Four shades of detachment: Regulation of floral organ abscission

Abstract

Abscission of floral organs from the main body of a plant is a dynamic process that is developmentally and environmentally regulated. In the past decade, genetic studies in Arabidopsis have identified key signaling components and revealed their interactions in the regulation of floral organ abscission. The phytohormones jasmonic acid (JA) and ethylene play critical roles in flower development and floral organ abscission. These hormones regulate the timing of floral organ abscission both independently and inter-dependently. Although significant progress has been made in understanding abscission signaling, there are still many unanswered questions. These include considering abscission in the context of reproductive development and interplay between hormones embedded in the developmental processes. This review summarizes recent advances in the identification of molecular components in Arabidopsis and discusses their relationship with reproductive development. The emerging roles of hormones in the regulation of floral organ abscission, particularly by JA and ethylene, are examined.

Abbreviations

JA

jasmonic acid

AZ

abscission zone

LRR

leucine-rich repeat

BOP1/2

BLADE ON PETIOLE 1/2

NPR1

NONEXPRESSOR OF PR GENES 1

PR1

Pathogenesis-related Protein 1

BTP/POZ

Broad-Complex, Tramtrack, and Bric-a-brac/Pox virus and Zinc finger

AGL15

AGAMOUS-LIKE 15

etr1-1

ethylene response1-1

FYF

FOREVER YOUNG FLOWER

AOS/DDE2

ALLENE OXIDE SYNTHASE/DELAYED DEHISCENCE 2

DDE1/OPR3

DELAYED DEHISCENCE 1/OXOPHYTODIENOATE-REDUCTASE 3

DAD1

DEFECTIVE ANTHER DEHISCENCE 1

LOX3/4

LIPOXYGENASE 3/4

FAD7/8/3

FATTY ACID DESATURASE 7/8/3

DAB4/ COI1

DELAYED ABSCISSION 4/CORONATINE INSENSITIVE 1

ein2-1

ethylene insensitive 2-1

XTH

XYLOGLUCAN ENDOTRANSGLUCOSYLASE/HYDROLASE

EXP

EXPANSIN

PG

POLYGALATURONASE

HAE/HSL2

HAESA/HAESA-LIKE 2

ida

inflorescence deficient in abscission

MKK4/5

MAP Kinase Kinase 4/5

MAPK3/6

MAP Kinase 3/6

KNAT1

KNOTTED-LIKE FROM ARABIDOPSIS THALIANA 1

NEV

NEVERSHED

ARF-GAP

ADP-ribosylation factor-GTPase activating protein

EVR

EVERSHED RECEPTOR-LIKE KINASE

SERK1

SOMATIC EMBRYO RECEPTOR-LIKE KIASE 1

CST

CAST AWAY RECEPTOR-LIKE KINASE

JAZ

JASMONATE-ZIM DOMAIN

IM

inflorescence meristem

TCP4

TEOSINTE BRANCHED/CYCLOIDEA/PCF4

CTR1

CONSTITUTIVE TRIPLE RESPONSE 1

Abscission, or organ separation, is a unique and critical cell separation process in plant growth and development. It is a developmentally programmed process that takes place in specific cell types. A typical abscission zone (AZ) is comprised of distinct layers of small cytoplasmic dense cells. In wild type Arabidopsis, the AZ is 2 to 4 cell layers; however, the AZ can vary considerably–more than 50 cell layers in some species.¹ A diverse set of signals have been identified that trigger a cascade of downstream events, culminating in separation after an organ has served its purpose on the parent plant, e.g., flower petals and anthers after fertilization, ripe fruit, senescent leaves, etc., or unfertilized, damaged or diseased organs. Although Arabidopsis does not display leaf or fruit abscission, it has become a model plant to study floral organ abscission.²

Over the past decade, genetic and mutational studies have identified key molecular components involved in the regulation of floral organ abscission. These include organ fate controlling transcription factors, leucine-rich repeat (LRR) receptor-like kinases, a small signaling peptide, a membrane traffic regulator, MADS-box transcription factors, and central components from several hormone pathways.^3–22 These studies have also uncovered distinct features of the abscission process associated with particular molecular components. Some of these components are considered to be essential for organ separation while others are responsible for controlling the timing of process. Cellular, physiological and gene expression studies have provided the framework for a working model in the regulation of floral organ abscission. The model divides abscission into 4 developmental phases that culminate in complete organ separation (Fig. 1A): Phase (1) establishment of AZ; Phase (2) acquisition of competence for activation of abscission; Phase (3) cell separation; and Phase (4) organ separation and trans-differentiation between the proximal and distal sides of the cells.^{2,16,19,23,24}

Figure 1.

