The Plant Cell, Vol. 9, 1169-1 179, July 1997 O 1997 American Society of Plant Physiologists

Last Exit: Senescence, Abscission, and Meristem Arrest in Arabidopsis

Anthony B. Bleeckerl and Sara E. Patterson Department of Botany, University of Wisconsin, Madison, Wisconsin 53706

INTRODUCTION

When creating a painting, an artist applies successive layers of pigment onto a blank canvas. A sculptor, on the other hand, may chip away at a block of stone until the form of a statue emerges. Although the painter's method can be ap- plied readily to developmental biology, the influence of the sculptor's hand should not be overlooked. The importance of programmed cell death in animal development is becom- ing increasingly appreciated (White, 1996), but it is in higher

potential in plants. To this end, we consider how Arabidopsis may be useful as a model for identifying and characterizing specific genes that influence these developmental traits.

THEoRETiCAL CoNSiDERATloNS

plants that the attrition of cells, tissues, and even whole or- gan systems plays a continuing role in shaping the organ- ism's architecture and its progression through its life cycle.

Vegetative development in plants is indeterminate and modular. The continuous generation of organ systems by the plant's meristems is countered by the programmed se- nescence and/or abscission (shedding) of existing organs throughout the life of the plant. Developmental potential is also influenced by the suppression of meristem activity, as in apical dominance, and by the loss of meristems that are used in the determinate development of flowers. Although all of these processes are genetically determined, the path- ways involved are also under the influence of environmental input. The resulting complex web of interactions contributes to the plasticity of plant development, which is evident in the wide variety of forms that individual, genetically identical plants can display.

The developmental mechanisms that determine the ulti- mate fates of somatic tissues and meristems in plants have been studied for decades, but it is not our intention to review this extensive volume of literature in this article (for recent reviews, see Noodén and Guiamet, 1996; Buchanan-Wollaston, 1997; Weaver et al., 1997). Instead, we turn our attention to the future and consider new avenues of research that may pro- vide insights into the genetic mechanisms that drive the pro- cesses of senescence, abscission, and loss of meristematic

When contemplating the longevity and ultimate fate of plant tissues, the second law of thermodynamics must be kept in mind. The complex systems that constitute a plant cell re- quire energy input for their creation and maintenance; in the- ory, an individual cell could be maintained indefinitely if it were provided with sufficient energy input. Yet, somatic tis- sues in plants have limited life spans.

The explanation for this is best illustrated by considering an individual leaf. The development of the leaf initially re- quires an input of nutrients from the plant. The payoff for this investment occurs when the leaf converts from a nutrient sink to a nutrient source as photosynthetic competence is achieved. After the leaf matures, its functional life span is limited by a variety of externa1 factors, including seasonal climate change, predation, disease, and invariably, increased shading by the continued growth of the plant. In principle, natural selection is unlikely to favor excessive energy invest- ment in the maintenance of such an organ past the average time of its usefulness to the plant, because such an invest- ment may compromise reproductive success. This expecta- tion is borne out by many studies indicating that the photosynthetic activity of individual leaves is highest at de- velopmental maturity and subsequently declines over time (see Hensel et al., 1993).

Given these constraints, it is 'not surprising that plants have evolved mechanisms for dealing with obsolete organ systems. The shedding of leaves by che process of abscis- sion is a very ancient trait-fossils from as early as the De- vonian period (i.e., -400 million years ago) show evidence of abscission SCarS (Addicott, 1982). However, photosyn- thetic organs are rich in nutrients, which would be lost

To whom correspondence should be addressed. E-mail bleeckerQ facstaff.wisc.edu; fax 608-262-7509.

11 70 The Plant Cell

without mechanisms that salvage them before the obsolete organs are cast off. We refer to this salvage process as the senescence syndrome. Although their underlying regulatory pathways are not understood in detail, both abscission and the senescence syndrome are genetically programmed pro- cesses that can be observed in all higher plants.

ARABIDOPSIS LlFE HISTORY

Although meristems provide plants with virtually unlimited growth potential, many plant species exhibit well-defined life spans. Most notable are the plants that senesce and die af- ter a single reproductive effort. So-called monocarpic or whole-plant senescence involves the senescence of existing somatic tissues and the suppression of meristematic devel- opment. Although the global aspects of the underlying regu- latory pathways have long intrigued plant physiologists, the developmental mechanisms involved are not well under- stood for either of these processes. Woolhouse (1983) has pointed out that the diversity of forms associated with monocarpic senescence indicates that this form of senes- cence has a polyphyletic origin, which may account for some of the controversy and confusion concerning the na- ture of the regulatory mechanisms involved. With this cau- tionary note in mind, we focus our attention on Arabidopsis, which provides an excellent example of the monocarpic habit. In particular, we consider the fates of both somatic and generative tissues of the plant throughout the course of development and comment on how the fates of individual tissues contribute to the demise of the parent plant.

After germination, the above-ground portion of the Arabi- dopsis plant initially develops as a compressed rosette of leaves formed by the primary shoot meristem (Clark, 1997; Kerstetter and Hake, 1997; Poethig, 1997, in this issue). Ini- tially, there is no anatomical evidence for axillary meristems in the axils of rosette leaves. Depending on genetic back- ground and environmental conditions, the primary apical meristem may undergo the transition to reproductive devel- opment after the initiation of six to 10 leaves (early flowering varieties). However, under short-day conditions and in ge- netically late-flowering lines, the transition to flowering is de- layed, and the apical meristem continues to produce leaves. When the transition to flowering occurs, the main axis elon- gates and produces a complex of higher order branches, which convert to reproductive development. After fertiliza- tion and as the seeds mature, the somatic tissues of the plant become increasingly senescent. However, if reproduc- tion is delayed, anatomically definable vegetative meristems can initiate in axils of older leaves (Grbic and Bleecker, 1996). This observation indicates that caution must be taken when studying somatic tissue senescence that the origin and age of the somatic tissue under study have been accurately as- sessed. The relevant phases of Arabidopsis development are depicted in Figure 1A.

