Seed dormancy and germination

Current Biology

Magazine Primer

Seed dormancy and germination Steven Penfield Reproduction is a critical time in plant life history. Therefore, genes affecting seed dormancy and germination are among those under strongest selection in natural plant populations. Germination terminates seed dispersal and thus influences the location and timing of plant growth. After seed shedding, germination can be prevented by a property known as seed dormancy. In practise, seeds are rarely either dormant or nondormant, but seeds whose dormancyinducing pathways are activated to higher levels will germinate in an ever-narrower range of environments. Thus, measurements of dormancy must always be accompanied by analysis of environmental contexts in which phenotypes or behaviours are described. At its simplest, dormancy can be imposed by the formation of a simple physical barrier around the seed through which gas exchange and the passage of water are prevented. Seeds featuring this so-called ‘physical dormancy’ often require either scarification or passage through an animal gut (replete with its associated digestive enzymes) to disrupt the barrier and permit germination. In other types of seeds with ‘morphological dormancy’ the embryo remains underdeveloped at maturity and a dormant phase exists as the embryo continues its growth post-shedding, eventually breaking through the surrounding tissues. By far, the majority of seeds exhibit ‘physiological dormancy’ — a quiescence program initiated by either the embryo or the surrounding endosperm tissues. Physiological dormancy uses germination-inhibiting hormones to prevent germination in the absence of the specific environmental triggers that promote germination. During and after germination, early seedling growth is supported by catabolism of stored reserves of protein, oil or starch accumulated during seed maturation. These reserves support cell expansion, chloroplast R874

development and root growth until photoauxotrophic growth can be resumed. The phenomenon of seed dormancy is best understood at three levels. At the population level, seed dormancy enables the formation of a soil seed bank from which plants can emerge at different times of year or in response to ecosystem disturbances. At the singleplant level, individual mothers have evolved mechanisms for maintaining control of progeny seed-germination behaviour and generating heterogeneity in progeny seed properties. This enables mothers to hedge their bets by producing seeds with different propensities to germinate or that are likely to germinate at different times and places. Finally, at the level of the individual progeny seed, mechanisms exist to maintain and break dormancy, often in response to environmental stimuli that limit germination to specific annual time windows, or enable seeds to wait for gaps in the canopy to appear. Control of seed dormancy and germination by the mother Although it is tempting to think of germination as a process that starts and ends after seed shedding, in reality, many of the most important aspects of seed germination behaviour are determined during seed maturation and are closely linked to the control of seed dispersal. In his pioneering work on resource conflicts during reproduction, W.D. Hamilton showed that in plants the mother benefits from diversity in seed properties. She favours wider dispersal of progeny, spreading individuals between safer zones closer to the mother herself but also to increasingly risky locations ever further away. Increasing dispersal has the duel benefits of discovering new locations suitable for plant reproduction and reducing resource conflict between individual progeny seeds. However, the fact that the chances of successful reproduction decrease with distance of germination from the mother plant (Figure 1) creates a conflict among individual progeny seeds who are best served by avoiding the fate of being included in the most risky and sacrificial cohort. This optimum maternal strategy therefore contrasts with the optimum strategy for any one individual seed,

which is to germinate closer to the location of the mother, where good chances of reproductive success have been convincingly demonstrated by the mother herself. The mother plant therefore has evolved a number of mechanisms for retaining control of the behaviour of progeny seeds. These include the maternal-derived hard outer tissues of seeds: either the seed coat, which is derived from the ovule integuments, or the pericarp, which is maternal fruit tissue. In addition, in angiosperms the endosperm also surrounds the embryo and has an increased dosage of the maternal genome. Similarly, mechanisms such as gene imprinting allow certain genes from the mother or father to be selectively silenced. Thus, the maternal contribution to the seed is substantial, and all these tissues have been shown to be important in the maternal control of seed dormancy. These maternal processes do not act alone, but instead work in concert with a gene-expression programme in the zygote that simultaneously blocks growth, initiates the accumulation of stored reserves and, in most species, induces a low-water tolerant state that enables the seed to survive wet or dry for long periods in the environment. Environmental signals are perceived by both the mother plant and the developing zygote, and are used to control the germination of progeny seed. The result is that the mother can impart seasonal cues to progeny and also use environmental noise to generate variation in progeny dormancy states. The temperature that the mother plant experiences throughout her life cycle, including whether or not the plant experiences vernalisation, both have major impacts on progeny seed dormancy. These effects are mediated by the same gene network that controls the transition to flowering; these ‘flowering time’ genes are highly expressed in fruit and seed tissues, as well as in leaves and the developing shoot apex. In tomato it is clear that photoperiod signals can be perceived by detached fruits and information passed to progeny seeds to control their behaviour. The manipulation of temperature, nitrate and light during seed maturation also have big impacts on progeny dormancy, and controlling these factors is often the biggest

Current Biology 27, R853–R909, September 11, 2017 Crown Copyright © 2017 Published by Elsevier Ltd.

