|Home | About | Journals | Submit | Contact Us | Français|
Ethylene (ET) is a notable signaling molecule in higher plants. In the year 1993 the ET receptor gene, ETR1, was identified; this ETR1 receptor protein being the first plant hormone receptor to be isolated. It is striking that there are six ET receptors in tomato instead of five in Arabidopsis, the two best-known signaling-model systems. Even though over the last few years great progress has been made in elucidating the genes and proteins involved in ET signaling, the complete pathway remains to be established. The present review examines the most representative successive advances that have taken place in this millennium in terms of the signaling pathway of ET, as well as the implications of the signaling in the reproductive organs of plants (i.e., flowers, fruits, seeds and pollen grains). A detailed comparative study is made on the advances in knowledge in the last decade, showing how the characterization of ET signaling provides clues for understanding how higher plants regulate their ET sensitivity. Also, it is indicated that ET signaling is at present sparking interest within phytohormonal molecular physiology and biology, and it is explained why several socio-economic aspects (flowering and fruit ripening) are undoubtedly involved in ET physiology.
Ethylene (ET), although it is the simplest olephine, plays determinant roles in coordinating and regulating a wide variety of growth and developmental processes in higher plants.1 This simple unsaturated hydrocarbon participates directly in controlling key plant-development processes, such as embryogenesis, the germination of pollen grains and seeds,2 floral senescence,3 fruit abscission and maturation,4–6 pathogen interactions7 and responses to a many types of stress.1,8 Recent studies on defence-signaling pathways reveal that ET, jasmonic acid, and salicylic acid play a dominant role in regulating the induced defences against microbial pathogens.9,10 These hormonal pathways can either positively or negatively interact with each other, depending on the type of pathogen,7,11–13 but how these defence-response pathways communicate with each other remains unclear.
To carry out such diverse functions the ET response is tightly regulated at multiple levels, from synthesis and perception to signal transduction and transcription regulation.8,14 Although most plant organs have the capacity to produce ET, its level is usually low. However, ET synthesis spikes with each of a series of developmental phases (i.e., flowering, maturity, germination) or due to environmental factors (e.g., flooding, drought, pathogen attack, noxious chemicals, ozone, mechanical damage). The presence of other hormones can also alter the synthesis and signaling of ET.15–19 The ET-biosynthesis pathway has been elucidated in recent decades, providing the basis for subsequent biochemical as well as molecular genetic analysis of this pathway.20–23 However, although knowledge of the ET-signaling pathway is considerable (for some reviews, see refs. 4, 5, 13 and 23–29), many questions remain to be resolved.
ET is synthesised from methionine through the intermediaries S-adenosyl-L-methionine (AdoMet) and 1-aminocyclo-propane-1-carboxylate (ACC).20 The transformation of AdoMet into ACC is catalysed by the enzyme ACC synthase (ACS)22,30,31 and is considered a rate-limiting step in the ET biosynthesis (Fig. 1). ACC is afterwards oxidized to ET by the enzyme ACC oxidase (ACO).22,32,33 ACS and ACO are encoded by multigene families with members that are differentially regulated at the transcriptional level by the developmental programme, hormonal signals, and by stimulants that induce ET production.6,8,22,33 Arabidopsis has at least six genes that encode ACS,22,34 with AtACS4, AtACS5 and AtACS6 being induced by auxin, cytokinin, and ozone, respectively.22 In species in which several ACS have been characterized, each gene has a single form of transcriptional and/or post-transductional regulation.23,35 ACS activity was thought to be key in controlling ET production and that ACO activity was constitutive. However, the role of ACO in regulating ET biosynthesis has become apparent in recent years. Thus, it is likely that the first step in catalytic ET biosynthesis from ripening climacteric fruits is the de novo synthesis of ACO, and the ET that is produced induces ACS gene expression, which in turn gives rise to more ACC.4,22 The biosynthesis of ET may also be regulated by a mechanism that involves the stability of the ACS and the C-terminal domain,36–38 as recently confirmed with the help of the mutant eto (ETO, ET overproduction).39 ETO2 was cloned and encodes ACS5 whereas the ETO1 protein appears to be a candidate for the repressor or inhibitor of ACS4 function.40 Likewise, accumulation of ACS protein controlled by phosphorylation (Fig. 1) may also occur.35 Evolutionarily, because the genes corresponding to the ACS are more similar between different species than between members of the same species,41 it has been suggested that the polymorphism must have occurred before the separation between mono- and dicotyledonous.6
Hormones are chemical signals that integrate internal developmental and external environmental inputs and transform them into appropriate responses. The responses require specific receptors and a signal-transduction pathway to coordinate downstream responses. It is logical to accept that a key point of signaling regulation is at the receptor level. Among those isolated receptors to date (e.g., ET,45 cytokinins,42 auxins43 and ABA44), the ET receptor (EtR) was isolated first. Thus, EtR is perhaps the first plant receptor kinase for which a ligand with hormonal properties (ET) has been identified.45,46 The genetic hierarchy among ET biosynthesis and signaling-pathway components in Arabidopsis (Table 1) has been established by epistasis analysis using mutants.47,48 Thus, Arabidopsis mutant seedlings that display either etiolated growth (plus ET) in the dark or show a constitutive triple response (plus air), have led to the identification of five EtR and a number of components of the signal-transduction pathway.48 The EtRs are sequence-related to the superfamily of catalytic receptors in bacteria referred to as two-component regulators.8,49 The two-component perception system is comprised of three domains: (a) NH2-terminal membrane-associated sensor which shows high-affinity ET binding; (b) the C-terminal domain of the receptors shows varying degrees of sequence identity to the His-kinase (HK) catalytic domains of bacterial two-component regulators; and (c) a C-terminal response regulator containing a receiver domain reminiscent of bacterial-response regulators.8,49 In bacterial systems, these proteins transduce signals via the autophos-phorylation of a His residue in the kinase-transmitter domain, followed by the transfer of a phosphate to an Asp residue in the receiver domain of a response-regulator protein.50 The family of EtRs in Arabidopsis was divided into two categories according to phylogenetic and sequence analyses:47,49,51 subfamily I, formed by AtETR1 and AtERS1, which presents three hydrophobic sub-domains at the NH2-terminal, shows HK activity and has the highest conservation of the HK elements; and subfamily II, consisting of AtETR2, AtERS2, and AtEIN4, has 4 hydrophobic subdomains and lacks HK activity. EtRs belong to one of four families of HK found in plants.52 The absence of phenotypes in single-receptor mutants suggests that despite the structural differences, there is functional redundancy (or compensation of function) among the EtRs.