Four phases of floral organ abscission in Arabidopsis. Progression of floral organ abscission is divided into 4 major phases (A or in wild type). Signaling components affecting the process are represented in the corresponding phase. In Phase 1, an abscission zone (AZ) is established to the developmental program. BOP1/2 are required for Phase 1 and essential for floral organ abscission (B(i)). In Phase 2, the established AZ needs to acquire competence for activation of abscission. JA and ethylene signaling pathways as well as ectopic AGL15, AGL18, and FYF (oxAGL15, 18 and oxFYF) affect the timing of abscission during Phase 2 (B(ii)). Activity of cell wall loosening and hydrolytic enzymes contribute to cell separation (B(ii)). Depending on the strength of the allele and the ecotype backgrounds, IDA and NEV pathways act in both Phase 3 and Phase 4 (B(iii)). In addition to IDA and NEV pathways, small cell wall proteins (see the text) may be involved in organ separation and trans-differentiation(B(iii)). Signaling components associated with Phase 1 and Phase 4 can affect abscission independent of senescence (dashed lines for senescent petals). Phase 2 and Phase 3 are more associated with the regulation of timing of abscission and senescence (B(ii)). X represents a loss of function and ox represents an overexpression of the components. Size of arrows indicates a relative extent of delay caused by corresponding components.

Although significant progress has been made in understanding the coordinated action of several signaling components that regulate floral organ abscission, there are many unresolved questions that need to be considered and addressed. To this end, this review will summarize the key components that have been identified and discuss their proposed roles in the 4 phases of abscission. This review focuses on the molecular components of floral organ abscission in relation to other processes associated with reproductive development, and also highlights the interplay between the plant hormones JA and ethylene in the regulation of floral organ abscission.

Four Phases of Floral Organ Abscission

Establishment of abscission zone

The first phase in abscission is the formation of the AZ. This is most often identified as a region of small, cytoplasmic dense cells at the base of an organ that will be shed. In Arabidopsis, 2 key genes have been identified that are required for AZ cell differentiation, BLADE ON PETIOLE 1/2 (BOP1/2) (Fig. 1B(i)).^4,25,26 BOP1/2 belong to the NONEXPRESSOR OF PR GENES 1 (NPR1) protein family that activates the Pathogenesis-related Protein 1 (PR1) gene expression. BOP genes encode BTP/POZ (Broad-Complex, Tramtrack, and Bric-a-brac/Pox virus and Zinc finger)-ankyrin transcription factors whose functions are recognized in a broad range of plant development and defense responses.²⁷ bop1/2 mutants display a lack of floral organ AZ formation (sepals, petals, and filaments). BOP1 and BOP2 are redundantly required for the structural changes leading to timely abscission.²⁵ The lack of an AZ is reflected in the increase in force required to remove the petals from the receptacle (petal break-strength). Interestingly, expression of abscission-specific marker genes (e.g., bean abscission cellulase and chitinase) in the bop1/2 mutants is not altered suggesting that part of the developmental program for abscission may be unchanged even though there is a defect in formation of the AZ. Whether abscission signals such as JA and ethylene can promote abscission in bop1/2 mutants is currently unknown. However, without correct AZ establishment, it is likely that signals (e.g., JA and ethylene) that accelerate abscission will be ineffective (Fig. 1B(i), NO SEPARATION).

Acquisition of competence for activation of abscission

Once an AZ has been established, the differentiated cells (target cells) need to acquire the competence to respond to developmental and hormonal signals that can induce or enhance the rate of abscission (Fig. 1A). The developmental process required for the acquisition of competence to respond to abscission signals remains elusive. However, studies on MADS-box transcription factors may provide some clues as to the developmental properties (or age-related properties) associated with the signal competence.^12,13 The AGAMOUS-LIKE 15 (AGL15) gene and protein are expressed at the base of young developing apical and flower primordia.¹² A decrease in AGL15 expression precedes the onset of abscission. However, it was found that constitutive expression of AGL15 caused a significant delay in floral organ abscission while changes in cell anatomy in the AZ were not observed. Ectopic expression of AGL15 did not severely alter biochemical properties in the AZ cells, as only a delay in the timing of process was recognized. Crosses of plants that overexpressed AGL15 with ethylene response mutants, ethylene response1–1 (etr1–1) indicated that the role of AGL15 in abscission is independent of ethylene.¹² A transient increase in the expression of AGL15 at the first flower opening was sufficient to cause a delay in floral organ abscission, which suggests a role for AGL15 in age-dependent reproductive development that subsequently influences abscission.¹³ Other MADS-box transcription factors, AGL18 and FOREVER YOUNG FLOWER (FYF/AGL42), had similar effects on abscission as AGL15. It was concluded that MADS-box transcription factors act as negative regulators to control the timing of abscission (Fig. 1B (ii), DELAYED SEPARATION).^14,15

The role of plant hormones may provide additional information on the nature of developmental competence. For instance, mutations in JA biosynthesis genes (e.g., ALLENE OXIDE SYNTHASE/DELAYED DEHISCENCE 2 (AOS/DDE2), DELAYED DEHISCENCE 1/OXOPHYTODIENOATE-REDUCTASE 3 (DDE1/OPR3), DEFECTIVE ANTHER DEHISCENCE 1 (DAD1), LYPOXIGENASE 3/4 (LOX3/4), and FATTY ACID DESATURASE 7/8/3 (FAD7/8/3)) and a JA receptor gene (DELAYED ABSCISSION 4 (DAB4)/CORONATINE INSENSITIVE 1 (COI1)) alter the development of flowers.^{18–20,28–32} Deficiency in JA levels (e.g., aos) and signaling (e.g., dab4–1/coi1–37) either delay or block the dehiscence of anthers.^19,20,30 In addition, when compared with wild type, development of floral buds is delayed until the first flower opening in the JA mutants (e.g., dad1).³² JA mutants are sterile due to the defective anthers and retarded flower development that appears to cause a delay in abscission: however, this can be complemented by exogenous JA.^19,30–32 In particular, the restoration of fertility that also accompanies abscission in JA deficient mutants (e.g., opr3) takes place at specific reproductive stage (i.e. flower stage 12), a similar stage for AGL15 effect (first flower opening or flower stage 13).^13,31 These results indicate that the properties of specific flower stages (i.e., flower stage 12 and 13) harbor developmental competence to respond to the abscission signal.

JA and ethylene activate separation processes, while the defects in JA or ethylene perception delay the progression of abscission.^16–20 Fracture plane observations of the petal AZ of a JA receptor mutant (dab4–1) and ethylene perception mutants, etr1–1 and ethylene insensitive 2–1 (ein2–1), reveal that the JA and ethylene mutants have a delay without the physical changes in the AZ cell, suggesting a role for JA and ethylene in the regulation of timing.^17,19 The results summarized above collectively suggest that the currently identified signaling components in hormone signaling pathways (i.e. JA and ethylene) and several MADS-box transcription factors are critical for the developmental competence (Phase 2) in abscission (Fig. 1B(ii), DELAYED SEPARATION).^{12–15,17,19,20}

