Dormancy cycling at the shoot apical meristem: Transitioning between self-organization and self-arrest

Abstract

To survive winter deciduous perennials of the temperate zones cease growth and acquire a cold-acclimated state. Timing of these events is guided by sensory systems in the leaves that register critical alterations in photoperiod. Growth cessation on its own is not sufficient to develop adequate freezing tolerance, which requires entry of the shoot apical meristem (SAM) into dormancy. To fully appreciate perennial dormancy as a precondition for cold acclimation it is necessary to assess how it is brought about in a timely fashion, what the nature of it is, and how it is released. Short day (SD) exposure results in growth cessation, bud set, dormancy establishment at the SAM, and a moderate to high level of freezing tolerance. Subsequent chilling releases the SAM from dormancy and enhances freezing tolerance further. Recent investigations indicate that dormancy is a state of self-arrest that is brought about by an enzyme-based system which disrupts the intrinsic signal network of the SAM. Release from this state requires a complimentary enzyme-based system that is preformed during SD and mobilized by chilling. These findings are in agreement with the paradigm of dormancy cycling, which defines the seasonal alternations at the SAM as transitions between states of self-organization and self-arrest [1]. The success of this survival strategy is based on the adequate scheduling of a complex array of events. The appreciation is growing that this involves signal cascades that are, mutatis mutandis, also recruited in floral evocation in many annuals, including Arabidopsis. A heuristic model is presented of dormancy cycling at the SAM, which depicts crucial molecular and cellular events that drive the cycle.

Introduction

Perennial plant life is attuned to and conditioned by the environment, particularly by photoperiod and temperature [2]. This opportunistic strategy rewards perennials with long life spans and growth over many seasons. In boreal and temperate zones this way of life is stretched to its limits because continuous growth is incompatible with the chilling and freezing conditions of winter. Although shoot apices of most growing perennials can cold acclimate due to exposure to chilling, just like herbaceous plants [3], [4], this level is usually insufficient to withstand winter [5]. It has been demonstrated for many perennial species that high levels of freezing tolerance require the prior acquisition of short day (SD) induced dormancy [6]. Intriguingly, chilling of dormant buds further increases their freezing tolerance, while simultaneously releasing them from dormancy [7], [8], [9]. Thus, it appears that dormancy assumption is necessary to prime a bud for cold acclimation [10] (Fig. 1). Curiously, the cessation of growth, terminal bud set, and entry into dormancy commence very early on in autumn, which may seem to unnecessarily limit growth and productivity. However, considering that sufficient reserves need to be stored in overwintering parts to facilitate cold acclimation [11], a timely cessation and storage accumulation may turn out to be an important part of the trees’ survival strategy.

A just-in-time approach is crucial, as perennials have to balance productivity and survival capacity in order to be competitive. This puts strong demands on the accuracy of the photoperiod-sensing mechanisms as well as on the scheduling of the multiple downstream events, including the initiation of bud scales (initially hidden between the young leaves), the formation of a compressed embryonic shoot within the confines of the developing bud scales, the gradual cessation of stem elongation, the rerouting of resources from the leaves to storage sites, the assumption of a dormant and freezing-tolerant state by the bud, and the eventual senescing of leaves [10], [12]. The identification and pinpointing of these distinct developmental events, which unfold partly in sequence, is still in its infancy. Phenologically, perennials show differences in the timing of this series of events. Ecotypes from the north respond earlier and faster than those of more southern origin [13], [14], [15], while individual ecotypes might have a different critical photoperiod for various subevents. For example, senescence of leaves appears to rely on a shorter critical photoperiod than growth cessation [12]. Release from dormancy takes place after exposure to a sufficient number of chilling hours. In northern perennials chilling requirements often are fulfilled already during autumn. Potentially, ecotypes also differ in their requirements for chilling and the timing of dormancy release. Fortunately this is less relevant to survival as buds remain freezing-tolerant under sub-zero temperatures [4]. Considering its crucial function in the survival strategy of northern perennials the concept of dormancy deserves a close examination.

The term dormancy is commonly used to denote a resting stage, which is characterized by the absence of visible growth in combination with a reduced level of metabolism [16], [17]. The question whether dormancy is a systemic or local phenomenon was addressed by Lang et al. [18], who concluded that dormancy only occurs in plant structures that contain a meristem. This viewpoint, which has become widely accepted [18], [19], [20], [21], [22], [23], [24], aimed to identify the loci of dormancy rather than specific dormancy mechanisms. However, the fact that dormancy can occur in various settings and under different conditions implies the existence of distinct dormancy-inducing mechanisms, and has led to the adoption of the categories para-, eco-, and endo-dormancy [19]. Despite the recognition that distinct dormancy-inducing processes might exist, dormancy was proposed to represent a universal state [18], implying that the various mechanisms may bring about an identical end-state. The traditional notion of dormancy as a hormone-mediated low-metabolic state where cell divisions are absent would fit such description [25]. Despite the attractiveness of such scenario, the differences between the dormant states in the three categories might be more informative than the similarities. For example, it is generally accepted that endo-dormancy is unique in that repression of growth and development persist under conditions that favor growth [18], [19]. Nonetheless, there is no consensus about what constitutes endo-dormancy and what its nature is.