The transition to reproductive development is associated with an enlargement of the apical meristem and the initiation of flower meristems instead of leaf primordia on the flanks of the apical meristem. Axillary meristems subsequently develop in the axils of leaf primordia in a basipetal wave from newly formed to older preexisting leaves (Hempel and Feldman, 1994). The youngest of these preexisting leaf primordia be- come the cauline leaves of the primary stem. The axillary meristems form the secondary shoots (coflorescences) of the flowering plant. These secondary shoots are initially veg- etative, and then they undergo the transition to flowering- repeating the basic pattern of primary shoot development (Grbic and Bleecker, 1996). Higher order coflorescence branches are generated in the same way, although they are smaller and less fecund than lower order branches-fertile fifth-order branches rarely develop in the Landsberg erecta (Ler) ecotype (Hensel et al., 1994). This observation is impor- tant in terms of life history because it indicates that a finite number of fertile branches are normally produced, despite the virtually unlimited potential of the system to generate higher order branches.

In principle, inflorescence development is also indetermi- nate in Arabidopsis. The inflorescence meristem produces between two and three new flowers per day in a spiral phyl- lotaxy. However, the primary inflorescence meristem pro- duces a predictable number of flowers before arresting (Shannon and Meeks-Wagner, 1991; Alvarez et al., 1992; Hensel et al., 1994). Moreover, in the Ler ecotype, the higher order coflorescence branches also stop producing new flowers within 48 hr of the arrest of the primary meristem, re- gardless of the number of flowers produced on the affected branch (Hensel et al., 1994). By this point, all effective shoot meristems have become inflorescence meristems. The glo- bal arrest of these meristems restricts the ability of the plant to produce new somatic tissues, so it must rely on existing photosynthetic tissues for survival. Proliferating and arrested inflorescence branches are shown in Figure 1 B.

Somatic tissues of the Arabidopsis shoot have a limited longevity. As the inflorescence develops, the sequential se- nescence of rosette leaves, from oldest to youngest, can be observed. This is characterized by a progressive yellowing of the leaves beginning at the leaf margins and spreading to the interior (Figure 1 D). The yellowing of the leaf is followed by necrosis and death of the tissues (Hensel et al., 1993; Lohman et al., 1994). A similar fate is observed for the pho- tosynthetic tissues of the inflorescence stem, cauline leaves, and siliques. Measurements of the longevity of somatic tis- sues indicated that a maximum life span of <25 days from the time of early development of the tissue is the rule for Ler plants growing in long-day conditions (Hensel et al., 1993).

The death of the senescing Arabidopsis plant can be viewed as a consequence of severa1 developmental deci- sions that favor reproduction at the expense of the soma. The proliferation of inflorescence meristems results in the timely production of many seeds. The arrest of these mer- istems after a predictable number of fruits have been set en-

Senescence and Abscission in Arabidopsis 1171

Figure 1. Features of Monocarpic Senescence, Inflorescence Meristem Arrest, and Leaf Senescence in Arabidopsis (Ler).

(A) Stages in the life cycle of the adult plant. Plants are pictured at 14, 21, 37, and 53 days after germination (left to right).(B) Primary inflorescence meristems. Proliferating (top) and arresting (bottom) meristems are shown.(C) Reinitiation of inflorescence meristems. A primary inflorescence branch that reinitiated proliferative activity after having arrested is shown.(D) Age-related changes in rosette leaf number 5. Leaves are pictured at 0, 3, 5, 7, 9, and 11 days after full leaf expansion (left to right).

sures that sufficient nutrients are available to complete thedevelopment of established embryos. In theory, the pro-gressive senescence of existing somatic tissues reflects theminimum energy spent on maintenance of these tissues-energy that is thus available for seed development. The pro-grammed nature of senescence, which involves the mobili-zation of nutrients released by the turnover of cellularconstituents, increases the supply of materials that are es-sential for proper seed fill at the time of maximum need. Allof these features may favor fecundity, but they leave theparent plant exhausted of resources and without the gener-ative capacity to renew itself somatically (for more completediscussions, see Hensel et al., 1993, 1994).

SOMATIC TISSUE SENESCENCE IN ARABIDOPSIS

The Senescence Syndrome

The senescence syndrome in photosynthetic plant tissuesinvolves complex cellular changes that exhibit a number ofdistinguishing features. Ultrastructural studies in severalspecies provide a consistent picture of the progressivechanges that occur during leaf senescence (Thomson andPlat-Aloia, 1987; Nooden and Leopold, 1988). As suggestedabove, these changes are most easily understood fromthe standpoint of nutrient salvage. For example, in a young

11 72 The Plant Cell

mesophyll cell, it has been estimated that >50% of the pro- tein and 70% of the lipid is associated with the chloroplasts (Forde and Steer, 1976; Dean and Leech, 1982). Consequently, the initial stages of the senescence syndrome involve a breakdown of membrane structure within the chloroplast. This is associated with the degreening of the affected tis- sues and the progressive loss of proteins associated with the chloroplast, such as ribulose bisphosphate carboxylase and chlorophyll alb binding proteins (Bate et al., 1990). The cytoplasm of cells undergoing the senescence syndrome is also affected, exhibiting a decrease in cytoplasmic volume and in the number of detectable cytoplasmic ribosomes. In Arabidopsis, these ultrastructural changes are reflected in biochemically detectable decreases in rRNA and protein in senescing tissues (Hensel et al., 1993; Lohman et al., 1994). These autolytic changes in the cell do not extend injtially to the mitochondria or nucleus, both of which remain unaf- fected until very late in the process of senescence (Noodén and Guiamet, 1996).

The loss of macromolecular constituents during plant tis- sue senescence is not associated with a concomitant rise in metabolite levels but rather with a decline in total carbon and nitrogen within the tissue. Indeed, the implied transport of metabolites out of the leaf has been substantiated in a number of cases (Noodén and Leopold, 1988). In this sense, the senescing leaf continues to function as a source of nutri- ents to the whole plant, but only at the expense of its own ability to survive.

