+ All Categories
Home > Documents > UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We...

UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We...

Date post: 23-Jul-2020
Category:
Upload: others
View: 1 times
Download: 0 times
Share this document with a friend
12
Review UTR-Dependent Control of Gene Expression in Plants Ashish Kumar Srivastava, 1,3, * Yuming Lu, 1 Gaurav Zinta, 1 Zhaobo Lang, 1 and Jian-Kang Zhu 1,2, * Throughout their lives, plants sense many developmental and environmental stimuli, and activation of optimal responses against these stimuli requires extensive transcriptional reprogramming. To facilitate this activation, plant mRNA contains untranslated regions (UTRs) that signicantly increase the coding capacity of the genome by producing multiple mRNA variants from the same gene. In this review we compare UTRs of arabidopsis (Arabidopsis thaliana) and rice (Oryza sativum) at the genome scale to highlight their complexity in crop plants. We discuss different modes of UTR-based regula- tion with emphasis on genes that regulate multiple plant processes, including owering, stress responses, and nutrient homeostasis. We demonstrate func- tional specicity in genes with variable UTR length and propose future research directions. Signicance of UTRs in Plants: Genome-Scale Evidence Because plants are sessile, they have to tolerate many abiotic and biotic stresses. Besides, plants undergo major growth transitions during their life cycle from seed germination to vegetative growth, owering, and seed development. Both the activation of stress tolerance mechanisms and developmental transitions require extensive transcriptional reprogramming, which is controlled at transcriptional and post-transcriptional levels. Transcriptional control is mediated by transcription factors while post-transcriptional control is regulated by a complex set of cis- and trans-acting elements [13]. Cis-acting elements for post-transcriptional control, which are generally present in UTRs (see Glossary) of mRNAs, can regulate various mRNA properties like stability, transport, and translation efciency and also the functioning and subcellular localization of the translated proteins [4]. Thus, UTRs represent important genetic components that regulate processes at the interface of mRNAs and proteins. In addition, owing to their characteristics of having multiple variants from the same gene, UTRs can increase the coding capacity of the genome. This provides versatility to the gene expression system and thereby helps plants adapt to environmental uctuations. In the genomes of both A. thaliana (a model dicot; genome size 129 Mb) and O. sativa (a model monocot; genome size 430 Mb), UTRs represent approximately 1819% of the transcribed sequence, which further substan- tiates their signicance in plants. Although the O. sativa genome is three-times larger than that of A. thaliana, the proportions of 5 0 and 3 0 UTRs are almost the same in the two species (Figure 1A); however, their length-wise distribution indicates greater complexity in rice. In general, UTRs are longer in rice than in arabidopsis, with the average lengths of 5 0 and 3 0 UTRs being 259 bp and 469 bp, respectively, in rice but only 155 bp and 242 bp in arabidopsis (Figure 1B). In arabidopsis, using reciprocal best hits that compare cross-species protein sequences, we identied a total of 7933 genes having orthologs in the rice genome. Since orthologs have common ancestors and diverged during speciation, they have high sequence similarity in the protein coding region. However, rice UTRs (especially 3 0 UTRs) are Highlights Plant mRNAs contain untranslated regions (UTRs) that signicantly enhance the coding capacity of the genome by producing multiple mRNA variants from the same gene. Approximately 1819% of transcribed sequences in Arabidopsis thaliana and Oryza sativa genomes are UTRs. The length-wise distribution of UTRs indi- cates greater complexity in rice than in A. thaliana. In general, UTRs are longer in rice than in Arabidopsis. At the whole-genome level, UTRs pro- vide functional specicity to genes in a length-dependent manner. Despite the conserved nature of many UTR-associated cis-regulatory ele- ments across plant families, the exis- tence of plant-specic mechanisms is expected. Multiple ways exist by which UTR ele- ments can exert control over gene expression and the functioning of translated proteins. 1 Shanghai Center for Plant Stress Biology and Center of Excellence for Molecular Plant Sciences, Chinese Academy of Sciences, Shanghai, China 2 Department of Horticulture and Landscape Architecture, Purdue University, West Lafayette, IN, USA 3 Permanent address: Nuclear Agriculture and Biotechnology Division, Bhabha Atomic Research Centre, Mumbai 400085, India *Correspondence: [email protected], [email protected] (A.K. Srivastava), [email protected] (J.-K. Zhu). TRPLSC 1624 No. of Pages 12 Trends in Plant Science, Month Year, Vol. xx, No. yy https://doi.org/10.1016/j.tplants.2017.11.003 1 © 2017 Elsevier Ltd. All rights reserved.
Transcript
Page 1: UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We discuss different modes of UTR-based regula-tion with emphasis on genes that regulate

TRPLSC 1624 No. of Pages 12

Review

UTR-Dependent Control of Gene Expressionin Plants

Ashish Kumar Srivastava,1,3,* Yuming Lu,1 Gaurav Zinta,1 Zhaobo Lang,1 and Jian-Kang Zhu1,2,*

HighlightsPlant mRNAs contain untranslatedregions (UTRs) that significantlyenhance the coding capacity of thegenome by producing multiple mRNAvariants from the same gene.

Approximately 18–19% of transcribedsequences in Arabidopsis thaliana andOryza sativa genomes are UTRs. Thelength-wise distribution of UTRs indi-cates greater complexity in rice than inA. thaliana. In general, UTRs are longerin rice than in Arabidopsis.

At the whole-genome level, UTRs pro-

Throughout their lives, plants sense many developmental and environmentalstimuli, and activation of optimal responses against these stimuli requiresextensive transcriptional reprogramming. To facilitate this activation, plantmRNA contains untranslated regions (UTRs) that significantly increase thecoding capacity of the genome by producing multiple mRNA variants from thesame gene. In this review we compare UTRs of arabidopsis (Arabidopsisthaliana) and rice (Oryza sativum) at the genome scale to highlight theircomplexity in crop plants. We discuss different modes of UTR-based regula-tion with emphasis on genes that regulate multiple plant processes, includingflowering, stress responses, and nutrient homeostasis. We demonstrate func-tional specificity in genes with variable UTR length and propose futureresearch directions.

vide functional specificity to genes in alength-dependent manner.

Despite the conserved nature of manyUTR-associated cis-regulatory ele-ments across plant families, the exis-tence of plant-specific mechanisms isexpected.

Multiple ways exist by which UTR ele-ments can exert control over geneexpression and the functioning oftranslated proteins.

1Shanghai Center for Plant StressBiology and Center of Excellence forMolecular Plant Sciences, ChineseAcademy of Sciences, Shanghai,China2Department of Horticulture andLandscape Architecture, PurdueUniversity, West Lafayette, IN, USA3Permanent address: NuclearAgriculture and BiotechnologyDivision, Bhabha Atomic ResearchCentre, Mumbai 400085, India

*Correspondence:[email protected],[email protected](A.K. Srivastava),[email protected] (J.-K. Zhu).