Cell separation

Biochemical changes in the AZ cells (e.g., disassembly in the cell wall matrix and cell wall loosening) are essential to facilitate floral organ abscission (Phase 3) (Fig. 1A). Although structural components (e.g., wax, suberin and proteins) deposited in the extracellular matrix beginning in Phase 3 and continuing into Phase 4 may be crucial for successful organ separation (see below), the cell separation process is largely attributed to the functions of cell wall hydrolytic and modifying enzymes. The suggested role for these enzymes (e.g., XYLOGLUCAN ENDOTRANSGLUCOSYLASE/HYDROLASE (XTH), EXPANSIN (EXP), CELLULASE and POLYGALATURONASE (PG)) in disassembly of the cell wall matrix triggered many scientists to specifically target these proteins and their genes to study cell separation processes. Several members in these gene families have been demonstrated by gene expression studies or genetic mutation to be involved in floral organ abscission as well as flower, fruit and leaf abscission, and also some in the dehiscence of anthers and seed pods.^{18,19,33–41} For example, there are 9 distinct expression profiles for Arabidopsis PGs expressed in the floral organ AZs.³⁴ All 4 phases of floral organ abscission are represented by one or more profiles for PG gene expression.³⁴ Based on the spatial-temporal patterns of expression, 4 out of 69 Arabidopsis PGs (At2g41850, At2g43880, At2g43890, At3g07970) have been identified as candidates for a role in cell separation (Phase 3) during floral organ abscission.^34,36 Due to the size of the PG gene family and underlying redundancy, pinpointing the functional relevance of a particular PG gene expressed in abscission remains challenging.^18,35 For example, one knock-out mutant for PG (At2g41850, SALK_035098) displayed a minor delay in the timing of abscission. The morphology of the separating cells in this mutant was indistinguishable from that of wild type. The results indicate that inhibiting one enzyme out of several expressed in the separation phase (Phase 3) can slow abscission down but not fully impede the process (Fig. 1B(ii), DELAYED SEPARATION).³⁵ In addition to functional redundancy, which appears to occur in the PG family as with other gene families, the substrate specificity of different isoforms of the enzymes adds to the complexity of understanding the role of each protein in organ separation.

Organ separation and trans-differentiation

Successful and complete abscission involves additional cellular mechanisms that result in a clean dispersal of organs. During abscission, the proximal side of the fracture plane engages in a trans-differentiation to repair and protect the exposed cells from potential pathogens and limit water loss (Phase 4) (Fig. 1A).^2,16,17,42 Abscission (or organ separation) demands a balance between cell wall disassembly, excessive cell expansion, and protection to optimize successful separation. Extracellular components that may be part of managing cell expansion, separation and protection are waxes, suberin, lignin and PR proteins that are deposited at the end of Phase 3 and throughout Phase 4 (Fig. 1A). Despite the obvious roles of these components in the formation of a protective layer, determining if they play a role in the actual separation process (Fig. 1B(iii)) would require new approaches and examination of the data from different perspectives.