The hydrolysis of proteins and nucleic acids that accom- panies senescence must be mediated by proteases and nu- cleases, respectively. Transaminases may function to shunt nitrogen into the most common mobile forms, glutamine and asparagine, and mobilization of lipids released from thyla- coid membranes presumably requires conversion to sugars through the glyoxylate cycle. Clearly, the salvage aspect of senescence requires the operation of a complex array of metabolic pathways, many of which may also operate to some degree in younger tissues. However, current knowl- edge of salvage pathway function during senescence is in- complete. For example, turnover of the most abundant protein in the cell, ribulose bisphosphate carboxylase, is a major aspect of salvage, but the mechanism involved has not been elucidated (Friedrich and Huffaker, 1980; Weaver et al., 1997). On the other hand, evidence for the increased activity of the glyoxylate cycle (DeBellis et al., 1990; Graham et al., 1992), amino acid metabolism (Kamachi et al., 1992), and RNase activity (Taylor et al., 1993) during leaf senes- cence has been reported.

Differential Gene Expression Associated with the Senescence Syndrome

Despite gaps in our understanding of the process, it is clear that the senescence syndrome is a genetically controlled process that requires continuing gene expression. A grow-

ing number of genes have been dubbed “senescence asso- ciated” on the basis of elevated levels of their mRNAs in senescing tissues (for recent reviews, see Noodén and Guiamet, 1996; Buchanan-Wollaston, 1997; Weaver et al., 1997). Differential hybridization screenings of cDNA libraries made from senescing leaf tissues of Arabidopsis have also yielded a number of clones. Such clones have been classi- fied as representing senescence-associated genes (SAGs) if they show a greater hybridization signal with RNA from se- nescing leaf tissues than with that from younger leaves (Hensel et al., 1993; Lohman et al., 1994).

The deduced amino acid sequences of at least some SAGs indicate possible roles in salvage. Indeed, cDNAs with sequence similarities to proteases, nucleases, and proteins involved in nitrogen and lipid metabolism have been identi- fied (see Noodén and Guiamet, 1996; Buchanan-Wollaston, 1997; Weaver et al., 1997). These studies have also identi- fied genes that decrease in expression during senescence. This is a common feature of genes associated with photo- synthesis (Hensel et al., 1993).

Continuing studies are likely to indicate that the conver- sion of a leaf from a photosynthetic system to a source of stored nutrients involves global changes in gene expression. However, some care must be taken in considering the regu- latory systems that mediate this conversion. Although tran- scriptional control may be important in some cases, message stability may also play an important role. For ex- ample, many of the above-mentioned SAGs are expressed at lower levels in younger leaf tissues, but their transcript levels may only increase two- to fivefold during senescence if measured on a leaf area basis (Hensel et al., 1993; Lohman et al., 1994). These modest changes may reflect changes in stability rather than changes in transcriptional activity. Ex- pression of these SAGs in young tissue may reflect the fact that salvage processes are always occurring in plant cells. Indeed, a defining feature of some SAGs may be that by a variety of mechanisms, these genes continue to be ex- pressed while much of the cell’s machinery is being disman- tled. This is not to say that transcriptional control should not be a focus of attention in particular cases but rather that this paradigm may reveal only a single facet of the regulatory mechanisms involved in the senescence syndrome and may not necessarily be the rate-limiting control mechanism.

Factors Controlling the Timing and Progression of the Senescence Syndrome

The overall process of the senescence syndrome may be very complex, but it is still worth considering whether there are definable regulatory systems that have an overriding in- fluente on the timing of the conversion from photosynthetic to salvage programs. There is no doubt that in certain con- texts, the timing of this conversion can be reversibly con- trolled by hormone application. In some cases, application of a cytokinin is sufficient to delay almost indefinitely the full

Senescence and Abscission in Arabidopsis 11 73

expression of the senescence syndrome (Woolhouse, 1983; Noodén and Leopold, 1988). A striking example of this ef- fect of cytokinin was provided by the phenotype of tobacco plants transformed with a chimeric gene consisting of a SAG promoter fused to the IPT gene (which encodes isopentyl transferase, a cytokinin biosynthetic enzyme) from Agrobac- terium tumefaciens (Gan and Amasino, 1995; see also Kende and Zeevaart, 1997, in this issue). On the other hand, cytokinin application to intact leaves had little effect on the timing of leaf senescence in Arabidopsis (A.B. Bleecker, un- published results). By contrast to the delaying effects of cy- tokinins, ethylene and abscisic acid often promote the premature onset of the senescence syndrome (Noodén and Leopold, 1988; Grbic and Bleecker, 1995). These observa- tions are consistent with the existence of master regulatory systems that can determine the timing of the senescence syndrome. If this is true, then a mutational approach using Arabidopsis could identify the genes involved.

We have already cited evidence that Arabidopsis leaves provide a reasonable model for the senescence syndrome in higher plants. Thus, in principle, mutations that influence the timing or progression of the senescence syndrome could be identified by screening mutagenized populations for plants that fail to undergo senescence. We are aware of severa1 groups (including our own) that have tried this approach. To our knowledge, no mutant has been found in which the se- nescence syndrome does not occur. This indicates either a lack of appropriate dedication or the possibility that no sin- gle gene is sufficiently crucial to the process to produce an obvious phenotype when mutated.