Significance of UTRs in Plants: Genome-Scale EvidenceBecause plants are sessile, they have to tolerate many abiotic and biotic stresses. Besides,plants undergo major growth transitions during their life cycle from seed germination tovegetative growth, flowering, and seed development. Both the activation of stress tolerancemechanisms and developmental transitions require extensive transcriptional reprogramming,which is controlled at transcriptional and post-transcriptional levels. Transcriptional control ismediated by transcription factors while post-transcriptional control is regulated by a complexset of cis- and trans-acting elements [1–3].Cis-acting elements for post-transcriptional control,which are generally present in UTRs (see Glossary) of mRNAs, can regulate various mRNAproperties like stability, transport, and translation efficiency and also the functioning andsubcellular localization of the translated proteins [4]. Thus, UTRs represent important geneticcomponents that regulate processes at the interface of mRNAs and proteins. In addition, owingto their characteristics of having multiple variants from the same gene, UTRs can increase thecoding capacity of the genome. This provides versatility to the gene expression system andthereby helps plants adapt to environmental fluctuations. In the genomes of both A. thaliana (amodel dicot; genome size 129 Mb) and O. sativa (a model monocot; genome size 430 Mb),UTRs represent approximately 18–19% of the transcribed sequence, which further substan-tiates their significance in plants. Although the O. sativa genome is three-times larger than thatof A. thaliana, the proportions of 50 and 30 UTRs are almost the same in the two species(Figure 1A); however, their length-wise distribution indicates greater complexity in rice. Ingeneral, UTRs are longer in rice than in arabidopsis, with the average lengths of 50 and 30

UTRs being 259 bp and 469 bp, respectively, in rice but only 155 bp and 242 bp in arabidopsis(Figure 1B). In arabidopsis, using reciprocal best hits that compare cross-species proteinsequences, we identified a total of 7933 genes having orthologs in the rice genome. Sinceorthologs have common ancestors and diverged during speciation, they have high sequencesimilarity in the protein coding region. However, rice UTRs (especially 30 UTRs) are

Trends in Plant Science, Month Year, Vol. xx, No. yy https://doi.org/10.1016/j.tplants.2017.11.003 1© 2017 Elsevier Ltd. All rights reserved.

Page 2: UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We discuss different modes of UTR-based regula-tion with emphasis on genes that regulate

TRPLSC 1624 No. of Pages 12

Glossary7-Methylguanosine (m7G):methylation of guanosine in the 50

terminus of mRNAs.Adenosine methylation (m6A): aconserved epitranscriptomicmodification of mRNA that is directedby a methyltransferase complexcontaining METTL3 as the SAM-binding subunit.Alternative polyadenylation (APA):

comparatively longer than arabidopsis UTRs, even for orthologous genes (Figure 1B). Thus,despite the conserved nature of many UTR-associated cis-regulatory elements across plantfamilies [5], the existence of plant species-specific mechanisms is expected. Here we provide acomprehensive overview of UTR-dependent regulatory mechanisms that control gene expres-sion as well as the fate of translated proteins in plants. Using enrichment analysis of UTR-containing genes at the whole-genome level (Box 1), we highlight the need to include UTR-related regulatory mechanisms in gene function studies, especially in crops. We also discussfuture research directions that will increase the application of UTR-based research to cropimprovement.

a process in which one pre-mRNAtranscript is processed to producemultiple mRNAs that generally differin their 30-UTR length.Alternative splicing (AS): a processby which pre-mRNAs are processeddifferentially to generate multipleisoforms that differ in biologicalactivity.Coding sequence (CDS): thesequence of a transcript that istranslated into protein. A CDS startswith a start codon (AUG) and stopsat one of the three stop codons(UAA, UGA, or UAG).COOLAIR: the collective term for agroup of antisense long noncodingtranscripts that undergo APA torepress the expression of theflowering time regulator FLC.Exon junction complex (EJC): aprotein complex formed at thejunction of two exons on a pre-mRNA strand joined together duringRNA splicing.Intron-mediated enhancement(IME): the ability of an intronsequence to enhance the expressionof a gene containing that intron.Nonsense-mediated decay(NMD): a translation-coupledmechanism that eliminates mRNAscontaining PTCs. Its main function isto reduce errors in gene expression.Open-reading frame (ORF): a partof the reading frame that haspotential to be translated; usuallycontains start (AUG) and stop (UAA,UAG, or UGA) codons.Termination codon (TC): a stopcodon (UAA, UAG, or UGA) is anucleotide triplet within mRNA thatsignals the termination of translationinto proteins.Untranslated region (UTR): refersto the two sections on each side of aCDS on an mRNA strand.Upstream open reading frame(uORF): an ORF within the 50 UTR ofan mRNA.

Different Modes of UTR-Mediated Control of Gene ExpressionUTRs are known to control gene expression and protein function via a wide range of mecha-nisms (Figure 2). The major mechanisms and their involvement in the regulation of multiple plantprocesses are briefly discussed below. Important genes that have been functionally charac-terized to be regulated by UTRs are listed in Table 1.

Alternative PolyadenylationIn alternative polyadenylation (APA) an mRNA transcript is processed to produce multipleisoforms that generally differ in their 30-UTR length. If the alternative polyA signal falls upstreamof the stop codon, APA can also produce a non-functional protein or a protein with an entirelydifferent function [6]. More than 80% of the genes in humans [7] and more than 75% of thegenes in arabidopsis [8] are known to produce multiple mRNA isoforms through APA. APAregulates multiple plant processes, and a well-studied example is flowering-time regulationthrough the FCA-, FPA-, and Flowering Locus C (FLC)-mediated autonomous pathway [9].FCA and FPA are RNA-binding proteins that promote proximal polyadenylation of FLC anti-sense RNA to decrease FLC expression and induce flowering [10]. In addition, FCA and FPAmRNAs undergo APA to generate either full-length coding transcripts (using distal polyA sites)or truncated transcripts (using proximal polyA sites). FCA and FPA provide feedback control totheir own levels by favoring the usage of proximal polyA sites [10,11]. The recently character-ized HLP1A regulates distal-to-proximal polyA site shift in FCA, which in turn upregulates FLCto delay flowering [12]. FLC expression is repressed by a group of antisense long noncodingtranscripts (collectively termed COOLAIR) that undergo APA. Promotion of proximal poly-adenylation of COOLAIR is directly linked to reduced FLC expression [13]. Besides flowering,APA also regulates mRNA transport to the surface of organelles to facilitate co-translationalimport of the protein into organelles. This can be observed in the case of VDAC3, which is amember of the voltage-dependent anion channel family in arabidopsis and produces twomRNA variants with a short or long UTR. The most-distal 140-nt 30 sequence of the long UTR isessential and sufficient for targeting VDAC3 mRNA or any other unrelated mRNA to themitochondrial surface for efficient co-translational import of the protein into mitochondria[14]. APA also regulates agriculturally important traits. A naturally occurring mutant, SEMI-DWARF AND HIGH-TILLERING (sdt), has been identified in which OsmiR156h precursorexpression is upregulated via APA to promote tillering capacity and grain yield in rice [15].

APA represents an important link between DNA-methylation- and histone-modification-depen-dent regulation of gene expression. Transposons and other repetitive elements are silenced byDNA methylation and heterochromatic histone modifications such as H3K9 dimethylation.Increase in Bonsai Methylation 1 (IBM1) is a histone demethylase that regulates the expressionof many genes by removing the repressive H3K9me2 mark from gene bodies [16]. Arabidopsisibm1 mutant plants display a wide range of developmental defects including decreased

2 Trends in Plant Science, Month Year, Vol. xx, No. yy

Page 3: UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We discuss different modes of UTR-based regula-tion with emphasis on genes that regulate

TRPLSC 1624 No. of Pages 12

Arabido

psis

thaliana

Oryza

sava

(A) (B)