Since ET is a diffusible ligand in both aqueous and lipid environments, ET theoretically could be located anywhere within the cell. Curiously, however, EtRs are membrane-bound. In 1993, the first gene candidate to EtR (ETR1) was isolated in A. thaliana by map-based cloning.45 Mutant alleles of ETR1 confer dominant ET insensitivity. Afterwards, the affinity of ETR1 for ET was confirmed and it was demonstrated that for this affinity the first 165 amino acids of the sequence are necessary and sufficient.53 The NH2-terminal hydrophobic region of the ETR1 contains three putative transmembrane-spanning subdomains that constitute the ET-binding site54 and serve to demonstrate the membrane location of receptor.53 All known ETR1 mutations are located within the ET-binding domain. After ETR1, another four membrane-bound receptors homologous to ETR1 (ERS1,55 ETR2,56 EIN4,51 ERS251) were isolated. Like ETR1, ERS1 has three membrane-spanning regions; whereas ETR2, EIN4 and ERS2 have four, the first being a signal sequence for which the function may be to target proteins to the secretory pathway.46,51 Although it appears beyond doubt that ETR1 and ERS1 have similar affinities for ET,57 this has not been effectively demonstrated for the EtRs with four membrane-spanning regions. The importance of the transmembrane domains is emphasized by the finding that, although the other members of the EtRs family share similar features with ETR1, the highest degree of amino-acid conservation is found within this region involved in ET binding.51
ETR153 and ERS157 form homodimers through disulphide bonds linking the NH2-terminal (for a review, see Bleecker and Kende, 2000).24 This fact was also demonstrated when both ETR1 and ERS1 genes were expressed in yeast.53,57 Although the capacity of the other EtRs to be dimerized still could not be demonstrated, it was in fact confirmed that ERS1 and ERS2 could be heterodimerized with proteins present in Arabidopsis.58 The possible heterodimerization between different members of the same family of EtRs may further increase the number of potential combinations among the receptor complexes, providing each with different properties, thereby explaining the fact that ET can bind to the receptor over a broad range of concentrations and with two different dissociation rates.59 Thus, in the absence of quantitative variations in the level of the EtRs, the changes in the ethylene-binding properties could regulate the activity of the receptors in a certain phase of plant growth and development. Each dimer contains a single ET-binding site.54 Any mutation in the transmembrane domain of ETR1 confers ET dominant insensitivity.45,51,56 Analysis of loss-of-function mutants (LOF) of ETR1, ETR2, EIN4 and ERS1 has shown that these receptors are functionally redundant and that they negatively regulate the ET response.60 All of A. thaliana-EtR LOF mutants have semi-dominant ET-insensitive phenotypes.48
The ET-binding domain in ETR1 possesses copper, the cofactor necessary to carry out the high-affinity binding of ET. The copper is coordinated by two conserved amino acids (Cys-65 and His-69) present in the second transmembrane domain.46,54 Silver, an inhibitor of ET perception, presumably acts by displacing copper at the active site of the ETR1 complex. The presence of copper and its requirement for ET binding was clearly established both genetically and biochemically.54,61 ETR1 could potentially interact with the membrane-associated protein RAN1 (Responsive to ANtagonist), a copper transporter. However, although RAN1 appears to be involved in ET signal transduction, its subcellular location is unknown.62 The mutants ran1-1 (Table 1) and ran1-2 have lower levels of copper, compared with the wild type, and show the triple response when treated with the ET antagonist trans-cyclooctene. On the other hand, plants without functional RAN1 protein (e.g., etr2) show a constitutive triple response.61 The RAN1 protein has a similarity with copper-transporting P-type ATPases such as the yeast ccc2.61 Also, as opposed to the other EtRs, ETR1 has no signal sequence nor domains indicative of a clear subcellular location.53 Evolutionarily, EtRs could have had an endosymbiotic origin. Thus, the symbiont cyanobacterium Synechocystis contains a protein slr1212, which resembles plant EtRs.54 The polypeptide slr1212 has an affinity for ET and has three transmembrane domains that are 38% identical to those of ETR1.54
