Introduction
The growth and development of multicellularorganisms is heavily dependent on communication between groups ofcells. Intercellular signaling pathways convey cell fate information,regulate cell division and differentiation processes, propagate andamplify specific signaling states, and coordinate tissue functions.Both plants and animals utilize systemic hormones as well as polypeptidesignaling molecules to mediate cell-to-cell communication. In animals,polypeptides such as epidermal growth factor (EGF) and transforminggrowth factor-beta (TGF-b) actas extracellular ligands that are generated in certain cell typesand perceived at the surface of neighboring cells, typically by transmembranereceptor kinases. Binding of the ligand to its receptor or receptorsinitiates a cascade of intracellular phosphorylation events that affectsthe activity of one or more nuclear transcription factors, resultingin the alteration of gene expression programs (
Bergeron et al., 2016).
Although plants lack canonical EGF,TGF-b, Wingless and other peptidesuperfamilies found in animals, they also make extensive use of polypeptidesignaling systems to mediate various biological processes (
Matsubayashi, 2014;
Tavormina et al., 2015). The genomeof
Arabidopsis thaliana, a memberof the mustard family related to food plants such as broccoli andcauliflower, encodes well over a thousand small proteins (<100amino acids) that may function as peptide signaling molecules as wellas more than 600 putative plasma membrane-bound receptor proteins(
Shiu and Bleecker, 2001;
Lease and Walker, 2006;
Tavormina et al., 2015). The CLAVATA3/EMBRYO SURROUNDING REGION-RELATED (CLE) family representsone of the largest families of plant polypeptides identified to date,consisting of 32 members in
Arabidopsis (
Cock and McCormick, 2001) and as many as 84 members in other species (
Hastwell et al., 2015). Members ofthe
CLE gene family are presentthroughout the land plant lineage and in some plant parasitic nematodes(
Cock and McCormick, 2001;
Wang et al., 2005).
The
CLE genes encode polypeptides of less than 15 kDa in molecular massthat share several structural features. Each peptide consists of eitheran N-terminal signal peptide or a membrane anchor sequence, a 40-to 90-amino acid variable domain, and a highly conserved 14-aminoacid motif near the carboxyl terminus called the CLE domain (
Cock and McCormick, 2001) (Fig. 1A).The signal peptide is sufficient to direct the CLE proteins throughthe secretory pathway (
Rojo et al.,2002;
Sharma et al.,2003) and is required for their
in vivo function (
Menget al., 2010). Full length CLE pre-propeptides are processed(
Casamitjana-Martínez et al.,2003;
Ni and Clark,2006) to produce biologically active 12-13 amino acidpolypeptides consisting of the CLE domain (
Kondo et al., 2006;
Ohyama et al., 2009). The roles ofthe
CLE genes are best understoodin plant development and particularly in the meristems, which aresmall stem cell reservoirs that provide cells for continuous organformation. Here we review recent insights into
CLE gene regulation and function in the shoot apical meristemas well as in the meristematic tissue of the shoot vascular cambium.
SAM maintenance requires intercellular communication
The shoot apical meristem (SAM) ofangiosperm plants is a small, highly organized structure at the growingshoot tip that provides all of the cells to generate the above groundarchitecture of the plant (Fig. 2). The SAM is established duringembryogenesis and is maintained throughout the life of the plant.Its two major functions are to continuously initiate organs such asleaves and flowers and to sustain a stem cell reservoir for futureorgan formation. The organs arise as primordia on the flanks of themeristem, while at the apex the self-renewing stem cell reservoirreplenishes the cells that have become incorporated into the organprimordia. To function as a site of ongoing organ formation, the SAMmaintains a continuous balance between loss of stem cells throughdifferentiation and their replacement through cell division.
The SAM is stratified into distinctcell layers called the tunica and corpus (
Satina et al., 1940). In
Arabidopsis and many other dicotyledonous plants, the tunica is comprised ofan overlying L1 epidermal layer and a sub-epidermal L2 layer (
Gifford, 1954). These layers area single cell thick and remain clonally distinct from one anotherdue to their specific cell division patterns (
Poethig, 1987). The corpus, or L3,lies beneath the tunica and consists of cells that divide in all planes.Because cells in each layer participate in both SAM maintenance andorgan formation (
Poethig and Sussex,1985a,
b), these activities must be coordinated between all of the cell layers.
The SAM is also organized into threedistinct functional domains. The central zone (CZ) at the very apexof the SAM consists of a reservoir of pluripotent stem cells withlow mitotic activity (
Steeves andSussex, 1989). Divisions of stem cells in the CZ continuouslydisplace their descendants outward into the surrounding peripheralzone (PZ) or downward into the interior rib zone (RZ). The PZ is atransitional region wherein more rapidly dividing stem cell descendantsacquire more specified fates and become incorporated either into organprimordia or into regions of stem between the organs. The upper regionof the RZ contains the organizing center (OC), which acts as a nichethat sustains the overlying stem cell population. Cells in the RZconstitute the meristem pith and contribute to the bulk of the stemand vascular tissue (
Steeves andSussex, 1989). Classical experiments have demonstratedthat the functional domains within the SAM exchange cell fate information(
Sussex, 1954) andthat the fate of each SAM cell is determined by positional informationfrom the surrounding cells rather than from a lineage-specific heritage(
Poethig et al., 1986;
Furner and Pumfrey, 1992;
Irish and Sussex, 1992). Thus SAM cells are in continuous communication with their neighborsin order to assess their relative positions in the meristem and behaveaccordingly.
A CLV3-mediated shoot apical meristem maintenance pathway
A molecular network called the CLAVATA(CLV)-WUSCHEL (WUS) pathway communicates cell fate decisions in theSAM and is essential for stem cell maintenance in higher plants (
Somssich et al., 2016) (Fig. 3).The
Arabidopsis CLV3 gene is afounding member of the
CLE genefamily that is expressed exclusively within the stem cell reservoirof shoot apical and floral meristems (
Fletcher et al., 1999).