<0.5k98%

126 bp

0.5–1k2%

614 bp

1–2k0.16%

1272 bp >2k0%

<0.5k96%

217 bp

0.5–1k4%

649 bp

1–2k0.39%

1330 bp >2k0%

<0.5k92%

155 bp

0.5–1k6%

673 bp

1–2k1.5%

1373 bp>2k

0.2%2656 bp

<0.5k74%

316 bp

0.5–1k20%

667 bp

1– 2k5%

1344 bp>2k1%

2588 bp

Length-wise distribu on and average UTR length

At whole-genome level In orthologs

CDS/UTR percentageat whole-genome

level

CDS81%

5′ UTR7%

3′ UTR12%

CDS82%

5′ UTR6%

3′ UTR12%

<0.5k96.5%136 bp

0.5–1k3.2%

627 bp

1–2k0.19%

1267 bp>2k

0.02%2521 bp

<0.5k95.1%215 bp

0.5–1k4.1%

649 bp

1–2k0.6%

1390 bp>2k

0.09%2338 bp

<0.5k87.7% 158 bp

0.5–1k8.4%

680 bp

1–2k3%

1366 bp>2k

0.8%2684 bp

<0.5k73%

293 bp

0.5–1k19%

672 bp

1–2k6%

1347 bp

>2k1.6%

2701 bp

5′ UTR

5′ UTR

3′ UTR

3′ UTR

5′ UTR

5′ UTR

3′ UTR

3′ UTR

Figure 1. Percentages of Coding Sequence (CDS) and Untranslated Region (UTR) Elements in Arabidopsis and Rice and their Lengths. (A) Genome-scale survey showing the percentages of transcript bases involved in protein coding (CDS) and in gene regulation (UTR) in Arabidopsis thaliana and Oryza sativa. (B)Percentages of 50 and 30 UTRs in different length categories along with average UTR length (bp) at the whole-genome level and in orthologs.

production of stomatal lineage cells, which is associated with reduced expression of ERECTA-family genes [17]. IBM1 has a large intron that harbors a heterochromatic repeat element [18].Anti-silencing 1 (ASl1), a bromodomain-containing RNA-binding protein and Enhanced DownyMildew 2 (EDM2), a chromatin regulator that recognizes both repressive and activation histonemarks in the intronic repeat, are required for the production of functional IBM1 transcripts . Inwild-type arabidopsis plants, a distal polyA signal in the IBM1 pre-mRNA is used, leading to theformation of functional IBM1 transcript. In asl1 and edm2 loss-of-function mutants, a proximalpolyA signal upstream of the intronic heterochromatic repeat is used, resulting in truncatedIBM1 transcripts [18,19]. Besides this epigenetic regulation, APA is also regulated geneticallyvia cleavage and polyadenylation specificity factor 30 (CPSF30). Because of the variation inpolyA signal selection in genes related to stress and defense, the arabidopsis oxt6 mutant(defective in CPSF30) has an oxidative stress-sensitive phenotype [20]. Thus, APA is regulated

Box 1. UTR Length-Dependent Functional Specificity in Arabidopsis and Rice

Arabidopsis and rice are the model plants for two major angiosperm subclasses: dicots and monocots, respectively. To date most genes that have beencharacterized for UTR-based regulation belong to arabidopsis (see Table 1 in main article). The cis-elements and up- and downstream sequence contexts formost UTR-based regulatory mechanisms are more or less conserved in plants [45,49,74,75]. However, UTR-based regulatory processes proposed forarabidopsis genes may not hold true for rice. In both arabidopsis and rice, distinct functional categories are enriched in genes with short (1–500 bp), medium(501–1000 and 1001–2000 bp), and long (>2000 bp) UTRs (Figure I). This suggests that, besides regulating gene expression, UTRs may provide functionalspecificity to genes in a length-dependent manner (Figure I. Recently, a subset of dehydration-stress-related genes has been demonstrated to undergo 30 UTRextension to regulation the expression of neighboring genes through long noncoding RNAs. This further substantiates the significance of UTR length in plantadaptation to stress conditions [76]. Furthermore, the functional enrichments within UTR length categories differ between arabidopsis and rice. For instance,genes with 50 or 30 UTRs in the 1–500-bp category mediate stress responses in arabidopsis, but genes in this category mediate transport and metabolism in rice.This further strengthens the notion that conclusions derived from UTR-related functional studies in arabidopsis cannot be directly applied to rice or other crops.Thus, large-scale projects like ‘RICE 2020’, which aims to understand the function of all proteins derived from the primary gene pool in rice [77], should becomplemented with parallel studies that determine the UTR-dependent regulation of all genes. Such a holistic approach is needed to ensure effective utilization offunctional genomics studies for rice genetic improvement.

Trends in Plant Science, Month Year, Vol. xx, No. yy 3

Page 4: UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We discuss different modes of UTR-based regula-tion with emphasis on genes that regulate

TRPLSC 1624 No. of Pages 12

0 5 10 15

DNA metabolic processCell cycle

Signal transduc�onCarbohydrate metabolic process

Cell deathCellular protein modifica�on process

AbscissionPost-embryonic development

Carbohydrate metabolic processNucleobase -containing compound metabolic process

Biosynthe�c process

Catabolic processCellular component organiza�on

TransportMetabolic process

Cellular process

0 10 20 30 40 50

Response to salt stressResponse to cadmium ion

Posi�ve regula�on of transcrip�onRegula�on of hypersensi�ve response

Protein targe�ng to membraneRegula�on of jasmonic acid signaling

Regula�on of systemic acquired resistancemRNA polyadenyla�on

Circadian rhythmResponse to symbio�c fungus

Unsaturated fa�y acid biosynthe�c processmRNA splicing via spliceosome

Anthocyanin accumula�on in �ssuesIsopentenyl diphosphate biosynthe�c process

Lipoate metabolic process Phototropism

Cold acclima�onRegula�on of anion channel ac�vity

Nega�ve regula�on of molecular func�onTranscrip�on from plas�d promoter

0 10 20 30 40 50

Response to salt stressResponse to cadmium ion

Response to coldCysteine biosynthe�c process

Posi�ve regula�on of transcrip�onRegula�on of transcrip�on

Ac�n nuclea�onDefense response by callose deposi�on

Determina�on of bilateral symmetryTrichome morphogenesis

Regula�on of ion transmembrane transportPhospholipid catabolic process

Leaf shapingRegula�on of anion transport

Fa�y acid elonga�onSister chroma�d cohesion

Cytokinin-ac�vated signaling pathwayMeio�c chromosome segrega�onChroma�n silencing by small RNA

Methyla�on-dependent chroma�n silencing

0 5 10 15

Carbohydrate metabolic processTransla�on

Cellular component organiza�onTransport

Cellular process

Carbohydrate metabolic processNucleobase-containing compound metabolic process

Signal transduc�onBiosynthe�c process

Cellular protein modifica�on process

Cell deathResponse to stress

Cellular protein modifica�on process

Carbohydrate metabolic processTransla�on

Cellular component organiza�onTransport

Cellular process

1–500 bp 501–1000 bp 1001–2000 bp >2000 bp

Arabidopsis thaliana Oryza sa�va

3′ U

TR le

ngth

5′ U

TR le

ngth

P value (−log10) P value (−log10)

Figure I. Functional Enrichment Analysis of Untranslated Region (UTR)-Containing Genes in Arabidopsis and Rice. UTR-containing genes ofarabidopsis and rice genomes were sorted according to their 30- and 50-UTR lengths before functional enrichment analysis was independently performed for geneswith short (1–500 bp), medium (501–1000 and 1001–2000 bp), and long (>2000 bp) UTRs using the Bioconductor package topGO in R software (version 3.1.1). Onthe basis of P values, the top five GO terms related to biological processes are represented independently for 30 and 50 UTRs. The colors of bars – blue (1–500 bp),black (501–1000 bp), green (1001–2000 bp), and red (>2000 bp) – signify variable UTR length. In rice relatively fewer GO terms were enriched, especially in themedium- (1001–2000 bp) and long- (>2000 bp) UTR categories.

by both genetic and epigenetic mechanisms and is important for plant processes like flowering,grain yield, and stress tolerance.