Studies on the essential components and their downstream cascades may provide insights into cellular mechanisms in the separation process (Fig. 1B(iii)). HAESA/HAESA-LIKE 2 (HAE/HSL2) were the first receptor-like kinases identified to be essential for floral organ abscission in Arabidopsis.^5,6,43 The 2 receptor-like kinases (RLK) act redundantly to control abscission. The ida mutant (inflorescence deficient in abscission) had a similar defect in the AZ compared to the hae hsl2 mutant, and it has been demonstrated that HAE HSL2 is epistatic to IDA. The IDA gene encodes a small, secreted peptide. These results suggested that IDA might be a ligand for the HAE and HSL2 kinases, and prompted further studies to characterize an IDA-HAESA signaling pathway.^6,43 The morphology of the fracture plane cells of AZs in the early positions of flowers in ida and hae hsl2 mutants appear to be indistinguishable from the wild type AZ (i.e., no swollen cells or uncontrolled cell expansion in the AZ).^6,7 However, constitutive expression of IDA in the wild type background caused early floral organ abscission and swollen AZ cells in the post-abscission phase (Phase 4). Moreover, when IDA was overexpressed in the hae hsl2 double mutant, the floral organs did not abscise, which resembled the hae hsl2 phenotype. This supported the hypothesis that IDA was a ligand for HAE and HSL2 and IDA-HAE/HSL2 signaling cascades are required for organ separation. Interestingly, a recent comparative study indicated that the role of IDA in cell separation (Phase 3) and organ separation (Phase 4) could differ based on the strength of the allele and the ecotype backgrounds.⁴⁴ Petal break-strength measurements uncovered that the alleles of ida-1 (C24) and ida-2 (Col) behaved differently in cell separation (Phase 3) and organ separation (Phase 4).⁴⁴ Incomplete organ separation in the ida-1 mutant (compared to ida-2) indicates that cell wall loosening (or cell wall disassembly) alone (Phase 3) may be insufficient to fully separate the organs from the parent plant. This study also indicates that successful organ separation requires a coordination of events that is absent from the ida-1 mutants, or simply implicates that the expression of IDA-like genes in floral AZs is higher in Col than in C24 that may not require the substantially different signaling cascades existing in Col than C24. Furthermore, the transcriptome analysis of floral receptacles in the hae hsl2 mutant compared to the wild type showed that the expression of cell wall loosening or hydrolytic enzymes (e.g., PGs, cellulases, and EXPs) was suppressed less than that of wax and/or suberin synthesis genes (>10 -fold less than in wild type), which suggests that in addition to IDA-HAE/HSL2 signaling pathway, complete organ separation requires the coordinated actions of genes associated with cell wall loosening (Phase 3) and the genes that may aid in the deposition of a modified extracellular matrix and the formation of a protective layer (Phase 4).^45,46 The transcriptome analysis unraveled invaluable information on the downstream cascades of IDA-HAE/HSL2-mediated signaling, however it should be noted that the sampling from floral receptacles may compromise true signaling events in the AZ due to a potential baseline level of noise that can make inter-gene expression comparisons problematic.

One model for IDA signaling postulates that IDA is a ligand for HAE/HSL2 that activates MAP Kinase Kinase 4/5 (MKK4/5) and MAP Kinase 3/6 (MAPK3/6) cascades, which inhibit the activity of KNOTTED-LIKE FROM ARABIDOPSIS THALIANA 1 (KNAT1), which in turn further de-represses downstream relatives, KNAT 2/6 to regulate cell wall hydrolytic and modifying enzymes (Fig. 1B(iii), NO SEPARATION), for organ separation.^5-7,47,48

Membrane trafficking, that may be part of the cellular mechanisms, is also critical for abscission. NEVERSHED (NEV) encodes an ADP-ribosylation factor-GTPase activating protein (ARF-GAP) that appears to regulate the movement of molecules leaving the trans-Golgi network and/or cycling between the cell surface and endosomes.^8-11 Mutations in NEV result in membrane trafficking defects and inhibition of abscission.⁸ Whereas nev mutants display distinct uncontrolled cell expansion late in abscission, the AZ cell morphology early in abscission is similar to ida mutants and appears to be normal.⁴⁴ Screening of suppressors for the nev phenotype has identified 3 receptor-like kinases (EVERSHED RECEPTOR-LIKE KINASE (EVR), SOMATIC EMBRYO RECEPTOR-LIKE KIASE 1 (SERK1), and CAST AWAY RECEPTOR-LIKE KINASE (CST)).^9–11 In addition to the abscission defects, the organization of the Golgi network of nev is restored in the evr, serk1, and cst backgrounds. It was concluded that these suppressors act downstream of NEV to spatially inhibit organ separation.^9-11 Notably, mutations in these 3 receptor-like kinases did not restore the phenotype in either the ida or hae hsl2 mutants.^9-11 Additionally, whereas the defects of abscission and the Golgi network in the nev mutants are rescued by these suppressors, the ectopic cell expansion of nev in late abscission remained unaffected in the double mutants (nev-3 evr, nev-3 serk1, nev-3 cst) and the triple mutants (nev-3 serk1 ida-2 and nev-3 ever-2 ida-2).^9–11 These results indicate that NEV signaling pathway is crucial for cell separation (Phase 3) and cell expansion (Phase 4), and may function in parallel with the IDA signaling pathway. Recent genetic study further revealed that organ separation mediated by HAE/HSL2 signaling may act downstream of NEV and EVR.⁴⁹ Collectively, both parallel and convergent signaling pathways mediated by NEV- and IDA-HAE/HSL2 modules contribute to floral organ abscission.^9-11,44,49 Based on the petal break-strength profiles and the collective observations of uncontrolled cell enlargement and expansion in nev mutants, the role of NEV-mediated intracellular trafficking is required for cell separation (Phase 3), and for organ separation and cell expansion (Phase 4) (Fig. 1B(iii)).^8-11,44,49

Reproductive Development Affecting Floral Organ Abscission

In many plants the onset of floral organ abscission is tightly linked to reproductive programs such as pollination, inflorescence meristem (IM) activity, and flower senescence. Studies on key signaling components identified in the 4 phases of abscission, whether they are essential or not, reveal that they also contribute to these programs.^{4,12-15,19,25-27,29,50-53} The relationships between these reproductive development programs and abscission are discussed next.

Post-pollination changes and floral organ abscission

In wild type Arabidopsis, floral organs are shed in a turgid state shortly after anthesis (after 2 to 3d).^16,17 Multiple researchers have reported that there are systematic signals relating to pollination, the ability of a plant to bear fruit, and floral organ abscission. Nutrient mobilization between flowers and the plant's main body occurs before and during abscission.⁵⁴ Ethylene has been suggested to be one of these systematic signals since a post-pollination ethylene burst can hasten the abscission process.^54,55 Studies of abscission mutants indicate that in addition to nutrition mobilization and ethylene production, there appear to be other developmental factors that underlie and coordinate these processes. For instance, in spite of the essential roles of BOP1/2, IDA and NEV in abscission, bop1/2, ida and nev mutants are normally fertile, suggesting that pollination signals including a local ethylene burst are not sufficient to overcome the cellular and downstream signaling defects in the AZ of these mutants.^7,8,56 It is clear that an ethylene signal is not essential to Arabidopsis abscission because, although abscission of the petals, sepals and anthers are delayed in the etr-1 and ein2–1, they do eventually abscise, and the flowers are fertile.^17,19,20 What is still unclear for Arabidopsis is whether ethylene is simply one of the signals evoked that indicate pollination has occurred, which then can trigger the onset of abscission, or if ethylene also plays a more direct role in regulating gene expression during the separation process as indicated for abscission in other species, e.g., tomato, bean, etc.^57,58

Alternatively, there are other plant hormones that have been routinely studied for their role in reproduction. One of the most studied roles of JA signaling is on plant fertility. Since fertilization elicits signals that can promote the floral organ abscission process, JA could be expected to play a role in the onset and/or timing of abscission. The phenotype of plants defective in JA biosynthesis and perception (aos/dde2, dde1/opr3, dad1, fad triple mutants as well as dab4–1) and the pattern of gene expression for JA biosynthesis genes (LOX2 and DAD1) in the AZ of wild type plants suggests JA's role in floral organ abscission.^{19,20,28,30–32,51,52,59} In addition, recent studies on the targets of JA signaling imply JA's encompassing involvement in abscission. Arabidopsis basic helix-loop-helix MYC3 and MYC4 transcription factors are novel activators of JA-regulated programs particularly in aerial tissues of the plant.⁵³ MYC3 and MYC4 are necessary for full responsiveness to JA, and their biochemical and functional roles in JA signaling are additive with MYC2 in plant development and defense responses. Furthermore, the expression patterns of MYC3 and MYC4 in the wild type AZ, and protein-protein interactions of these 2 new targets with JAZ (JASMONATE-ZIM DOMAIN) repressors, indicate that other identified JA signaling components may tightly be associated with the regulation of abscission.⁵³ To what extent JA is involved in the regulation fertility and floral organ abscission still remains to be determined.

Proliferative activity in inflorescence meristem and floral organ abscission

Abscission is a developmentally programmed process and can be linked to an age-related progression. Floral organ abscission is associated with a latter stage of reproductive development that can be influenced by the proliferative activity of the inflorescence meristem (IM).^16,19 Abscission progresses from the oldest to the youngest flowers in a spatial-temporal manner. Thus, the rate of floral organ abscission is determined based on the position of the flower along the inflorescence at which point the organs abscise or don't abscise.¹⁷ Interpreting the timing and rate of abscission for the abscission mutants can be complex. For instance, JA (aos, dad1, lox3 lox4, and dab4–1) and ethylene (etr1–1 and ein2–1) abscission mutants exhibit altered development of the IM.^{17,19,20,29,32} Particularly, mutations in JA and ethylene signaling components can influence the proliferation (number) of flowers and the development (or age) of flowers initiated in the IM. This proliferative activity in the IM may indirectly influence the position-specific timing of abscission, which is reflected in the petal break-strength in JA and ethylene abscission mutants. Force to remove petals from the early positions of flowers in JA and ethylene mutants (e.g.,, dab4–1, etr1–1, and ein2–1) is significantly greater than that required for wild type plants, indicating that although flowers on the inflorescence between mutants (e.g., dab4–1, etr1–1, and ein2–1) and wild type plants are same for positions, their developmental properties are not identical.^{4,7,17,19,32,44} Moreover, the altered development of inflorescence and flowers in additional JA biosynthesis mutants (lox3 lox4), and the expression of genes critical for proper maintenance of the IM in a JA receptor abscission mutant (dab4–1) and a biosynthesis abscission mutant (aos) suggest that the activity of the IM that can be affected by JA signaling may ultimately influence the rate of abscission.^19,29

Flower senescence and floral organ abscission

Flower senescence is often coupled with floral organ abscission. It has been shown that the role of flower senescence in plants is associated with nutrient recycling and reproductive success.⁵⁴ Studies of the signaling components controlling the timing of abscission have provided molecular insights into the developmental relationship between senescence and abscission. Ectopic expression of MADS-box transcription factors (AGL15, AGL18, and FYF) delays flower senescence as well as flower organ abscission in Arabidopsis.^12,13,15 Interestingly, a transient increase of AGL15 expression at the first flower opening caused a delay in both processes.¹³ The results suggest that there are inherent properties of development (or age-related properties) that can be controlled by AGL15 to regulate both abscission and senescence.¹³

Recently a regulatory role of JA in flower senescence has also been demonstrated. Rubio-Somoza and Weigel showed that the regulation of flower senescence includes a network of plant hormones with miRNA-transcription factors.⁵² This study demonstrated that in the later stages of flower development, miR519, mi319, and miR167 coupled with MYB33 and TEOSINTE BRANCHED/CYCLOIDEA/PCF4 (TCP4) coordinately controlled flower senescence by modulating hormone actions including JA biosynthesis. It is worth noting that the inflorescence architecture of mutants affected by this regulatory module also displayed a delay in floral organ abscission, implying that these 2 processes can be targeted by the same machinery.

Currently, the modulators for the timing of abscission, MADS-box transcription factors (AGL15, AGL18, and FYF) and the regulatory network of miRNA-transcription factors with JA signaling pathway, are also considered to control the timing of flower senescence. Conversely, essential signaling components such as BOP1/2, IDA and NEV regulate floral organ abscission independent of senescence, and the loss of function in these genes appears to separate flower senescence from floral organ abscission (Fig. 1B(i) and B(iii): dashed lines for senescent petals).^4,7,8

Timing of Detachment Balanced by Hormones; Emerging Roles of JA and Ethylene in Floral Organ Abscission

Depending on the developmental and physiological contexts, it has been shown that JA and ethylene can act antagonistically or synergistically.^60–63 JA and ethylene are critical for diverse aspects in reproductive development including floral organ abscission. During floral organ abscission, JA and ethylene regulate the timing of process in both independent and inter-dependent manners.^19,20

Independent regulation of floral organ abscission by JA and ethylene

One of the first hormones implicated for the regulation of floral organ abscission in Arabidopsis was ethylene.^64,65 Ethylene -insensitive mutants (etr1–1 and ein2–1) exhibit a delay in abscission but with no defects in the cellular morphology in the AZ suggesting that the perception of ethylene is critical for the timing of abscission.^17,19 Study on a JA receptor (DAB4) and a JA biosynthesis gene (AOS) has revealed that JA biosynthesis and signaling also modulate the timing of process.¹⁹ The extent of delay caused by defects in perception and biosynthesis of JA was significantly greater than delays caused by etr1–1 or ein2–1. Mutational studies using ethylene or JA insensitive plants as well as plants with lower JA biosynthesis revealed that these hormones act partly in parallel to regulate the timing of abscission. Particularly interesting is the observation that there is an ethylene-independent pathway that is mediated by JA (see below).^19,20 Figure 2 (left) shows a model for how the JA and ethylene signaling pathways are thought to act independently and in parallel to activate abscission.^17,19

Figure 2.

A proposed model for regulation of floral organ abscission by JA and ethylene. JA and ethylene can regulate floral organ abscission in independent and inter-dependent manners. JA and ethylene act independently to control the timing of abscission (left). Inter-dependent regulation of JA and ethylene is mediated in the downstream signaling of the ethylene receptor, ETR1 (right). The model predicts that when JA levels are low and EIN2 activity is eliminated in Arabidopsis, there is an EIN2-independent component (Xs) or signaling pathway (dotted line) that is inhibited by JA or DAB4/COI1^SCF.

Inter-dependent regulation of floral organ abscission by JA and ethylene

EIN2 is a central component in the ethylene signaling pathway, and is thus important in the regulation of abscission by ethylene.^19,66 Previous study demonstrated that the ein2–1 mutant could respond to JA with an accelerated rate of abscission, indicating that there is an ethylene-independent pathway.¹⁹ A recent study has unraveled a novel interplay between JA and ethylene in the control of the timing of abscission. In particular, this study found that lowering JA levels (either mutationally, or pharmacologically) caused ein2 ethylene-insensitive mutants to respond to ethylene with accelerated abscission.²⁰ Additionally, when a JA receptor mutant, dab4–1, was genetically introgressed into ein2–1, the dab4–1 ein2–1 also gained responsiveness to ethylene in floral organ abscission (unpublished data, Kim and Patterson).²⁰ Currently, the pathway by which JA levels affect ethylene signaling is unclear.^19,20 The results summarized above suggest that there is an interplay between the signaling pathways to control abscission. A working model posits that there may be an EIN2-independent signaling pathway downstream of CTR1 (Fig. 2, right). The detailed genetic and molecular networks for an EIN2-independent signaling pathway in the regulation of abscission remain to be filled in.

Concluding Remarks and Perspectives

Genetic, physiological and cellular studies from diverse floral organ abscission mutants in Arabidopsis have identified key molecular components in the progression of abscission. The proposed working model has begun to shed light on the regulation of floral organ abscission.

Molecular components critical for early establishment of the AZ (BOP1/2) and late abscission cellular mechanisms in the AZ (IDA and NEV) are essential for abscission, while other components modulate the timing of process. Abscission signaling pathways involving BOP1/2, IDA and NEV can separate the process of floral organ abscission from flower senescence. Of interest is that the functions of a small signaling peptide IDA and a membrane trafficking NEV share similar effects in the progression of abscission. Genetic interactions between IDA-mediated and NEV-mediated signaling pathways have begun to unravel their roles in the molecular network. Although recent comparative studies propose overlapping and non-overlapping regulations in cell separation (Phase 3) and organ separation (Phase 4), further epistatic relationships still need to be substantiated. Additionally, comprehensive biochemical assays that can support the genetic interactions and proposed molecular networks still await. In particular, physical interaction studies including ligand-receptors binding and co-receptor interactions need to be confirmed. On the other hand, simple but effective peptide-based assays in various abscission mutants flowers may further validate the currently established genetic relationships.⁴³

Central components from JA and ethylene signaling pathways are emerging for their roles in the regulation of floral organ abscission. Although one of the first molecular components to be identified was the ethylene receptor, ETR1, much of the downstream signaling cascades remain uncovered. The recent identification of JA's roles in floral organ abscission has started to unravel the complex interactions of JA and ethylene; however, the details of this interplay remain to be elucidated.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

I gratefully acknowledge Mark L. Tucker, Brad M. Binder, Caren Chang, and Sara E. Patterson for critical reading of this manuscript and their assistance.