In the NH2-terminal hydrophobic region of the EtRs, there exists, in addition to the sensor domain described above, a highly conserved domain called GAF (found in cGMP phosphodiesterases, Adenylate cyclases and Fh1a transcription factors).63,64 The function of the GAF domain is currently unknown. After the GAF sequence is the kinase domain (KD). In ETR1, only EtR with demonstrated HK activity, this putative KD was expressed in yeast as a fusion protein with glutathione S-transferase and demonstrated to have histidine kinase (HK) activity.65 The role of this HK activity in ET signal transduction has not been determined. It is known that bacterial HKs have in their catalytic centre five consensus motifs termed H, N, G1, F and G2. In relation to the presence of these subdomains in EtRs, the same findings indicate that HK activity may not be necessary for an ET response, because some receptors lack the His within the kinase domain that is predicted to be phosphorylated.51,64,66 The heterodimer formation could permit signal transduction by proceeding via cross-phosphorylation. The lack of conservation of HK domain in EtRs is a confirmed finding. The final part of the EtR corresponds to the response regulator (RR) domain and possesses, as in the two-component system of bacteria, an Asp that can be phosphorylated by the His residue of the HK domain.47 In the case of AtETR1, AtETR2 and AtEIN4, the RR seems to modulate the activity of a downstream factor (e.g., CTR1).45,51,56 However, some members of EtRs family of A. thaliana and Licopersicom esculentum (e.g., AtERS1, AtERS2, and LeETR3) are missing the RR.24,25 This fact suggests an important but unidentified function of this domain and may indicate nonredundant functions for a subset of receptors. In tomato, of which a great number of aspects are known concerning the ET regulation of fruit development,4,67,68 six members of the EtR family have been described (LeETR1 and LeETR2;69,70 LeETR3 (Nr, Never ripe);67 LeETR4 and LeETR5,66 LeETR670). All of these LeEtRs are ortholog to those described in Arabidopsis. These six proteins are very divergent (exhibiting less than 50% identity in the most extreme case); and, as in Arabidopsis, the NH2-terminal hydrophobic region of the LeETR1 contains three putative transmembrane-spanning subdomains that constitute the ET-binding site. All the tomato EtRs (except LeETR6) have affinity for ET. Five consensus motifs (e.g., H, N, G1 F and G2) which are key features of bacterial HK are present in LeETR1, LeETR2 and NR. With respect to the HK activity in tomato EtRs, some HK domains lack the His involved in phosphorylation (e.g., LeTR5).66 When all the results for tomato are taken together, ET perception becomes quite complex, with at least five structurally divergent, presumed receptor family members exhibiting significant variation in expression levels throughout development.66 It has been noted that AtETR1 is not regulated by ET,45 whereas putative EtRs in tomato and Rumex palustris are upregulated by ET.71
The EtRs are thought to transmit ET signals through interaction with RAF-related Ser/thr kinase, CTR1 (Constitutive Triple Response 1), the first ET signaling pathway gene cloned in higher plants.72,73 The triple response was observed in the mutant ctr1, which has a recessive character and constitutive phenotype. The role of RAF Ser/thr protein kinase in mammals acts as a MAPKKK (Mitogen Activated Protein Kinase-Kinase-Kinase). The similarity of CTR1 to MAPKKK and genetic epistatic analysis suggests a downstream MAPK cascade in the ET signaling. To date, no gene with homology to MAPK in A. thaliana has been associated with ET signaling and no intermediate components have been identified to act between the EtRs and CTR1 kinase. Thus, yeast two-hybrid and in vitro data indicate that the kinase domain of ETR1 and ERS1 can directly interact with CTR1.73 CTR1 show in vitro Ser/thr kinase activity. However, there is no evidence in vivo to demonstrate the cascade role of MAPK nor to identify the substrate of CTR1. In the absence of ET, EtRs stimulate the kinase activity of CTR1 and repress the signaling pathway. However, ET signaling does not appear to be completely dependant upon CTR1 signaling.72 Arabidopsis has various CTR1-related proteins, but their function and relationship with ET transduction is still unknown. While EtR genes have been isolated from tomato, few presumed members of the ET transduction pathway have been described.74,75
EIN2 (ET INsensitive), an integral membrane protein with 12 predicted transmembrane domains at its NH2-terminal and a novel hydrophilic C-terminal, is an essential positive regulator of the ET signaling, since null mutations in EIN2 result in a complete loss of ET responsiveness.62,76 Over 25 EIN2 alleles have been discovered, more than of any other ET signaling-pathway component. Of all the ein mutants obtained (Table 1), ein2 is the only known recessive mutation that leads to complete ET insensitivity.62 Epistasis analysis places EIN2 downstream of CTR1, and upstream of EIN3.62,76 The hydrophobic NH2-terminal domain shares homology with the NRAMP family of mammalian metal-ion transporters,77 implying that in upstream EIN2, there is a metal ion.62 However, no studies have been able to identify a metal (calcium?) transporting activity for EIN2. The mechanism by which EIN2 is activated remains unresolved. The C-terminal domain of EIN2 was overexpressed and partially activated the downstream nuclear ET signaling.62 The results to date suggest that the NRAMP domain of EIN2 is needed for sensing the upstream ET signal, while the C-terminal domain serves to send the signal to other nuclear components situated in downstream EIN2.