Current Biology

Magazine challenge in the design and execution of experiments analysing germination behaviour. Seeds can be sensitive even to 1°C changes in temperature, and can use natural noisy environmentaltemperature variation as an important way to generate diversity in germination propensity in seeds on the same inflorescence. In this way, the mother produces cohorts of seeds with varying strategies, for immediate germination or to populate the soil seed bank. There are several mechanisms through which maternal signals control seed properties. Firstly, they affect the developmental and metabolic processes in the fruit- and seed-coat tissues during seed development and maturation, resulting in changes to fruit morphology (affecting dispersal), seed-coat thickness and the deposition of seed-coat polymers such as tannin and suberin that in turn affect the depth of seed dormancy. Secondly, through specific regulation of the maternal copies of genes in the endosperm, the mother can control gene expression patterns in the endosperm during seed maturation and even after seed shedding and imbibition, affecting dormancy loss, storage-protein breakdown and germination rate. Maternal effects are particularly important in species that exhibit what is known as coatimposed dormancy — that is, dormancy that can be removed by excising the embryo from the seed, or even by simply puncturing or scarring the seed — because of the dominant influence of the maternal genome in the development and physiology of these tissues. In these cases the mechanism of dormancy imposition by the seed coat and endosperm barrier is not always entirely clear; but a feature of these seeds is a low oxygen environment, and increasing the partial pressure of oxygen often results in faster germination and higher germination rates. Oxygen levels are essential for the production of reactive oxygen species (ROS) and nitric oxide (NO) in seeds, both of which have important signalling roles that promote seed germination. For instance, both ROS and NO promote the stabilisation of a group of germination-promoting VII Ethylene Response Factor transcription factors, which otherwise are rapidly targeted for degradation.

Dispersal Competition

Risk

Dormancy

Current Biology

Figure 1. Kin selection and parent–offspring conflict theory applied to plant reproduction. The optimal maternal strategy (brown seeds) is to reduce competition among kin by producing progeny with a range of dormancy characters, enabling dispersal in time and space, and populating the soil seed bank. In contrast, the optimum strategy for individual progeny (white seeds) is to germinate close to the mother, where the environment must be conducive for successful reproduction. Because germination terminates dispersal, seed dormancy and dispersal characters are closely linked.

Heterogeneity in seed behaviour is achieved through varying mechanisms. First, production of seeds may occur at different times of year, or at times of year with changing environments, leading to variation in maternal processes controlling seed development and maturation. Alternatively, there can be variation in the degree to which seeds reach full maturity at shedding, with seed dormancy often declining slightly in the later stages of seed maturation. Also, specific developmental programs that produce seeds of different morphs, often including production of seeds of different types in different locules of the fruit, leads to distinct germination behaviors. It has also been hypothesised that stochasticity in the molecular networks that control germination behavior may contribute

to heterogeneity, but in practise it is challenging to separate truly stochastic events from environmental effects because of the extreme sensitivity of seeds to micro-environmental variation. The induction of dormancy during seed maturation When embryogenesis is complete, the maturation programme commences and the seed begins to accumulate reserves of carbon and nitrogen that will fuel seedling establishment after germination. During the maturation phase, seeds of many species also acquire the means to survive desiccation, and enter primary dormancy. In the dormant state, the cells lack hydrated vacuoles, do not break down stored reserves and the cell cycle is suppressed. This developmental state is promoted by a

Current Biology 27, R853–R909, September 11, 2017 R875

Current Biology

Magazine Cutin

O2

Tannin

H2O

Me lan in

ABA synthesis

ABA transport

Suberin

Maternal effects

Suberin

ABA

ROS, NO

O2

Wall

Cell expansion

GA

GA RAM

Wall loosening

loosening ABA

Reserve

tion

mobilisa

Current Biology

Figure 2. Major events in imbibed seeds. Abscisic acid (ABA) is produced primarily in the endosperm and maintains dormancy via transport to the embryo. If ABA levels fall then the root apical meristem (RAM) produces gibberellin (GA), which stimulates water uptake through vacuolisation, combined with cell wall loosening of the hypocotyl and micropylar endosperm. Hypocotyl cell expansion is driven partly by endoreduplication and provides a mechanical force opposed only by the weakening endosperm layer. ABA signalling is attenuated by ROS and NO as seed coat integrity is lost, and the oxygen content of the seed rises.

group of B3-family transcription factors known from Arabidopsis as LEAFY COTYLEDON 1 (LEC1), ABSCISIC ACID INSENSITIVE 3 (ABI3), FUSCA3 (FUS3) and LEAFY COTYLEDON 2 (LEC2), often abbreviated as the ‘LAFL’ transcription factors. This same group is also responsible for the homeotic conversion of leaves to cotyledons, and when any of these genes are compromised desiccation tolerance is reduced or absent, dormancy is lost and accumulation of storage reserves is incomplete. A key function of the LAFL genes in dormancy imposition is the induction of a high abscisic acid (ABA) state, which is necessary for the prevention of germinationrelated processes. If ABA signalling in developing seeds is insufficient, they can begin to germinate on the mother plant before shedding, a property R876

known as vivipary. Vivipary is common in major crops that have been bred for vigorous stand establishment, and is worse in environmental conditions that lead to low dormancy, such as warm, wet conditions. In cereals this vivipary is known as pre-harvest sprouting, and is accompanied by premature production of -amylase and a sharp reduction in grain quality. Pre-harvest sprouting can be a major problem for cereal production in areas of the world with higher humidity and rainfall. Hormones and control of seed germination The key roles of ABA and another, antagonistic plant hormone, gibberellin (GA), in the control of seed dormancy and germination are well established. In Arabidopsis, tomato and wheat, ABA-deficient and GA-biosynthetic and

Current Biology 27, R853–R909, September 11, 2017

signalling mutants have strong seed dormancy phenotypes. LAFL mutants, which abrogate ABA responses, alongside ABA-deficient mutants, are completely unable to produce dormant seeds whatever environment seeds are set in. The DELLA proteins are absolutely required for seed dormancy and growth inhibition, and as such plants harboring mutations in these genes are also strongly non-dormant. In contrast, strong GA-deficient mutants and the GA-signalling mutant sleepy 1 are unable to germinate unless the seed coat and endosperm are removed. This highlights the central role of GA and ABA responses in seed dormancy and germination control. However, other hormones are also involved and may be particularly important in certain species. Ethylene is a germination stimulant that likely promotes

Current Biology

Magazine germination by degradation of ERF transcription factors that mediate reactive oxygen and reactive nitrogen signalling in seeds. And in the parasitic plant species Striga, germination is strongly contingent on the perception of secreted strigolactone from host plant roots. It seems that evolution has modified a common signalling network in different species in distinct ways, allowing the importance of different aspects of the germination process to vary. Natural variation in seed dormancy also seems to be affected by two important loci, REDUCED DORMANCY 5 (RDO5) and DELAY OF GERMINATON 1 (DOG1), which appear to primarily interact with ABA signalling pathways to modulate germination. In temperate environments, temperature is the dominant seasonal signal controlling seed dormancy. Some species have simple temperature requirements for germination, such as cool temperatures or a period of cold or warm stratification. Others have complex requirements for alternate daily temperature fluctuations, or prolonged incubations at cool and warm temperatures to promote germination. A good example is the European ash (Fraxinus excelsior), which requires a period of cold to remove physiological dormancy and a period of warm to overcome morphological dormancy. Thus, there is a period of the year that precedes seedling emergence in which the seeds lose dormancy and can germinate if given light. If they don’t see light during this period they will re-enter dormancy, then known as secondary dormancy, and wait in the seed bank for the next emergence window. A second dormancy-breaking process, and one frequently in a laboratory setting, is dry after-ripening. This is an obscure process by which dormancy is lost over weeks or months during dry storage, and opinion is divided as to whether dry after-ripening is an adaptive response relevant to seeds in real soils, or whether it is an artefact of storage of seeds in very dry conditions not found in nature, except in the most extreme habitats. The mechanism of dry afterripening is unclear, but after imbibition a signal specific to after-ripened seeds activates ABA catabolism. A common feature of environmental signals that break primary dormancy,