The growing number of existing mutants in Arabidopsis provide an alternative approach for examining effects of sin- gle genes on the senescence process. Ironically, this ap- proach has told us more about what does not control senescence than what does. For example, in some mono- carpic plants, such as pea and soybean, vegetative senes- cence is coupled to the development of fruit (see Noodén and Leopold, 1988). However, single-gene mutations in Ara- bidopsis that cause male sterility (e.g., male sterile7 [ms7-7]), delayed flowering (e.g., constans [co]), or early termination of the inflorescence (e.g., terminal flower7 [ t f l l ] ) have no ef- fect on the timing of senescence in rosette leaves 5 and 6 (Hensel et al., 1993). Instead, the age of the individual leaf was the best predictor of the timing of leaf senescence. The timing of rosette leaf senescence has also been examined in hormone-deficient and hormone-insensitive mutants. Again, mutations affecting auxin, gibberellin, and abscissic acid synthesis or response had no measurable effects on the tim- ing or progression of leaf senescence (Hensel et al., 1993).

Of all the hormone mutants tested, only the ethylene- insensitive etr7-7 (for gfhy/ene [esponse) mutant line showed a measurable delay in the timing of leaf senescence (Grbic and Bleecker, 1995). This was associated with a delay in the timing of expression of SAG markers but not in the leve1 of expres- sion of these markers once senescence occurred. Results of this study demonstrated that ethylene enhances the expression

of SAGs while suppressing that of photosynthesis-associ- ated genes. Based on these data, it was suggested that ethylene plays a role in coordinating the timely transition of the leaf to the senescent state but it is not essential to the process itself.

Future Avenues of Research on the Senescence Syndrome

We end this section in an optimistic vein by considering the prospects for elucidating the regulatory mechanisms that govern the senescence syndrome, using Arabidopsis as a model system. Because straightforward screens for senes- cence-altered mutants in Arabidopsis have not yet borne fruit, more creative screens involving individual facets of the senescence syndrome are needed. For example, by using reporter genes fused to SAG regulatory elements as mark- ers, regulatory pathways that drive expression of subsets of SAGs could be targeted by mutagenesis. Combining differ- ent mutations obtained in this way in a single genetic back- ground may reveal developmental phenotypes of interest. Moreover, the growing number of SAGs that have been cloned from Arabidopsis by differential hybridization (Weaver et al., 1997) provide starting points for functional analyses such as antisense suppression and for genetic experiments, including screening for knockouts in T-DNA populations (Krysan et al., 1996).

SAGs of particular interest include SAG 72, which is spe- cifically expressed in senescing tissues at high levels (Lohman et al., 1994). The protein encoded by SAG 72 ap- pears to lack a canonical signal sequence, indicating a pos- sible cytoplasmic location (Weaver et al., 1997). It is also a member of the cathepsin superfamily of proteases, which includes CED-3 (for E// death) and ICE (for interleukin-7/3- - converfing gnzyme). These proteases trigger programmed cell death in animals (White, 1996; see also Pennell and Lamb, 1997, in this issue). SAG 73 codes for a protein re- lated to short-chain alcohol dehydrogenases, a family that includes the protein encoded by tasselseed2, which is re- quired for the programmed degeneration of floral primordia that is associated with reproductive development in maize (Delong et al., 1993). SAG 75 (fRD7) shows homology to the ClpA regulatory subunit of a bacterial protease system. This system has been identified in plant chloroplasts (Shanklin et al., 1995) and could be involved in the turnover of chloro- plast proteins. Targeted mutagenesis of these and additional SAGs may provide important insights into the regulation of the senescence syndrome in plants.

THE FATE OF INFLORESCENCE MERISTEMS IN ARABIDOPSIS

The cessation of proliferative activity of shoot apical meri- stems that is often associated with monocarpic senescence

11 74 The Plant Cell

has been referred to as apical senescence. This is some- thing of a misnomer with respect to Arabidopsis. As mentioned above, Arabidopsis inflorescence meristems cease prolifer- ative activity after a predictable number of fruit has been produced. However, these meristems appear to be quiescent rather than senescent in that they may be reactivated after the subtending fruits have developed (Figure 1 C). Surgical removal of developing fruits also results in the reactivation of an arrested meristem (Hensel et al., 1994). For those scientists who are interested in cell-cycle control, this developmental system is worth looking into as a natural system in which proliferative activity is developmentally controlled and can be reliably manipulated in a defined way (see Jacobs, 1997, in this issue, for a review of cell-cycle control in plants).

The proliferative arrest of inflorescence meristems in Ara- bidopsis poses some interesting biological questions. The observation that the arrest is global-it occurs at all mer- istems of the inflorescence within a 24-hr period-indicates that it is a whole-plant phenomenon. Moreover, inflorescence meristem arrest is directly correlated with seed production; male-sterile and even reduced-fertility plants do not undergo arrest (Hensel et al., 1994). In these cases, inflorescence meristems continue to proliferate, although not indefinitely. Detailed studies of the male-sterile mutant msl-7 indicate that meristems eventually cease producing flowers as a result of defects in patterning at the inflorescence apex. This ultimately results in the loss of the apical meristem (Hensel et al., 1994).

The nature of the global signal that triggers proliferative arrest in Arabidopsis is unknown. Depletion of a key nutrient is a possibility, but the production of an inhibitoty signal by devel- oping fruits cannot be ruled out. Mutations conferring hormone deficiency or insensitivity did not appreciably affect the global arrest process (Hensel et al., 1994), although cytokinins were not tested in this way. However, a screen of a mutagenized pop- ulation of Arabidopsis resulted in the identification of a reces- sive, single-gene mutation that disrupts proliferative arrest. The proliferous7 ( p l f l ) mutant was initially indistinguishable from the isogenic Ler but continued to proliferate after Ler arrested, ultimately producing twice as many seeds as did Ler, with no measurable decrease in seed weight or viability (A.B. Bleecker and L.H. Hensel-Burke, unpublished results). These results imply that genetic determinants may mediate the conflicting resource allocation requirements of existing fruits and the meristems responsible for producing more fruits. This developmental system in Arabidopsis provides an interesting opportunity to apply a genetic approach to the poorly under- stood phenomenon of correlative control systems in plants.