The focus on inducing conditions and putative signal cascades perhaps underlies the strong tendency in the literature to view endo-dormancy as a process rather than a state. However, we can discriminate between the processes and the state they invoke. Up to certain point on the path to endo-dormancy growth can be reinitiated by a return of favorable conditions. Consistent with this, the empirical definition defines endo-dormancy as the end point of a process, the point-of-no-return, which is reached when the bud meristem becomes intrinsically arrested and unable to resume growth under growth-promoting conditions [1], [15], [18], [19], [20], [26]. This clearly depicts endo-dormancy as a state which establishes itself via an intrinsic mechanism and in response to appropriate input signals. Although elicited by SD-induced processes, endo-dormancy is neither a process nor a condition that is maintained or imposed from the outside, hence the term innate dormancy [18].

The consensus view that meristem-containing structures are the loci of dormancy should not be understood to imply that all such structures can establish endo-dormancy. It is well known that roots of overwintering trees can start growing when the soil is locally warmed, showing that root meristems are not endo-dormant but in a state of suppressed growth. Cambium might pass through an endo-dormant phase [10], [21], but this is difficult to establish unambiguously. Only in case of buds the presence or absence of endo-dormancy can be assessed unambiguously in bud break or sprouting tests. Bud formation and cessation of elongation growth precede the establishment of dormancy, but they are separate developmental events which can occur independently (Section 4.1). The most plausible hypothesis therefore is that endo-dormancy is a state in which primary morphogenesis at the embryonic shoot within the bud is obstructed by an intrinsic mechanism. The fact that the shoot apical meristem (SAM) is the central player in primary morphogenesis may single it out as the locus of an intrinsic mechanism [1], [8], [10], [28], [31]. In spite of some clear phenological and physiological similarities with para- and eco-dormancy, as for example the cessation of cell division, endo-dormancy (or ‘true dormancy’) is distinct in nature [32], [33], [34], [35], [36]. The remainder of this review addresses the phenomenon of endo-dormancy or the state of self-arrest, which is a condicio sine qua non for the development of cold acclimation and survival. In the following, we will refer to this state simply as ‘dormancy’, distinguishing it from the quiescence of eco-dormancy and the correlative inhibition of para-dormancy [10], [26].

For many tree species, the assumption that perennials are most freezing-tolerant when they are in deep dormancy is arguably incorrect. Rather, dormancy primes for cold acclimation in anticipation of winter stress, while freezing tolerance is often greater in the subsequent quiescent phase at sub-zero temperatures [10] (Fig. 1). This raises the question if dormancy as a trait might have evolved in the context of stress responses, particularly so as there are similarities in terms of metabolism and stress-induced factors between stressed and overwintering plants [7], [37], [38], [39]. It seems plausible that ad hoc cellular defense programs have become linked to mechanisms that sense the environment and instruct the buildup of resistance to freezing temperatures [28]. The fact that the same basic defense responses are elicited by a variety of agents or triggers shows that plants initially deal with acute threats in an unspecific fashion. For example, the hypersensitive-response is a first defense against viruses, impeding systemic spread, but it can also be induced by abiotic factors like low and high temperatures, aluminum, plasmolysis, wounding, and ozone. All of these stress factors induced the rapid intra- and extracellular depositions of 1,3-β-d-glucans or callose at plasmodesmata (PD) and in cell walls [40], [41]. In its wake, 1,3-β-d-glucanases are produced, probably reflecting a tandem function of 1,3-β-d-glucansynthase and 1,3-β-d-glucanase [28], [42], [43], [44], [45]. In perennials, these functions are temporarily uncoupled at the end of the season when more permanent callose deposits form in the sieve tubes of winter phloem [46], [47], [48] and in the SAM during dormancy establishment (Section 4.2). In the dormant SAM callose depositions are present inside the PD channels, and in ring-like extracellular spaces at the PD collars. As a consequence of these callose deposits the cooperatively generated signal network that sustains SAM function is interrupted and development ceases. The isolated cells might subsequently embark on a path of genetic and metabolic adjustment that promotes SAM acclimation [27]. In perennials, 1,3-β-d-glucansynthase thus not only responds to stress signals, but is part of a mechanism that serves to anticipate future stress. Anticipation requires a mechanism that detects environmental cues foretelling the arrival of winter. This timing mechanism is predominantly located in the leaves, and monitors photoperiod in a way similar to that in annuals.