including both temperature and afterripening, is a transcriptional effect on the CYTOCHROME P450 707A (CYP707A) gene family, members of which catabolise the ABA necessary to maintain dormancy. The corresponding drop in ABA levels permits GA synthesis and the events surrounding germination to subsequently begin. In addition to pure seasonal cues, other signals are used to detect changes in the canopy that suggest an opportunity to establish. Nitrate is a key signal molecule that increases in the soil in the absence of a growing canopy. Applying nitrate to seeds — either during seed maturation or after imbibition — strongly promotes germination. As with responses to other environmental signals, nitrate appears to act by interacting with hormone metabolism in seeds. In Arabidopsis, NIN-like protein 8 (NLP8) is a key nitrate sensor and promotes germination in the presence of nitrate by the direct transcriptional regulation of CYP707A gene expression. Fire is a second important canopy signal, and specific compounds in smoke, known as karrikins, have a strong germination-promoting activity. In ecosystems where fire is a common disturbance, plants have evolved to recognise these karrikins as signals that indicate an opportunity for seedling establishment. Karrikin signalling appears to share elements of the strigolactone signalling pathway, but also appear to have their own specific receptor. Thus, it has been hypothesised that the karrikin receptor recognises an as yet unknown plant hormone, as well as compounds from smoke. Light is a final important cue for germination promotion in seeds with low dormancy. Absence of light may indicate burial, and many plants use the red/far red ratio detected by phytochromes to signal the presence of a leafy canopy and prevent germination. Phytochrome signalling control of germination is contingent on the interaction of phytochromes with specific transcription factors that control ABA metabolism and GA production. In contrast, for many species darkness is more favourable for germination, and germination can be prevented by blue light, perceived by cryptochromes. Such a

mechanism is common in monocots, including wheat and barley, where cryptochrome signalling represses CYP707A gene expression and lowers seed germination rates in the light. Functions of different seed tissues during the decision to germinate After imbibition of the seed, the different seed tissues have distinct roles in the decision to germinate (Figure 2). For seeds with coatimposed dormancy, ABA is primarily produced in the endosperm tissue; specific transporters then export ABA from the endosperm into the embryo. In cereal seeds, it is especially clear that during germination GA is produced by the embryo, and its arrival in the aleurone layer of the endosperm initiates the breakdown and mobilisation of stored endospermic carbon reserves. Indeed, metabolic activation, vacuolisation and reserve mobilisation of endosperm cells are the first visible signs that a seed has transitioned to germination. These events are followed by expansion of embryo cells that lead to testa rupture, endosperm rupture and radicle emergence. In Arabidopsis, the root apical meristem appears to be an important site of GA production in the embryo. There is good evidence from a number of studies that a GA-derived signal moves from the root to activate water uptake and metabolism in the basal cells of the hypocotyl. The subsequent expansion of hypocotyl cells is driven by the GA-dependent regulation of homeodomain transcription factors, and is the principle cellular event in the embryo that provides the mechanical force for radicle emergence. At this time, ribosome synthesis increases and an apparent translational block is released, allowing translation of the key proteins at the heart of basic metabolic processes and cell growth. The endosperm also responds by secreting cell wall-loosening enzymes from the micropylar pole that weaken the attachments between micropylar endosperm cells to facilitate radicle emergence. All these events are under the inhibitory control of ABA in dormant seeds and are promoted by GA in germinating seeds. The relative importance of these events in the control of germination is likely to differ

Current Biology 27, R853–R909, September 11, 2017 R877

Current Biology

Magazine in seeds with different ratios of embryo and endosperm in the mature seed. Much of our recent understanding of germination control comes from work in Arabidopsis, or other, highly unnatural systems such as our major crops. Considerably less is known about the mechanisms underlying more exotic seed behaviour and seasonaldormancy regulation in seeds present in the soil bank. The biology of seed germination is also highly relevant to seed technology, the field of seedperformance enhancement that underpins the consistent production of high-quality crop seeds, particularly for the vegetable seed market. In the future it should be possible to harness advances in our understanding of seed germination to make improvements to seed performance for farmers and growers. FURTHER READING Bassel, G.W., Stamm, P., Mosca, G., Barbier de Reuille, P., Gibbs, D.J., Winter, R., Janka, A., Holdsworth, M.J., and Smith, R.S. (2014). Mechanical constraints imposed by 3D cellular geometry and arrangement modulate growth patterns in the Arabidopsis embryo. Proc. Natl. Acad. Sci. USA 111, 8685–8690. Chen, M., MacGregor, D.R., Dave, A., Florance, H., Moore, K., Paszkiewicz, K., Smirnoff, N., Graham, I.A., and Penfield, S. (2014). Maternal temperature history activates Flowering Locus T in fruits to control progeny dormancy according to time of year. Proc. Natl. Acad. Sci. USA. 111, 18787–18792. Chiang, G.C., Barua, D., Kramer, E.M., Amasino, R.M., and Donohue, K. (2009). Major flowering time gene, flowering locus C, regulates seed germination in Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA 106, 11661–11666. Gibbs, D.J., Lee, S.C., Isa, N.M., Gramuglia, S., Fukao, T., Bassel, G.W., Correia, C.S., Corbineau, F., Theodoulou, F.L., BaileySerres, J., and Holdsworth, M.J. (2011). Homeostatic response to hypoxia is regulated by the N-end rule pathway in plants. Nature 479, 415–418. Hamilton, W.D. and May, R.M. (1977). Dispersal in stable habitats. Nature 269, 578–581. Kang, J., Yim, S., Choi, H., Kim, A., Lee, K.P., Lopez-Molina, L., Martinoia, E., and Lee, Y. (2015). Abscisic acid transporters cooperate to control seed germination. Nat. Commun. 6, 8113. Piskurewicz, U., Iwasaki, M., Susaki, D., Megies, C., Kinoshita, T., and Lopez-Molina, L. (2016). Dormancy-specific imprinting underlies maternal inheritance of seed dormancy in Arabidopsis thaliana. Elife 5, e19573. Springthorpe, V., and Penfield, S. (2015). Flowering time and seed dormancy control use external coincidence to generate life history strategy. eLife 4, e05557. Yan, D., Easwaran, V., Chau, V., Okamoto, M., Ierullo, M., Kimura, M., Endo, A., Yano, R., Pasha, A., Gong, Y., et al. (2016). NIN-like protein 8 is a master regulator of nitratepromoted seed germination in Arabidopsis. Nat. Commun. 7, 13179.