ABSClSSlON

Developmental Basis

During the course of development, plants shed entire organ systems through the process referred to as abscission. Ab-

scission provides a mechanism for the removal of senescing or otherwise damaged organs and for the release of fruits as they ripen. In some cases, intact, healthy organ systems may be shed as a component of development. The abscis- sion of petals, sepals, and stamen as fruits begin to develop provides an example of this process.

Abscission of organ systems occurs at anatomically dis- tinct bands of cells referred to as abscission zones (AZs). AZs are often located in stems between the abscising organ and the body of the plant. They are generally created during the development of the associated organ system and are of- ten characterized as a band of small, densely cytoplasmic cells from a few to many cell layers thick (Addicott, 1982; Sexton and Roberts, 1982; Osborne, 1989). The AZs are usually defined early in the development of organ systems as a band of cells that fail to enlarge and vacuolate along with surrounding tissues. Under the appropriate develop- mental or environmental conditions, cells in the AZ begin to enlarge, and then the middle lamellae between the cells dis- solves, which results in the development of a fracture plane across the stem (Addicott, 1982; Sexton and Roberts, 1982). Shedding of the organ is followed by the continued enlarge- ment of the cells on the proximal face of the fracture plane and, ultimately, the differentiation of these cells into suber- ized scar tissue (Addicott, 1982). The basic developmental pro- cesses associated with abscission are depicted in Figure 2A.

Although the abscission process generalized above is consistent with a wide range of observations on a variety of abscising systems, our understanding of the specific mech- anisms that determine the timing and progression of abscis- sion is based largely on detailed studies of a few model systems. Studies with Phaseolus vulgaris (bean) explants containing leaf AZs led to a model in which the plant hor- mones ethylene and auxin control the timing of abscission (Sexton and Roberts, 1982). According to this model, ethyl- ene is the primaty signal that drives the abscission process, whereas auxin reduces the sensitivity of AZ cells to ethylene and thus prevents or delays abscission. Subsequent studies in a variety of species support the idea that applied ethylene can stimulate abscission and that the inhibition of ethylene synthesis or action can delay abscission in many cases (Sexton and Roberts, 1982; Abeles, 1992).

The question of how the abscission process is driven by ethylene or other developmental signals has been primarily addressed by the hypothesis that abscission is achieved through the induction of genes that code for cell wall hydro- lytic enzymes (Morre, 1968). For example, specific cellulase (i.e., endo-p-l,4-glucanase) genes are induced in AZs in as- sociation with the abscission process (Varner and Taylor, 1989; de1 Campillo and Lewis, 1992; Lashbrook et al., 1994). In bean, a specific cellulase isoform is ,induced at the sepa- ration layer but also in a less localized pattern in adjacent vascular tissue (de1 Campillo et al., 1990; Tucker et al., 1991). Specific cellulase genes are also differentially ex- pressed in tomato flower AZs (Lashbrook et al., 1994; de1 Campillo and Bennett, 1996).

Senescence and Abscission in Arabidopsis 1175

AbscisedOrgan

B

Figure 2. Progressive Changes Associated with Floral Organ Abscission in Arabidopsis.

(A) Schematic representation of developmental processes associated with abscission. Heavily shaded circles represent small, densely cytoplas-mic cells; open circles represent large vacuolated cells of mature tissues; lightly shaded ovals represent expanding cells on the proximal face ofthe fracture plane that transdifferentiate into the peridermal scar tissue.(B) Scanning electron micrographs of the proximal fracture plane of petal abscission zones of wild-type plants. Panels i and ii show fractureplanes after petals were forcibly removed from flowers at positions 3 and 5 on the inflorescence, respectively, where position 1 represents theflower at anthesis. Panels ill and iv show fracture planes after petals were shed naturally and represent positions 7 and 9 on the inflorescencestem, respectively.(C) Scanning electron micrographs of the proximal fracture plane of petal abscission zones of etr1-1 mutant plants. Panels i to iii show fractureplanes after petals were forcibly removed and represent positions 3, 5, and 7 on the inflorescence stem, respectively. Panel iv shows the fractureplane after the petal was naturally shed and represents position 15 on the inflorescence stem.

Because abscission appears to involve a dissolution of thepectin-rich middle lamella, a number of studies have focusedon enzymes that act on pectins. Results have varied, de-pending on the system (Sexton and Roberts, 1982; Osborne,1989), but an AZ-specific polygalacturonase has recently

been cloned from tomato (Kalaitzis et al., 1995). This gene isexpressed in AZs in response to ethylene (Kalaitzis et al., 1997).

Although cell wall hydrolytic enzymes and their respec-tive genes may be expressed in activated AZs, the specificroles they play in the cell separation process have not been

1176 The Plant Cell

determined. Similarly, the signal transduction systems thatregulate these genes and the abscission process itself arealso not well understood. Although ethylene appears to playan important role, del Campillo and Bennett (1996) havefound differences between cellulase gene expression pat-terns in AZs of naturally abscising tomato flowers and those oftomato flowers treated with ethylene. The possibility thatother signal molecules are essential for abscission has beenintroduced by experiments on the bean leaf system that indi-cate that a diffusable signal from the stele is necessary forcell separation in the cortex (Thompson and Osborne,1994). These results suggest that the current models of ab-scission are oversimplified.

Floral Organ Abscission in Arabidopsis

The identification of a developmentally regulated abscissionevent in Arabidopsis would open the possibility of institutinga thorough genetic analysis of the process. As shown in Fig-ures 3A and 3B, the shedding of turgid floral organs shortlyafter anthesis of wild-type Arabidopsis flowers provides thedesired model. An attractive feature of this system is thatthe architecture of the inflorescence branch provides a gra-dient of flowers at all stages of development.

The hypothesized central role for ethylene in the abscis-sion process can be examined in Arabidopsis by using avail-able ethylene-insensitive mutants (see Kende and Zeevaart,

Figure 3. Aspects of Floral Organ Abscission in Wild-Type Arabidopsis and the Ethylene-insensitive Mutant etr1-1.