Riboswitching and Other UTR-Mediated Regulation by Small MoleculesRiboswitches are metabolite-sensing gene-control elements that are typically located in non-coding portions of mRNAs. In eukaryotes a well-characterized riboswitch senses the vitamin B1derivative thiamine pyrophosphate (TPP). TPP biosynthesis is regulated through THIC, whichcatalyses the conversion of 5-aminoimidazole ribotide to hydroxymethyl pyrimidine phosphate,a key step in the synthesis of thiamine and ultimately TPP [21]. THIC represents a typicalexample of gene expression control through the combination of a riboswitch and APA. Anintron element outside the stop codon of THIC contains a primary polyA site followed by a

4 Trends in Plant Science, Month Year, Vol. xx, No. yy

Page 5: UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We discuss different modes of UTR-based regula-tion with emphasis on genes that regulate

TRPLSC 1624 No. of Pages 12

5ʹ UTR

3ʹ UTR

ExonIntronpolyA site

Metabolite

CAP

m6A

uORF

Transla�on repression

Stalled ribosome

mRNA decay

PABPC1

eRFs

NMD factors

Metabolite binding promotes long 3ʹ UTR

Metabolite absence promotes short 3ʹ UTR

(B): Riboswitching

(A): Alterna�ve polyadenyla�on

Distal polyA produces long 3ʹ UTR

Proximal polyA produces short 3ʹ UTR

polyA signal in exon produces truncated transcript

(E): Nonsense-mediated decay

AAAAAAAAAAAA TC

Transla�ontermina�on NMD ac�va�on

AAAAAAAAAAAA

PTC

m6A in 5ʹ UTR promotes CAP-independent transla�on

(C): Adenosine methyla�on

(D): Short-pep�de transla�on

uORF stalls ribosomes and represses transla�on

uORF promotes mRNA decay through NMD

(F): Alterna�ve splicing

Intron reten�on in 5ʹ UTR can promote/repress transla�on

Intron reten�on in 3ʹ UTR modulates miRNA-mediated cleavage

Figure 2. Various Modes of Untranslated Region (UTR)-Dependent Regulation of Genes. (A) Alternative polyadenylation. Either more than one polyA site isutilized to produce mRNA variants that differ in 30-UTR length or, if the polyA is located upstream of the stop codon, truncated transcripts are produced. (B)Riboswitching. 30 or 50 UTRs form folding structures that sense a particular metabolite andmodulate transcript stability. (C) Adenosine methylation (m6A). The presenceof m6A in the 50 UTR promotes CAP-independent translation. (D) Short-peptide translation. A short open reading frame in the 50 UTR (uORF) can either repress thetranslation of the main ORF or induce mRNA decay via the nonsense-mediated decay (NMD) pathway. (E) NMD. A premature termination codon (PTC) upstream of theregular termination codon (TC) recruits NMD factors that mediate mRNA decay. (F) Alternative splicing. Retention of intronic elements in the 50 UTR can either promoteor repress translation, while retention of intronic elements in the 30 UTR can modulate miRNA-mediated cleavage.

riboswitch element that can bind TPP, and then a secondary polyA site at the 30 end of thetranscript. The primary polyA site is also flanked by a splicing signal whose accessibility isregulated through TPP binding with the riboswitch. In the absence of TPP the splicing site isinaccessible; hence, the primary polyA site remains activated. The resulting transcript has ashorter 30 UTR that is more stable; as a consequence THIC expression is increased, whichfacilitates TPP biosynthesis. When TPP is available, it binds to the riboswitch and, because ofconformational change, the splice site becomes accessible. As a result the primary polyA site isspliced out, which extends transcription until the secondary polyA site located 1 kb away from

Trends in Plant Science, Month Year, Vol. xx, No. yy 5

Page 6: UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We discuss different modes of UTR-based regula-tion with emphasis on genes that regulate

TRPLSC 1624 No. of Pages 12

Table 1. Summary of Genes Functionally Characterized for UTR-Dependent Regulation

Gene Organism UTR-dependent regulation Refs

APA

Antisense FLC Arabidopsisthaliana

Proximal polyA in antisense FLCRNAs represses FLC expression [10]

FCA and FPA A. thaliana Activation of distal polyA in FCA and FPA produces functionalFCA and FPA to suppress FLC expression

[10,11]

COOLAIR A. thaliana Proximal polyA in COOLAIR reduces FLC expression [13]

VDAC3 A. thaliana Most-distal 140-nt sequence in 30 UTR is essential and sufficientfor targeting VDAC3 mRNA into mitochondria

[14]

OsmiR156h precursor Oryza sativa Shortening of 30 UTR induces expression of both precursor andmature forms of OsmiR156h

[15]

IBM1 A. thaliana ASl1-mediated activation of distal polyA produces functionalIBM1 transcripts

[18]

Riboswitching

THIC A. thaliana Activation of proximal polyA produces stable THIC transcriptswith short 30 UTRs that promote vitamin B1 biosynthesis

[22]

EBF1 and EBF2 A. thaliana Binding of EIN2 to 30 UTRs of EBF1 and EBF2 repressestranslation to promote ethylene signaling

[25]

Short-peptide translation

bZIP11 A. thaliana Translation of uORF in response to sucrose reduces bZIP11protein level

[37]

NAC096 A. thaliana uORF represses the expression of the main ORF [39]

Polyamine oxidase-2 A. thaliana uORF mediates translational repression of the main ORF [40]

SAC51 A. thaliana uORF mediates translational repression of the main ORF [41]

AdoMetDC1 A. thaliana uORF causes ribosomal arrest and promotes mRNA decaythrough NMD

[44]

NIP5;1 A. thaliana Translation of minimum AUG-stop uORF affects translationinitiation in a B-dependent manner

[45]

NMD

FLM A. thaliana High temperature reduces FLM expression by producing NMD-sensitive variants

[59]

Alternative splicing

PSY A. thaliana PSY variant with long 50 UTR can sense carotene flux, whichattenuates translation to reduce PSY protein level

[62]

ZIF2 A. thaliana Intron retention in 50 UTR favors the translation of ZIF2 transcripts [63]

FRD3 A. thaliana Alternative transcription initiation sites are utilized to produceFRD3 transcripts with shorter 50 UTRs for increased translationefficiency

[73]

MHX A. thaliana An intronic element in the 50 UTR of MHX considerably increasestranslation efficiency

[66]

BES1 A. thaliana Alternative transcriptional start site is used to produce long BES1variant with high nuclear localization signal (NLS) activity

[67]

Osa-miR7695 O. sativa AS in the target gene OsNramp6 produces variants with short 30

UTRs that are immune to Osa-miR7695-mediated cleavage[71]

Ath-miR400 A. thaliana AS in the 50 UTR controls removal of the miR400 hairpin from thehost gene and thereby affects the expression of mature miR400

[72]

6 Trends in Plant Science, Month Year, Vol. xx, No. yy

Page 7: UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We discuss different modes of UTR-based regula-tion with emphasis on genes that regulate

TRPLSC 1624 No. of Pages 12

the primary polyA site. The resulting transcript with a longer 30 UTR is relatively unstable andconsequently THIC expression is decreased [22,23].