CLV3 expression in the SAM initiates during the early heart stage ofembryogenesis and continues throughout the life cycle. Loss-of-functionmutations in
CLV3 cause an accumulationof supernumerary stem cells that leads to progressive SAM enlargement,resulting in the formation of strap-like fasciated stems that producemany more flowers than normal (
Clarket al., 1995). Flowers arise from transient stem cellreservoirs in floral meristems (FM), which are also enlarged in
clv3 plants and produce extra floral organs.Thus
CLV3 restricts above groundstem cell accumulation throughout the life of the plant. Live imagingexperiments revealed that
CLV3 performsthis function both by restricting stem cell fate to the CZ and alsoby non-cell autonomously limiting cell division rates in the PZ (
Reddy and Meyerowitz, 2005).
CLV3 encodes a secreted signaling molecule that is localized to the extracellularspace (
Fletcher et al., 1999;
Rojo et al., 2002). As with many animal polypeptides, the CLV3 ligand is generatedfrom a larger pre-propeptide that undergoes proteolytic cleavage ofthe signal peptide and the pro-domain before displaying biologicalactivity. Cleavage of CLV3 occurs between the Leu
69 and Arg
70 residues (
Ni and Clark, 2006;
Ni et al., 2011;
Xu et al., 2013), and requires arecognition domain of five amino acids flanking the N terminus ofthe CLE domain (
Xu et al., 2013). It has been suggested that CLV3 cleavage is catalyzed by a serineprotease (
Ni et al., 2011), but this remains to be confirmed experimentally.
The active form of CLV3 was firstidentified as a 12 amino acid glycopeptide consisting of Arg
70 to His
81 of the CLE motif,in which the first two proline residues are modified to hydroxyproline(
Kondo et al., 2006).When applied to
Arabidopsis seedlingsthis synthetic MCLV3 peptide generates a SAM termination phenotypecharacteristic of
CLV3 gain-of-functionplants, demonstrating its biological activity. Two studies have examinedthe contributions of individual residues within the CLV3 peptide toits function in restricting meristem cell accumulation. The transformationof constructs encoding Alanine-substituted CLV3 peptides into
clv3 null mutants revealed that, in orderof importance, the Asp
8, His
11, Gly
6, Hyp
4, Arg
1 and Pro
9 residues are the most critical for CLV3 activityin the SAM (
Song et al., 2012). However, the hydroxyproline residue at position 7 has a minimalimpact on CLV3 function, as do the flanking sequences outside thecore CLE motif. A follow up study applying synthetic CLV3 peptidesto cultured
clv3 null mutant seedlingsdemonstrated that the Pro
9 and His
11 residues of the CLV3 peptide (Fig. 3) are the mostcritical for restricting SAM size
in vitro (
Song et al., 2013). The presence of these two residues positively correlates withCLV3 protein stability
in vitro, suggesting that CLV3 stability may be important for its role inSAM maintenance.
Sugar modification of the CLV3 peptideis also important for its activity in SAM maintenance. A 13 aminoacid hydroxylated and arabinosylated secreted peptide was biochemicallyidentified from
Arabidopsis CLV3 overexpressing plants (
Ohyama etal., 2009), in which the Hyp
7 residue of CLV3 is post-translationally modified with three L-arabinoseresidues (Fig. 3). This modification was shown to enhance CLV3 activityin the seedling SAM (
Ohyama et al.,2009). NMR spectroscopy revealed that arabinosylationinduces a conformational change in the carboxyl-terminal half of theCLV3 peptide and enhances receptor binding affinity (
Shinohara and Matsubayashi, 2013). In tomato, three arabinosyltransferase genes,
FASCIATED INFLORESCENCE (FIN), REDUCED RESIDUAL ARABINOSE 3 (RRA3A) and
FASCIATED AND BRANCHED2 (FAB2), are implicated in the arabinosylation of CLV3 peptides (
Xu et al., 2015). Mutations in anyof these genes cause increase inflorescence branching and the formationof fasciated flowers with extra floral organs.
SlCLV3 null mutants generated using the CRISPR-Cas9 genomeediting method display a fasciated SAM phenotype and closely resemble
fin plants. Synthesized arabinosylated SlCLV3peptides partially rescue the
fin fasciated phenotype, confirming the importance of CLV3 arabinosylation
in vivo. A similar effect was obtained usingarabinosylated SlCLE9 peptides, implying a role for SlCLE9 in SAMsignaling, although this is yet to be confirmed using mutational analysis.Together, the results indicate that the progressive addition of arabinosechains to CLV3, and potentially to related CLE peptides, by a cascadeof arabinosyltransferases is required to fully maintain stem cellhomeostasis in the SAM.
CLV3 orthologs are also present in a variety of other crop plants, whichover the past ten thousand years have undergone intense selectionby humans (
Kuittinen and Aguadé,2000;
Doebley et al.,2006) for yield traits such as larger and more numerousinflorescences, fruits, and seeds. The
CLV3 locus has been a target of selection during the domestication ofseveral crop species to enhance agricultural yields (
Somssich et al., 2016). A naturallyoccurring mutation in the mustard
(Brassicarapa) CLV3 gene
MULTILOCULAR4 (ML4) leads to the formation of fruits with four chambers instead of two,which increases seed production (
Fanet al., 2014). Likewise, the mild branching and fasciatedflower and fruit phenotype of the classical tomato
fasciated (fas) allele results from a regulatorymutation at the
SlCLV3 locus thatreduces the size of the
CLV3 expressiondomain without affecting peptide function (
Xu et al., 2015). These studiessuggest that fine-tuning CLV signaling in the SAM by modulating
CLV3 mRNA expression levels and/or peptideactivity may also be exploited in other crops to improve productivity.