(A) Loss of floral organs as fruits develop on the primary inflorescence branch of Columbia wild type.(B) Abscised floral organs and silique shortly after abscission in Columbia wild type.(C) Retention of floral organs as fruits develop on the primary inflorescence branch of etr1-l.(D) Developing fruit of etr1-1 showing senescence of retained floral organs.(E) Longitudinal thin section through an Arabidopsis flower receptacle. Tissues protruding to the upper right are the filament (top) and sepal (bot-tom). Note the small, densely cytoplasmic cells that constitute the abscission zones at the bases of the organs.(F) Dark-field view of the section shown in (E). Note the abscission zone-localized expression (red particles) of a basic chitinase-p-glucu-ronidase (CHIT::GUS) reporter gene.(G) Time course of CHIT::GUS expression in abscission zones. The progressive changes in CHIT::GUS expression in Columbia wild-type (top)and etr1-1 (bottom) flowers are shown. Flowers pictured represent positions 3, 4, 5, 6, 7, and 8 (left to right) on the inflorescence stem, whereposition 1 represents the flower at anthesis.

Senescence and Abscission in Arabidopsis 1177

CONCLUDING REMARKS

1997, in this issue). To illustrate the floral organ abscission scission; clearly, the data from analyses of the etrl-7 and process in more detail and, at the same time, to demon- ein2 mutants demonstrate that ethylene is not required for strate the utility of the mutational approach in this system, a the abscission process to occur. These observations are comparative analysis of the etr7-7 mutant and the isogenic consistent with the possibility that developmental pathways wild-type Columbia ecotype is illustrated in Figure 3. that are not dependent on ethylene regulate floral organ Whereas wild-type floral organs are shed shortly after anthe- abscission. sis (Figures 3A and 3B), the organs of etrl-1 flowers are re- tained longer and tend to become senescent before being shed (Figures 3C and 3D). These results are consistent with a role for ethylene in the timing of abscission.

One approach for tracking the abscission process is the use of molecular markers. In the course of our studies on ethylene regulation of the basic chitinase gene (CHIJ) in Somatic tissue senescence, abscission, and the cessation Arabidopsis (Chen and Bleecker, 1995), we observed recep- of meristematic activity are all examples of life-and-death tacle-specific expression of a CHIT::p-glucuronidase (GUS) decisions made by plants throughout their life cycles. These reporter gene that was associated with organ abscission. decisions may affect specific organ systems or, when act- This expression occurs specifically in the small, densely cy- ing in concert, may lead to the demise of the entire plant. toplasmic cells at the base of floral organs (Figures 3E and Within the affected tissues, the physiological processes 3F) and is consistent with reports that the basic chitinases of involved are only partially understood and, in the case of severa1 species are expressed in AZs (de1 Campillo and senescence, may involve global changes in gene expres- Lewis, 1992; Eyal et al., 1993). The expression of this marker sion. How these life history traits are coordinated at the reached a peak shortly after anthesis and correlated with the whole-plant leve1 is a mystery. Yet, decisions to stop pro- timing of abscission (Figure 3G). In contrast, in the e t r l - l ducing new organ systems, to cull existing reproductive background, expression of the reporter gene was signifi- structures by activating pedicle AZs, and to activate the se- cantly lower and delayed in time but did occur. nescence syndrome in otherwise healthy tissues are all de-

Scanning electron microscopy was used to examine the terminants of yield in crop plants. The mutational approach proximal fracture plane of the peta1 AZ after petals were ei- afforded by Arabidopsis provides a method for examining ther forcibly removed or abscised naturally (Figures 2B and these complex processes without prejudice. On the other 2C). When petals were forcibly removed from wild-type hand, the ever-growing pool of existing mutants also pro- flowers at anthesis, a smooth fracture plane was observed, vides unique opportunities to test long-standing hypotheses indicating that the tissues had separated along the middle in novel ways. lamellae between cells (Figure 26, i). The fracture plane in The studies of factors affecting leaf senescence in Arabi- etrl-7 at this stage consisted of a jagged plane of broken dopsis are somewhat disappointing in that the dramatic influ- cell walls, indicating that forcible separation involves the ences of hormones and correlative influences by reproductive breaking of primary walls (Figure 2C, i). At later stages of de- structures that are characteristic of other monocarpic plants velopment, the cells at the wild-type fracture plane became such as soybean and tobacco (Noodén and Leopold, 1988) increasingly enlarged and rounded, finally forming the char- have little influence on the timing of Arabidopsis leaf senes- acteristic AZ scar (Figure 28, ii to iv). The same developmen- cence (Hensel et al., 1993). It may be that correlative and tal progression was observed in the etrl-7 line, although it hormonal factors are most influential only within the con- occurred later in flower development, and the enlargement straints of an intrinsic maximum life span for the somatic tis- and rounding of cells was less pronounced than it was in sue. These influences are apparent in systems in which the wild-type AZs (Figure 2C, ii to iv). intrinsic life span is a full season or more, but they are less

As a result of these observations, it is clear that the shed- obvious when the intrinsic life span is as short as it appears ding of floral organs in Arabidopsis is a well-defined exam- to be in Arabidopsis. ple of the abscission process. An unexpected result was the The influence of reproductive structures on leaf longevity appearance of all the markers for abscission in the ethylene- may be lacking in Arabidopsis, but the evidence for correla- insensitive etrl-7 mutant. However, similar results have tive control of inflorescence meristem activity by developing been obtained for the ethylene-insensitive2 (ein2) mutant, fruits in this species is compelling (Hensel et al., 1994). which is considered to carry a null mutation that completely Thus, this system offers a unique opportunity to study 109s- blocks ethylene responses farther down the ethylene signal distance communication in plants. Mutational analysis day transduction pathway (Ecker, 1995). Thus, although much of reveal whether the underlying mechanisms are nutritional or the literature on the subject assumes a primary role for eth- are a function of specific but undiscovered signaling mecha- ylene in the abscission process, our results do not neces- nisms. The identification of novel signaling systems for sys- sarily justify this ethylocentric point of view. At least in the temic acquired resistance (Ryals et al., 1996) and the case of Arabidopsis floral organs, ethylene may play only a systemic induction of protease inhibitors (Pearce et al., subsidiary role in determining the timing of the onset of ab- 1991) provide the impetus for this kind of investigation.