A UTR element was recently found to be involved in regulating ethylene-dependent signalingthrough translational repression. The mechanism of ethylene signaling has been well describedin both arabidopsis and rice [24]. In brief, the ethylene signal is propagated through theendoplasmic reticulum (ER)-located membrane protein ETHYLENE INSENSITIVE 2 (EIN2).On activation the C-terminal domain of EIN2 (CEND) is cleaved and transported to the nucleuswhere it stabilizes two master transcription factors, ETHYLENE INSENSITIVE 3 (EIN3) and itsparalog EIN3-LIKE 1 (EIL1), by repressing their proteasomal degradation caused by two F-boxproteins, EIN3-BINDING F-BOX 1 (EBF1) and EBF2. EIN3 and EIL1 are mainly responsible forethylene-responsive gene expression. In addition to the EIN2–EBF1/2 interaction at the proteinlevel in the nucleus, an independent mode of regulation also exists wherein EIN2 interacts withthe 30 UTR of EBF1/2 to modulate their translation. The first clue for this came from a study inwhich ectopic expression of the EBF1 or EBF2 30 UTR generated strong ethylene-insensitivephenotypes by promoting the translation of endogenous EBF1/2 mRNAs [25]. When EIN2binds to the EBF1/2 30 UTR, the complex directs the transcripts to cytoplasmic processing-bodies; therefore, higher expression of the 30-UTR element alone is sufficient to deregulateEIN2-mediated translational repression of EBF1/2 [25].

By using a combination of genome and transcriptome sequencing approaches, 31 newriboswitches (including TPP) were identified from a marine phytoplankton [26]. Given the largeproportion of UTR elements in the model plants arabidopsis and rice (Figure 1), it is likely thatmany more metabolites and small molecules that can be sensed in a UTR-dependent mannerremain to be identified in higher plants.

Adenosine MethylationAdenosine methylation (m6A) was recently identified as a conserved epitranscriptomicmodification of mRNAs. In the 50 UTR, m6A facilitates CAP-independent protein translation.The translation of mRNAs begins with the binding of eukaryotic initiation factor 4E (eIF4E) to the50 7-methylguanosine (m7G) cap, which in turn recruits the small ribosomal subunit (40S) toenable efficient translation. Using a combination of in vitro reconstitution and translation assays,it has been shown that a single 50-UTR m6A can directly bind to eukaryotic initiation factor 3(eIF3), which is sufficient to recruit 43S ribosomal complex to initiate translation even in theabsence of eIF4E [27]. Due to its dramatic increase under cellular stress, m6Amodification in 50

UTRs is considered a defense strategy to promote CAP-independent protein translation [27].Using the Hsp70 mRNA, the m6A effect has been shown to be sequence dependent. Hence,not all m6A modifications are equally efficient in stimulating protein translation [28].

How m6A regulates plant gene expression is an active area of research. MTA (At4g10760), aMETTL3 ortholog in arabidopsis, is an active component of the m6A methyltransferasecomplex [29]. MTA is expressed at high levels in dividing tissues, such as developing seeds,shoot meristems, and emerging lateral roots. MTA disruption causes multiple growth defects,including reduced apical dominance, organ abnormality, and increased trichome branching[30]. FKBP12 INTERACTING PROTEIN37 KD (FIP37) is another core component of the m6Amethyltransferase complex and fip37 mutants exhibit massive over-proliferation of shootmeristems and transcriptome-wide loss of m6A [31]. Using comparative transcriptome-widem6A profiling in two arabidopsis accessions (Can-0 and Hen-16), it has been confirmed thatm6A modification is enriched in both 50 and 30 UTRs, as was previously observed in mammals[32]. However, little information is available about how m6A modification regulates plant

Trends in Plant Science, Month Year, Vol. xx, No. yy 7

Page 8: UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We discuss different modes of UTR-based regula-tion with emphasis on genes that regulate

TRPLSC 1624 No. of Pages 12

development and responses to various abiotic/biotic stress conditions. Thus, additionalresearch involving more precise measurement of m6A is required to clarify the function ofthis dynamic epi-modification in the regulation of various aspects of plant growth anddevelopment.

Short-Peptide TranslationThe 50-UTR region of many eukaryotic mRNAs contains a short open-reading frame (ORF)termed an upstream ORF (uORF). Initiation of translation from an uORF can repress transla-tion from a downstream ORF by stalling ribosome movement. Also, if the uORF stop codon islocated upstream of the exon junction complex (EJC), it can be recognized as a prematuretermination codon (PTC) to activate mRNA degradation through the nonsense-mediateddecay (NMD) pathway [33]. In arabidopsis�35% genes (>9000) have at least one uORF [34].Recent research using high-resolution ribosome profiling confirmed that 187 genes with anuORF having an AUG start codon are translated [35]. However, this number might be anunderestimate, because uORF translation can also start from unconventional start codons. Thetranslation efficiency of uORFs is sequence dependent and is also regulated under stressconditions. In maize (Zea mays), uORF translation is globally enhanced under drought stress,which helps to shape the proteome under stress [36]. The functional relevance of uORFs is welldemonstrated for sucrose-induced repression of translation (SIRT) of basic leucine zipper(bZIP)-type transcription factor family genes. For instance, the arabidopsis ATB2/AtbZIP11 hasa highly conserved uORF within its exceptionally long 50 UTR that is translated in response tosucrose resulting in the reduction of AtbZIP11 protein [37]. Based on this information, SIRTderegulation was attempted in tomato and 50UTR-free versions of SlbZIP1 and SlbZIP2 wereover-expressed for increasing sweetness [38]. Similar uORFs have been identified in othergenes – like NAC096 [39], Polyamine oxidase-2 [40], suppressor-of-ACAULIS5 (SAC51) familymembers [41], TL1-binding factor-1 (TBF1) [42], and homeodomain–leucine zipper transcrip-tion factor (AtHB1) [43] – that can generate regulatory peptides to repress the translation ofmain ORFs. The translation arrest can also be coupled with mRNA decay through the NMDpathway. This has been demonstrated using the polyamine-responsive AdoMetDC1 (S-adeno-sylmethionine decarboxylase) uORF-based reporter system [44]. Recently, a role has beenassigned for a minimal possible uORF termed an ‘AUG-stop’ in the context of boron (B)deficiency. In arabidopsis, NIP5;1, which encodes a boric acid channel required for normalgrowth under B-deficient conditions, contains an AUG-stop-like uORF. Under normal con-ditions, ribosomes halt at the AUG-stop, which suppresses translation and promotes mRNAdegradation. A conserved sequence at the 50 edge of the stalled ribosome is required for mRNAdegradation. Under B-deficient conditions, however, translation re-initiation is promoted, whichleads to the synthesis of functional NIP5;1 [45]. Thus, translation of short peptides from uORFsprovides versatility to the translation machinery and thereby helps plants to adapt to changingenvironmental conditions.