CLV3 signal perception
Genetic analyses have uncovered asmall suite of membrane-associated receptors that mediate CLV3 signalingin shoot and floral meristems (Fig. 3). However, the contributionsof the various receptors to CLV3 signal transduction have remainedunclear, as has the functional relationships between them and theirrelative effects on downstream signaling outputs. Several recent studiesprovide new insights into these questions.
The first receptor gene shown toplay a role in SAM stem cell homeostasis was
CLV1. Loss-of-function mutations in
CLV1 cause progressive shoot and floral meristem enlargementphenotypes that are similar to but weaker than
clv3 phenotypes, and the two genes act in the same geneticpathway (
Clark et al., 1993,
1995;
Diévart et al., 2003). CLV1encodes a leucine-rich repeat (LRR) receptor serine/threonine kinase(Fig. 1B) that is produced in shoot and floral meristem cells interiorto the
CLV3-expressing stem celldomain (
Clark et al., 1997). CLV1 is localized to the plasma membrane, where it forms homodimers(Bleckmann et al., 2010) and binds the CLV3 ligand (
Ogawa et al., 2008). In contrastwith other intercellular signaling pathways in plants, CLV3 appearsto bind pre-formed receptor complexes at the plasma membrane (
Somssich et al., 2015). Ligand bindingtriggers activation of the CLV1 kinase domain on the cytosolic surfaceof the cell, which is thought to lead to recruitment of accessoryproteins and to result in CLV1 internalization and trafficking tothe lytic vacuole for degradation (
Nimchuk et al., 2011b). Signaling through the CLV3-CLV1ligand-receptor pair limits SAM stem cell accumulation by negativelyregulating the
WUS expression domainin the underlying RZ cells (
Brand etal., 2000). The CLV1 receptors are sequestered withinplasma membrane microdomains following CLV3 perception, attenuatingtheir signaling activity to prevent complete repression of
WUS transcription and SAM termination (
Somssich et al., 2015).
Genetic and biochemical studies alsoprovide evidence for a second distinct receptor complex involved inCLV3-mediated stem cell signaling, consisting of the CLV2 and CORYNE(CRN) proteins (
Guo et al., 2010;
Durbak and Tax, 2011).
CLV2 encodes a receptor-likeprotein with extracellular LRRs, a transmembrane domain and a shortcytoplasmic tail (
Jeong et al., 1999). Like
CLV3 and
CLV1,
CLV2 restricts shoot and floral stem cell accumulation (
Kayes and Clark, 1998), as does
FASCIATED EAR2 (FEA2), the maize orthologof
CLV2 that maps to a quantitativetrait locus (QTL) for kernel row number (
Bommert et al., 2013).
CRN encodes a membrane-associated protein with a cytoplasmic serine/threoninekinase domain, and
crn mutantsdisplay
clv-like enlarged SAM phenotypes(
Müller et al., 2008). Unlike
CLV1, both
CLV2 and
CRN are widely expressed in many plant tissues and have broad effectson plant development (
Jeong et al.,1999;
Müller etal., 2008).
CLV2 and CRN proteins localize tothe plasma membrane and form heterodimers (Bleckmann et al., 2010;
Zhu et al., 2010). However, CRNlacks kinase activity and is likely to be a pseudokinase that functionsas a CLV2 co-receptor (
Nimchuk etal., 2011a). Overexpression of
CLV3 in
clv2 plantsfails to rescue the enlarged SAM phenotype, indicating that CLV2 isinvolved in CLV3 signal transduction (
Brand et al., 2000). However, the observation that
clv1 phenotypes are enhanced by mutationsin either
CLV2 or
CRN shows that the CLV2-CRN complex functionsindependently of CLV1 in this process (
Müller et al., 2008;
Zhu et al., 2010). Whether the CLV2-CRN complex bindsCLV3 peptide is unresolved, though, as immunoprecipitation experimentsindicate that CLV2 generates a CLV3 binding activity in tobacco leaves(
Guo et al., 2010),whereas photo affinity labeling experiments show that CLV2 does notdirectly bind to arabinosylated CLV3 peptide (
Shinohara and Matsubayashi, 2015).
The transcriptional regulation of
CRN is important for SAM maintenance (
Yue et al., 2013).
CRN transcription is directly repressed bySKB1/PRMT5, a member of the type II arginine methyltransferase familythat in animals regulate chromatin remodeling, transcription and pre-mRNAsplicing (
Bedford and Clarke, 2009). SKB1 directs symmetric dimethylation of histone H4R3 at the
CRN locus, which leads to upregulation of
CLV3 and
WUS transcription in their native domains and maintenance of properSAM size (
Yue et al., 2013).
CLV1 forms a monophyletic group with three other LRR-RLK genes,
BARELY ANY MERISTEM1, 2 and
3 (BAM1-3), which are predominantly expressedon the flanks of the SAM (
DeYounget al., 2006). Plants carrying higher order combinationsof
bam alleles have reduced SAMsize, indicating that the
BAM genesredundantly promote stem cell maintenance (
DeYoung et al., 2006). However,
clv1 null mutant phenotypes can be enhancedby mutations in
BAM1 or
BAM2 (
DeYoungand Clark, 2008); thus BAM1 and BAM2 also function asredundant CLV3 receptors in the PZ. Indeed, both BAM1 and BAM2 bindCLV3 peptide (
Guo et al., 2010;
Shinohara and Matsubayashi, 2015). Because
clv1 bam1 plants areinsensitive to exogenous arabinosylated CLV3 peptide treatment, CLV1and BAM1 activity is sufficient to regulate CLV3-mediated stem cellhomeostasis in the SAM (
Shinoharaand Matsubayashi, 2015). A recent study has clarifiedthe relationship between
CLV1 andthe
BAM genes in SAM maintenance.CLV1 signaling was found to repress the expression of
BAM1 and
BAM3 in the RZ, such that in
clv1 mutants,ectopic
BAM expression partiallycompensates for loss of CLV1 activity (
Nimchuk et al., 2015). Interestingly,
clv1 bam123 quadruple mutants have strongervegetative SAM phenotypes than
clv3 null mutants (
Nimchuk et al., 2015), indicating that at least one other ligand that acts partiallyredundantly with CLV3 in SAM maintenance remains be identified.