11 78 The Plant Cell

Abscission plays a very minor role in the life history of Arabidopsis. Nevertheless, the abscission of floral organs in Arabidopsis shows a range of characteristics in common with the abscission process in most plants in which the pro- cess has been studied. The discovery that abscission was delayed but not blocked in ethylene-insensitive mutants was an unexpected result that may indicate the need for a re- evaluation of the role of ethylene in regulating abscission. This finding also opens the possibility for using a mutational approach to dissect ethylene-independent pathways that in- fluente the timing of abscission.

ACKNOWLEDGMENTS

We thank all members of the Bleecker, Sussman, and Amasino labo- ratories for stimulating discussions; we especially thank Jeff Esch, Anne Hall, Michelle Nelson, and Todd Richmond for assistance in compiling the manuscript. A special thanks to Claudia Lipke, the photographer in the Botany Department, and Heide Barnhill, the scanning electron microscopy specialist. We acknowledge support for S.E.P. from the Department of Energy-National Science Founda- tion-U.S. Department of Agriculture Collaborative Research in Plant Biology Program (Grant No. BlR92-20331) and from the U.S. Depart- ment of Agriculture (Grant No. 94-37304-1 105).

REFERENCES

Abeles, F.B., Morgan, P.W., and Saltveit, M.E., Jr. (1992). Ethylene in Plant Biology, 2nd ed. (San Diego, CA: Academic Press).

Addicott, F.T. (1 982). Abscission. (Berkeley: University of California Press).

Alvarez, J., Guli, C.L., Yu, X., and Smyth, D.R. (1992). terminal flower, a gene affecting inflorescence development in Arabidopsis thaliana. Plant J. 2, 103-1 16.

Bate, N.J., Rothstein, S.J., and Thompson, J.E. (1990). Expression of nuclear and chloroplast photosynthesis-specific genes during leaf senescence. J. Exp. Bot. 239,801-81 1.

Buchanan-Wollaston, V. (1997). The molecular biology of leaf senescence. J. Exp. Bot. 48,181-199.

Chen, G.Q., and Bleecker, A.B. (1995). Analysis of ethylene signal- transduction kinetics associated with seedling-growth response and chitinase induction in wild-type and mutant Arabidopsis. Plant Physiol. 108, 597-607.

Clark, S.E. (1997). Organ formation at the vegetative shoot mer- istem. Plant Cell9,1067-1076.

Dean, C., and Leech, R.M. (1982). Genome expression during nor- mal leaf development. Plant Physiol. 69, 904-910.

DeBellis, L., Picciarelli, P., Pisteli, L., and Alpi, A. (1 990). Localiza- tion of glyoxylate-cycle marker enzymes in peroxisomes of senes- cent leaves and green cotyledons. Planta 180,435-439.

de1 Campillo, E., and Bennett, A.B. (1996). Pedicel breakstrength and cellulase gene expression during tomato flower abscission. Plant Physiol. I1 1,813-820.

de1 Campillo, E., and Lewis, L.N. (1 992). ldentification and kinetics of accumulation of proteins induced by ethylene in bean abscis- sion zones. Plant Physiol. 98, 955-961.

de1 Campillo, E., Reid, P.D., Sexton, R., and Lewis, L.N. (1990). Occurrence and localization of 9.5 cellulase in abscising and non- abscising tissues. Plant Cell 2, 245-254.

Delong, A., Calderonurrea, A., and Dellaporta, S.L. (1993). Sex determination gene tasselseed2 of rnaize encodes a short chain alcohol-dehydrogenase required for stage-specific floral organ abortion. Cell 74, 757-768.

Ecker, J.R. (1 995). The ethylene signal transduction pathway in plants. Science 268, 667-674.

Eyal, Y., Meller, Y., Lev-Yadun, S., and Fluhr, R. (1993). A basic- type PR-1 promoter directs ethylene responsiveness, vascular and abscission zone specific expression. Plant J. 4, 225-234.

Forde, J., and Steer, M.W. (1976). The use of quantitative electron microscopy in the study of lipid composition of membranes. J.

Friedrich, J.W., and Huffaker, R.C. (1 980). Photosynthesis, leaf resistance, and ribulose-I ,5-bisphosphate carboxylase degrada- tion in senescing leaves. Plant Physiol. 65, 11 03-1 107.

Gan, S., and Amasino, R.M. (1995). lnhibition of leaf senescence by auto-regulated production of cytokinin. Science 270, 1986-1 988.

Graham, I.A., Leaver, C.J., and Smith, S.M. (1992). lnduction of malate synthase gene expression in senescent and detached organs of cucumber. Plant Cell4,349-357.

Grbic, V., and Bleecker, A.B. (1995). Ethylene regulates the timing of leaf senescence in Arabidopsis. Plant J. 8, 595-602.

Grbic, V., and Bleecker, A.B. (1996). An altered body plan is con- ferred on Arabidopsis plants cartying dominant alleles of two genes. Development 122,2395-2403.

Hempel, F.D., and Feldman, L.J. (1 994). Bi-directional inflores- cence development in Arabidopsis thaliana: Acropetal initiation of flowers and basipetal initiation of paraclades. Planta 192,276-286.

Hensel, L.L., Grbic, V., Baumgarten, D.A., and Bleecker, A.B. (1 993). Developmental and age-related processes that influence the longevity and senescence of photosynthetic tissues in Arabi- dopsis. Plant Cell 5, 553-564.

Hensel, L.L., Nelson, M.A., Richmond, T.A., and Bleecker, A.B. (1 994). The fate of inflorescence meristems is controlled by devel- oping fruits in Arabidopsis. Plant Physiol. 106, 863-876.