Nonsense-Mediated DecayNMD is a surveillance mechanism that detects and degrades transcripts with a PTC to preventthe formation of non-functional or aberrant proteins. Although the molecular mechanism ofNMD is conserved between mammals and plants [46,47], some plant-specific features havealso been reported [48]. Various models have been proposed to explain NMD activation[49,50]. In brief, during steady-state translation (which follows the first or pioneer round oftranslation), the cap-binding proteins eIF4E and eIF4G form a complex with polyA-bindingprotein cytoplasmic 1 (PABPC1), which results in a circular conformation of the mRNA. In thecircular state, when a ribosome encounters a stop codon eukaryotic release factors like eRF1andeRF3 are recruited; these interact with eIF4G–PABPC1 and terminate translation. This

8 Trends in Plant Science, Month Year, Vol. xx, No. yy

Page 9: UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We discuss different modes of UTR-based regula-tion with emphasis on genes that regulate

TRPLSC 1624 No. of Pages 12

interaction, however, is inhibited by long 30 UTRs and consequently eRF3 interacts with up-frameshift1 (UPF1) and other NMD factors to make the SMG1–Upf1–eRF1–eRF3 (SURF)complex, which triggers translation repression and mRNA decay. Introns located �50–55nt downstream of termination codons (TCs) can also activate NMD, even in the first round oftranslation. After intron splicing, EJC proteins are deposited about 24 nt upstream of exon–exon junctions. If the termination occurs >50–55 nt upstream of EJC, terminating ribosomescannot physically remove EJC, which therefore interacts with UPF1 via UPF3 and UPF2 andtriggers NMD. UPF1 can also directly bind mRNA before translation in a 30-UTR length-dependent manner and activate NMD. Thus, multiple signals can activate NMD. Other mech-anisms, like translation read-through (translation elongation until the next in-frame stop codon),cause mRNA to escape NMD. In arabidopsis eRF1 is the only gene with all of the cis-elementsrequired for read-though. A low eRF1 level promotes read-through, which masks the NMDsignal at the 30 UTR of eRF1 mRNA and hence increases the eRF1 protein level [51]. Thus, anautoregulatory feedback mechanism of eRF1 is important for the regulation of NMD and overallprotein translation.

NMD plays a critical role in integrating plant responses to various environmental cues [52–56].An interesting example concerns high-temperature-induced flowering, which is controlled bythe MADS-box transcription factors FLOWERING LOCUS M (FLM) and SHORTVEGETATIVEPHASE (SVP). FLM produces two main FLM protein splice variants, FLM-b and FLM-d, whichcompete with the floral repressor SVP. At normal growth temperatures, the SVP–FLM-bcomplex predominates and prevents flowering. At high temperatures FLM-d increases atthe expense of FLM-b, and because the SVP–FLM-d complex is impaired in DNA binding,flowering is accelerated [57]. This model has not been supported, however, by recent findingssuggesting that temperature differences in flowering time can be explained by FLM-b onlyrather than the FLM-b:FLM-d ratio [58]. Another study has demonstrated that thermal inductionof flowering is abolished in upfmutants, which suggests active involvement of NMD [59]. Novelthermosensitive splice sites have been identified in the FLM locus and temperature-dependentalternative splicing (AS) was shown to produce NMD-sensitive variants of FLM that can bedegraded through the NMD pathway thereby reducing FLM expression to accelerate flowering.The temperature sensitivity of various arabidopsis natural accessions was also correlated withthe NMD sensitivity of their FLM 30 UTR sequences [59]. Many more examples of NMD-mediated regulation of various plant processes are likely to be discovered in the future.

Alternative Splicing in UTR ElementsAS refers to the process by which a pre-mRNA is processed differentially to generate multipleisoforms that differ in biological activity [60]. AS in UTRs, especially in 50 UTRs, can regulatetranslation efficiency by producing mRNA variants that differ in terms of uORFs or riboswitches[61]. For instance, the arabidopsis Phytoene synthase (PSY) gene, which catalyzes the rate-limiting step of carotene biosynthesis, produces AS variants with long (ASV1) and short (ASV2)50 UTRs. The long 50 UTR of ASV1 forms riboswitch-like conformations that can sense caroteneflux to attenuate translation. This structure is absent in ASV2, which facilitates rapid carotenesynthesis in response to salt stress or changes in light intensity [62]. A similar mechanism wasreported for zinc-induced facilitator 2 (ZIF2)-mediated control of Zn uptake in arabidopsis; inthis case, however, intron retention in the 50 UTR favors translation [63]. 50 UTRs may alsocontain introns that affect translation through a process called intron-mediated enhance-ment (IME), a conserved but poorly understood mechanism [64,65]. In the arabidopsisMg2+/H+ exchanger (MHX) gene, an internal intronic element has been identified that consid-erably increases translation efficiency [66]. AS in UTRs can also affect protein localization. Forexample, the transcription factor BRI1-EMS-SUPPRESSOR1 (BES1), which promotes

Trends in Plant Science, Month Year, Vol. xx, No. yy 9

Page 10: UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We discuss different modes of UTR-based regula-tion with emphasis on genes that regulate

TRPLSC 1624 No. of Pages 12

Outstanding QuestionsIs UTR length important for the activa-tion of stress-tolerance mechanismsand developmental transitions? Ifyes, what are themechanistic links thatconnect UTR length with geneexpression?

Can we extend the knowledge gainedfrom UTR-related studies in Arabidop-sis thaliana to other crop plants likerice?

Why do plants spend energy to tran-scribe such a large proportion of UTRelements and do not translate it? Is itpossible that UTR function can beextended beyond gene expressioncontrol?

What is the physiological relevance ofepitranscriptomic modifications likem6A at the level of single genes?

At the whole-genome level, can weidentify a subset of genes having allthe required features to undergo mul-tiple modes of UTR-dependentregulation?

brassinosteroid signaling, has short (BES1-S) and long (BES1-L) isoforms. BES1-L, which isproduced from an alternative transcriptional start site located upstream of BES1-S, undergoesAS such that it includes the 50 UTR of BES1-S. This adds 22 amino acid residues to the Nterminus of BES1-L, which provides it with high nuclear localization activity. BES1-L isconsidered a recently evolved isoform that increases plant fitness [67]. 30 UTR-mediatedcontrol of translation is well studied in animals [68] but little information is currently availablein plants except for the few studies in which transgene effectiveness is increased on expressionwith a 30 UTR [69]. AS in UTRs can also produce variants that are insensitive to miRNA-mediated cleavage [70]. To prove this notion, osa-miR7695 has been recently identified in ricethat specifically cleaves a natural resistance-associated macrophage protein 6 variant (OsN-ramp6) having an miRNA cleavage site within its 30-UTR region to provide resistance againstfungal infection; other variants with short 30 UTRs are immune to osa-miR7695-mediatedcleavage [71]. In addition, AS in UTRs can affect miRNA biogenesis. An interesting example ofthis is miR400, which is an intronic miRNA from the 50 UTR of its host gene. Under controlconditions the entire 306-bp intron harboring the miR400 hairpin is excised and processed togenerate mature miR400. Under heat stress, only a small, 100-bp fragment is excised and theremaining 206-bp portion is retained within the host gene. This negatively affects miR400processing and consequently reduces the expression of maturemiR400 [72]. Thus, AS in UTRscan fine-tune gene expression as well as protein levels in multiple ways that help plants adapt tochanging environmental conditions.

Concluding Remarks and Future PerspectivesWe have evaluated UTR-containing genes in arabidopsis and rice at the genome-scale andfound that, although the relative proportions of coding sequence (CDS) and UTRs are similarin arabidopsis and rice, the UTRs of rice genes are longer than those of arabidopsis. UTR-length-dependent functional specialization of genes is also demonstrated. We have reviewed abroad range of UTR-dependent regulatory mechanisms by looking at genes with multiplemRNA variants that differ in UTR length and that are involved in the regulation of flowering,stress responses, nutrient homeostasis, and other processes. Given these findingswe proposea holistic approach to crop functional genomics wherein protein-coding CDS and regulatoryUTR elements are studied in parallel. For newly sequenced complex plant genomes, like thoseof chickpea (Cicer arietinum), tomato (Solanum lycopersicum), Chenopodium, and other cropplants, genome-wide searches are needed to understand how UTRs code their regulatoryfunctions. Although the mechanisms described here can be studied in more detail, it is alsoimportant to discover new mechanisms. For example, a range of 30-UTR-binding proteins inanimals are known to modulate protein translation but this area has been little explored inplants. Also needed are in-depth studies concerning m6A function in the regulation of specificgenes. Overall, it is obvious that our understanding of UTR function remains inadequate andthat future research, especially in crop plants, will provide many new angles on the ways thatUTRs modulate gene expression and protein function (see Outstanding Questions).

AcknowledgmentsThe authors’work has been supported by the Chinese Academy of Sciences and by USNational Institutes of Health grants

R01GM070795 and R01GM059138 (to J-K.Z.). A.K.S. acknowledges the receipt of a postdoctoral fellowship from the

Chinese Academy of Sciences through the President International Fellowship Initiative program. The authors thank X.Q.

Chai, Shanghai Center for Plant Stress Biology for doing GO enrichment analysis.

Supplemental InformationSupplemental information associated with this article can be found, in the online version, at https://doi.org/10.1016/j.

tplants.2017.11.003.

10 Trends in Plant Science, Month Year, Vol. xx, No. yy

Page 11: UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We discuss different modes of UTR-based regula-tion with emphasis on genes that regulate

TRPLSC 1624 No. of Pages 12

References

1. Hernando, C.E. et al. (2017) Transcriptional and post-transcrip-

tional control of the plant circadian gene regulatory network.Biochim. Biophys. Acta 1860, 84–94

2. Floris, M. et al. (2009) Post-transcriptional regulation of geneexpression in plants during abiotic stress. Int. J. Mol. Sci. 10,3168–3185

3. Petrillo, E. et al. (2014) Let there be light: regulation of geneexpression in plants. RNA Biol. 11, 1215–1220

4. Mignone, F. et al. (2002) Untranslated regions of mRNAs.Genome Biol. 3, reviews0004.1–reviews0004.10

5. Vaughn, J.N. et al. (2012) Known and novel post-transcriptionalregulatory sequences are conserved across plant families. RNA18, 368–384

6. Mitra, M. et al. (2015) Alternative polyadenylation can regulatepost-translational membrane localization. Trends Cell Mol. Biol.10, 37

7. Mayr, C. (2016) Evolution and biological roles of alternative30UTRs. Trends Cell Biol. 26, 227–237

8. Guo, C. et al. (2016) A genome-wide study of “non-3UTR” poly-adenylation sites in Arabidopsis thaliana. Sci. Rep. 6, 28060

9. Hepworth, J. and Dean, C. (2015) Flowering locus C’s lessons:conserved chromatin switches underpinning developmental tim-ing and adaptation. Plant Physiol. 168, 1237–1245

10. Hornyik, C. et al. (2010) The spen family protein FPA controlsalternative cleavage and polyadenylation of RNA. Dev. Cell 18,203–213

11. Macknight, R. et al. (2002) Functional significance of the alterna-tive transcript processing of the Arabidopsis floral promoter FCA.Plant Cell 14, 877–888

12. Zhang, Y. et al. (2015) Integrative genome-wide analysis revealsHLP1, a novel RNA-binding protein, regulates plant flowering bytargeting alternative polyadenylation. Cell Res. 25, 864–876

13. Csorba, T. et al. (2014) Antisense COOLAIR mediates the coor-dinated switching of chromatin states at FLC during vernalization.Proc. Natl. Acad. Sci. U. S. A. 111, 16160–16165

14. Michaud, M. et al. (2014) Differential targeting of VDAC3 mRNAisoforms influences mitochondria morphology. Proc. Natl. Acad.Sci. U. S. A. 111, 8991–8996

15. Zhao, M. et al. (2015) Regulation of OsmiR156h through alterna-tive polyadenylation improves grain yield in rice. PLoS One 10,e0126154

16. Miura, A. et al. (2009) An Arabidopsis jmjC domain protein pro-tects transcribed genes from DNA methylation at CHG sites.EMBO J. 28, 1078–1086

17. Wang, Y. et al. (2016) Demethylation of ERECTA receptor genesby IBM1 histone demethylase affects stomatal development.Development 143, 4452–4461

18. Wang, X. et al. (2013) RNA-binding protein regulates plant DNAmethylation by controlling mRNA processing at the intronic het-erochromatin-containing gene IBM1. Proc. Natl. Acad. Sci. U. S.A. 110, 15467–15472

19. Lei, M. et al. (2014) Arabidopsis EDM2 promotes IBM1 distalpolyadenylation and regulates genome DNA methylation pat-terns. Proc. Natl. Acad. Sci. U. S. A. 111, 527–532

20. Thomas, P.E. et al. (2012) Genome-wide control of polyadeny-lation site choice by CPSF30 in Arabidopsis. Plant Cell 24,4376–4388

21. Sudarsan, N. et al. (2003) Metabolite-binding RNA domains arepresent in the genes of eukaryotes. RNA 9, 644–647

22. Wachter, A. et al. (2007) Riboswitch control of gene expression inplants by splicing and alternative 30 end processing of mRNAs.Plant Cell 19, 3437–3450

23. Bocobza, S. et al. (2007) Riboswitch-dependent gene regulationand its evolution in the plant kingdom. Genes Dev. 21,2874–2879

24. Yang, C. et al. (2015) Ethylene signaling in rice and Arabidopsis:conserved and diverged aspects. Mol. Plant 8, 495–505

25. Li, W. et al. (2015) EIN2-directed translational regulation of ethyl-ene signaling in Arabidopsis. Cell 163, 670–683

26. McRose, D. et al. (2014) Alternatives to vitamin B1 uptakerevealed with discovery of riboswitches in multiple marine eukary-otic lineages. ISME J. 8, 2517–2529

27. Meyer, K.D. et al. (2015) 50 UTR m6A promotes cap-independenttranslation. Cell 163, 999–1010

28. Zhou, J. et al. (2015) Dynamic m6A mRNA methylation directstranslational control of heat shock response. Nature 526,591–594

29. Zhong, S. et al. (2008) MTA is an Arabidopsis messenger RNAadenosine methylase and interacts with a homolog of a sex-specific splicing factor. Plant Cell 20, 1278–1288

30. Bodi, Z. et al. (2012) Adenosine methylation in ArabidopsismRNAis associated with the 30 end and reduced levels cause develop-mental defects. Front. Plant Sci. 3, 48

31. Shen, L. et al. (2016) N6-Methyladenosine RNA modificationregulates shoot stem cell fate in Arabidopsis. Dev. Cell 38,186–200

32. Luo, G.Z. et al. (2014) Unique features of the m6A methylome inArabidopsis thaliana. Nat. Commun. 5, 5630

33. Hellens, R.P. et al. (2016) The emerging world of small ORFs.Trends Plant Sci. 21, 317–328

34. von Arnim, A.G. et al. (2014) Regulation of plant translation byupstream open reading frames. Plant Sci. 214, 1–12

35. Hsu, P.Y. et al. (2016) Super-resolution ribosome profiling revealsunannotated translation events in Arabidopsis. Proc. Natl. Acad.Sci. U. S. A. 113, E7126–E7135

36. Lei, L. et al. (2015) Ribosome profiling reveals dynamic transla-tional landscape in maize seedlings under drought stress. Plant J.84, 1206–1218

37. Wiese, A. et al. (2004) A conserved upstream open reading framemediates sucrose-induced repression of translation. Plant Cell16, 1717–1729

38. Sagor, G.H. et al. (2016) A novel strategy to produce sweetertomato fruits with high sugar contents by fruit-specific expressionof a single bZIP transcription factor gene. Plant Biotechnol. J. 14,1116–1126