The BAM1 protein has been shown tophysically associate with a LRR receptor-like kinase encoded by the
RECEPTOR-LIKE PROTEIN KINASE2 (RPK2) gene(
Kinoshita et al., 2010). Plants carrying
rpk2 mutationsdisplay slightly enlarged SAMs and are insensitive to CLV3 peptidetreatment, indicating that RPK2 is involved in CLV3 ligand perception.
RPK2 is expressed uniformly throughout theSAM (
Kinoshita et al., 2010), and forms homomers as well as interacting with BAM1. However,it neither associates with CLV1 or CLV2 (
Kinoshita et al., 2010;
Shimizu et al., 2015) nor binds directly to CLV3 peptide(
Shinohara and Matsubayashi, 2015). RPK2 is therefore likely to regulate meristem maintenance by transmittingthe CLV3 signal through the BAM1 pathway rather than the CLV1 or CLV2/CRNpathways.
The relationship between CLV1 andthe other SAM receptors has been investigated using genetic analysis.Like
CLV1, CLV2 and
CRN mediate stem cell regulation exclusivelyin the
WUS-expressing cells ofthe RZ (
Nimchuk, 2017). However,
CLV2, CRN and
RPK2 are dispensable for the repression ofthe
BAM receptor kinase genes byCLV3-CLV1 signaling. The CLV1-mediated repression of
WUS transcription and consequent restrictionof stem cell accumulation was determined to be genetically separablefrom its regulation of
BAM geneexpression. CLV1 therefore controls two distinct signaling outputs– the repression of
WUS transcriptionand the repression of
BAM transcription– in SAM stem cell niches in response to the CLV3 ligand independentlyof the other receptors.
CLV3 signal transduction
Several classes of cytosolic componentsfunction in CLV3 signal transduction downstream of ligand binding(Fig. 3). In
Arabidopsis, the kinaseassociated protein phosphatase KAPP and a Rho GTPase-related proteinphysically associate with the cytosolic CLV1 kinase domain (
Williams et al., 1997;
Trotochaud et al., 1999), whilethe related protein phosphatase 2C proteins POLTERGEIST (POL) andPOL-LIKE1 (PLL1) act downstream of CLV1 to promote stem cell maintenanceby regulating
WUS expression (
Song et al., 2006). A mitogen-activatedprotein kinase (MAPK) activity (
Betsuyakuet al., 2011) and a E3 ubiquitin ligase called PLANTU-BOX4 (PUB4) (
Kinoshita et al., 2015) have also been implicated in signaling downstream of the CLV receptors,although their roles in the signaling network remain to be preciselydefined.
In maize, mutations in the
COMPACT PLANT2 (CT2) gene, which encodesthe alpha-subunit of a heterotrimeric GTP binding protein, cause
clv-like SAM phenotypes (
Bommert et al., 2013). HeterotrimericGTP binding proteins, which are composed of alpha, beta and gammasubunits, are signaling molecules that link extracellular signalsto intracellular readouts (
Uranoand Jones, 2014). The CT2 protein localizes to the plasmamembrane and physically interacts with the FEA2 receptor protein invitro
, suggesting a molecular mechanismthrough which receptor-like proteins that lack a kinase domain cantransmit information inside the cell (
Bommert et al., 2013). Similarly, mutations in the
Arabidopsis G protein beta-subunit1 gene
AGB1 produce enlarged SAMs similar to
clv mutant SAMs and
AGB1 acts upstream of
WUS in stem cell homeostasis (
Ishidaet al., 2014). AGB1 protein physically associates withRPK2 at the plasma membrane, although not with CLV1 or CLV2. In contrastto the situation in maize,
Arabidopsis Ga activity does not affect SAMfunction, although Gg activityhas a minor role in limiting SAM size (
Ishida et al., 2014). Together these observations indicatea role for heterotrimeric GTP binding proteins in transducing CLV-dependentsignals within the recipient cells.
CLV3-independent signaling pathways in SAM regulation
Members of the
ERECTA (ER) receptor kinase gene family also influencestem cell homeostasis in the SAM.
ER, ERL1 and
ERL2 act redundantly to restrictvegetative SAM activity (
Uchida etal., 2013). The promoters of all three genes are activein the SAM, and
er erl1 erl2 seedlingsform enlarged SAMs in which the L1 and L2 cells are wider than normal(
Chen et al., 2013).Hormone induction experiments suggest that ER family members regulatestem cell homeostasis in the SAM by buffering its responsiveness tocytokinin, which promotes cell proliferation and stem cell activity(
Gordon et al., 2009). The ER pathway negatively regulates
WUS transcription (
Chen et al., 2013), although this occurs independently of the CLV pathway (
Mandel et al., 2014). In fact, ER,CLV and a third pathway consisting of class III HOMEODOMAIN-LEUCINEZIPPER (HD-ZIP III) transcription factors (
Prigge et al., 2005) act in parallelto regulate SAM size (
Mandel et al.,2016). The three pathways seem to affect different aspectsof SAM activity, as
CLV3 preferentiallyrestricts SAM cell accumulation along the longitudinal axis whereas
ER and the
HD-ZIPIII genes restrict its growth along distinct lateral orientations.RNA-seq analysis provides evidence that the CLV pathway limits theaccumulation of stem cells in the CZ, whereas the ER pathway regulatesmitotic activity in the PZ (
Mandelet al., 2016). Thus the coordination of cell behaviorswithin the SAM appears to be orchestrated by distinct signaling pathwaysacting along discrete growth vectors.
