development during the allelopathic process (Goga et al.2017).

In addition to reversible changes in gene expression related to respiration,photosynthesis,starch and sucrose metabolism,amino acid metabolism,purine metabolism,and ROS metabolism,we also found some differences in gene expression among the different treatments,such as energy metabolism,cell wall modification,development,stimulus response and secondary metabolism.These findings suggested that some unknown regulation underlying plant allelopathy remain unknown and should be investigated in future studies.

PARE analysis revealed extensive posttranscriptional regulation of plant allelopathy inPopulus×euramericana‘Neva’ leaves.We detected uncapped transcripts for 42.84%of the expressed genes in leaves.InBrachypodium distachyonseedlings,67% of all expressed genes produce uncapped transcripts (Zhang et al.2013).InCunninghamia lanceolataseeds,54.89% of expressed genes produce uncapped transcripts (Cao et al.2016).These results indicated that mRNA degradation participated broadly in gene expression regulation of plants.

Numerous miRNAs participate in plant development and stress responses (Yuan et al.2015).Many miRNAs that have been identified in leaves through high-throughput sequencing may participate in regulating various biological processes in leaves (Gao et al.2015).Moreover,miRNA399 (miR399) is involved in the regulation of phosphate homeostasis (Huen et al.2017).Here,we detected a close association between miR399 and its target SUC2,which is a sucrose transport protein (Eom et al.2016),and accumulating evidence suggests the potential involvement of sucrose in plant allelopathy (Grove et al.2012).Eight miRNAs (aly-miR397a-5p_R+1_1ss21GTmiR397,athmiR397a_R+1,ptc-miR397a,ptc-MIR1444c-p3_1ss3CT,ptc-miR1444a_R+1,ptc-miR7833,PC-3p-79202_59,ptcmiR6424) target genes involved in oxidoreductase activity,supporting critical roles of miRNAs in regulating reactive oxygen species in ‘Neva’ leaves.Considering that the expression levels of these miRNAs were closely related to plant allelopathy and regulated by reactive oxygen species,these findings feedback regulation of reactive oxygen species in plant allelopathy.

On the basis of our present results and previously published findings,we illustrate the networks of interactions between miRNAs and their targets underlying plant allelopathy inPopulus×euramericana‘Neva’ (Fig.12).Growth and photosynthesis inEucalyptus grandisseedlings are affected by allelopathy (Huang et al.2015),and respiration and cell division in watermelon is affected by allelopathic compounds in corn pollen (Ortega et al.1988).The microRNAs of barnyardgrass are involved in the rice response to allelopathy (Fang et al.2015).Gene modules are co-regulated with biosynthetic gene clusters involved in allelopathy between rice and barnyardgrass (Sultana et al.2019).Clearly,under certain allelopathyic conditions,photosynthesis and respiration are disturbed,and distinct gene expression programs are induced.Allelochemicals may induce a decrease in photosynthesis and increase in respiration,leading to allelopathy.

In addition to the central roles played by photosynthesis and respiration during allelopathic responses,mRNA degradation is also required to reduce RNA levels during allelopathy.

The present sequential transcriptome and degradome analyses have provided important novel finding into allelopathy in seedlings ofPopulus×euramericana‘Neva’.(1) The expression of mRNAs that are related to alterations in photosynthesis and respiratory balance changes;(2) mRNA is degraded during the regulation of gene expression during allelopathy;and (3)some miRNAs are involved in plant allelopathy by targeting crucial genes in reactive oxygen species regulation.