In summary, when ET is absent, the EtR/CTR1 complex negatively regulates the ethylene-response pathways by suppressing the activity of the putative ion channel, EIN2.62 The EtR complex inhibits signaling, leading to the activation of response pathway, to an increase in EIN2 activity, and to an upregulation of ET-response pathways.60 The uncharacterized component EIN5, near EIN2, may also positively regulate the pathway.76 The EIN2 protein is thought to stimulate ET responses through the activation of a transcriptional cascade mediated by the EIN3 family of transcription factors.78 As in Arabidopsis, LeEIN2 is encoded by a single gene.
The ET response in higher plants involves a transcriptional cascade. The cloning of EIN3 provided direct evidence for early nuclear regulation of ET signal transduction.79 EIN3 is a nuclear protein included in a family of 6 members in Arabidopsis and is a transcription factor that works downstream from EIN2.62,76,79 EIN3, EIL1 (EIN3-Like) and EIL2 (three members of EIN3 family) are involved in ET signal transduction since they can rescue the ein3 mutant phenotype. Homodimers of EIN3, EIL1 and EIL2 bind to the promoter of ERF1 (ET Response Factor) gene and activate its transcription in an ET-dependent way.78,79 Transcription factor ERF1 and other EREBPs (also called ERFs) can interact with the GCC box in the promoter of target genes and activate downstream ET responses.80 EIN3 is both necessary and sufficient to stimulate ERF1 expression, and there are data indicating that ERF1 may regulate one branch of the ET response pathway downstream of EIN3.79 EIN3 gene expression is not induced by ET. Experiments carried out with mutants indicates that EIN3 may be regulated by ET at the protein level.78,79
The updated model of ET signaling in higher plants, based mainly on genetic analysis of A. thaliana is reflected in Figure 2. Though knowledge of the perception of ET in dicots advanced considerably in the first six years of the present millennium (reviews in refs. 5, 7, 13, 14, 28, 29, 70, 81 and 82), the signaling pathway downstream EtR is still far from being completely elucidated. Recently, in mono-cots such as rice, five genomic sequences have been isolated from several other putative EtRs (OS-ERS1, OS-RS2, OS-ETR2, OS-ETR3 and OS-ETR4), the autophosphorylated His residue being absent in both OS-ETR2 and OS-ETR3.83,84 Likewise, ETR1 homologues have also been isolated from plants other than Arabidopsis and tomato: carnation,85 geranium,86 peach,87 tobacco,88,89 pear90 and damson-plum.91 It has been confirmed in tomato that, although all the EtRs isolated may be bonded to ET in vitro, some of the them do not have HK activity.26 The epistatic analysis and LOF continues to be one of the most widely used tools in trying to complete the ET signaling framework.23 Cancel and Larsen (2002)92 showed that etr1 LOF had increased sensitivity to ET and that a greater ET response results not only in increased sensitivity but also in exaggerated response. Moreover, Wang et al. (2003)93 demonstrated that the etr1/ers1 double mutant was the only LOF that reduces levels of the Subfamily I receptors, exhibiting constitutive ET hypersensitivity; these results suggest that both subfamilies I and II may not have entirely redundant functions in vivo. Although etr1 LOF display a severe ET phenotype, these mutants remain ET responsive,94 suggesting that an alternative mechanism bypassing CTR1 in ET signaling may exist in Arabidopsis.92 Null mutants of each EtR gene are phenotypically wild-type, with the exception of etr1-7, which displays increased sensitivity to ET.92 Many of the genetic data compiled to date lead to the conclusion that the responses to ET in higher plants could be modulated by an alteration in the expression of the EtRs. However, a model in which clustered EtRs act cooperatively was proposed for Arabidopsis.95 By contrast, analysis of the expression patterns of the different EtRs genes in tomato suggests that different tissues contain different pools of receptor proteins.96 When this revision was being made, a previously undescribed gene (RTE1; Reversion-To-Ethylene-Sensitivity) that regulates ET responses in Arabidopsis was identified.97 RTE1 is a regulator of ETR1 function and negative regulator of ET signaling. ET treatment induces RTE1 expression, and overexpression confers reduced ET sensitivity that partially depends on ETR1. In parallel to RTE1, its homologue in tomato, the Gr/Nr-2 (GR) locus, was identified by means positional cloning.98 Gr/Nr-2 encodes an evolutionarily conserved protein of unknown function and that the authors associate with ET signaling. It is possible that GR may function to disrupt ET signaling from specific receptors and modulate ET responses in a tissue-specific manner. It is noteworthy that LOF of an Arabidopsis GR homologue, RTE1/At2626070, suppresses ETR1-2-mediated ET insensitivity.97
The subcellular location of EtRs is still subject to debate. Recently, it was demonstrated that At-ETR1 is located predominantly in the endoplasmic reticulum (ER) membrane (Fig. 2), these data supporting a central role of this cellular compartment in hormonal perception and signaling.99 However, NT-HK1 and OS-ER1, OS-ERS2 and OS-ETR2 receptors of tobacco and rice, respectively, were localized mainly in the plasma membrane in a transient system.83,89 On the other hand, an indirect phosphotransfer from ETR1 to CTR1 was suggested,100 and weak associations between CTR1 and some EtRs have also been recognized.92 Regarding the EtR location, it was noted that CTR1 can also be found in the ER membrane due to its interaction with ETR1.101 Thus, it has been demonstrated that the HK and the receiver domain of ETR1 can interact with the Raf-like kinase CTR1.101,102 The NH2-terminal of CRT1 has been shown to interact in vivo directly with the ETR1, via a putative CTR1-specific protein-protein interaction domain called the CN box and this possible association is required to turn off the ET-signaling path-way.102 Moreover, Huang's work demonstrated that CTR1 have intrinsic Ser/thr protein kinase activity.102
More data on HK activity of EtRs and the formation of the EtR-CRT1 complex were reported by Mason and Schaller (2005).103 In the year 2004, it was observed that the kinase domain of the EtR of the subfamily II of Arabidopsis had Ser/thr protein kinase activity,104,105 and that the His autokinase activity of ETR1 demonstrated in vitro is not required for ET signaling in vivo.93,104,105 Notably, evidence was found for Ser/thre and HK activities in NTHK2 of N. tabacum, another EtR gene. The NTHK2 heterologous expression in yeast demonstrated that the Ser/thre and HK activities were found in presence of Mn+2 and Ca+2, respectively.105 Even though HK activity of ETR1 has been demonstrated,65 the role of HK activity in ET signal transduction is still not clear.93,106 A mutation that eliminates HK activity did not affect the ability of etr1 to confer ET insensitivity. A truncated version of etr1 that lacks the HK domain also conferred ET insensitivity.107 In relation to the possible need for ETR1 to have kinase activity for the CTR1-binding, the heterologous expression of ETR1 demonstrated that this enzymatic activity was not necessary.101 However, recent studies appear to point to their essentiality for signal output, this enzymatic activity together with receiver domains playing a role in modulating the repression of ET signal.105 In addition, because CTR1 can be isolated from the ETR1-CTR1 complex and that ETR1 and CTR1 appear to be related to the ER membrane,99,101 the CTR1-ER bond possibly occurs in the ER. For the formation of this complex, the presence of EtRs is necessary, as the mutation in EtRs induces the appearance de CTR1 in the cytosol.101 In tomato, there are at least four genes encoding proteins with homology to At-CTR1 and all of these have been shown to functionally complement the A. thaliana etr1 mutation.5,108
MAPKs are involved in regulating plant growth and development. However, the underlying mechanisms are unknown because of the lack of information about their substrates. Recently, Ouaked et al. (2003)109 suggested that CTR1 (MAPKKK) acts in the “MAPK module” and show a very direct association of this module with ET signaling. They propose that in absence of ET, CTR1 is activated, negatively regulating SIMKK (a MAPKK from Medicago sativa). When CTR1 is inactivated by ET, SIMKK becomes activated and in turn activates two M. sativa MAPKs (SIMK and MMK3) or the presumed Arabidopsis orthologues of SIMK (MPK6 and MPK13, respectively) (Fig. 2). However, it remains unclear how the MAPKs activate ET targets genes through EIN2 and EIN3. EIN2 is unable to complement metal-uptake-deficient yeast strains, as shown by authentic Arabidopsis Nramps genes.110 The ein2 mutants have been found (Table 1), into others processes, in screens for defects in ABA hypersensitivity.111,112 It bears noting that the activation of SIPK, a tobacco MAPK, induces ET biosynthesis, whereas MPK6 (orthologue of SIPK in A. thaliana) is required for ET induction in Arabidopsis. In this induction, phosphorylation of ACS2 and ACS6 by MAPK are involved (Fig. 1).113 These investigations have demonstrated that the primary role of the MPK6 is the regulation of ET biosynthesis, not signaling.113,114
In Arabidopsis, the EIN3 family is composed by five members115 and the increased expression of EIN3/EIL (EIL is an EIN3-like) is sufficient to induce an ET response,116 the transcription of EIN3/EIL having an essential role.115,117,118 The remaining members of EIN3 family participate in certain developmental phases.117,118 On the other hand, in tomato fruits, all members of the LeEIL gene family are functionally redundant and regulate the ET responses throughout plant development.117 In Arabidopsis and tomato, that transcription is not regulated by ET, whereas the protein EIN3/EIL is.119,120 It is striking that the increased EIN3 protein levels were correlated with decreased glucose sensitivity.121 In the absence of ET, EIN3 undergoes continuous degradation.119–124 The signaling mechanisms through which ET acts to stabilize the EIN3 protein is unknown. The EIN3 binding factor (EBF) 1 and 2 (two F-box proteins) control the EIN3 levels. Both EBF factors form part of the E3 complex in which EIN3 is confined, poly-ubiquitinated, and degraded by proteasome (Fig. 3). That is, EIN3 is regulated post-translationally by a proteasome-mediated protein-degradation pathway.119,120,122 ETO1 is a component of the E3 complex,35 and ET stimulates la expression de EBF2.115,122 In tomato, the family EIN3 is represented by three members (e.g., LeEIL1–3).117 The tobacco EIN3-like gene, TEIL, has been cloned and its overexpression induced constitutive triple-response phenotypes.123 The expression of EBF genes confers a constitutive ET response in ein3 backgrounds.47 Thus, EIN2 and EIN3 or EIN5 may act through EBF1 to regulate ET-dependent processes.23,119 However, EIN3 also interacts with specific F-boxes, which function to suppress ET action and promote plant growth.120,122 How the downstream divergences in the ET pathway occur and subsequently lead to the diverse plant phenotypes is unclear.
Fruit ripening is a complex process of great physiological and economical relevance, the function of which is first to form a receptacle for developing seeds, giving protection at this stage, and later to help their dispersal by undergoing changes that make this structure amenable to predators or by starting a dehiscence/shattering process. The coordination between many biotic and abiotic factors and the genetic programme has to be precise, and this coordination relies significantly upon plant-hormone signals.124,125 With regard to ripening characteristics, fruits have been divided into climacteric and nonclimacteric, the former being especially dependant on ET for a successful ripening.
Most advances of our knowledge about ET signal transduction comes from Arabidopsis used as a model plant, but this species does not develop fleshy fruits. Orthologue genes were found in tomato, and ET signal transduction has been studied thoroughly in this model system for fleshy fruits. As described before, a total of six EtR have been found in tomato. This set of receptors is diverse in structure and sequence, suggesting the possibility of different physiological roles; indeed, each receptor has a distinct pattern of expression through development.66,126 LeETR1 and LeETR2 are expressed in all tissues, LeETR1 expression being higher and constant, while LeETR2 peaks in its expression during germination.66 It is noteworthy that the expression of the other four tomato receptors is highly regulated; LeETR4, LeETR5 and LeETR6 present higher expression in flowers and fruits than in vegetative tissues; NR and LeETR4 are pathogen-inducible. Nevertheless, to date, there is no evidence for specific developmental functions for each receptor. NR antisense fruits show normal ripening. A notable functional compensation was found when the mRNA levels of the receptors were investigated in NR antisense lines, reflecting that LeETR4 levels compensate for the loss of NR transcripts; by contrast, no other receptor showed increased expression in the LeETR4 antisense lines and these fruits have accelerated ripening.126
Orthologues to the genes found in Arabidopsis and tomato were found in many other fruits, including climacteric and non-climacteric: cucumber (Cucumis sativus),127 pears (Pyrus communis),90 muskmelon (Cucumis melo),128 passion fruit (Passiflora edulis),129 peach (Prunus persica),130 strawberry (Fragaria ananassa),131 damson-plum (Prunus domestica) (Fernández-Otero CI, de la Torre F, Iglesias-Fernández R, Rodríguez-Gacio MC, Matilla AJ. 2006; unpublished results). In most cases, these studies revealed that the expression levels of the EtRs increased with the onset of ripening. The increase in the synthesis of negative-regulating receptors at this stage may seem paradoxical, but there are some possible explanations. ET levels rise dramatically during ripening, and this probably explains how ripening may proceed in the presence of high levels of receptors. Given the long EtR dissociation times,46 perhaps by adjusting receptor levels, the plant can modulate the ET response even when high levels of ET are present. It has been proposed that the regulation of ET perception allows the plant to initiate an ET response in one tissue while suppressing the response in others, and synthesis of new receptors is probably a way to do so.5,96,126,132 It has also been suggested that EtRs could function as molecular clocks, sensing the total amount of exposure to ethylene through the development until the fruit enters ripening.5,96 EtRs are expressed in good correlation with ET production in strawberry,131 and the involvement of ET in the regulation of ADH expression, an important gene for ripening, was demonstrated in grapevine.133 These and other findings suggest that alterations in ET responsiveness might be able to mediate physiological changes in nonclimacteric fruits.
Some additional ET-signaling components were isolated from tomato; a CTR1-like gene was found to function in ET signaling by complementation in Arabidopsis.108 Several other CTR genes were identified in tomato (LeCTR2, LeCTR3 and LeCTR4), and this multi-gene family was shown to be differentially regulated with an increase in its expression during fruit ripening.134 The existence of several CTR genes opens some intriguing possibilities. Each LeCTR could interact preferentially with one receptor or with one type of receptor; and the varying ratio of receptors: CTRs might represent a mechanism for modulating ET response in tomato and other species.134 Downstream of CTR, several homologues to the transcription factors implicated in ET signal transduction were isolated from tomato; of these, LeEIL4 exhibits ripening associated expression,135 as is the case with LeERF2, one of the four tomato members of the ERF family for which expression is not found in several ripening mutants.136
Although transcriptional regulation of receptors and transcription factors are thought to regulate the response of the fruits to ET, the intricacies of this system are far from understood. Analysis of the growing information of databases, genomics experiments, functional characterization of these genes, and examination of new candidate genes will undoubtedly shed light on the subject in the future.
ET intervenes in such important processes as flower growth, development, senescence, and abscission;1,137 the two latter provoke serious losses in the flower industry. ET-sensitive flowers offer a unique model system for studying biological responses mediated by ET. Research on orchid,138 geranium,139 petunia140 and carnation141 among other species, has revealed some clues about the complex regulation that occurs with both signaling of ET biosynthesis and perception. In all of these systems, flowers treated with ET either wilt or their flower petals undergo premature abscission.
During the mutant screenings in Arabidopsis, ET-response mutants such as etr1, ein2, ein3 and ers2 showed changes in the progression of floral abscission.142 In tomato, the mutant Nr was impaired in flower abscission and senescence,143 and the induction of LeCTR1 is associated with the increase of ET during peduncle abscission and petal senescence.5 Other mutants such as Gr and Nr-2 (recently identified as alleles of the same NR gene98) also affect the ET response, regulating floral abscission and senescence.144
In geranium (Pelargonium hortorum) two genes homologous to ETR1 from Arabidopsis (PhETR1 and PhETR2) were isolated, in which the level of transcripts remains constant over floral development. The control of ET-induced petal abscission in geranium florets may be mediated by another uncharacterized member of the PhETR gene family, at the post-transcriptional level, or via a downstream component of the signal-transduction pathway.145 In cucumber (Cucumis sativus L.) CS-ETR1, CS-ETR2, and CS-ERS were also isolated. A major accumulation of CS-ETR2 and CS-ERS mRNAs were observed in the gyneceum, due probably to the higher level of endogenous ET.146 Delphinium florets are sensitive to ET, which is produced in the pistil and receptacle and causes sepal abscission, posing a serious marketing problems.137,147 Three EtR (DI-ERS1, Dl-ERS2 and Dl-ERS1–3) were isolated from Delphinium florets, and it was observed that the abscission of the sepals induced by ET could be controlled by a complex formed by these three proteins.148,149 The manipulation of the EtR genes may be a useful tool to prolong the mean life of the organ and delay its senescence. Thus, the mutation etr1-1 has been used in Arabidopsis to reduce the sensitivity to ET in petunia, carnation, and tomato.140,150 In Coriandrum sativum, a medicinal plant, this strategy was used to achieve a longer mean life cycle for the plant and delay flower senescence.151
Four EtR genes were cloned from Rosa hybrida, these being expressed differently.152,153 The expression of RhETR1 varied over development, RhETR2 was expressed in a constitutive way, and RhETR3 increased its expression in senescent flowers of the cultivar “Bronze” (this having a short life cycle) and its expression is constitutive in flowers of the cultivar “Vainilla” (which has a long floral life). RhETR1 and RhETR3 appeared to be rate-limiting for ET perception and determinants of flower longevity.154 Two genes homologous to the gene CTR1 of Arabidopsis were also isolated during flower senescence of R. hybrida (RhCTR1 and RhCTR2). The expression of RhCTR1 increased during floral senescence, coinciding with the increase in ET production, while RhCTR2 was expressed constitutively during senescence. In this system, contrast was found with respect to the negative regulation model that assumes that ET sensitivity increases during floral senescence and in response to ET.155 In petunia (Petunia hybrida) the gene PhEIN2, homologous to EIN2 of Arabidopsis was isolated, and it was found that transgenic plants with reduced levels of PhEIN2 have reduced sensitivity to ET, indicating the essential role of PhEIN2 in the signaling of ET involved in floral senescence.156
In Passiflora edulis, three cDNAs (PeETR1, PeERS1 and PeERS2) homologous to At-EtRs were isolated. The level of transcripts PeERS1 and PeERS2 increased during floral senescence; but ET production did not rise. Therefore, flower senescence should be regulated in an ET-independent manner.157 In Catharanthus roseus, the expression of a gene homologous to ETR1 of Arabidopsis was analysed, CrETR1, the transcription of which presents a high expression in the petals and ovaries.158
In carnation (Dianthus caryophyllus), three EtR genes DC-ERS1,159 DC-ETR1160 and DC-ERS285 were identified and their expression studied during senescence. The DC-ERS2 mRNA decreased in petals, increased slightly in the ovaries and remained unchanged in styles. All of these genes are regulated in a tissue-specific manner and independently of ET.85 On the other hand, in D. caryophyllus, 4 genes homologous to EIN3 were also isolated, these being regulated by development and ET; the expression of DC-EIL1 fell in the petals during natural senescence and that induced on applying ET.85,161 The levels of DC-EIL1/2 and DC-EIL4 mRNAs were very similar during floral development in the presence of ET and during pollination162 while DC-EIL3 mRNA accumulated in petals and styles exposed to ET during floral development and pollination.
In addition to the senescence and abscission, the expression of some genes belonging to the ET transduction chain was studied in different parts of the flower during their development. In Arabidopsis, a high expression of ETR1, ETR2 and ERS1 was observe-d in flower primordia, while in the stamens the ETR1 expression was greater than for ETR2 or ERS1.163 ETR2 and ERS2 were expressed during carpel development, especially in the funicle and in the ovules.56 In addition, it was confirmed that the double mutants ers1 and etr1 presented a severe phenotype with a major delay in flowering and sterility. This confirms the importance of ETRS and ERS in floral development.164
The Angiosperm flower represents the culmination of reproductive evolution in plants. In recent years, increasing studies have analyzed the regulation and expression of signaling molecule genes within the developing male gametophyte and the signaling in pollen-tube growth.165–168 The role of ET synthesis in the development and germination of tobacco pollen has previously been demonstrated.2 Moreover, the involvement of ET in reproductive physiology has also been elucidated in early pistil development.169 However, scarce work has been devoted to the study of ET signaling in pollen development. In rice, a substantial amount of OS-ERS1 mRNA was accumulated during pollen development when compared with other tissues studied and OS-ERS2 and OS-ETR2 expression.83 Similar results were demonstrated for At-EIN4 in pollen and tapetum cells when compared with the At-EIN4 mRNA level in leaves and roots.110 The expression of several At-EtRs was analysed at the cellular level by means of in situ hybridization.51 Very strong expression of At-ERS1 was observed in the developing pollen grains, whereas weak expression of At-ETR2 was detected in the stamens. At-EIN4 was highly expressed in pollen and tapetum cells, with At-ERS2 in developing pollen grains.51 Recently, a mutated melon EtR gene (Cm-ETR1/H69A) was introduced into tobacco plants.170 Some transformants had reduced sensitivity to ET, low seed yields due to abnormal stamen development, and reduced pollen production.171 Similar construction induced extended flower longevity in plants of Nemesia strumosa.172 In flowers of damson-plum, the expression of Pd-ERS1 and Pd-CTR1 was highly constitutive during stamen development, this expression decreasing during ovary development (Fernández-Otero CI, de la Torre F, Iglesias-Fernández R, Rodríguez-Gacio MC, Matilla AJ, 2006; unpublished data). All above results taken together in a recent review (e.g., Klee and Clark, 20043) indicate that the active involvement of the EtR gene family during early pollen development implies that ET may play an important role in regulating the initial formation stages of anther.171 On the other hand, limited knowledge exists about downstream EtR signaling in pollen. A petunia homologue of the Arabidopsis EIN2 gene (PhEIN2) has been characterized. Their expression is spatially regulated and controlled by ET in a tissue-specific manner, the PhEIN2 mRNA level being lower in anther than other tissues studied.173
Knowledge of the role of ET during seed development and germination is still fragmentary.2,22 Genetic and biochemical analysis of ET-response mutants in Arabidopsis has provided invaluable information on ET signaling. The mutation etr1-1 insensitizes Arabidopsis seeds to ET and intensifies sensitivity to ABA, while seeds that produce ctr1 are less sensitive to ABA than are those of the wild type.111 The involvement of ET in determining the time to radicle protrusion was studied in ET-insensitive gain-of-function (GOF) receptor mutants in tomato and Arabidopsis, as well as single and double LOF receptor mutant in Arabidopsis.174 The results support a role for ET perception in determining the length of time Arabidopsis seeds remain in the lag phase prior to radicle protrusion. It appears that EIN2 is a negative regulator of seed sensitivity with respect to ABA. Thus, the mutation ein2 in Arabidopsis produces seeds with great sensitivity to ABA and very little to E T. Consequently, both signaling pathways appear to be closely related.111 These facts indicate that seed dormancy in Arabidopsis is negatively regulated by endogenous ET, a fact that has been suspected1 but not demonstrated until now.111,112 In short, ET counteracts the effect of ABA by diminishing sensitivity of the seed to endogenous ABA. The few cases indicate that ET alone is incapable of positively regulating seed germination and, at some point of the process, an interaction between the signaling pathways of ABA and ET appears to be needed.175,176 More recently, it was suggested that ET signaling modulates the metabolism of the other plant hormone pathways in seeds, and that the hormone profiles during germination of seeds obtained from etr1-2 mutant (dominant ET insensitivity and higher level of dormancy) also suggest a requirement for higher than wild-type levels of gibberellin to promote germination in absence of a functional ET signaling pathway.16 Data on ET signaling in seeds of fleshy fruits are very scant. In damson-plum, the expression of Pd-ETR1 gene was typically constitutive throughout seed development, Pd-ERS1 was stimulated during late-green, maturation and ripening phases in relation to the early-green one, and Pd-CTR1 reached the highest level during ripening (Fernández-Otero CI, de la Torre F, Iglesias-Fernández R, Rodríguez-Gacio MC, Matilla AJ, 2006; unpublished data).
ET, a key signaling molecule of higher plants, plays a notable and pivotal role in the modulation and integration of both developmental and environmental cues into overall growth. In recent years, evidence have been accumulated that both ET biosynthesis and signaling are integrated in a broad network. Thus, regulation of ET synthesis by different hormonal and developmental signals provided an excellent model for the study of hormonal interactions and served to demonstrate that ET synthesis is regulated by a hormone-dependent mechanism during reproductive plant development, and that ET modulates responses to other hormones. Taken all this knowledge together, ET signaling is the most well-defined signaling pathway in plants. However, the mechanisms underlying these hormone-hormone interactions are largely unknown. The identification of the complete “hormone-modulated transcriptome” will uncover key regulatory genes as candidates for participation in the integration of different signals. In adittion, the avalilability of LOF and GOF mutations in all Arabidopsis genes will greatly facilitate testing of their in planta requirement in the different hormonal signaling pathways. To fully understand the various levels of communication among hormone pathways, it is essential that the biochemical activities of each of the individual signaling components are determined at the cellular level in response to a diversity of stimuli.
The experimental work that has been done over the last decade to delimite the ET signaling in Arabidopsis and tomato has been spectacular. However, the role that ET plays in plant reproductive processes is as notable that other models are needed. Understanding molecular events that lead to the alteration of ET sensitivity in transgenic plants with ET receptor genes may be a clue to elucidate the mechanism of how plants regulate the sensitivity to ET. It is to say, it is necessary to know how plants regulate the expression of ET receptor genes. Recently, important advances on ET signaling have been made using genetic and molecular-biology approaches. However, detailed aspects of each signaling component are far to be understood. Thus, is not evidenced which activity of EtR is regulated by ET nor how ET binding modulates the function of these receptors. A more deep characterization of the ET-ETR complex and its repercussion in the CTR1 regulation will provide new insights about ET perception. On the other hand, the knowledge and action of MAPK cascade is weak. Other major unresolved steps in ET signaling are the function of EIN2, a protein that plays an essential role in mediating all known ET responses, and the nuclear processes coordinated by transcription factors like EIN3/EIL and ET. In summary, taking in account the great advance in the study of ET signaling during the last decade, in the upcoming years, more and more interaction points and the molecular mechanisms involved will be revealed.
We than the Editor and two Reviewers for valuable comments that led to the improvement of this review. This work was supported by grants from Xunta de Galicia (Community of Galicia, Spain) (PGIDIT04RAG203010PR) and Ministerio de Educación y Ciencia (CGL2004-01996).
Previously published online as a Plant Signaling & Behavior E-publication: http://www.landesbioscience.com/journals/psb/abstract.php?id=3390