A novel CLE ligand-receptor signaltransduction pathway that regulates maize shoot apical meristem activityhas been revealed by the recent study of the
FASCIATED EAR3 (FEA3) gene (
Je et al., 2016).
FEA3 encodes a LRR receptor-like protein with 12 extracellular LRR motifs,a transmembrane domain and a short cytoplasmic tail. FEA3 functionsto limit maize SAM size and suppresses
ZmWUS expression in cells below the OC. However, FEA3 does not perceivea CLV3 signal. Rather it responds to a CLE peptide encoded by the
FON2-LIKE CLE PROTEIN 1 (ZmFCP1) gene, whichis orthologous to the rice
FCP1 gene. Mutations in
ZmFCP1 causeenlarged SAM phenotypes, and
ZmFCP1 and
FEA3 function in the samegenetic pathway. Interestingly,
ZmFCP1 is not expressed in the SAM itself but in the initiating organ primordiaon the SAM flanks. The authors propose that a ZmFCP1 signal originatingfrom differentiating cells within organ primordia is perceived byFEA3 in the interior of the SAM where it acts to restrict stem cellproliferation by negatively regulating
ZmWUS expression in the RZ cells beneath the OC. The
Arabidopsis FEA3 ortholog,
AtFEA3, also appears to restrict SAM activity and
AtFEA3 RNAi lines are resistant to CLE27 peptide application,although
CLE27 is not the
ZmFCP1 ortholog. Thus stem cell homeostasisin plants is mediated by multiple CLE peptides that originate fromdifferent cell types and associate with distinct transmembrane receptors.
WUS-CLV3 stem cell homeostasis feedback loop
The key biologically relevant targetof the CLV3 stem cell signaling pathway is the
WUSCHEL (WUS) gene. WUS is the founding member of theWUSCHEL-LIKE HOMEOBOX (WOX) family of transcription factors that containa homeodomain superficially resembling that found in animal homeodomainproteins (
Mayer et al., 1998) (Fig. 1C). In addition, the protein contains three conserved shortsequence motifs at the carboxyl terminus: an acidic domain that mayfunction in transcription activation, a canonical WUS box, and anEAR domain that can mediate transcriptional repression (
Ohta et al., 2001).
WUS expression is restricted to a small setof cells just beneath the stem cells (
Mayer et al., 1998), which is called the organizingcenter (OC) based on its functional correspondence to animal stemcell niches (Fig. 2).
Although
WUS is dispensable for the establishment of the
Arabidopsis shoot stem cell reservoir (seebelow), it is required to sustain stem cell activity in shoot andfloral meristems throughout the life of the plant (
Laux et al., 1996). WUS promotesstem cell fate in a non-cell autonomous fashion, with the proteinmoving from the OC into the overlying stem cells where it accumulatesat a lower level than in the OC cells themselves (
Yadav et al., 2011). This movementoccurs via cytoplasmic channels between neighboring cells called plasmodesmataand is essential for SAM maintenance (
Daum et al., 2014). WUS protein accumulation in theCZ induces the expression of
CLV3 (Fig. 4), activating its own negative regulator in a dynamic feedbackloop that regulates stem cell homeostasis in the SAM (
Brand et al., 2000;
Schoof et al., 2000).
The regulation of
CLV3 transcription by WUS occurs in a dosage-dependentmanner. Lower concentrations of WUS protein activate
CLV3 transcription whereas higher concentrationsrepress
CLV3 transcription (
Perales et al., 2016). WUS proteinbinds with different affinities to six
cis elements in the regulatory region of the
CLV3 locus, five of which occur in a module in the 3′region, and these six elements mediate both the activation and repressionof
CLV3 expression. WUS binds the
CLV3 cis elements as monomers at lower concentrations,and as homodimers at higher concentrations (
Perales et al., 2016). Structure-functionanalysis indicates that reduced WUS protein accumulation in the stemcell reservoir may occur through a combination of potent nuclear exportand weak nuclear retention of the protein within these cells, potentiallydue to a lower affinity for DNA and reduced dimerization activity(
Rodriguez et al., 2016). In this manner, lower levels of WUS protein in the nucleus ofthe stem cells leads to
CLV3 activation,whereas higher levels of WUS protein in the nucleus of OC cells leadsto
CLV3 repression (
Perales et al., 2016).
WUS-dependent gene regulatory network
WUS is a bi-functional protein thatcan act as both an activator and a repressor of transcription (
Ikeda et al., 2009), and regulatesthe expression of hundreds of genes in the shoot apical meristem.A genome-wide identification of WUS response genes using
Arabidopsis ATH1 arrays yielded a total of675 genes (
Busch et al., 2010), including 4 hormone responsive type-A
ARABIDOPSIS RESPONSE REGULATOR (ARR) genes previouslydescribed as WUS targets (
Leibfriedet al., 2005). Gene ontogeny analysis revealed an over-representationof WUS responsive genes in three categories: regulation of development,metabolic processes, and hormone signaling. The
CLV1 gene was found to be directly repressed by WUS despitetheir overlapping expression patterns in the interior of the SAM,suggesting that WUS acts to fine-tune
CLV1 transcription rather than acting as a binary switch. WUS also directlyrepresses the transcription of
TPR1 and
TPR2, members of the
TOPLESS/TOPLESS RELATED (TPL/TPR) familyof transcriptional co-repressor genes that play key roles in embryopatterning and auxin responses (
Longet al., 2006;
Szemenyeiet al., 2008). WUS protein was shown to directly bindtwo distinct DNA motifs, one of which is a G-Box motif with strikingsimilarity to binding sites for proteins involved in stem cell renewalin animals, the zinc-finger homeodomain transcription factor Zeb-1(
Grooteclaes and Frisch, 2000) and the bHLH-ZIP transcription factor MYC (
Blackwell et al., 1990).
A second study using an inducibleWUS system also identified over 600 WUS-responsive genes in SAM tissue,among which 49 upregulated and 140 downregulated genes are directWUS targets (
Yadav et al., 2013). The majority of WUS-activated genes are expressed in the CZ andOC of the SAM, whereas the majority of WUS-repressed genes are expressedin the PZ. Among the latter, WUS directly binds to the regulatoryregions of key transcription factor genes such as
KANADI1 (KAN1), KAN2, ASYMMETRIC LEAVES2 and
YABBY3 that promote organ identity and celldifferentiation. Thus WUS controls stem cell homeostasis in part byrepressing the expression of differentiation-inducing transcriptionfactor genes in the central regions of the SAM to prevent prematurestem cell differentiation.
An important direct target of WUSrepression in the OC is the bHLH transcription factor gene
HECATE1 (HEC1) (
Schuster et al., 2014).
HEC1 is expressed in the PZ of the SAM as well as in developing organprimordia, and functions redundantly with the related
HEC2 and
HEC3 genes to promote SAM cell accumulation. HEC1 activity represses
CLV3 and
WUS expression while elevating the expression of cell-cycle regulatorygenes to stimulate cell proliferation (Fig. 4). In addition, transcriptomeanalysis indicates that WUS and HEC1 oppositely regulate suites ofmetabolic and hormone signaling genes, including the type-A
ARR7 and
ARR15 genes. These
ARR genes are involvedin negative feedback regulation of cytokinin signaling (
To et al., 2004) and can arrestSAM function when constitutively activated (
Leibfried et al., 2005). WhereasWUS directly represses type-A
ARR gene transcription to enhance cytokinin signaling in the SAM (
Leibfried et al., 2005), HEC1 inducestheir expression and thereby acts as a negative regulator of downstreamcytokinin signaling outputs (
Schusteret al., 2014). The opposing activities of these twotranscription factors in hormone regulation are thought to representan important mechanism for coordinating a balance between cell proliferationand differentiation in distinct functional domains of the SAM.
WUS-associated factors
WUS does not regulate gene expressionin isolation but physically associates with members of the HAIRY MERISTEM(HAM) family of GRAS domain transcriptional regulators (
Zhou et al., 2015). Members of thisfamily promote stem cell maintenance in
Arabidopsis and petunia (
Stuurman et al., 2002;
Engstrom et al., 2011), and
Arabidopsis ham1234 plantsarrest at early seedling stage with terminated SAMs (
Zhou et al., 2015). The
HAM gene expression patterns overlap withthat of
WUS in the SAM, with both
HAM1 and
HAM2 being expressed in the RM, including the OC cells. In addition
HAM1 expression, like
WUS expression, is repressed by CLV3 signaling. Becausea weak
wus-7 allele displays dose-dependentgenetic interactions with
ham nullalleles, the strong
wus-1 mutantis epistatic to
ham123 null mutants,and the WUS and HAM proteins share some common downstream regulatorytargets (
Zhou et al., 2015), the HAM transcriptional regulators are proposed to act as conservedinteracting cofactors with WUS in the OC of the SAM (Fig. 4).
WUS also physically associates withTPL and several TPR co-repressor proteins via the acidic domain, WUSbox, and EAR motif in the carboxyl-terminal region of the WUS protein(
Kieffer et al., 2006;
Dolzblasz et al., 2016). The TPL and TPR proteins associate with HISTONE DEACETYLASE19(HDA19) to form a transcription repression complex (
Szemenyei et al., 2008), suggestinga mechanism through which WUS may repress the expression of differentiation-inducinggenes within the OC by recruitment of histone modifying complexes.Whether members of either the HAM or TPL/TPR families act togetherwith WUS to regulate
CLV3 transcriptionremains to be determined.
CLV-WUS activity in early development
The CLV-WUS signal transduction pathwayis essential for shoot apical meristem maintenance, but recent studieshave revealed unexpected nuances in the pathway during early stemcell initiation and activation. Surprisingly
WUS, although its expression commences prior to SAM formationat the 16-cell stage of embryogenesis (
Mayer et al., 1998), is dispensable for embryonic stemcell initiation and
CLV3 activation(
Zhang et al., 2017). Instead the
WUS paralog
WOX2, along with the redundant
WOX1, WOX3 and
WOX5 genes, is required for the initiation of the embryonic stem cellprogram, as assayed by
CLV3 expression.
WOX2 mRNA expression initiates in the zygoteand then becomes restricted to the apical lineage (
Breuninger et al., 2008). The
WOX2 module downregulates the expressionof cotyledon-specific genes and limits the distribution of the syntheticauxin response reporter
DR5:GFP in apical embryo cells, suggesting that these
WOX genes block cell differentiation in the presumptivestem cell domain (
Zhang et al., 2017). The
WOX2 module also contributesdirectly to SAM formation by promoting expression in the presumptiveSAM region of
HD-ZIP III genesthat induce cytokinin biosynthesis gene transcription and are requiredfor shoot stem cell identity (
Priggeet al., 2005;
Smithand Long, 2010). The authors propose that the functionof the
WOX2 module in embryonicstem cell formation is largely mediated through its upregulation of
HD-ZIP III gene expression, thereby balancingthe activities of the cytokinin and auxin pathways to suppress celldifferentiation and promote cell proliferation (
Zhang et al., 2017).
Plants undergo distinct post-embryonicdevelopmental programs depending on whether they germinate in thedarkness or in the light. In the darkness, the SAM of the germinatingseedling remains dormant and growth occurs predominantly via cellelongation in the hypocotyl. Only in the light is the SAM activatedand the above ground organ development program triggered. Recent evidenceindicates that SAM stem cell identity, as monitored by
CLV3 expression, is sustained even in thedormant state, independent of growth conditions (Pfeiffer et al.,2016). In contrast,
WUS expressionis not detectable in dark-grown seedlings and its induction dependson both light and sucrose, which acts as an energy source for stemcell activity. Shoot stem cell activation by sugars is dependent ona TARGET OF RAPAMYCIN (TOR) kinase-dependent signaling pathway (
Dobrenel et al., 2016; Pfeiffer etal., 2016). Light activation of
WUS expression also occurs via TOR kinase-dependent signaling, independentof photosynthesis, and thus TOR acts as a central regulator of post-embryonicstem cell activation in response to environmental cues. The findingssuggest that TOR activation by light enables plant cells to anticipatethe amount of energy available for photomorphogenesis and efficientlyadapt their growth and development to the local environmental conditions(Pfeiffer et al., 2016).
CLE peptide function during shoot vascular development
Several CLE genes also play an important role in vascular development in shoottissues. The wedge-shaped vascular bundles of the plant stem consistof two conducting tissues, the phloem and the xylem (Fig. 5). Thephloem lies laterally and contains sieve tubes that transport sugarsand amino acids, whereas the vessels and tracheids of the interiorxylem transport water and ions absorbed from the soil. Between thesetwo mature tissues lies a narrow strip of meristematic cells calledthe procambium or vascular cambium. Procambial cells divide parallelto the plane of the stem to generate phloem cells in one directionand xylem cells in the other direction. This secondary growth propertyof plant stems allows them to grow radially over long periods of timeand is regulated by both systemic hormone signals and localized CLEpeptide activity.
TDIF regulates the organization of the stem vascular bundle
TRACHEARY ELEMENT DIFFERENTIATIONINHIBITORY FACTOR (TDIF) is an active peptide derived from the
CLE41 and
CLE44 coding sequences that has two related functions: to promote theproliferation of
Arabidopsis procambialcells and to inhibit their differentiation into xylem (
Ito et al., 2006). The effect ofTDIF peptide application to procambial cell proliferation is enhancedby simultaneous treatment with CLE6 peptide (
Whitford et al., 2008), althoughthe biological significance of this is as yet unknown. The TDIF ligandis bound by the LRR receptor-like kinase PHLOEM INTERCALATED WITHXYLEM (PXY) (
Hirakawa et al., 2008;
Etchells and Turner, 2010), also known as TDIF RECEPTOR (TDF), which controls the rate andorientation of procambial cell division (
Fisher and Turner, 2007;
Etchells and Turner, 2010;
Hirakawa et al., 2010). The CLE42 peptide, which differsfrom TDIF in one amino acid, also has partial TDIF activity (
Hirakawa et al., 2008) and showsa weak interaction with the PXY extracellular domain in
in vitro assays (
Zhang et al., 2016).
PXY also interacts genetically with the
ER receptor kinase gene to regulate vascularorganization, acting to prevent the intercalation of phloem and xylemin the inflorescence stem vascular bundles (
Etchells et al., 2013).
Like its counterparts in the shootapical meristem, the TDIF/PXY ligand-receptor duo appears to act asa short-range signaling module.
CLE41 and
CLE44 are both expressedin the phloem (
Hirakawa et al., 2008;
Etchells and Turner, 2010), and TDIF protein can be detected in the extracellular space aroundphloem precursor cells (
Hirakawa etal., 2008). In contrast,
PXY is specifically expressed in the adjacent vascular procambium (
Fisher and Turner, 2007). Thus TDIFsignals in a non-cell autonomous fashion from phloem cells to inducethe proliferation and suppress the differentiation of the neighboringprocambial cells. At the molecular level CLE41 signaling negativelyregulates
PXY expression in inflorescencestems (
Etchells and Turner, 2010). Such ligand-mediated repression of receptor gene expression alsooccurs in animal systems to shape the gradient of ligand activityacross tissues (
Cadigan et al., 1998).
Several groups have recently reportedthe crystal structure of the TDIF-PXY ligand-receptor complex (
Morita et al., 2016;
Zhang et al., 2016;
Li et al., 2017). The TDIF peptideconsists of 12 amino acids, including two hydroxyproline (Hyp) residues(
Ito et al., 2006),whereas PXY belongs to the class XI subfamily of LRR-RLKs and is composedof an extracellular LRR domain, a transmembrane domain and a cytoplasmickinase domain. Resolution of the PXY extracellular domain crystalstructure revealed that it forms a twisted superhelix consisting of22 LRR motifs, with N-terminal and C-terminal capped ends that areinvolved in stabilizing the structure (
Morita et al., 2016;
Zhang et al., 2016). TDIF was found to adopt an extendedconfirmation that fits in a shallow groove along the LRR4-LRR15 motifson the inner concave surface of the PXY extracellular domain. A horseshoe-shapedkink around the Gly
6-Hyp
7 residues in the middle of the peptide is recognized by a complimentarilyshaped pocket formed by residues in LRR8, LRR9 and LRR11 and is criticalfor the TDIF-PXY interaction.
The crystal structures provide importantinsights into the mechanisms of CLE ligand-receptor binding. The kink-formingresidues of TDIF are highly conserved among the CLE family peptides,and the Gly
6 residue is required for thebinding of TDIF to PXY (
Li et al.,2017). Mutations in Gly
6 reducethe activity of TDIF as well as several other CLE peptides, includingCLV3 and CLE3 (
Fletcher et al., 1999;
Ito et al., 2006).Further, the residues in the kink recognition pocket of PXY are conservedamong CLE receptors such as CLV1 and BAM1/2/3 (
Morita et al., 2016;
Li et al., 2017), suggesting thatthis interface may function as a feature for general recognition ofCLE ligands by their receptors. In contrast, the His
1 and Asn
12 residues at the ends of thepeptide appear to be important for the discrimination of TDIF fromother CLE molecules by the PXY receptor (
Morita et al., 2016;
Li et al., 2017).
Downstream components of TDIF signaling
The dual functions of TDIF-PXY signalingin shoot vascular development are mediated by distinct downstreamcomponents. A key downstream target of TDIF-PXY signaling to directprocambial cell proliferation is
WOX4, which is expressed in the vascular procambium and cambium (
Hirakawa et al., 2010;
Etchells et al., 2013).
WOX4 is rapidly induced by exogenous applicationof TDIF in a PXY-dependent fashion (
Hirakawa et al., 2010), and promotes procambial celldivision (
Ji et al., 2010;
Suer et al., 2011). The
WOX14 gene acts redundantlywith
WOX4 to promote procambialcell division but not vascular organization (
Etchells et al., 2013). This alongwith the data that the PXY-ER genetic interaction affects vascularorganization but not vascular cell division indicates that these aregenetically separable processes that may be regulated by CLE-WOX signalingmodules with some shared and some unique constituents.
WOX4 is not required, however, for the suppression of xylem differentiationby TDIF (
Hirakawa et al., 2010). Instead, at the plasma membrane of procambial cells, the PXY receptorkinase physically associates with and promotes the kinase activityof BRASSINOSTEROID-INSENSITIVE 2 (BIN2) and other members of the GLYCOGENSYNTHASE KINASE 3 (GSK3) family of proteins in a TDIF-dependent manner(
Kondo et al., 2014).The GSK3 proteins, which also function in brassinosteroid signaling,act redundantly to inhibit procambial cell differentiation into xylemby repressing the activity of the transcription factor BES1. Giventhat brassinosteroids also promote xylem cell differentiation (
Caño-Delgado et al., 2004;
Yamamoto et al., 2007), furtherstudies should uncover the extent of crosstalk between the differentsignal transduction pathways.
Finally, the role of the TDIF/PXYpathway in shoot vascular development in trees, which produce woodvia the differentiation of procambium cells into xylem, has been investigatedby cloning
PtCLE41 and
PtPXY from hybrid aspen (
Etchells et al., 2015). Molecularcomplementation experiments showed that both
PtCLE41 and
PtPXY arefunctional orthologs of the corresponding
Arabidopsis genes. Tissue-specific overexpression of
PtCLE41 and
PtPXY in hybrid aspen produced taller trees with a twofold increase inthe rate of wood formation and increased overall woody biomass, indicatingthat the CLE41 signaling pathway functions to regulate secondary growthin trees by controlling procambial activity. Such knowledge may beexploited to enhance secondary growth and wood formation in commerciallygrown tree species.
Perspectives and frontiers
Functional analysis of CLE peptideligands will become increasingly important as systematic genome-wideanalyses continue to identify
CLE gene families in agriculturally valuable crop species. Recently,84
CLE peptide-encoding genes wereidentified in soybean
(Glycine max) and 44 in common bean
(Phaseolus vulgaris) (
Hastwell et al., 2015). Phylogenetic analyses of the soybean, common bean and
Arabidopsis pre-propeptide sequences yieldedseven distinct groups based on their CLE domain sequence and predictedfunction, enabling the distinguishing of soybean and common bean orthologsof the
Arabidopsis CLV3, CLE40and the TDIF peptides. In poplar
(Populustrichocarpa), genome-wide analysis identified a total of50
CLE genes (
Han et al., 2016), adding 24 genesto those found in a previous study (
Oelkers et al., 2008). The first systematic analysisof
CLE genes in gymnosperms identified93
CLE genes among eight coniferspecies (
Strabala et al., 2014). In this case, only the TDIF peptide sequence was completely conservedbetween gymnosperms and angiosperms. Two TDIF orthologs from
Pinus radiata were shown to be expressedin the root and in the phloem of the inflorescence stem, suggestinga possible conserved role for TDIF peptides in regulating vascularcambium development between dicots and conifers whose ancestors divergedover 270 million years ago (
Bowe etal., 2000).
In addition, a phenetic method wasused to identify 1628
CLE genesfrom 57 different plant genomes (
Goadet al., 2016). This study found two additional
CLE genes in soybean, two in poplar, and19 more in maize
(Zea mays) thanpreviously reported (
Je et al., 2016). Up to nine
CLE genes were identifiedin mosses and lycophytes, but none were detected in green algae. Clusteringanalysis based on the full pre-propeptide sequences generated 12 groupsof CLE proteins sequences, with CLE peptides known to be involvedin meristem activity, vascular development or nodulation clusteringtogether. However, these studies underscore how little informationis available about the biological functions of the vast majority ofCLE peptide ligands.
The advent of genome engineeringthrough the CRISPR-Cas9 system (
Fenget al., 2013;
Nekrasovet al., 2013) has the potential to dramatically accelerateour understanding of
CLE gene functionin plants. Unlike other methods such as TILLING (
McCallum et al., 2000) or transposonmutagenesis, the small size of the
CLE coding sequences is not an impediment to generating null mutationsusing the CRISPR-Cas9 system. Moreover, multiple
CLE genes that show tight genetic linkage and/or stronglyoverlapping expression patterns can be targeted simultaneously, helpingto surmount the widespread functional redundancy that occurs among
CLE family members (
Jun et al., 2010). The applicationof genome editing to
CLE genesin both model plants and crop systems will provide valuable new insightsinto the mechanisms of cell-to-cell communication in plants as wellas an expanded toolkit for augmenting crop plant growth and resiliencein response to global climate change.
Higher Education Press and Springer-Verlag BerlinHeidelberg
PDF
(1130KB)