39. Noh, A.L. et al. (2015) An upstream open reading frame repressesexpression of a tomato homologue of Arabidopsis ANAC096, aNAC domain transcription factor gene, in a peptide sequence-dependent manner. Plant Biotechnol. 32, 157–163

40. Guerrero-Gonzalez, M.L. et al. (2016) Arabidopsis Polyamineoxidase-2 uORF is required for downstream translational regula-tion. Plant Physiol. Biochem. 108, 381–390

41. Cai, Q. et al. (2016) The SAC51 family plays a central role inthermospermine responses in Arabidopsis. Plant Cell Physiol. 57,1583–1592

42. Xu, G. et al. (2017) uORF-mediated translation allows engineeredplant disease resistance without fitness costs. Nature 545,491–494

43. Ribone, P.A. et al. (2017) A uORF represses the transcriptionfactor AtHB1 in aerial tissues to avoid a deleterious phenotype.Plant Physiol. 175, 1238–1253

44. Uchiyama-Kadokura, N. et al. (2014) Polyamine-responsive ribo-somal arrest at the stop codon of an upstream open readingframe of the AdoMetDC1 gene triggers nonsense-mediatedmRNA decay in Arabidopsis thaliana. Plant Cell Physiol. 55,1556–1567

45. Tanaka, M. et al. (2016) The minimum open reading frame, AUG-stop, induces boron-dependent ribosome stalling and mRNADegradation. Plant Cell 28, 2830–2849

Trends in Plant Science, Month Year, Vol. xx, No. yy 11

Page 12: UTR-Dependent Control of Gene Expression in Plants...Dec 01, 2017  · complexity in crop plants. We discuss different modes of UTR-based regula-tion with emphasis on genes that regulate

TRPLSC 1624 No. of Pages 12

46. He, F. and Jacobson, A. (2015) Nonsense-mediated mRNAdecay: degradation of defective transcripts is only part of thestory. Annu. Rev. Genet. 49, 339–366

47. Hug, N. et al. (2016) Mechanism and regulation of the nonsense-mediated decay pathway. Nucleic Acids Res. 44, 1483–1495

48. Shaul, O. (2015) Unique aspects of plant nonsense-mediatedmRNA decay. Trends Plant Sci. 20, 767–779

49. Brogna, S. et al. (2016) The meaning of NMD: translate or perish.Trends Genet. 32, 395–407

50. Zhang, X. and Guo, H. (2017) mRNA decay in plants: bothquantity and quality matter. Curr. Opin. Plant Biol. 35, 138–144

51. Nyiko, T. et al. (2017) Expression of the eRF1 translation termi-nation factor is controlled by an autoregulatory circuit involvingreadthrough and nonsense-mediated decay in plants. NucleicAcids Res. 45, 4174–4188

52. Filichkin, S.A. et al. (2015) Environmental stresses modulateabundance and timing of alternatively spliced circadian tran-scripts in Arabidopsis. Mol. Plant 8, 207–227

53. Gloggnitzer, J. et al. (2014) Nonsense-mediated mRNA decaymodulates immune receptor levels to regulate plant antibacterialdefense. Cell Host Microbe 16, 376–390

54. Ding, F. et al. (2014) Genome-wide analysis of alternative splicingof pre-mRNA under salt stress in Arabidopsis. BMC Genomics15, 431

55. Shikata, H. et al. (2014) Phytochrome controls alternative splicingto mediate light responses in Arabidopsis. Proc. Natl. Acad. Sci.U. S. A. 111, 18781–18786

56. Vexler, K. et al. (2016) The Arabidopsis NMD factor UPF3 isfeedback-regulated at multiple levels and plays a role in plantresponse to salt stress. Front. Plant Sci. 7, 1376

57. Pose, D. et al. (2013) Temperature-dependent regulation of flow-ering by antagonistic FLM variants. Nature 503, 414–417

58. Lutz, U. et al. (2015) Modulation of ambient temperature-depen-dent flowering in Arabidopsis thaliana by natural variation ofFLOWERING LOCUS M. PLoS Genet. 11, e1005588

59. Sureshkumar, S. et al. (2016) Nonsense-mediated mRNA decaymodulates FLM-dependent thermosensory flowering response inArabidopsis. Nat. Plants 2, 16055

60. Syed, N.H. et al. (2012) Alternative splicing in plants – coming ofage. Trends Plant Sci. 17, 616–623

61. Roy, B. and von Arnim, A.G. (2013) Translational regulation ofcytoplasmic mRNAs. Arabidopsis Book 11, e0165

62. Álvarez, D. et al. (2017) 50UTR-mediated translational control ofsplice variants of phytoene synthase. Plant Physiol. 172,2314–2326

12 Trends in Plant Science, Month Year, Vol. xx, No. yy

63. Remy, E. et al. (2014) Intron retention in the 50UTR of the novelZIF2 transporter enhances translation to promote zinc tolerancein arabidopsis. PLoS Genet. 10, e1004375

64. Parra, G. et al. (2011) Comparative and functional analysis ofintron-mediated enhancement signals reveals conserved featuresamong plants. Nucleic Acids Res. 39, 5328–5337

65. Gallegos, J.E. and Rose, A.B. (2015) The enduring mystery ofintron-mediated enhancement. Plant Sci. 237, 8–15

66. Akua, T. and Shaul, O. (2013) The Arabidopsis thalianaMHX geneincludes an intronic element that boosts translation when local-ized in a 50 UTR intron. J. Exp. Bot. 64, 4255–4270

67. Jiang, J. et al. (2015) A recently evolved isoform of the transcrip-tion factor BES1 promotes brassinosteroid signaling and devel-opment in Arabidopsis thaliana. Plant Cell 27, 361–374

68. Jia, J. et al. (2013) Regulation and dysregulation of 30UTR-medi-ated translational control. Curr. Opin. Genet. Dev. 23, 29–34

69. Amin, U.S. et al. (2016) Enhanced salt tolerance conferred by thecomplete 2.3 kb cDNA of the rice vacuolar Na+/H+ antiportergene compared to 1.9 kb coding region with 50 UTR in transgeniclines of rice. Front. Plant Sci. 7, 14

70. Yang, X. et al. (2012) Alternative mRNA processing increases thecomplexity of microRNA-based gene regulation in Arabidopsis.Plant J. 70, 421–431

71. Campo, S. et al. (2013) Identification of a novel microRNA(miRNA) from rice that targets an alternatively spliced transcriptof the Nramp6 (Natural resistance-associated macrophage pro-tein 6) gene involved in pathogen resistance. New Phytol. 199,212–227

72. Yan, K. et al. (2012) Stress-induced alternative splicing provides amechanism for the regulation of microRNA processing in Arabi-dopsis thaliana. Mol. Cell 48, 521–531

73. Charlier, J.B. et al. (2015) Zinc triggers a complex transcriptionaland post-transcriptional regulation of the metal homeostasisgene FRD3 in Arabidopsis relatives. J. Exp. Bot. 66, 3865–3878

74. Vandivier, L.E. et al. (2016) The conservation and function of RNAsecondary structure in plants. Annu. Rev. Plant Biol. 67, 463–488

75. Hunt, A.G. et al. (2012) Plant polyadenylation factors: conserva-tion and variety in the polyadenylation complex in plants. BMCGenomics 13, 641

76. Sun, H.-X. et al. (2017) Dehydration stress extends mRNA 30

untranslated regions with noncoding RNA functions in Arabidop-sis. Genome Res. 27, 1427–1436

77. Zhang, Q. et al. (2008) Rice 2020: a call for an internationalcoordinated effort in rice functional genomics. Mol. Plant 1,715–719


Recommended