Department of Crop Genetics, John Innes Centre, Norwich, NR4 7UH, UK. E-mail: [email protected]

R878

Primer

Radial plant growth Nina Tonn and Thomas Greb* One of the extraordinary features of plants is their growth capacity. Depending on the species and the environment, body forms are manifold and, at the same time, constantly reshaped. An important basis of this plastic variation and life-long accumulation of biomass is radial growth. Here, we use this term to describe the ability to grow in girth by the formation of wood, bast and cork. The more technical term for radial growth is secondary growth, which distinguishes the process from primary growth taking place at the tips of stems and roots during plant elongation. After reaching a certain size, most plants start producing wood at the periphery of their organs, which transports water and nutrients and is also essential to support the increasing body weight. At the same time, bast is produced, which transports sugars and growth hormones throughout the entire body. As a protective tissue, cork prevents water loss, microbial infections and physical damage. Illustrating the impact of radial growth on plant performance, there are distinct examples having even a world-wide reputation. The ‘Hyperion’ redwood in Northern California is, at 115 m tall, one of the record holders in organismal height; the trunk girth of the ‹Árbol del Tule› in Mexico is 11 m; and the age of the Great Basin bristlecone pine in the White Mountains of California is around 5000 years. On the cellular level, all these huge and long-lasting structures derive from radial growth, and seemingly non-impressive tissues running through stems and roots — the vascular and the cork cambium. In this Primer, we highlight the evolutionary history of radial plant growth, which underlines the urgent need for plants to be able to modulate long-distance transport capacities throughout their whole life cycle. We explain how the vascular

Current Biology 27, R853–R909, September 11, 2017 © 2017 Elsevier Ltd.

cambium determines this growth process and why earlier cambial systems may have disappeared during evolution. We also give examples for the diversity of plant growth forms due to variation in cambium activity, and provide a brief insight into the molecular players regulating fundamental aspects of cambium-derived tissue production. A short glimpse on the importance of the process for our daily life concludes this overview. The invention of a new skill Nowadays, most plants grow radially throughout their entire life. It has been proposed that the evolution of radial growth began during the Early Devonian (~400 Mya), supporting the transition from sea to land (Figure 1). Long-distance transport of water and nutrients and mechanical stability in a gaseous environment were only a few of the many challenges to be faced during this transition. Some extant land plants — including liverworts, hornworts and mosses (bryophytes) — still exclusively rely on primary growth and, in this respect, resemble the first land plants. Fossils indicate that those plants had one central strand of water-conducting tissue which was notably small compared with the diameter of the whole organ. They also formed only very simple water-conducting cells with smooth, thickened cell walls and pores, as found in hydroids of today’s bryophytes. Hydroids are elongated, dead cells often located in the centre of organs, which lack the more sophisticated cell-wall structure of modern vascular elements. They also do not deposit lignin, a hydrophobic polymer providing resistance against compression and microbial decomposition in cell walls of higher plants. Consequently, body stability of early land plants was predominantly based on turgor pressure of parenchymatous cells rather than on cell wall rigidity on its own. Furthermore, since roots did not yet exist, anchorage was rather poor and uptake of water and nutrients was achieved by simple structures named rhizoids. Other features present in modern land plants were likewise missing, including stomata as specialized structures for gas