Jacobs, T. (1 997). Why do plant cells divide? Plant Cell9,1021-1029.

Kalaitzis, P., Koehler, S.M., and Tucker, M.L. (1995). Cloning of a polygalacturonase expressed in abscission. Plant MOI. Biol. 28,

Kalaitzis, P., Solomos, T., and Tucker, M.L. (1997). Three different polygalacturonases are expressed in tomato leaf and flower abscission, each with a different temporal expression pattern. Plant Physiol. I 13, 1303-1 308.

Kamachi, K., Yamaya, T., Hayakawa, T., Mae, T., and Ojima, K. (1 992). Changes in cytosolic glutamine synthetase polypeptide and its mRNA in a leaf blade of rice plants during natural senes- cence. Plant Physiol. 98, 1323-1 329.

Kende, H., and Zeevaart, J.A.D. (1997). The five “classical” plant hormones. Plant Cell 9, 11 97-1 21 O.

EXP. BOt. 27,1137-1 141.

647-656.

Senescence and Abscission in Arabidopsis 11 79

Kerstetter, R.A., and Hake, S. (1997). Shoot meristem formation in vegetative development. Plant Cell 9, 1001-1 O1 O.

Krysan, P.J., Young, J.C., Tax, F., and Sussman, M.R. (1996). ldentification of transferred DNA insertions within Arabidopsis genes involved in signal transduction and ion transport. Proc. Natl. Acad. Sci. USA 93,8145-8150.

Lashbrook, C.C., Gonzalez-Bosch, C., and Bennett, A.B. (1994). Two divergent endo-p-l,4-glucanase genes exhibit overlapping expression in ripening fruit and abscising flowers. Plant Cell 6,

Lohman, K.N., Gan, S., John, M.C., and Amasino, R.M. (1994). Molecular analysis of natural leaf senescence in Arabidopsis thaliana. Physiol. Plant. 92, 322-328.

Morré, J. (1 968). Cell wall dissolution and enzyme secretion during leaf abscission. Plant Physiol. 43, 1545-1 559.

Noodén, L.D., and Guiamet, J.J. (1996). Genetic control of senes- cence and aging in plants. In Handbook of the Biology of Aging, 4th ed, E.L. Schneider and J.W. Rowe, eds (San Diego, CA: Aca- demic Press), pp. 94-1 18.

Noodén, L.D., and Leopold, A.C., eds (1988). Senescence and Aging in Plants. (San Diego, CA: Academic Press).

Osborne, D.J. (1989). Abscission. Curr. Rev. Plant Sci. 8, 103-129. Pearce, G., Strydom, D., Johnson, S., and Ryan, C.A. (1991). A

polypeptide from tomato leaves induces wound-inducible protein- ase inhibitor proteins. Science 253, 895-898.

Pennell, R.I., and Lamb, C. (1997). Programmed cell death in plants. Plant Cell9, 11 57-1 168.

Poethig, R.S. (1997). Leaf morphogenesis in flowering plants. Plant Cell 9, 1077-1 087.

Ryals, J.A., Neuenschwander, U.H., Willits, M.G., Molina, A., Steiner, H.-Y., and Hunt, M.D. (1996). Systemic acquired resis- tance. Plant Cell8,1809-1819.

Sexton, R., and Roberts, J.A. (1982). Cell biology of abscission. Annu. Rev. Plant Physiol. 33, 133-162.

1485-1 493.

Shanklin, J., DeWitt, N.D., and Flanagan, J.M. (1 995). The stroma of higher plant plastids contain ClpP and ClpC, functional homologs of Escherichia coli ClpP and ClpA An archetypal two- component ATP-dependent protease. Plant Cell7, 1713-1 722.

Shannon, S., and Meeks-Wagner, D.R. (1991). A mutation in the Arabidopsis TFL7 gene affects inflorescence meristem develop- ment. Plant Cell 3, 877-892.

Taylor, C.B., Bariola, P.A., delcardayré, S.B., Raines, R.T., and Green, P.J. (1993). RNS2: A senescence-associated RNase of Arabidopsis that diverged from the S-RNases before speciation. Proc. Natl. Acad. Sci. USA90, 5118-5122.

Thompson, D.S., and Osborne, D.J. (1994). A role for the stele in intertissue signaling in the initiation of abscission in bean leaves (Phaseolus vulgaris L.). Plant Physiol. 105,341-347.

Thomson, W.W., and Plat-Aloia, K.A. (1987). Ultrastructure and senescence in plants. In Plant Senescence: Its Biochemistry and Physiology, W.W. Thomson, E.A. Nothnagel, and R.C. Huffaker, eds (Rockville, MD: American Society of Plant Physiologists), pp. 20-30.

Tucker, M.L., Baird, S.L., and Sexton, R. (1991). Bean leaf abscis- sion: Tissue specific accumulation of a cellulase mRNA. Planta

Varner, J.E., and Taylor, R. (1989). New ways to look at the archi- tecture of plant cell walls. Localization of polygalacturonase blocks in plant tissues. Plant Physiol. 91,31-33.

Weaver, L.M., Himelblau, E., and Amasino, R.M. (1997). Leaf senescence: Gene expression and regulation. In Genetic Engi- neering, Vol. 19, Principles and Methods (New York: Plenum Press), pp. 215-234.

White, E. (1996). Life, death, and the pursuit of apoptosis. Genes Dev. 10, 1-15.

Woolhouse, H.W. (1983). Hormonal control of senescence allied to reproduction in plants. In Beltsville Symposia in Agricultura1 Research-Strategies of Plant Reproduction, W.J. Meudt, ed (Totowa, NJ: Allanheld, Osmun, and Co.), pp. 201-236.

186,52-57.

DOI 10.1105/tpc.9.7.1169 1997;9;1169-1179Plant Cell

A B Bleecker and S E PattersonLast exit: senescence, abscission, and meristem arrest in Arabidopsis.

 This information is current as of August 26, 2018

 

Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists