Desiccation Tolerance and Sensitivity in Medicago Truncatula and Inga Vera Seeds

Desiccation tolerance and sensitivity in Medicago truncatula and Inga vera seeds

José Marcio Rocha Faria

Promotoren:

Prof. Dr. A.M.C. Emons Hoogleraar bij de leerstoelgroep Plantencelbiologie, Wageningen Universiteit

Prof. Dr. L.H.W. van der Plas Hoogleraar bij de leerstoelgroep Plantenfysiologie, Wageningen Universiteit

Co-promotoren:

Dr. H.W.M. Hilhorst Universitair hoofddocent bij de leerstoelgroep Plantenfysiologie, Wageningen Universiteit Dr. A.A.M. van Lammeren

Universitair hoofddocent bij de leerstoelgroep Plantencelbiologie, Wageningen Universiteit

Promotiecommissie:

Prof. Dr. ir. M. Koornneef (Wageningen Universiteit/MPI, Keulen, Duitsland) Prof. Dr. L.A.C.J. Voesenek (Universiteit Utrecht) Prof. Dr. O. Leprince (Institut National d'Horticulture, Angers, France) Dr. W. Finch-Savage (Warwick HRI, University of Warwick, Wellesbourne, Warwick, UK)

Dit onderzoek is uitgevoerd binnen de onderzoekschool Experimentele Plantenwetenschappen (EPW)

José Marcio Rocha Faria

Desiccation tolerance and sensitivity in Medicago truncatula and Inga vera seeds

Proefschrift

ter verkrijging van de graad van doctor op gezag van de rector magnificus van Wageningen Universiteit Prof. Dr. M. Kropff in het openbaar te verdedigen op maandag 1 mei 2006 des namiddags te vier uur in de Aula

Faria, J.M.R. (2006). Desiccation tolerance and sensitivity in Medicago truncatula and Inga vera seeds. PhD thesis, Wageningen University, Wageningen, The Netherlands. With summaries in English, Dutch and Portuguese.

ISBN: 90-8504-417-0

I dedicate this thesis to the three women of my life: Lucília, my mother; Regiane, my wife; and Alissa, my daughter. God bless you!

Contents

Chapter 1. General introduction 1

Chapter 2. Desiccation sensitivity and cell cycle aspects in seeds of Inga vera 17 subsp. affinis

Chapter 3. Changes in DNA and microtubules during loss and re-establishment of 43 desiccation tolerance in germinating Medicago truncatula seeds

Chapter 4. Changes in gene expression during loss and re-establishment of 69 desiccation tolerance in germinated Medicago truncatula seeds

Chapter 5. Improvement of storability of Inga vera subsp. affinis embryos 87

Chapter 6. General discussion 109

Summary 117

Samenvatting 121

Resumo 125

Acknowledgements 129

Curriculum vitae 133

Chapter 1

General introduction

Introduction

Desiccation tolerance is one of the most outstanding features found in the plant kingdom. Seeds that possess such an attribute, the so-called orthodox seeds, can be dried and stored for many years without significant loss of viability. Seeds that lack this characteristic, the so- called recalcitrant seeds, represent a big challenge for those who need to keep them in seed banks for germplasm conservation purposes. While many studies have been devoted to reveal the secrets of desiccation tolerance, less effort has been put into clarifying the recalcitrance phenomenon. In order to gain insight into this theme, this thesis deals with physiological, cytological and molecular aspects of desiccation sensitivity in seeds of Inga vera (a recalcitrant-seeded species) and in germinating seeds of Medicago truncatula (an orthodox-seeded species).

Seed storage behavior

Seeds are the basis of human feeding and the main way of plant propagation. For those reasons farmers have acquired the habit of storing seeds for planting in the subsequent years, since remote ages. To date, storage of crop seeds is an important step in the agricultural productive chain. In recent years, awareness of the loss of plant diversity1 has captured worldwide attention and germplasm conservation has become necessary as a means of maintaining species diversity to prevent genetic erosion (Marzalina and Krishnapillay, 1999). In such context, stored seeds can be used for restoration of degraded ecosystems. Plants produce seeds that respond differently to desiccation and storage and, based on such distinct behavior, Roberts (1973) classified the seeds as orthodox and recalcitrant. Orthodox seeds acquire desiccation tolerance during their development and undergo substantial drying in the final developmental phase (Fig. 1). After shedding they may be dried even further by artificial means and their viability may be extended in a predictable

1 During the 1990s, the loss of natural forests was in average 16.1 million hectares per year, of which 15.2 million occurred in the tropics (FAO, 2000). Chapter 1

way by storage at low temperature and low relative humidity (Roberts, 1973; Roberts and Ellis, 1989; Berjak and Pammenter, 1997). Most crop species, such as wheat, rice, bean, soybean, maize and barley, produce orthodox seeds.

Dispersion/harvest Imbibition Start of DT loss

Re-establishment of DT

tolerance Desiccation

Histodiffe- Maturation Desiccation Storage Germination and rentiation seedling growth

Figure 1. Acquisition and loss of desiccation tolerance (DT) in orthodox seeds. DT is acquired during development enabling the seeds to withstand maturation drying. After dispersion or harvest, DT is maintained during storage and even for some time after imbibition (beginning of germination). With the progress of germination DT starts to decline until it is totally lost upon radicle protrusion or with further seedling growth. Even after DT is totally lost, it can still be re-established (dotted line) by applying mild stresses. Adapted from McKersie (1996).

In contrast with the orthodox types, recalcitrant seeds do not undergo maturation drying and are shed at relatively high moisture contents (0.8-4.0 g H2O/g dry matter). They are desiccation-sensitive both before and after shedding and do not survive if desiccated to water potentials below -15 MPa. Recalcitrant seeds have a very limited post-harvest lifespan in the hydrated condition (Berjak and Pammenter, 1997, 2001). They remain metabolically active upon shedding and, consequently, germinate readily after abscission (Berjak et al., 1989). Moreover, many recalcitrant seeds, particularly those of tropical origin, are also chilling sensitive, and, thus, cannot be stored at temperatures below approximately 15oC (Pammenter and Berjak, 1999; Dussert et al., 1999). Recalcitrant seeds are mainly produced by tree species. In terms of habitat, most recalcitrant species (89%) grow in wet-forest, riverine, flooded, or coastal environments. Most species (79%) are native to the tropics (Farnsworth, 2000). The best known examples are the economically important species cacao (Theobroma cacao), rubber-tree (Hevea brasiliensis), avocado (Persea americana), mango (Mangifera indica) and coconut (Cocus nucifera). Although not abundant as in the tropics,

2 General introduction

recalcitrance is also found in the Temperate Zone in some important tree species belonging to the genera Castanea, Quercus, Aesculus and Acer (Connor and Bonner, 2001). A third category of seeds, called intermediate, was introduced by Ellis et al. (1990) to group seeds that show intermediate post-harvest behavior, i.e. they are relatively desiccation tolerant, but do not withstand removal of water to levels as low as orthodox seeds. They survive desiccation to about -150 MPa, and can generally be stored for periods of intermediate length. Such seeds are often chilling sensitive, even in the dehydrated state. Coffee - Coffea spp (Ellis et al., 1990) and neem - Azadirachta indica (Sacandé, 2000) are examples of species with intermediate seeds. It is thus clear that while it is relatively easy to store orthodox and intermediate seeds, it is still a challenge to store recalcitrant seeds for longer periods. To progress towards a protocol for storage of this category of seeds, one of the first steps is to gain insight into the causes of their desiccation sensitivity. This can be done by studying both the recalcitrant seeds themselves and germinating/germinated orthodox seeds, since the latter are a useful tool for understanding desiccation sensitivity (Farrant et al., 1997; Sun, 1999).

Desiccation tolerance and sensitivity

Desiccation tolerance (DT) in plants can be considered as the ability to rehydrate successfully after the removal of 80-90% of protoplasmic water, reaching moisture contents below 0.3 g H2O/g dry matter (Oliver et al., 2000; Hoekstra et al., 2001). There are three criteria which a plant or plant structure must meet in order to survive such severe loss of protoplasmic water: 1) limitation of the damage suffered by the tissues during desiccation to a repairable level, 2) maintenance of its physiological integrity in the dry state, 3) mobilization of repair mechanisms upon rehydration aiming to revert the damages caused by desiccation and/or rehydration (Bewley, 1979). While few plants are capable of withstanding full desiccation, orthodox seeds undergo this event as a programmed and final stage in their development (Kermode, 1997). Desiccation tolerance enables seed survival during storage or environmental stress and ensures better dissemination of the species (Leprince et al., 1993) A number of processes and mechanisms have been implicated in the acquisition of DT in orthodox seeds, including a) the accumulation of putatively protective molecules, such as the late-embryogenesis-abundant proteins (LEAs), sucrose and certain oligosaccharides, b) reduction of the degree of vacuolisation, c) switching off of metabolism, d) presence of active antioxidant systems, and e) presence and operation of repair mechanisms during rehydration. The absence or ineffective operation of one or more of these processes or mechanisms could determine the desiccation sensitivity in recalcitrant seeds (Pammenter and Berjak, 1999). Figure 2 shows the relationship between developmental/germinative stages

3 Chapter 1

and the hydration level in orthodox and recalcitrant seeds. It also shows the desiccation damages and the mechanisms to avoid or repair them.

5 orthodox seed very recalcitrant seed -1.5 MPa 2.3 g/g mechanical stress vacuolation vacuole filling lessens surface area to volume hydrolysis of 4 LEA proteins synthesized storage proteins reduced monosaccharides and initiation of increased oligosaccharides repair processes synthesis of proteins from stored RNA -3 MPa dedifferentiation of 0.82 g/g organelles unregulated catabolism mobilization of sugars 3 HYDRATION LEVEL HYDRATION recalcitrant seed membrane -11 MPa disruption membrane 0.33 g/g protection catabolic activity 2 aqueous glass -150 MPa GERMINABLE 0.09 g/g destruction of glass, loss of protectant phase 1

GENES

VACUOLATION QUIESCENCE * DESICCATION * OF EXPRESSION AND DEVELOPMENT SEEDLING GROWTH PRE-DESICCATION * AXIS DIFFERENTIATION GERMINATION-RELATED ROOT PROTRUSIONAND DRY MATTER ACCUMULATION MATTER DRY POST-VASCULAR SEPARATION * SEPARATION POST-VASCULAR RESUMPTION OF METABOLISM * REACTIVATION AND IMBIBITION OF MEMBRANES AND ENZYMES * DEVELOPMENT GERMINATION

Figure 2. The desiccation tolerance of developing and imbibing/germinating seeds. Five hydration levels are depicted on the Y axis. The solid and the dashed lines represent the moisture level below which drying is lethal for orthodox and recalcitrant seeds, respectively. Text outside the solid line represents different types of desiccation damage for developing seeds and the processes occurring in imbibing/germinating seeds. Text within the solid line represents mechanisms to resist the damages. Developmental and germinative processes are described along the abscissa. Those that are marked (*) do not occur in recalcitrant seeds, and so it is expected that the corresponding protective mechanisms also do not occur. Redrawn from Vertucci and Farrant (1995).

4 General introduction

The cell cycle and seed desiccation

The cell cycle comprises the events necessary for cell division and can be divided into two stages: interphase and mitosis. Interphase comprises the phases G1 (gap 1), S (DNA synthesis) and G2 (gap 2), while mitosis (M phase) is subdivided in prophase, metaphase, anaphase, telophase and cytokinesis. During interphase the cell undergoes changes (duplication of DNA, synthesis of organelles and an increase in volume), which are essential for further cell division. It has been suggested that cells at G2 phase, with duplicated DNA, are less resistant to stress when compared to cells at G1, pre-synthetic phase. Furthermore, cells in G1 maintain a high viability for extended periods of time (Deltour, 1985). Orthodox seeds are normally shed in a quiescent or dormant state, with most (or all) cells in the axis exhibiting nuclei with 2C DNA content, indicating the arrest of the cell cycle predominantly in the G1 phase (Deltour, 1985). Thus it can be hypothesized that high 4C DNA content may be related to the desiccation sensitivity shown by recalcitrant seeds as well as by germinating/germinated orthodox seeds.

Microtubules and seed desiccation

Another component of the cell cycle that may be involved in the desiccation sensitivity of recalcitrant seeds and germinating/germinated orthodox seeds is the microtubular cytoskeleton, which is markedly sensitive to desiccation stress (Sargent et al., 1981). The cytoskeleton is a major structural feature common to all eukaryotic cells. It is an extensive system of protein fibers. The dynamic cytoskeleton of plants consists of microtubules and actin filaments. Actin filaments are long fibrous structures made of the globular protein actin. Microtubules are tubular elements composed of the proteins α-tubulin and β-tubulin. Microtubules and actin filaments work in concert and provide structural support, and are involved in intracellular transport, cell division and cell shaping (Schmidt and Lambert, 1990, Ketelaar et al., 2001; Kost et al., 2002; Wasteneys, 2002; Wasteneys and Galway, 2003). Microtubules (MTs) play an important part in cell elongation, determination of the division site, chromosome separation and cytokinesis in plant cells (Wasteneys and Galway, 2003). In desiccation-sensitive seeds MTs can be irreversibly deranged by dehydration (Berjak and Pammenter, 2000). Microtubule reassembly is among the cellular repair processes that have been linked to DT (Oliver, 1996; Mycock et al., 2000). Since MTs dynamics can be affected by desiccation (Sargent et al., 1981), the study of their behavior during dehydration of sensitive seeds may help to better characterize DT.

5 Chapter 1

Studying orthodox seeds to understand recalcitrance

The research on recalcitrant seeds faces a number of obstacles, the main one being the very limited time of availability of fresh seeds. Tree species can make the situation even more difficult, since they can alternate years of massive seed production with years of low or no production. In fact, the lack of a model system easy to manipulate was held responsible for the slow progress in studies on recalcitrant seeds (Bray, 1993). For this reason, Sun (1999) suggested that germinated orthodox seeds can be a useful model system for studies on recalcitrance, based on the fact that upon germination, orthodox seeds lose DT progressively (Fig. 1) and become comparable to the recalcitrant types. The main advantage of the use of orthodox seeds is the continuous supply of experimental material (Groot et al., 1996). Many processes, at the physiological, cellular and molecular levels, which occur during the loss of DT in germinated orthodox seeds, may be similar to those responsible for the desiccation sensitivity shown by recalcitrant seeds. Another advantage of the use of orthodox seeds is the feasibility of the re-establishment of DT after its loss by applying mild stresses (Bruggink and van der Toorn, 1995). Putative protective substances such as sucrose and dehydrins, which accumulate during seed development and deplete during germination, accumulate again during re-establishment of DT in germinated seeds (Buitink et al., 2003). The re- induction of DT in germinated seeds enables the comparison of different levels of DT in seeds of the same species, and emerges as an outstanding tool for studies on the mechanisms of desiccation tolerance and sensitivity in seeds.

Use of the technique of re-establishment of DT in orthodox seeds as an attempt to improve storability of recalcitrant seeds

Various stresses, such as drying, osmoticum, cold- and heat-shock, and exogenous ABA, when applied moderately, have shown to be capable to confer or increase DT not only to germinated orthodox seeds, but also to somatic embryos of Medicago sativa (Senaratna et al., 1989; Anandarajah and McKersie, 1990), aba, abi3 mutant seeds (desiccation sensitive) of Arabidopsis thaliana (Ooms et al., 1994), developing orthodox seeds (Kermode and Finch- Savage, 2002), orchid protocorms (Wang et al., 2002) and the cyanobacterial lichen Peltigera polydactylon (Beckett et al., 2005). It is thus reasonable to hypothesize that such treatments could have similar effects on recalcitrant seeds, reducing partially their desiccation sensitivity and/or increasing their storability. To date there are few reports on this approach (Goldbach, 1979; Barbedo and Cicero, 2000; Beardmore and Whittle, 2005).

6 General introduction

Gene expression in seeds

In seeds, the pathways of morphogenesis, maturation and germination are loosely integrated and largely independent developmental programs (McCarty, 1995). Several thousands of genes are believed to be active during seed development, among which those that encode enzymes for housekeeping activities and mobilization of food reserves, structural proteins of the cytoskeleton and membranes, storage proteins and embryo-specific genes that regulate specific aspects of embryo development (Nakabayashi et al., 2005). The maturation drying, which has a regulatory role, switching the seed developmental program to the germination-oriented program (Kermode, 1990) is not simply a passive loss of water, but is also marked by the expression of a number of new genes (Delseny et al., 1994).

Seed developmental genes

The ABSCISIC ACID-INSENSITIVE (ABI3) gene of Arabidopsis, the deletion mutant of which was first described by Koornneef et al. (1984), is specifically expressed in seeds (Parcy et al., 1994). ABI3 proteins are members of a large group of transcription factors that act as intermediates in regulating ABA-responsive genes during seed development and appear to control important processes at the later stages of seed development, viz reserve deposition, dormancy imposition and the acquisition of DT (Zeng et al., 2003). Seeds of the Arabidopsis abi3 mutant show no dormancy and display reduced DT and poor longevity (Léon- Kloosterziel, 1997). ABI3 acts in concert with various other transcription factors such as LEAFY COTYLEDON (LEC) in controlling gene expression during seed maturation (Parcy et al., 1997; Zeng et al., 2003). LEAFY COTYLEDON (LEC) genes, defined by mutations at three loci in Arabidopsis, LEC1, LEC2 and FUSCA3 (FUS3), have major effects on embryo development (Harada, 2001). In Arabidopsis lec1 mutant embryos, post-germinative development is initiated prematurely, simultaneously to the expression of the embryonic program. Cotyledons are partially converted into leaves and the normally quiescent shoot apical meristem is activated (West et al., 1994). Mutant seeds exhibit defects in DT to varying degrees and vivipary occasionally occurs (Harada, 2001). In seeds the transition from the opposing developmental programmes of dormancy and germination is a critical control point regulating the initiation of vegetative growth. The Arabidopsis COMATOSE (CTS) locus is required for this transition, and cts mutant embryos show perturbation in their fatty acid metabolism (Footitt et al., 2002). The CTS gene regulates germination potential by enhancing after-ripening, sensitivity to gibberellins and pre-chilling, and by repressing the activities of loci that activate embryo maturation (Holdsworth et al., 2001). CTS expression transiently increases shortly after imbibition during germination, but not in imbibed dormant seeds (Footitt et al., 2002). The cts mutation

7 Chapter 1

requires the prior action of ABA1, ABI3, FUS3 and LEC1 to induce embryo dormancy, suggesting an interaction between these loci and CTS (Russell et al., 2000).

Stress-related genes

Besides being induced in some stress conditions, several stress-related genes are also naturally expressed during the late stages of orthodox seed development, like the so called LEA (late embryogenesis abundant) genes. It is generally assumed that LEA genes play a role in the establishment of DT during seed development (Delseny et al., 2001). LEA proteins have physical properties consistent with a role in DT, e.g. they are extremely hydrophilic and resistant to denaturation (Oliver et al., 2000). The Em (Early methionine) genes correspond to the class I LEA genes and the first one was isolated from wheat in 1984. Since then, Em genes have been described in a large number of species (monocots, dicots and gymnosperms). Arabidopsis has two Em genes, named AtEm1 and AtEm6. The AtEm6 gene (among others) is completely suppressed in the abi3-4 mutant and is tightly controlled by ABI3, LEC1, ABI5, ABI4 and FUS3 (Delseny et al., 2001). ABA-responsive protein kinases (PKABAs) are critical components in signal transduction pathways leading to cellular adjustments in response to changes in extracellular conditions (Holappa and Walker-Simmons, 1995). PKABA1 transcript levels increase in response to ABA in scutellar, root and shoot tissues (Gómez-Cadenas et al., 2001). Peroxiredoxins (Per) are enzymatic antioxidants involved in the mitigation and repair of the damage initiated by free radicals (Mowla et al., 2002). The plant 1-Cys Per genes from brome grass, barley, Arabidopsis and buckwheat are expressed in developing seeds with the highest levels detected during the stages of massive water loss at the end of seed development and also in the mature dry seed (Haslekas et al., 2003). When non-dormant seeds are imbibed, the transcript level decreases dramatically and transcripts disappear completely after seed germination (Mowla et al., 2002). In Arabidopsis, ABI3 seems to be a prime regulator of the expression of AtPer1 (Haslekas et al., 2003). Small heat-shock proteins (sHSP) are thought to be involved in correctly refolding other proteins partially denatured by heat shock (Delseny et al., 1994). However their importance extends beyond the protection from high-temperature stress (Vierling, 1991), since their role is also seen as either preserving or repairing macromolecular structures during dehydration or rehydration, respectively (Helm and Abernethy, 1990).

Cytoskeleton and cell cycle genes

The microtubular organization in plant cells changes with the progression of the cell cycle. In G2 phase a preprophase band prepares the eventual site of cell plate attachment. At mitosis, spindle microtubules function in chromosome positioning in the equatorial plane and in

8 General introduction

chromosome movement. Phragmoplast microtubules function in cell plate formation and as cells enter G1, endoplasmic microtubules and cortical microtubules are formed. The microtubules in the cells’ periphery play a critical role in controlling growth direction in expanding interphase cells (Wasteneys 2002). In most plant cells cytoskeletal proteins are abundant but not in embryo cells of dry seeds. In seeds of bean (Phaseolus vulgaris L.), actin is greatly diminished but its expression increases dramatically after 24h upon imbibition both at the protein and mRNA levels (Villanueva et al., 1999). Tubulins were hardly detectable in dry tomato seeds but accumulated upon imbibition from 24h onwards and were highest in embryos after germination (de Castro, 1998). In coffee (Coffea arabica L.), tubulin was present in dry seeds but again accumulated upon inbibition and upon germination (da Silva, 2002). Arabidopsis contains eight actin genes, of which ACT7 is the most strongly expressed in young plant tissues and shows the greatest response to physiological cues. Seeds from a homozygous act7-1 mutant showed delayed and lower germination, relative to wild type (Gilliland et al., 2003). Cell proliferation is controlled by a molecular machinery whose core key players are the cyclin-dependent kinases, CDKs (Vandepoele et al., 2002), which respond to both external and internal stimuli (Vazquez-Ramos and Sanchez, 2003). CDK2 is required for the transition from G1 to S (Shen, 2001) and is necessary for DNA replication (Omelyanchuk et al., 2004).

Species chosen for the study

Inga vera Willd. subsp. affinis (DC.) T.D. Penn. (Leguminosae, Mimosoideae), native to South America (Brazil, Paraguay, Uruguay, Argentina, Bolivia, Peru and Colombia) is a common riverbank tree species in lowland rainforest and gallery forest in Cerrado (Savannah) regions, growing up to 25m high (Pennington, 1997). In Southeastern Brazil, it is one of the most abundant species found in the riparian forests and the main one in programs of ecological restoration of those ecosystems (Davide et al., 1996). Its seeds, shed at high water contents, are recalcitrant (desiccation-sensitive), and to date there is no protocol to safely dry or store seeds of this species. Seeds of Inga vera are extremely recalcitrant and are only available during a very limited period of the year. Furthermore, they cannot be stored longer than a few weeks without a rapid loss of viability. Therefore we chose another legume, Medicago truncatula, as a model species for parallel studies. Model species provide insight into processes previously poorly understood (Mandoli and Olmstead, 2000). Most of the recent progress in seed biology at the molecular and physiological levels is owed to studies with model species, such as Arabidopsis thaliana and tomato. Medicago truncatula Gaertn. (Leguminosae, Papilionoideae), which grows in the

9 Chapter 1

Mediterranean Basin (Cook, 1999), emerged as a model system for legume biology and, although the primary focus is on symbiotic nitrogen fixation (Barker et al., 1990), the increasing availability of molecular data (Lamblin et al., 2003) and mutants (Penmetsa and Cook, 2000) is useful for many other aspects of plant sciences, including seed biology. It has been chosen as a model species for genomic studies because of its small, diploid genome, short generation time, self-fertility and high transformation efficiency (Bell et al., 2001). Several research institutes around the world are carrying out projects toward the complete gene inventory and function of the M. truncatula genome (Vandenbosch and Stacey, 2003) and much information has already been published. Although the ultimate objective of this thesis is to better understand desiccation sensitivity, the use of M. truncatula seeds, which are orthodox (desiccation tolerant), is justified by the fact that comparisons of orthodox seeds during or after germination with recalcitrant seeds may be useful in understanding desiccation sensitivity (Farrant et al., 1997; Sun, 1999). The significance of the plant model systems for the advance of knowledge is thus indubitable, but the study of wild, lesser known species should not be precluded. Firstly, because results from model species must be validated in other species to allow general conclusions and, secondly, the results are crucial for people who work with those species. The acquisition of more fundamental knowledge of the recalcitrance phenomenon is crucial for the advancement of ex situ conservation of germplasm from endangered species. Only with successful ex situ conservation protocols will the biodiversity of recalcitrant species be preserved.

Objectives

The processes and mechanisms responsible for the absence of desiccation tolerance and the short longevity of recalcitrant seeds are still far from being understood. For this reason, the general aim of this thesis was to study physiological, cytological and molecular aspects of desiccation sensitivity in seeds, in order to enlarge our understanding on this topic. More specifically, the objectives of this thesis were: 1. To characterize the development, desiccation sensitivity and germination of Inga vera seeds, and their relation with cell cycle and the microtubular cytoskeleton. 2. To investigate the relationship of microtubular cytoskeleton and the relative content and integrity of DNA with the loss and re-establishment of desiccation tolerance in germinating seeds and seedlings of Medicago truncatula. 3. To study the expression of genes related to seed development, desiccation tolerance, cell cycle and cytoskeleton during loss and re-establishment of desiccation tolerance in germinating seeds and seedlings of Medicago truncatula. 4. To improve the storability of Inga vera embryos through the use of osmoticum medium, abscisic acid and sealed storage.

10 General introduction

Scope of the thesis

Chapter 1. General introduction.

An overview of desiccation tolerance and sensitivity in seeds is presented, as well as the approaches considered in this thesis to better understand these important seed traits.

Chapter 2. Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis.

The relation between DNA content and desiccation sensitivity in developing Inga vera seeds is investigated, as well as the effect of dehydration on the microtubular cytoskeleton in embryonic axes. Characterization of the I. vera seed and its physiology of germination are also presented.

Chapter 3. Changes in DNA and microtubules during loss and re-establishment of desiccation-tolerance in germinating Medicago truncatula seeds.

The changes in DNA content and microtubules during and after germination of Medicago truncatula seeds are studied in this chapter. These aspects plus DNA integrity are also used in order to characterize the loss of desiccation tolerance after germination and its re- establishment through osmotic treatment.

Chapter 4. Changes in gene expression during loss and re-establishment of desiccation-tolerance in germinating Medicago truncatula seeds.

By using Real-Time PCR, the expression of various genes related to seed development, desiccation tolerance, cell cycle and cytoskeleton during loss and re-establishment of desiccation tolerance in germinating seeds of Medicago truncatula is investigated.

Chapter 5. Improvement of storability of Inga vera subsp. affinis embryos.

Mature embryos of Inga vera were subjected to several treatments in order to improve their storability. The concomitant changes in physiology and ultrastructure during storage are investigated, aiming at better understanding of the viability loss of stored recalcitrant seeds and to find directions for the development of storage protocols.

11 Chapter 1

Chapter 6. General discussion.

The results achieved in the other chapters are consolidated in order to obtain a clearer picture of the desiccation sensitivity in seeds.

References

Anandarajah, K. and McKersie, B.D. (1990) Manipulating the desiccation tolerance and vigor of dry somatic embryos of Medicago sativa L. with sucrose, heat shock and abscisic acid. Plant Cell Reports 9, 451-455. Barbedo, C.J. and Cicero, S.M. (2000) Effects of initial quality, low temperature and ABA on the storage of seeds of Inga uruguensis, a tropical species with recalcitrant seeds. Seed Science and Technology 28, 793-808. Barker, D.G., Bianchi, S., London, F., Dattee, Y., Duc, G., Essad, S., Flament, P., Gallusci, P., Genier, G., Guy, P., Muel, X., Tourneur, J., Denarie, J. and Huguet, T. (1990) Medicago truncatula, a model plant for studying the molecular genetics of the Rhizobium-legume symbiosis. Plant Molecular Biology Reporter 8, 40-49. Beardmore, T. and Whittle, C.A. (2005) Induction of tolerance to desiccation and cryopreservation in silver maple (Acer saccharinum) embryonic axes. Tree Physiology 25, 965–972. Beckett, R.P., Mayaba, N., Minibayeva, F.V. and Alyabyev, A.J. (2005) Hardening by partial dehydration and ABA increase desiccation tolerance in the cyanobacterial lichen Peltigera polydactylon. Annals of Botany 96, 109-115. Bell, C.J., Dixon, R.A., Farmer, A.D., Flores, R., Inman, J., Gonzales, R.A., Harrison, M.J., Paiva, N.L., Scott, A.D., Weller, J.W. and May, G.D. (2001) The Medicago genome initiative: a model legume database. Nucleic Acids Research 29, 114-117. Berjak, P., Farrant, J.M. and Pammenter, N.W. (1989) The basis of recalcitrant seed behaviour. Cell biology of the homoiohydrous seed condition, in Recent Advances in the Development and Germination of Seeds (ed R.B. Taylorson), Plenum Press, New York, pp. 89-108. Berjak, P. and Pammenter, N.W. (1997) Progress in the understanding and manipulation of desiccation-sensitive (recalcitrant) seeds, in Basic and Applied Aspects of Seed Biology (eds R.H. Ellis, M. Black, A.J. Murdoch and T.D. Hong). Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 689-703. Berjak, P. and Pammenter, N.W. (2000) What ultrastructure has told us about recalcitrant seeds. Revista Brasileira de Fisiologia Vegetal 12, 22-55. Berjak, P. and Pammenter, N.W. (2001) Seed recalcitrance - current perspectives. South African Journal of Botany 67, 79-89. Bewley, J.D. (1979) Physiological aspects of desiccation tolerance. Annual Review of Plant Physiology 30, 195-238. Bray, E.A. (1993) Molecular responses to water deficit. Plant Physiology 103, 1035-1040. Bruggink, T. and van der Toorn, P. (1995) Induction of desiccation tolerance in germinated seeds. Seed Science Research 5, 1-4. Buitink, J., Vu, B.L., Satour, P. and Leprince, O. (2003) The re-establishment of desiccation tolerance in germinated radicles of Medicago truncatula Gaertn. seeds. Seed Science Research 13, 273-286.

12 General introduction

Connor, K.F. and Bonner, F.T. (2001) The effects of desiccation on seeds of Acer saccharinum and Aesculus pavia: recalcitrance in temperate tree seeds. Trees – Structure and Function 15, 131- 136. Cook, D.R. (1999) Medicago truncatula – a model in the making! Current Opinion in Plant Biology 2, 301-304. da Silva, E.A.A. (2002) Coffee seed (Coffea arabica cv. Rubi) germination: mechanism and regulation. PhD thesis, Wageningen University, Wageningen, The Netherlands. Davide, A.C, Botelho, S.A., Faria, J.M.R. and Prado, N.J.S. (1996) Comportamento de espécies florestais de mata ciliar em área de depleção de reservatório da Usina Hidrelétrica de Camargos, Itutinga, MG. Cerne 2, 20-34. de Castro, R.D. (1998) A functional analysis of cell cycle events in developing and germinating tomato seeds. PhD thesis, Wageningen Agricultural University, Wageningen, The Netherlands. Delseny, M., Gaubier, P., Hull, G., Saez-Vasquez, J., Gallois, P., Raynal, M., Cooke, R. and Grellet, F. (1994) Nuclear genes expressed during seed desiccation: relationship with responses to stress, in Stress-induced Gene Expression in Plants (ed A.S. Basra), Harwood Academic Publishers, Switzerland, pp. 25-59. Delseny, M., Bies-Etheve, N., Carles, C., Hull, G., Vicient, C., Raynal, M., Grellet, F. and Aspart, L. (2001) Late Embryogenesis Abundant (LEA) protein gene regulation during Arabidopsis seed maturation. Journal of Plant Physiology 158, 419-427. Deltour, R. (1985) Nuclear activiation during early germination of the higher plant embryo. Journal of Cell Science 75, 43-83. Dussert, S., Chabrillange, N., Engelmann, F. and Hamon, S. (1999) Quantitative estimation of seed desiccation sensitivity using a quantal response model: application to nine species of the genus Coffea L. Seed Science Research 9, 135-144. Ellis, R.H., Hong, T.D. and Roberts. E.H. (1990) An intermediate category of seed storage behaviour? I. Coffee. Journal of Experimental Botany 41, 1167-1174. FAO - Food and Agricultural Organization of the United Nations (2000) Global Forest Resources Assessment. FAO Forestry Paper 140. FAO, Rome. Farnsworth, E. (2000) The ecology and physiology of viviparous and recalcitrant seeds. Annual Review of Ecology and Systematics 31, 107-138. Farrant, J.M., Pammenter, N.W., Berjak, P. and Walters, C. (1997) Subcellular organization and metabolic activity during the development of seeds that attain different levels of desiccation tolerance. Seed Science Research 7, 135-144. Footitt, S., Slocombe, S.P., Larner, V., Kurup, S., Wu, Y., Larson, T., Graham, I., Baker, A. and Holdsworth, M. (2002) Control of germination and lipid mobilization by COMATOSE, the Arabidopsis homologue of human ALDP. EMBO Journal 21, 2912-2922. Gilliland, L.U., Pawloski, L.C., Kandasamy, M.K. and Meagher, R.B. (2003) Arabidopsis actin gene ACT7 plays an essential role in germination and root growth. The Plant Journal 33, 319–328. Goldbach, H. (1979) Imbibed storage of Melicoccus bijugatus and Eugenia brasiliensis (E. dombeyi) using abscisic acid as a germination inhibitor. Seed Science and Technology 7, 403-406. Gomez-Cadenas, A., Zentella, R., Walker-Simmons, M.K. and Ho, T.H.D. (2001) Gibberellin/abscisic acid antagonism in barley aleurone cells: site of action of the protein kinase PKABA1 in relation to gibberellin signaling molecules. The Plant Cell 13, 667-679.

13 Chapter 1

Groot, S.P.C., van Pijlen, J.G., Bergervoet, J.H.W., Kraak, H.L. and Bino, R.J. (1996) Seed storage behaviour in relation to cell cycle progression, in Proceedings of a Workshop on Improved Methods for Handling and Storage of Intermediate/Recalcitrant Tropical Forest Tree Seeds (ed K. Ouedraogo, K. Poulsen and F. Stubsgaard), IPGRI, Rome and DANIDA Forest Seed Centre, Humlebaek, Denmark, pp. 98-102. Harada, J.J. (2001) Role of Arabidopsis LEAFY COTYLEDON genes in seed development. Journal of Plant Physiology 158, 405-409. Haslekas, C., Grini, P.E., Nordgard, S.H., Thorstensen, T., Viken, M.K., Nygaard, V. and Aalen, R.B. (2003) ABI3 mediates expression of the peroxiredoxin antioxidant AtPER1 gene and induction by oxidative stress. Plant Molecular Biology 53, 313-326. Helm, K.W. and R.H. Abernethy (1990) Heat shock proteins and their mRNAs in dry and early imbibing embryos of wheat. Plant Physiology 93, 1626-1633. Hoekstra, F.A., Golovina, E.A. and Buitink, J. (2001) Mechanisms of plant desiccation tolerance. Trends in Plant Science 6, 431-438. Holappa, L.D. and Walker-Simmons, M.K. (1995) The wheat abscisic acid-responsive protein kinase mRNA, PKABA1, is up-regulated by dehydration, cold temperature, and osmotic stress. Plant Physiology 108, 1203-1210. Holdsworth, M., Lenton, J., Flintham, J., Gale, M., Kurup, S., McKibbin, R., Bailey, P., Larner, V. and Russell, L. (2001) Genetic control mechanisms regulating the initiation of germination. Journal of Plant Physiology 158, 439-445. Kermode, A.R. (1990) Regulatory mechanisms involved in the transition from seed development to germination. Critical Reviews in Plant Sciences 9, 155-195. Kermode, A.R. (1997) Approaches to elucidate the basis of desiccation-tolerance in seeds. Seed Science Research 7, 75-95. Kermode, A.R. and Finch-Savage, B.E. (2002) Desiccation sensitivity in orthodox and recalcitrant seeds in relation to development, in Desiccation and Survival in Plants: Drying without Dying (eds M. Black and H.W. Pritchard), Cabi Publishing, Wallingford, Oxon, UK, pp. 149-184. Ketelaar, T. and Emons, A.M.C. (2001) The cytoskeleton in plant cell growth: lessons from root hairs. New Phytologist 152, 409-418. Koornneef, M., Reuling, G. and Karssen, C.M. (1984) The isolation and characterization of abscisic acid-insensitive mutants of Arabidopsis thaliana. Physiologia Plantarum 61, 377-383. Kost, B., Mathur, J. and Chua, N.H. (1999) Cytoskeleton in plant development. Current Opinion in Plant Biology 2, 462-470. Kost, B., Bao, Y.Q. and Chua, N.H. (2002) Cytoskeleton and plant organogenesis. Philosophical Transactions of the Royal Society of London. Series B 357, 777–789. Lamblin, A.F.J., Crow, J.A., Johnson, J.E., Silverstein, K.A.T., Kunau, T.M., Kilian, A., Benz, D., Stromvik, M., Endre, G., VandenBosch, K.A., Cook, D.R., Young, N.D. and Retzel, E.F. (2003) MtDB: a database for personalized data mining of the model legume Medicago truncatula transcriptome. Nucleic Acids Research 31, 196-201. Léon-Kloosterziel, K.M. (1997) Genetic Analysis of Seed Development in Arabidopsis thaliana. PhD thesis, Wageningen Agricultural University, Wageningen, The Netherlands. Leprince, O., Hendry, G.A.F. and McKersie, B.D. (1993) The mechanisms of desiccation tolerance in developing seeds. Seed Science Research 3, 231-246.

14 General introduction

Mandoli, D.F. and Olmstead, R. (2000) The importance of emerging model systems in plant biology. Journal of Plant Growth Regulation 19, 249–252. Marzalina, M. and Krishnapillay, B. (1999) Recalcitrant seed biotechnology applications to rain forest conservation, in Plant Conservation Biotechnology (ed E.E. Benson), Taylor and Francis, London, pp. 265-276. McCarty, D.R. (1995) Genetic control and integration of maturation and germination pathways in seed development. Annual Review of Plant Physiology and Plant Molecular Biology 46, 71–93. McKersie, B.D. (1996) Desiccation stress. The Pennsylvania State University, Department of Crop and Soil Sciences. http://cropsoil.psu.edu/Courses/AGRO518/ Desiccat.htm (accessed on Aug 25, 2005). Mowla, S.B., Thomson, J.A., Farrant, J.M. and Mundree, S.G. (2002) A novel stress-inducible antioxidant enzyme identified from the resurrection plant Xerophyta viscosa Baker. Planta 215, 716–726. Mycock, D.J., Berjak, P. and Finch-Savage, W.E. (2000) Effects of desiccation on the subcellular matrix of the embryonic axes of Quercus robur, in Seed Biology: Advances and Applications (eds M. Black, K.J. Bradford and J. Vazquez-Ramos), Cabi Publishing, Wallingford, Oxon, UK, pp. 197-203. Nakabayashi, K., Okamoto, M., Koshiba, T., Kamiya, Y., and Nambara, E. (2005). Genome-wide profiling of stored mRNA in Arabidopsis thaliana seed germination: epigenetic and genetic regulation of transcription in seed. The Plant Journal 41, 697-709. Oliver, M.J. (1996) Desiccation tolerance in vegetative plant cells. Physiologia Plantarum 97, 779-787. Oliver, M.J., Tuba, Z. and Mishler, B.D. (2000) The evolution of vegetative desiccation tolerance in land plants. Plant Ecology 151, 85–100. Omelyanchuk, L.V., Trunova, S.A., Lebedeva, L.I. and Fedorova, S.A. (2004) Key Events of the cell cycle: Regulation and organization. Russian Journal of Genetics 40, 219-234. Ooms, J.J.J., van der Veen, R. and Karssen, C.M. (1994) Abscisic acid and osmotic stress or slow drying independently induce desiccation tolerance in mutant seeds of Arabidopsis thaliana. Physiologia Plantarum 92, 506-510. Pammenter, N.W. and Berjak, P. (1999) A review of recalcitrant seed physiology in relation to desiccation-tolerance mechanisms. Seed Science Research 9, 13-37. Parcy, F., Valon, C., Raynal, M., Gaubier-Comella, P., Delseny, M. and Giraudat, J. (1994) Regulation of gene expression programs during Arabidopsis seed development: roles of the ABI3 locus and of endogenous abscisic acid. The Plant Cell 6, 1567-1582. Parcy, F., Valon, C., Kohara, A., Misera, S. and Giraudat, J. (1997) The ABSCISIC ACID- INSENSITIVE3, FUSCA3, and LEAFY COTYLEDON1 loci act in concert to control multiple aspects of Arabidopsis seed development. Plant Cell 9, 1265-1277. Penmetsa, R.V. and Cook, D. R. (2000) Production and characterization of diverse developmental mutants of Medicago truncatula. Plant Physiology 123, 1387-1397. Pennington, T.D. (1997) The Genus Inga: Botany. The Royal Botanic Gardens, Kew, London, UK. Russell, L., Larner, V., Kurup, S., Bougourd, S. and Holdsworth, M. (2000) The Arabidopsis COMATOSE locus regulates germination potential. Development 127, 3759-3767. Roberts, E.H. (1973) Predicting the storage life of seeds. Seed Science and Technology 1, 499-514. Roberts, E.H. and Ellis, R.H. (1989) Water and seed survival. Annals of Botany 63, 39-52.

15 Chapter 1

Russell, L., Larner, V., Kurup, S., Bougourd, S. and Holdsworth, M. (2000) The Arabidopsis COMATOSE locus regulates germination potential. Development 127, 3759-3767. Sacandé, M. (2000) Stress, Storage and Survival of Neem Seed. PhD thesis, Wageningen Agricultural University, Wageningen, The Netherlands. Sargent, J.A., Sen Mandi, S. and Osborne, D.J. (1981) The loss of desiccation tolerance during germination: an ultrastructural and biochemical approach. Protoplasma 105, 225-239. Schmit, A. and Lambert, A. (1990) Microinjected fluorescent phalloidin in vivo reveals the F-actin dynamics and assembly in higher plant mitotic cells. The Plant Cell 2, 129-138. Senaratna, T., McKersie, B.D. and Bowley, S.R. (1989) Desiccation tolerance of alfalfa (Medicago sativa L.) somatic embryos. Influence of abscisic acid, stress pretreatments and drying rates. Plant Science 65, 253-259. Shen, W.H. (2001) The plant cell cycle: G1/S regulation. Euphytica 118, 223-236. Sun, W.Q. (1999) Desiccation sensitivity of recalcitrant seeds and germinated orthodox seeds: Can germinated orthodox seeds serve as a model system for studies of recalcitrance?, in Proceedings of Iufro Seed Symposium 1998: Recalcitrant Seeds (eds M. Marzalina, K.C. Khoo, N. Jayanthi, F.Y. Tsan and B. Krishnapillay), FRIM, Kuala Lumpur, Malaysia, pp. 29-42. Vandenbosch, K.A. and Stacey, G. (2003) Summaries of legume genomics projects from around the globe. Community resources for crops and models. Plant Physiology 131, 840-865. Vandepoele, K., Raes, J., De Veylder, L., Rouze, P., Rombauts, S. and Inze, D. (2002) Genome-wide analysis of core cell cycle genes in Arabidopsis. The Plant Cell 14, 903-916. Vazquez-Ramos, J.M. and Sanchez, M. de la P. (2003) The cell cycle and seed germination. Seed Science Research 13, 113-130. Vertucci, C.W. and Farrant, J.M. (1995) Acquisition and loss of desiccation tolerance, in Seed Development and Germination (eds J. Kigel and G. Galili), Marcel Dekker Inc., New York, pp. 237- 271. Vierling, E. (1991) The roles of heat shock proteins in plants. Annual Review of Plant Physiology and Plant Molecular Biology 42, 579-620. Villanueva, M.A., Campos, F., Diaz, C., Colmenero-Flores, J.M., Dantan, E., Sanchez, F. and Covarrubias, A.A. (1999) Actin expression in germinating seeds of Phaseolus vulgaris L. Planta 207, 582-589. Wang, X.J., Loh, C.S., Yeoh, H.H. and Sun, W.Q. (2002) Drying rate and dehydrin synthesis associated with abscisic acid-induced dehydration tolerance in Spathoglottis plicata orchidaceae protocorms. Journal of Experimental Botany 53, 551-558. Wasteneys, G.O. (2002) Microtubule organization in the green kingdom: chaos or self-order? Journal of Cell Science 115, 1345-1354. Wasteneys, G.O. and Galway, M.E. (2003) Remodeling the cytoskeleton for growth and form. Annual Review of Plant Biology 54, 691-722. West, M.A.L., Yee, K.M., Danao, J., Zimmerman, J.L., Fischer, R.L., Goldberg, R.B. and Harada, J.J. (1994) LEAFY COTYLEDON1 is an essential regulator of late embryogenesis and cotyledon identity in Arabidopsis. The Plant Cell 6, 1731-1745. Zeng, Y., Raimondi, N. and Kermode, A.R. (2003) Role of an ABI3 homologue in dormancy maintenance of yellow-cedar seeds and in the activation of storage protein and Em gene promoters. Plant Molecular Biology 51, 39-49.

16 Chapter 2

Desiccation sensitivity and cell cycle aspects * in seeds of Inga vera subsp. affinis

José Marcio Rocha Faria, André A. M. van Lammeren and Henk W. M. Hilhorst

Abstract

The desiccation sensitivity of seeds of Inga vera Willd. subsp. affinis, a recalcitrant-seeded tree from Brazil, was analysed, focusing on water relations and cell cycle aspects, including DNA content and the microtubular cytoskeleton. Seeds were collected at four developmental stages, dried to different moisture contents (MCs), assessed regarding water activity and set to germinate. Samples of fresh (non-dried) developing and mature seeds were used for assessment of DNA content by flow cytometry. Immunohistochemical detection of microtubules (MTs) was done in mature seeds at different MCs. Slight desiccation of immature seeds increased germination, but further drying resulted in a quick decline of germinability. During seed development the desiccation sensitivity decreased slightly, but DNA content of the embryonic axis cells remained constant, suggesting no relation between those two parameters. Embryonic axis cells of mature seeds showed abundant cortical microtubule arrays, which were not affected by mild desiccation, but broken down by further drying. It appeared that, upon rehydration, damaged cells were not able to reconstitute the microtubular cytoskeleton. The failure of germination of Inga vera seeds after drying could not be attributed to cellular damage to the DNA synthesis and mitosis, since the radicle protruded by means of cell elongation, without a need for cell division. However, the breakdown of MTs during desiccation and their subsequent inability to reassemble upon rehydration, may be related to the decreased germination, since MTs are required for cell elongation.

* This chapter has been published in Seed Science Research (2004) 14, 165-178. Chapter 2

Introduction

The genus Inga belongs to the family Leguminosae (subfamily Mimosoideae) and comprises circa 300 species of trees restricted to tropical and subtropical America. It is a ubiquitous component of lowland and montane rainforests from 24oN in Mexico to 34oS in Uruguay. Inga species generally exhibit in general rapid growth, tolerance of acid soils and high production of leafy biomass, which helps to control weeds and erosion. Inga vera Willd. subsp. affinis (DC.) T.D. Penn., native to South America (Brazil, Paraguay, Uruguay, Argentina, Bolivia, Peru and Colombia) is a common riverbank species in lowland rainforests and gallery forests in Cerrado (savannah) regions, growing up to 25m high and 40cm stem diameter. It associates with Rhizobium and mycorrhizae (Davide et al., 1995; Pennington, 1997). In south-eastern Brazil, I. vera is used in restoration of riparian forests, and its fast- growing seedlings are able to withstand submersion for up to 3 months. Fruits can float on water, which is the most effective agent in Inga seed dispersal (Bilia et al., 1999). Its seeds are shed at high water contents, they are considered recalcitrant (desiccation-sensitive) and germinate immediately upon shedding. Vivipary is a common feature within the genus (Pennington, 1997; Farnsworth, 2000). Because of its habitat, Inga seeds are not naturally exposed to drying, which, even to a small extent, is injurious to them. Such sensitivity to drying makes the storage of Inga vera seeds, as well as those of other recalcitrant species, very difficult. Seeds that can be dried to very low moisture content (MC) and have their viability extended by storage at low temperature and relative humidity have been termed ‘orthodox’ by Roberts (1973). With opposite behaviour, seeds that lose viability when dried to relatively high MC, were termed ‘recalcitrant’. Such seeds, particularly those of tropical origin, are generally also chilling sensitive and cannot be stored at temperatures below about 15oC (Dussert et al., 1999; Pammenter and Berjak, 1999). A third category of seeds, called ‘intermediate’, was proposed by Ellis et al. (1990), for grouping seeds that have intermediate postharvest behavior, i.e. those that are relatively desiccation tolerant, but do not withstand removal of water to levels as low as those tolerated by orthodox seeds. They can generally be stored for periods of intermediate length. Such seeds are often chilling sensitive, even in the dehydrated state. Development of orthodox seeds can be divided in three major phases: histo- differentiation, maturation and maturation drying (Bewley and Black, 1994). During the phase of maturation, orthodox seeds acquire desiccation tolerance, which is maintained after shedding. In contrast, recalcitrant seeds do not complete maturation drying, are shed at relatively high water contents and are always sensitive to desiccation (Berjak and

Pammenter, 2000). At shedding, MCs of recalcitrant seeds (1.00 to 2.33 g H2O/g dry matter) are much higher than those of orthodox types (0.18 to 0.25 g/g) (Chin et al., 1989), except those that develop within a wet fruit. A suite of mechanisms and processes has been

18 Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis

implicated in the tolerance to desiccation of orthodox seeds, e.g. insoluble reserve accumulation, metabolic switching off, and presence and operation of repair systems during rehydration. The absence or inadequate expression of one or more of these mechanisms is considered to be the cause of desiccation sensitivity in recalcitrant types (Pammenter and Berjak, 1999). Water is, perhaps, the key to understanding seed biology, since it is involved in every dynamic process in a living cell (Sun and Gouk, 1999). The loss of water from plant cells is an important environmental stress, influencing the structural stability and all aspects of biological functions (Sun, 2002). After shedding, the extent of water loss tolerated by recalcitrant seeds without decreased viability varies among species, but it is always much less than in the orthodox types. Furthermore, the level of desiccation sensitivity is influenced by the developmental stage of the seed and the dehydration conditions (Hong and Ellis, 1990; Berjak et al., 1993). According to Sun and Gouk (1999) in studies on water relations in seeds, thermodynamic parameters, such as water activity (aw) and water potential, should be preferred to the mass-based expression of MC, because they relate directly to the biophysical and physiological events that govern the desiccation sensitivity in recalcitrant seeds, while the latter is influenced largely by various components of the seeds. Probert and Longley (1989) stated that the relationship between moisture content and viability on recalcitrant seeds should be independent of drying rate. However there has been some disagreement about the influence of drying rate on desiccation sensitivity (Berjak et al., 1993). In many species, drying rate affects desiccation sensitivity (Farrant et al., 1985; Pritchard, 1991; Pammenter et al., 1991, 1998, 2000; Liang and Sun, 2000; Pammenter and Berjak, 2000; Wesley-Smith et al., 2001). In contrast, very small or no effects have also been reported (Finch-Savage, 1992; Berjak et al., 1993; Pritchard et al., 1995). From the great majority of reports on drying rate influencing desiccation sensitivity, it appears that the faster the desiccation rate, the lower the MC that can be reached before viability starts to decrease. It is obvious though that, no matter how rapid desiccation- sensitive tissue is dried, there is a lower limit below which it cannot survive, and this MC is always higher than the water content to which orthodox and intermediate seeds can be dried (Pammenter and Berjak, 1999). What can be drawn from the literature is that the effect of drying rate on viability loss in recalcitrant seeds is species specific. When recalcitrant seeds are subjected to artificial drying, viability is first slightly reduced (Hong and Ellis, 1996) or is not affected (Pritchard et al., 1995), depending on the species. Further desiccation leads to a quick reduction of viability. The point at which a rapid drop of viability starts to be observed is termed the “critical moisture content” (King and Roberts, 1979) or “lowest safe moisture content” (Tompsett, 1984). The point where viability is completely lost is called “lethal moisture content” (Hong and Ellis, 1992). The cell cycle is the orderly sequence of events by which a cell duplicates its contents and divides into two (Alberts et al., 1998). It has been suggested that cells at the G1 phase of the cell cycle (2C DNA content) are more resistant to stress, when compared to cells at G2

19 Chapter 2

(4C DNA content), and also have higher viability for extended periods of time (Deltour, 1985). This was confirmed by Saraco et al. (1995), who found that primed seeds of pepper (Capsicum annuum), with a high frequency of 4C nuclei, were more sensitive to ageing. Cells in embryos of desiccation-tolerant seeds of various species predominantly contain 2C DNA, implying that, during embryo formation, developmental control imposes an arrest of the cell cycle, predominantly in the G1 phase (reviewed by Vazquez-Ramos and Sanchez, 2003). The inhibition of DNA synthesis might be linked to the drop in water content during seed maturation (Deltour, 1985) and also to the presence of abscisic acid (Bouvier-Durand et al., 1989). Cell cycling has been related to metabolic activity in hydrated tissues (Pammenter and Berjak, 1999). Since recalcitrant seeds show no developmental arrest and are metabolically active and highly hydrated at shedding (Berjak et al., 1992, 1993; Uniyal and Nautiyal, 1996; Berjak and Pammenter, 2000) it can be expected that the cell cycle is not totally arrested at shedding, leading to a higher proportion of cells at G2 phase (4C DNA content), when compared to orthodox types. This might be one of the causes of desiccation sensitivity, as proposed by Boubriak et al. (2000). Another component of the cell cycle, the microtubular cytoskeleton, should also be considered, since its reassembly is markedly sensitive to desiccation stress (Sargent et al., 1981). Microtubules (MTs) are elongated tubular structures, made of α and β-tubulin, which are capable of rapid changes of length, number and location by assembling from a cytoplasmic tubulin pool and later disassembling (Dustin, 1978). In higher-plant cells, the MT configurations (cortical, preprophase band, spindle and phragmoplast) change dynamically during the cell cycle (Vantard et al., 2000). MTs and their associated motor proteins play an important part in intracellular transport and positioning of a wide variety of macromolecules, vesicles, organelles and other macromolecular structures essential to plant growth and development (Goddard et al, 1994; Alberts et al., 1998). Such spatial architecture within the cell can be irreversibly deranged by stresses, such as dehydration of desiccation-sensitive seeds (Berjak and Pammenter, 2000). MTs also function in the morphogenesis of plant cells, i.e. cell elongation, determination of the division site and cytokinesis (Kumagai and Hasezawa, 2001). MT reassembly is among the cellular repair processes that have been linked to desiccation tolerance (Oliver, 1996; Mycock et al., 2000). In order to proceed toward a long-term storage of recalcitrant seeds, e.g. for seed bank purposes, a better understanding of the basis of desiccation sensitivity is a prerequisite. For this reason the present study was carried out to characterize the development, desiccation sensitivity and germination of I. vera seeds, focusing on such cell cycle aspects as DNA content and the microtubular cytoskeleton. General information about the characteristics of the seed and the species was also included, to have a more complete picture of the phenomenon of recalcitrance, as suggested by Berjak and Pammenter (1994).

20 Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis

Material and Methods

Plant material

Seeds of I. vera at four developmental stages [6, 9, 10 and 14 weeks after flowering (WAF)] were obtained from fruits hand-harvested from ten adult, healthy trees, growing in riparian forests along the Rio Grande river surrounding the city of Lavras, Minas Gerais State, Brazil. The coordinates at the collection site are 21o09’S and 44o53’W, and the altitude is 808 m. The climate is classified as a transition between Cwb and Cwa, i.e. hot and humid summers with mild and dry winters, according to the Köppen classification (Köppen, 1936). Mature fruits were collected in January 2002 and January 2003. Immature fruits were collected in November and December 2002. Just after collection, seeds were extracted manually from the fruits, packed into a Styrofoam box and shipped to The Netherlands by an express service. The seeds arrived after 48 h and were immediately processed, sampled and treated, as indicated for each experiment. During processing of the seeds, the sarcotesta was removed, which caused removal of the seed coat as well (Fig. 1a-d).

Storability of fresh embryos

To determine the lifespan of fresh (non-dried) mature I. vera embryos, a small storage trial was carried out. Embryos were stored in glass containers covered with Parafilm, at three different temperatures (5, 10 and 20oC), in the dark. After 10 and 18 days, germination tests were performed at 35oC on moist sand in constant light to assess embryo viability (4 replications of 20 embryos).

Desiccation methods and assessment of moisture content, water activity and water potential

As a preliminary test, mature embryos were dried in a single layer, in closed cabinets at 20oC under circulating air at different relative humidities (RH%), achieved by different saturated salt solutions. Three drying rates were tested: fast (10% RH, achieved by NaOH;), intermediate (35% RH; MgCl2) and slow (89% RH; KCl). Since there was no effect of drying rate on desiccation sensitivity, the intermediate rate was chosen for drying seeds for the subsequent experiments. MCs of the embryo and axis were determined gravimetrically on four replications of ten embryos by oven drying for 17 h at 103oC (ISTA, 1996). Water activity (aw) measurements of the embryos were performed with a water activity meter

(HygroLab 3, Rotronic, Germany) in four replications of 5-8 embryos. Water potential (Ψw)

21 Chapter 2

was calculated from water activity using the equation Ψw = RT ln (aw)/V, where R is the gas constant (8.314 J mol-1 K-1), T is the temperature (Kelvin) and V is the partial molal volume of water (18.048 mL/mole). Fresh and dried embryos were used for germination tests and immunohistochemical detection of the microtubular cytoskeleton.

Germination tests

Germination tests of fresh and dried embryos were carried out on moistened sand in plastic boxes (11 x 11 x 3.5 cm), under constant light. In order to determine the best temperature for germination, the mature seeds collected in January 2002 were set to germinate at eight temperatures, from 5 to 40oC, at 5oC intervals, following a recommendation of Hong and Ellis (1996). The subsequent germination tests (for immature seeds and for mature seeds collected in January 2003) were conducted at 35oC, which was found to be the best temperature for germination. Germination was scored at least daily, with visible growth of the radicle being the criterion for germination.

DNA content assessment

Relative DNA content was assessed by flow cytometry, using suspensions of intact nuclei prepared from embryonic axes. Each treatment was composed of ten replications of one axis, divided in root and shoot parts. Samples were prepared according to Arumuganthan and Earle (1991). All analyses were performed with a Beckman-Coulter EPICS XL-MCL flow cytometer (Beckman-Coulter, Miami, Florida, USA) equipped with an argon ion laser at 488 nm. At least 5000 nuclei were analysed in each replication. Histograms were processed using ModFit LT (Verity Software House, Topsham, Maine, USA) for data analysis and correction of the background noise. Statistical analyses were performed by using the software SPSS 11.0.1.

Immunohistochemical detection of microtubular cytoskeleton (β-tubulin)

Embryonic axes were excised from fresh and dried mature embryos, cut longitudinally, plunged into liquid propane and transferred to cryo-tubes containing 1.5 mL of frozen freeze- substitution medium (water-free methanol + 0.1% glutaraldehyde). The cryo-tubes were put in a freeze substitution unit (Cryotech, Benelux, Schagen, The Netherlands) for 48 h. After freeze substitution the freeze-fixation medium was replaced by ethanol (series of increasing concentrations), followed by embedding in butylmethylmethacrylate (BMM) and UV polymerization at –20oC, according to Baskin et al. (1992). Four axes were analyzed for each

22 Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis

treatment. Longitudinal sections with a thickness of 3µm were placed on slides, and BMM was removed by washing in acetone. The slides were then rinsed in full-strength phosphate- buffered saline (PBS), pH 7.3, and sections were blocked in 0.1 M hydroxyl tetra ammonium chloride (HAH) and in 26 mM bovine serium albumine (BSA). Next, they were incubated with the primary antibody [mouse anti-β-tubulin (Sigma, Zwijndrecht, The Netherlands), diluted 1:200 v/v] and with the secondary antibody [goat anti-mouse IgG conjugated with fluorescein-5-isothiocyanate (FITC) (Molecular Probes, Leiden, The Netherlands), diluted 1:200 v/v]. As a control, slides without the first antibody were used for every treatment.

Results

General characteristics of Inga vera seeds

Flowering started in the beginning of October, and mature fruits began to fall by the end of December. Mature fruits are yellowish-brown, indehiscent elongated pods, 7 cm length on average (3-15cm). Seeds are elliptical, green and non-endospermic, with the surface entirely covered by a white edible sarcotesta (Fig. 1a). The sarcotesta is rich in sugars and originates from the hypertrophied layer of Malpighian cells. It forms an attractant for dispersing animals, mostly primates (Pennington, 1997). Some physical characteristics of the embryos are shown in Figs. 1 and 2. At shedding, the MC of the embryos was around 1.5 g H2O/g dry matter (hereafter g/g). Surrounding tissues, composed of sarcotesta and pericarp, showed even higher MCs; respectively 7.33 g/g and 2.68 g/g. Often germination was detected within the unopened legume (vivipary), as already reported for Inga spp. by Farnsworth (2000). Polyembryony was found in a relatively high proportion, with 23% of the seeds with two, 11% with three and 1% with four embryos. The embryonic axis constitutes 1.3% of the total mature embryo mass.

Storability of fresh embryos

After 10 days of storage, germination of fresh embryos stored at 5, 10 and 20oC decreased from 100% to 50, 30 and 0%, respectively. At 18 days, only the embryos stored at 5oC still showed some germination, albeit very low (5%) (results not shown).

23 Chapter 2

Figure 1. General aspects of Inga vera seeds. (a) Seed with sarcotesta; (b) after removal of the sarcotesta, the inner part of the seed coat is visible; (c) seed coat partially removed;

(d) embryo; (e) embryo opened, showing one cotyledon (Cot) and the embryonic axis (arrow); (f) detail of the axis, with the shoot (St) and root (Rt) apex, and the cotyledonary scar (Sc). Bar in (a) indicates 10 mm and is valid for (a)-(e). Bar in (f) indicates 5 mm.

100 3.1 Fresh weight 2.9

80 MC 2.7 2.5 60 2.3 Dry weight 2.1 O/g matter) dry

40 2 1.9

20 1.7

H (g MC Weight of 100 embryos (g) 1.5

0 1.3

0 6 7 8 9 10 11 12 13 14 15

Weeks after flowering

Figure 2. Variation in fresh and dry weight and in moisture content (MC) of Inga vera embryos during development (6, 9 and 10 weeks after flowering; WAF) and at shedding (14 WAF). 24 Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis

Germination, water relations and loss of viability during seed desiccation

To determine the best temperature for germination, mature embryos of I. vera (fresh or dried to different MCs) were put to germinate at eight temperatures, ranging from 5 to 40oC. Fresh embryos attained 100% germination in a wide range of temperature, from 15 to 40oC (Table 1 and Fig. 3), with faster germination occurring at 35 and 30oC. Under such temperatures, embryos started to germinate as soon as 12 hours after the beginning of the test and reached circa 100% within three days (Fig. 3). At 10oC the final germination of fresh embryos was also high (85%) (Table 1), although slower (from 9 to 47 days) (Fig. 3, inset). At 5oC germination was only 6% (Table 1). Desiccation of mature embryos from 1.38 to 1.0 g/g did not affect the final germination (Table 1), nor the germination speed (not shown). However, further removal of water from the embryos quickly reduced germination, particularly at lower temperatures (Table 1).

Temperature (oC): 15 20 25 30 35 40

100

80

100 60 10oC 75

40 50

(%) Germination 25 20 0 0 1020304050

0 01234567 Time (days)

Figure 3. Germination time courses of fresh, mature Inga vera embryos at different temperatures. Each data point is the mean of four replications of 20 seeds. Inset: o germination at 10 C.

25 Chapter 2

It is worth noting the high variation in MC among individual embryos in the same batch. In ten fresh, mature embryos, the MC ranged from 1.05 to 1.58 g/g, with an average of 1.38 g/g and a standard deviation of 3%. When dried to 0.75 g/g, the differences among embryos became even greater, ranging from 0.57 to 0.99 g/g, with a standard deviation of 4.2%. The optimal germination temperature (35oC) was used for germination tests of fresh and dried developing embryos. Embryos from seeds collected at 6 WAF attained 80% germination, which did not change when they were dried from 2.95 to 1.84 g/g. Further desiccation led to a drop in germination, with 0.43 g/g being the lethal moisture content, i.e. when all embryos died (Fig. 4). Seeds collected at 9 WAF showed similar behaviour, i.e. an initial germination of 81%, not affected by mild drying up to 1.25 g/g. Further removal of water decreased germination, with 0.43 g/g as the limit for the maintenance of some viability (Fig. 4). Seeds collected at 10 WAF attained 86% of germination, which is close to the values shown by the two previous developmental stages. However, germination increased to 99% with desiccation to 1.07 g/g. Continuation of drying led to a quick drop in germination, with the same lethal MC as observed for the other developmental stages (Fig. 4).

Table 1. Final germination percentages of mature Inga vera embryos with different moisture contents and at different temperatures.

Moisture content (g H2O/g dry matter)

Germination 1.38 1.00 0.75 0.59 0.39 temperature

5oC 6 0 0 0 0 10oC 85 90 0 0 0 15oC 98 98 35 0 0 20oC 95 100 64 13 3 25oC 98 100 69 15 0 30oC 100 100 71 19 1 35oC 100 100 69 23 3 40oC 100 100 64 18 0

A decreasing desiccation sensitivity with development was observed (Table 2). For instance, drying the embryos to 1.0 g/g reduced germination by 71% and 16%, for seeds collected at 6 and 10 WAF, respectively. For mature seeds, such desiccation had no effect on

26 Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis

germination. The MC at which germination of 50% of the embryos (G50) was observed, dropped from 1.17 to 0.69 g/g for seeds at 6 and 14 WAF, respectively (Table 2). Although the critical moisture content decreased with development, the lethal moisture content was the same (around 0.4 g/g) for all developmental stages studied (Fig. 4). For mature embryos the three dehydration conditions tested provided very different drying rates. Drying at fast, intermediate and slow rates decreased the MC from 1.38 g/g (initial MC) to circa 0.39 g/g (the lowest MC considered) after 27, 51 and 106 hours, respectively (Fig. 5a). However, germination was not influenced to any great extent by those different drying rates (Fig. 5b).

100

80

60

40

Germination (%)Germination

20

0 3.0 2.8 2.5 2.3 2.0 1.8 1.5 1.3 1.0 0.8 0.5 0.3

MC (g H O/g dry matter) 2

Figure 4. Final germination of developing and mature Inga vera embryos dried to various moisture contents. Seeds were collected at 6 (◊), 9 (), 10 (∆) and 14 (ο) weeks after flowering (WAF). Each data point is the mean of four replications of 20 seeds. Bars represent standard deviation. Germination was carried out at 35oC and constant light.

Embryos at different developmental stages dried under the same condition (35% RH; 20oC) exhibited similar rates of water loss (Fig. 6), although the water potential varied with the maturation stage (Fig. 7). The more mature the embryo, the greater (less negative) its water potential for the same MC. For instance at 1.0 g/g the water potentials of embryos with 6, 9 and 14 WAF were -9.0, -6.3 and -4.1 MPa, respectively (Fig. 7). The sorption isotherms of immature and mature embryos showed a hyperbolic shape (Fig. 8), which is in

27 Chapter 2

accord with isotherms of other desiccation-sensitive seeds and plant tissues, such as cacao (Leopold and Vertucci, 1986), germinated axes of soybean and leaves of Polypodium vulgare (Vertucci and Leopold, 1987a).

Table 2. Germination (G) of developing (6 and 10 WAF) and mature (14 WAF) Inga vera embryos at different moisture contents (MC). Some values for G at MC = 1.0 g H2O/g dry matter and MC at G50 were estimated roughly from Fig. 4.

Initial embryo G (%) G at MC = 1.0 g H O/g dry MC (g H O/g WAF Max 2 2 MC (g H2O/g dry matter (and % decrease in dry matter) matter) relation to GMax) at G50

6 2.95 85 25 (71) 1.17 10 1.99 99 83 (16) 0.89 14 1.43 100 100 (0) 0.69

WAF = weeks after flowering; GMax = maximal germination attained; G50 = 50% of the maximal germination attained.

Drying rate: Fast Intermediate Slow

1.4 100

1.2 80 1.0 60 0.8 40 O/g dry matter)

2 0.6 Germination (%) (%) Germination Moisture content 20 (g H 0.4 a b 0.2 0 0 20406080100 1.4 1.2 1.0 0.8 0.6 0.4 Time of desiccation (h) Moisture content (g H2O/g dry matter)

Figure 5. (a) Drying curves for mature embryos of Inga vera submitted to three desiccation conditions (fast: 10% RH; intermediate: 35% RH; and slow: 89% RH); (b) Germination of mature embryos at different MCs, reached by the three drying rates. Bars represent standard deviation.

28 Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis

2.8 6 WAF 9 WAF

2.4 10 WAF 14 WAF

2.0

1.6

O/g dry matter) dry O/g

2 1.2 Moisture content (g H

0.8

0.4

0 1020304050607080 Time of desiccation (h)

Figure 6. Drying rates of developing Inga vera embryos at 20oC and 35% RH. Seeds were collected at 6, 9, 10 and 14 weeks after flowering (WAF).

2.3 6 WAF 9 WAF 14 WAF

1.8 O/g dry matter) O/g 2

1.3

0.8

Moisture content (g H 0.3 0 -2 -4 -6 -8 -10 -12 -14 -16 -18 -20 -22 Water potential (MPa)

Figure 7. Water potentials of developing and mature Inga vera embryos dried to various moisture contents at 20oC and 35% RH. Seeds were collected at 6, 9 and 14 weeks after flowering (WAF). Each data point is the mean of four replications of 5-8 embryos. Lines were fitted by the following equations: 6 WAF: Y = 2.4681 e0.0995 X (R2 = 97%); 9 WAF: Y = 1.7321 e0.0875 X (R2 = 83%); 14 WAF: Y = 1.3461 e0.0718 X (R2 = 92%).

29 Chapter 2

3.0 3.0 6 WAF 9 WAF 2.5 2.5

2.0 2.0

1.5 1.5

1.0 1.0

0.5 0.5

O/g dry matter) 0.0 0.0 2 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

3.0 3.0 10 WAF 14 WAF 2.5 2.5 2.0 2.0

1.5 1.5

1.0 1.0

Moisture content (g H 0.5 0.5

0.0 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Water activity

Figure 8. Desorption isotherms of developing [6, 9 and 10 weeks after flowering (WAF)]

and mature (14 WAF) Inga vera embryos at 20oC.

DNA content

At shedding, 4C DNA content was 7.8 and 14.3% in the shoot and root apices, respectively (Fig. 9). Although quite different numerically, those averages are statistically comparable (t- test; P = 0.05), because of the high variation in the individual values within the ten shoots and the ten roots used (3.0 to 14.1 and 9.3 to 22.4%, respectively). Both 2C and 4C DNA contents remained constant in the shoot and in the root apex during the developmental period studied (Fig. 9). Upon germination (13 h of imbibition at 35oC), 4C DNA contents in the shoot and root apex were 11.2 and 17.5%, respectively, statistically still similar to values at shedding.

30 Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis

100

80

60

40

DNA content (%) 20

0 5 6 7 8 9 10 11 12 13 14 15 Weeks after flowering

Figure 9. 2C and 4C DNA content of embryonic axes cells of developing [6, 9 and 10 weeks after flowering (WAF)] and mature (14 WAF; at shedding) Inga vera seeds. Each data point is the mean of ten root ( , 2C; „, 4C) or shoot ({, 2C; z, 4C) apices. Bars represent standard deviation.

Microtubules

Embryonic axes of fresh mature I. vera seeds exhibited an extensive cortical microtubular cytoskeleton. In the root apex MTs appeared transversely oriented (Fig. 10a) while in the shoot apex they were less ordered (Fig. 10b). No mitotic configurations (preprophase band, spindle or phragmoplast) were found, indicating that at shedding there is no cell division. Slight desiccation of the embryos (from 1.38 to 1.0 g/g) had no effect on MT abundance (Figs. 10c and 10d), but further drying (to 0.75 g/g) caused a dramatic decrease in their abundance and the appearance of tubulin granules or clusters (Fig. 10e). The damage became even worse with drying to 0.59 g/g, which led to a complete disappearance of the microtubular cytoskeleton and a great reduction in the tubulin granules (Fig. 10f). When the embryos were dried to 0.39 g/g, no tubulin granules were observed (Fig. 10g). No differences were found between root and shoot regarding the rate of the collapse of MTs during dehydration. No mitotic configurations of MTs were found in radicle cells of fresh axes upon protrusion (Fig. 10h), indicating that cell division is not required for completion of germination.

31 Chapter 2

32 Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis

Discussion

Fresh and dry masses of I. vera embryos continued to accumulate until shedding. This is a common feature of recalcitrant seeds (Pammenter and Berjak, 1999). In the eight weeks of development analysed (from 6 to 14 WAF), the gain in fresh and dry mass of the whole embryo was more intense during the last four weeks. Axes also accumulated fresh and dry mass until shedding, but the gain was more intense in the first half of the period analysed. Since fresh and dry mass of the embryos accumulated until shedding, the decline in water content observed during development was not due to water loss, but the result of dry mass accumulating more than water (Berjak and Pammenter, 2000). The almost total loss of germinability of mature I. vera embryos after only 18 days of storage at 5oC is consistent with the poor storability of recalcitrant seeds in general (Roberts, 1973; Chin, 1989) and the remarkably short lifespan reported for seeds of other Inga species (Bilia et al., 1999). Although the storage conditions are hardly the same among the several storage trials reported in the literature for various recalcitrant species, the present results might rank I. vera among the species with the shortest storability known. In the present study, microbial proliferation could not account for the loss of viability of the embryos during storage, since it did not occur. One of the possible causes is that, during storage, hydrated seeds remain metabolically active with extensive vacuolation and increase in size of cells. There would be a necessity for additional water, which is not supplied, resulting in increasing water stress to the seeds (Pammenter et al., 1994). When dried at 35% RH, developing and mature I. vera embryos showed similar rates of water loss. Different situations have been reported for other species, such as Landolphia kirkii and Camellia sinensis, which displayed increasing drying rates with development (Berjak et al., 1992; 1993), and Avicennia marina, with opposite behaviour (Farrant et al., 1993). It appears that there is no relation between drying rate and developmental stages of recalcitrant seeds. Interestingly, drying rates of I. vera were not influenced by the increasing embryo water potential verified during development.

Figure 10 (opposite page). Fluorescent micrographs of Inga vera embryonic axes cells, labelled with antibody to β-tubulin and fluorescent conjugated secondary antibody. (a) Root and (b) shoot apex of non-dried embryos (1.38 g H2O/g dry matter), showing abundant cortical MTs, transversely and randomly oriented, respectively; germination is 100%; (c) Root and (d) shoot apex of embryos dried to 1.0 g/g, showing abundant cortical MTs, similar to the control; germination is also 100%; (e) Root apex of embryos dried to

0.75 g/g. Very few MTs can be seen. Tubulin granules (arrows) appeared, as a result of MT breakdown. At this point, germination dropped to 69%; (f) Root apex of embryos dried to 0.59 g/g. MTs totally disappeared and very few tubulin granules (arrows) can still be seen; germination is only 23%; (g) At 0.39 g/g, neither MTs nor tubulin granules can be visualized; here, germination is close to zero; (h) Root apex upon radicle protrusion (non- dried embryos) showing only cortical MTs. Bars indicate 25µm.

33 Chapter 2

Seeds of many tropical recalcitrant species cannot be cooled below about 10-15oC without causing chilling injury (Roberts and Ellis, 1989). The high germination attained at 10oC (85-90%) by fresh or slightly dried I. vera embryos (Table 1) indicates that the species is quite resistant to low temperatures, at least to 10oC, since at 5oC germination was practically absent. Similar results were found by Pritchard et al. (1995) for four Inga species, which showed high germination at 11oC, but very low germination or none at all at 8oC. In seeds of Shorea roxburghii, chilling injury also occurs at between 5 and 10oC (Corbineau and Côme, 1988). The decrease in desiccation sensitivity with development, shown by I. vera seeds, appears to be the normal feature for recalcitrant seeds, since it has been reported for many other species, including Camellia sinensis (Berjak et al., 1993), Acer pseudoplatanus (Hong and Ellis, 1990) and Quercus robur (Finch-Savage, 1992). Increasing desiccation-sensitivity with maturation seems to be rare and has been reported only for seeds of Landolphia kirkii (Berjak et al., 1992). It should be stressed, however, that even with the decrease in desiccation sensitivity during development, I. vera seeds are still highly desiccation sensitive at shedding. The observed increase in germinability of developing I. vera seeds after slight drying has also been reported for developing seeds of other species, e.g. Quercus robur (Erdey et al., 2003) and Ekebergia capensis (Pammenter et al., 1998; Erdey et al., 2003) and has been attributed to a continuation of maturation processes (Pammenter and Berjak, 1999). The critical moisture content of mature embryos of I. vera, between 0.75 and 1.0 g/g, corresponding to 1.08 and 1.50 g/g for the embryonic axes, is comparable to that found by Pritchard et al. (1995) for embryonic axes of five Inga species (1.00 to 1.22 g/g) and higher than the values for axes of Quercus robur (0.82 g/g) (Poulsen and Eriksen, 1992) and Q. rubra (0.47 to 0.82 g/g) (Pritchard, 1991).

The critical MC for mature I. vera embryos was reached while values of aw were still high, between 0.97 and 0.94, which correspond to water potentials of -4.0 to -7.7 MPa. These values fall into (or partially overlap) the intervals reported for recalcitrant seeds as being the critical MCs (0.30 and 1.60 g/g), and their corresponding water activities (0.96- 0.98) and water potentials (-1.5 to -5.0 MPa) (Hong and Ellis, 1996 and references therein). The lethal MC, both in immature and in mature embryos of I. vera, circa 0.43 g/g, was very close to the results reported by Pritchard et al. (1995) for several Inga species.

In the present study, total loss of viability occurred at an aw around 0.88. In such a high humidity range (above 0.60 aw), the water is considered to be free water, which is easily removed during drying (Vertucci and Leopold, 1987a, b; Grabe, 1989). It has already been shown that loss of viability in recalcitrant seeds occurs at hydration levels far higher than those at which water would be removed from cellular membranes and other macromolecular structures (Pammenter et al., 1991; Berjak et al., 1992). The water that is associated (strongly bound) with molecular surfaces, also called non-freezable or structural

34 Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis

water is found at a very low humidity range (0-0.2 aw) (Vertucci and Leopold, 1987a, b; Berjak et al., 1993). It has been suggested that the drastic drop in water at maturation drying might inhibit DNA synthesis, resulting in accumulation of cells at the pre-synthetic G1 phase (Deltour, 1985). However, I. vera seeds, which do not undergo maturation drying, are shed with most of the cells (81%) in the root apex arrested at the G1 phase with nuclei containing 2C DNA. Castanea sativa, another recalcitrant species, is also shed with most of the cells

(94%) in the root tip arrested at the G1 phase (Bino et al., 1993). In root meristems of seeds of the highly recalcitrant species Avicennia marina, Boubriak et al. (2000) have shown that DNA synthesis decreased during development, ceased at shedding, but was resumed 48 h later, whether or not the seeds had access to exogenous water. The authors suggested that this temporary replication arrest at the time of shedding is, in fact, an extended interphase. In seeds of neem (Azadirachta indica) at 10 weeks after pollination, when full germinative capacity is attained, the MC is still high (1.04 g/g), but 4C DNA content is only 6% (Sacandé, 2000). In a study with seeds of Norway maple (Acer platanoides; orthodox) and sycamore (A. pseudoplatanus; recalcitrant), which are shed with a MC of 1.17 and 0.35 g/g, respectively, Finch-Savage et al. (1998) found no difference in the 4C DNA content in radicle tips of mature seeds of both species (37 and 38%, respectively). All these results strongly support the suggestion of Bino et al. (1993), that the arrest of the cell cycle in G1 is not directly linked to the seed water status. From 6 WAF onwards, we observed neither DNA replication activity nor mitosis in the root apex of I. vera embryos; 2C and 4C DNA contents were constant. This suggests that there are two blocks in the cell cycle: one at G2, which prevents the progress of cells with 4C nuclei through mitosis, and another at G1, which keeps the cells at pre-synthetic phase. The abundant cortical microtubules in cells of fresh axes of I. vera might be an indication that, at shedding, those cells are metabolically active. The different orientation of MTs in cells of root (transverse) and shoot (no predominant orientation) could be related to their differential growth rate and patterning upon germination, with root growth occurring before shoot elongation. Transversely oriented MTs enable root cells to prompt elongation, while less-ordered MTs in the shoot cells would need to be rearranged before cell elongation. In elongating cells, cortical MTs are generally transverse to the direction of cell expansion (Yuan et al., 1994) and act by channeling cellulose-synthesizing complexes across the plasma membrane (Lloyd, 1994). During dehydration of I. vera embryos, the microtubular cytoskeleton became disrupted at around -4.5 MPa, similar to the values reported for seeds of Trichilia dregeana (- 3.8 MPa) (Berjak et al., 1999) and Quercus robur (-3.5 MPa) (Mycock et al., 2000). There is a general trend towards contraction or dismantling of organelles during water stress to about -5 MPa (Walters et al., 2002). Under normal conditions, the tubulin released from the disassembly (depolymerization) of microtubules is not catabolized and returns to the cytoplasmic pool,

35 Chapter 2

maintaining an equilibrium between MTs and free tubulin (Dustin, 1978). In the present study the appearance of tubulin granules was observed during drying of the embryos as a consequence of the breakdown of MTs. However, it appears that the severity of the desiccation not only disassembled the MTs, but also led to a degradation of the tubulin. Tubulin granules were also observed in dried seeds of tomato (de Castro et al., 2001) and coffee (da Silva, 2002). Breakdown of MTs and the subsequent inability of reassembly, may be regarded as one of the causes of desiccation sensitivity (Berjak and Pammenter, 2000). The relation between MT dismantling and cell death during desiccation has been attributed to the loss of functional biochemical pathways associated with the cytoskeleton (Mycock et al., 2000). Although in plants the cytoskeleton is more involved with intracellular transport and cell division and less in the maintenance of cell shape, which is guaranteed by the relatively rigid cell walls (Goddard et al, 1994), the cellular collapse during desiccation has also been linked to damage to the cytoskeleton (Berjak and Pammenter, 2000). As in every biotic or abiotic stress, it is plausible that there is a limit to the magnitude of the damage to the MTs caused by seed desiccation, which can still be reverted by the cells when seeds are put to germinate. This could explain why 69% of the I. vera embryos dried to 0.75 g/g could still germinate. Hence, the damage in these embryos was not too severe, as it was in the other 31% of the population. The most prominent reason for this variation is the great difference in MC among individual embryos in a same batch, which leads to different drying rates and different MCs at the end of the drying period. Our data show that axes from fresh (non-dried) seeds with 100% germination display a well-developed MT cortical cytoskeleton, whereas seeds dried to the lethal MC exhibit neither MTs nor tubulin granules (cf. Figs. 10a and 10g). Regarding the intermediate condition, the almost complete disappearance of MTs in embryonic axes cells of embryos dried to 0.75 g/g (Fig. 10e) coincided with a drop from 100 to 69% germination, and was observed in all axes analyzed for that treatment. The probability that only non-germinable axes have been sampled is less than 1%, suggesting therefore that the damage to the MTs, although severe (Fig. 10e), could still be reverted when seeds were rehydrated. Using the same reasoning, it is probable that the absence of MTs and reduction of tubulin granules in cells of embryos dried to 0.59 g/g (23% germination) (Fig. 10f) represent seeds that would not germinate. Thus, the visible threshold for the reversibility of the damage caused to the MTs during drying of I. vera embryos is probably between the situations shown in Fig. 10e and in Fig. 10f. It is clear that there is a strong correlation between disappearance of MTs and loss of viability during drying of I. vera embryos, suggesting that, upon rehydration, severely damaged cells are no longer able to rebuild the microtubular cytoskeleton. Although very few studies have been devoted to desiccation damage of the cytoskeleton in recalcitrant seeds (e.g. Mycock et al., 2000; Gumede et al., 2003), it seems that this might be a general feature among such seeds.

36 Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis

Acknowledgements

We thank CNPq (National Council for Scientific and Technological Development, from the Ministry of Science and Technology, Brazil) for financial support of the studies of J.M.R. Faria. Henk Kieft (Laboratory of Plant Cell Biology, Wageningen University) is acknowledged for his assistance with the microscopic studies. Jan Bergervoet and Jeroen Peters from Plant Research International (PRI, Wageningen) are acknowledged for their support with the flow cytometry. We also thank Antonio Claudio Davide, Edvaldo Amaral da Silva, José Carlos Martins and José Pedro de Oliveira, of the Lavras Federal University, Brazil, for seed collecting and handling, and shipping to The Netherlands.

References

Alberts, B., Bray, D., Johnson, A., Lewis, J., Raff, M., Roberts, K. and Walter, P. (1998) Essential cell biology: an introduction to the molecular biology of the cell. Garland Publishing, Inc. New York. Arumuganthan, K. and Earle, E.D. (1991) Estimation of nuclear DNA content of plants by flow cytometry. Plant Molecular Biology Report 9: 229-233. Baskin, T.I., Busby, C.H., Fowke, L.C., Sammut, M. and Gubler, F. (1992) Improvements on immunostaining samples embedded in methacrylate: localization of microtubules and other antigens throughout developing organs in plants of diverse taxa. Planta 187, 405-413. Berjak, P. and Pammenter, N.W. (1994) Recalcitrance is not an all-or-nothing situation. Seed Science Research 4, 263-264. Berjak, P. and Pammenter, N.W. (2000) What ultrastructure has told us about recalcitrant seeds. Revista Brasileira de Fisiologia Vegetal 12 (Edição Especial): 22-55. Berjak, P., Pammenter, N.W. and Vertucci, C. (1992) Homoiohydrous (recalcitrant) seeds: developmental status, desiccation sensitivity and the state of water of Landolphia kirkii Dyer. Planta 186, 249-261. Berjak, P., Vertucci, C.W. and Pammenter, N.W. (1993) Effects of developmental status and dehydration rate on characteristics of water and desiccation-sensitivity in recalcitrant seeds of Camellia sinensis. Seed Science Research 3, 155-166. Berjak, P., Kioko, J., Walker, M., Mycock, D.J., Wesley-Smith, J., Watt, P. and Pammenter, N.W. (1999) Cryopreservation: an elusive goal, in IUFRO Seed Symposium 1998. Recalcitrant Seeds. Proceedings of the Conference (eds M. Marzalina et al.), Forestry Research Institute Malaysia, Kuala Lumpur, pp. 96-109. Bewley, J.D. and Black, M. (1994) Seeds. Physiology of Development and Germination, 2nd edn, Plenum Press, New York. Bilia, D.A.C., Marcos Filho, J. and Novembre, A.D.C.L. (1999) Desiccation tolerance and seed storability of Inga uruguensis (Hook. et Arn.). Seed Science and Technology 27, 77-89. Bino, R.J., Lanteri, S., Verhoeven, H.A. and Kraak, H.L. (1993) Flow cytometric determination of nuclear replication stage in seed tissues. Annals of Botany 72, 181-187.

37 Chapter 2

Boubriak, I., Dini, M., Berjak, P. and Osborne, D.J. (2000) Desiccation and survival in the recalcitrant seeds of Avicennia marina: DNA replication, DNA repair and protein synthesis. Seed Science Research 10, 307-315. Bouvier-Durand, M., Real, M. and Côme, D. (1989) Changes in nuclear activity upon secondary dormancy induction by abscisic acid in apple embryos. Plant Physiology and Biochemistry 27, 511-518. Chin, H.F. (1989) Recalcitrant seeds. Extension Bulletin no. 288. ASPAC: Food and Fertilizer Technology Center, Taipei City, Taiwan. Chin, H.F., Krishnapillay, B. and Stanwood, P.C. (1989) Seed moisture: recalcitrant vs. orthodox seeds, in Seed Moisture. CSSA Special Publication No. 14. (eds P.C. Stanwood and M.B. McDonald), Crop Science Society of America, Madison, USA, pp. 15-22. Corbineau, F. and Côme, D. (1988) Storage of recalcitrant seeds of four tropical species. Seed Science and Technology 16, 97-103. da Silva, E.A.A. (2002) Coffee seed (Coffea arabica cv. Rubi) germination: mechanism and regulation. PhD Thesis, Wageningen University, The Netherlands. Davide, A.C., Faria, J.M.R. and Botelho, S.A. (1995) Propagação de espécies florestais. Technical Bulletin. CEMIG/UFLA/FAEPE, Belo Horizonte, Brazil. de Castro, R.D., Bino, R.J., Jing, H.C., Kieft, H. and Hilhorst, H.W.M. (2001) Depth of dormancy in tomato (Lycopersicon esculentum Mill.) seeds is related to the progression of the cell cycle prior to the induction of dormancy. Seed Science Research 11, 45-54. Deltour, R. (1985) Nuclear activiation during early germination of the higher plant embryo. Journal of Cell Science 75, 43-83. Dussert, S., Chabrillange, N., Engelmann, F. and Hamon, S. (1999) Quantitative estimation of seed desiccation sensitivity using a quantal response model: application to nine species of the genus Coffea L. Seed Science Research 9, 135-144. Dustin, P. (1978) Microtubules. Springer-Verlag, Berlin. Ellis, R.H., Hong, T.D. and Roberts. E.H. (1990) An intermediate category of seed storage behaviour? I. Coffee. Journal of Experimental Botany 41, 1167-1174. Erdey, D.P., Pammenter, N.W., Finch-Savage, W.E. and Berjak, P. (2003) Physiological assessments of vigour enhancement by mild dehydration stress in recalcitrant seeds: effect of seed maturity, in Proceedings of the 4th International Workshop on Desiccation Tolerance and Sensitivity of Seeds and Vegetative Plant Tissues. Blouwaterbaai, South Africa, p. 16. Farnsworth, E. (2000) The ecology and physiology of viviparous and recalcitrant seeds. Annual Review of Ecology and Systematics 31, 107-138. Farrant, J.M., Berjak, P. and Pammenter, N. (1985) The effect of drying rate on viability retention of recalcitrant propagules of Avicennia marina. South African Journal of Botany 51, 432-438. Farrant, J.M., Pammenter, N.W. and Berjak, P. (1993) Seed development in relation to desiccation tolerance: a comparison between desiccation-sensitive (recalcitrant) seeds of Avicennia marina and desiccation-tolerant types. Seed Science Research 3, 1-13. Finch-Savage, W.E. (1992) Embryo water status and survival in the recalcitrant species Quercus robur L.: evidence for a critical moisture content. Journal of Experimental Botany 43, 663-669.

38 Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis

Finch-Savage, W.E., Bergervoet, J.H.W., Bino, R.J., Clay, H.A. and Groot, S.P.C. (1998) Nuclear replication activity during seed development, dormancy breakage and germination in three tree species: Norway Maple (Acer platanoides L.), Sycamore (Acer pseudoplatanus L.) and Cherry (Prunus avium L.). Annals of Botany 81, 519-526. Goddard, R.H., Wick, S.M., Silflow, C.D. and Snustad, D.P. (1994) Microtubule components of the plant cell cytoskeleton. Plant Physiology 104, 1-6. Grabe, D.F. (1989) Measurement of seed moisture, in Seed Moisture. CSSA Special Publication No. 14. (eds P.C. Stanwood and M.B. McDonald), Crop Science Society of America, Madison, USA, pp. 69- 92. Gumede, Z., Merhar, V. and Berjak, P. (2003) Effect of desiccation on the microfilament component of the cytoskeleton in zygotic embryonic axes of Trichilia dregeana Sond, in Proceedings of the 4th International Workshop on Desiccation Tolerance and Sensitivity of Seeds and Vegetative Plant Tissues. Blouwaterbaai, South Africa, p. 22. Hong, T.D. and Ellis, R.H. (1990) A comparision of maturation drying, germination, and desiccation tolerance between developing seeds of Acer pseudoplatanus L. and Acer platanoides L. New Phytologist 116, 589-596. Hong, T.D. and Ellis, R.H. (1992) Optimum air-dry storage environments for arabica coffee. Seed Science and Technology 20, 547-560. Hong, T.D. and Ellis, R.H. (1996) A protocol to determine seed storage behaviour. IPGRI Technical Bulletin No. 1. International Plant Genetic Resources Institute, Rome. ISTA (International Seed Testing Association) (1996) International rules for seed testing. Seed Science and Technology 24 (suppl.). King, M.W. and Roberts, E.H. (1979) The Storage of Recalcitrant Seeds: Achievements and Possible Approaches. International Board for Plant Genetic Resources, Rome. Köppen, W. (1936) Das geographische system der climate, in Handbuch der Klimatologie, Vol. 1, part C. (eds W. Köppen and R. Geiger), Gebrüder Borntraeger, Berlin, pp. 1-44. Kumagai, F. and Hasezawa, S. (2001) Dynamic organization of microtubules and microfilaments during cell cycle progression in higher plant cells. Plant Biology 3, 4-16. Leopold, A.C. and Vertucci, C.W. (1986) Physical attributes of desiccated seeds, in Membranes, Metabolism and Dry Organisms (ed. A. C. Leopold), Comstock Publishing, London, pp. 23-34. Liang, Y. and Sun, W.Q. (2000) Desiccation tolerance of recalcitrant Theobroma cacao embryonic axes: the optimal drying rate and its physiological basis. Journal of Experimental Botany 51, 1911-1919. Lloyd, C. (1994) Why should stationary plant cells have such dynamic microtubules? Molecular Biology of the Cell 5, 1277-1280. Mycock, D.J., Berjak, P. and Finch-Savage, W.E. (2000) Effects of desiccation on the subcellular matrix of the embryonic axes of Quercus robur, in Seed biology: Advances and Applications (eds M. Black, K.J. Bradford and J. Vazquez-Ramos), CABI Publishing, Wallingford, pp. 197-203. Oliver, M.J. (1996) Desiccation tolerance in vegetative plant cells. Physiologia Plantarum 97, 779-787. Pammenter, N.W. and Berjak, P. (1999) A review of recalcitrant seed physiology in relation to desiccation-tolerance mechanisms. Seed Science Research 9, 13-37. Pammenter, N.W. and Berjak, P. (2000) Aspects of recalcitrant seed biology. Revista Brasileira de Fisiologia Vegetal 12, 56-69.

39 Chapter 2

Pammenter, N.W., Vertucci, C.W. and Berjak, P. (1991) Homeohydrous (recalcitrant) seeds: dehydration, the state of water and viability characteristics in Landolphia kirkii. Plant Physiology 96, 1093-1098. Pammenter, N.W., Berjak, P., Farrant, J.M., Smith, M.T. and Ross, G. (1994) Why do stored hydrated recalcitrant seeds die? Seed Science Research 4, 187-191. Pammenter, N.W., Greggains, V., Kioko, J.I., Wesley-Smith, J., Berjak, P. and Finch-Savage, W.E. (1998) Effects of differential drying rates on viability retention of recalcitrant seeds of Ekebergia capensis. Seed Science Research 8, 463-471. Pammenter, N.W., Berjak, P. and Walters, C. (2000) The effect of drying rate on recalcitrant seeds: ‘lethal water contents’, causes of damage, and quantification of recalcitrance, in Seed biology: Advances and Applications (eds M. Black, K.J. Bradford and J. Vazquez-Ramos), CABI Publishing, Wallingford, pp. 215-221. Pennington, T.D. (1997) The genus Inga: Botany. The Royal Botanic Gardens, Richmond. Poulsen, K.M. and Eriksen, E.A.N. (1992) Physiological aspects of recalcitrance in embryonic axes of Quercus robur L. Seed Science Research 2, 215-221. Pritchard, H.W. (1991) Water potential and embryonic axis viability in recalcitrant seeds of Quercus rubra. Annals of Botany 67, 43-49. Pritchard, H.W., Haye, A.J., Wright, W.J. and Steadman, K.J. (1995) A comparative study of seed viability in Inga species: desiccation tolerance in relation to the physical characteristics and chemical composition of the embryo. Seed Science and Technology 23, 85-100. Probert, R.J. and Longley, P.L. (1989) Recalcitrant seed storage physiology in three aquatic grasses (Zizania palustris, Spartina anglica and Porteresia coarctata). Annals of Botany 63, 53-63. Roberts, E.H. (1973) Predicting the storage life of seeds. Seed Science and Technology 1, 499-514. Roberts, E.H. and Ellis, R.H. (1989) Water and seed survival Annals of Botany 63, 39-52. Sacandé, M. (2000) Stress, storage and survival of neem seed. PhD Thesis, Wageningen Agricultural University, The Netherlands. Saraco, F., Bino, R.J., Bergervoet, J.H.W. and Lanteri, S. (1995) Influence of priming-induced nuclear replication activity on storability of pepper (Capsicum annuum L.) seed. Seed Science Research 5, 25-29. Sargent, J.A., Sen Mandi, S. and Osborne, D.J. (1981) The loss of desiccation tolerance during germination: an ultrastructural and biochemical approach. Protoplasma 105, 225-239. Sun, W.Q. (2002) Methods for the study of water relations under desiccation stress, in Desiccation and Survival in Plants: Drying without Dying (eds M. Black and H.W. Pritchard). Cabi Publishing, Wallingford, pp. 47-91. Sun, W.Q. and Gouk, S.S. (1999) Preferred parameters and methods for studying moisture content of recalcitrant seeds, in IUFRO Seed Symposium 1998. Recalcitrant Seeds. Proceedings of the Conference (eds M. Marzalina et al.), Forestry Research Institute Malaysia, Kuala Lumpur, pp. 404-430. Tompsett, P.B. (1984) Desiccation studies in relation to the storage of Araucaria seed. Annals of Applied Biology 105, 581-586. Uniyal, R.C. and Nautiyal, A.R. (1996) Physiology of seed development in Aesculus indica, a recalcitrant seed. Seed Science and Technology 24, 419-424.

40 Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis

Vantard, M., Cowling, R. and Delichère, C. (2000) Cell cycle regulation of the microtubular cytoskeleton. Plant Molecular Biology 43, 691-703. Vazquez-Ramos, J.M. and Sanchez, M. de la P. (2003) The cell cycle and seed germination. Seed Science Research 13, 113-130. Vertucci, C.W. and Leopold, A.C. (1987a) The relationship between water binding and desiccation tolerance in tissues. Plant Physiology 85, 232-238. Vertucci, C.W. and Leopold, A.C. (1987b) Water binding in legume seeds. Plant Physiology 85, 224- 231. Walters, C., Farrant, J.M., Pammenter, N.W. and Berjak, P. (2002) Desiccation stress and damage, in Desiccation and Survival in Plants: Drying without Dying (eds M. Black and H.W. Pritchard). Cabi Publishing, Wallingford, pp. 263-291. Wesley-Smith, J., Pammenter, N.W., Berjak, P. and Walters, C. (2001) The effects of two drying rates on the desiccation tolerance of embryonic axes of recalcitrant jackfruit (Artocarpus heterophyllus Lamk.) seeds. Annals of Botany 88, 653-664. Yuan, M., Shaw, P.J., Warn, R.M. and Lloyd, C.W. (1994) Dynamic reorientation of cortical microtubules, from transverse to longitudinal, in living plant cells. Proceedings of the National Academy of Sciences, USA 91, 6050-6053.

41

Chapter 3

Changes in DNA and microtubules during loss and re- establishment of desiccation tolerance in germinating Medicago truncatula seeds*

José Marcio Rocha Faria, Julia Buitink, André A. M. van Lammeren and Henk W. M. Hilhorst

Abstract

Desiccation-tolerance (DT) in orthodox seeds is acquired during seed development and lost upon imbibition/germination, purportedly upon the resumption of DNA synthesis in the radicle cells. In the present study, flow cytometric analyses and visualisation of microtubules (MTs) in radicle cells of seedlings of Medicago truncatula showed that up to a radicle length of 2 mm, there is neither DNA synthesis nor cell division, which were first detected in radicles with a length of 3 mm. However, DT started to be lost well before the resumption of DNA synthesis, when germinating seeds were dried back. By applying an osmotic treatment with polyethylene glycol (PEG) before dehydration, it was possible to re-establish DT in seedlings with a radicle up to 2 mm long. Dehydration of seedlings with a 2 mm radicle, with or without PEG treatment, caused disassembly of MTs and appearance of tubulin granules. Subsequent pre-humidification led to an almost complete disappearance of both MTs and tubulin granules. Upon rehydration, neither MTs nor tubulin granules were detected in radicle cells of untreated seedlings, while PEG-treated seedlings were able to reconstitute the microtubular cytoskeleton and continue their normal development. Dehydration of untreated seedlings also led to an apoptotic-like DNA fragmentation in radicle cells, while in PEG- treated seedlings DNA integrity was maintained. The results showed that for different cellular components, desiccation-tolerant seedlings may apply distinct strategies to survive dehydration, either by avoidance or further repair of the damages.

* This chapter has been published in Journal of Experimental Botany (2005) 56, 2119-2130. Chapter 3

Introduction

Desiccation tolerance (DT) in plants can be considered as the ability to rehydrate successfully after the removal of 80 to 90% of protoplasmic water, leading to moisture content (MC) below 0.3 g H2O/g dry matter (or 23% on a wet basis), when the hydration shell of molecules is lost (Oliver et al., 2000; Hoekstra et al., 2001). There are three criteria that a plant or plant structure must meet in order to survive such severe loss of protoplasmic water: (i) limitation of the damage suffered by the cells during desiccation to a repairable level; (ii) maintenance of its physiological integrity in the dry state; and (iii) mobilisation of repair mechanisms upon rehydration aiming to revert the damages caused by desiccation and/or rehydration (Bewley, 1979). In orthodox seeds, DT is acquired during seed development, enabling them to withstand maturation drying, when more than 90% of the water may be lost (Adams and Rinne, 1980). DT is maintained after shedding, allowing further drying of seeds, when MC may be diminished to circa 0.05 g H2O/g dry matter, without loss of viability. When dried seeds are imbibed, DT remains unchanged for some time, so they can be dried back to their original MC without irreversible damage. However, if seeds are allowed to imbibe longer, DT is gradually lost. The point at which DT starts to be lost varies among species if analysed in terms of imbibition time (Senaratna and McKersie, 1983, 1986; Hong and Ellis, 1992; Reisdorph and Koster, 1999; Koster et al., 2003; Ren and Tao, 2003) or protruded radicle length (Lin et al., 1998; Leprince et al., 2000; Pukacka, 2001; Buitink et al., 2003). However, if germination is assessed regarding the activation of the cell cycle, loss of DT coincides, irrespective of species, with the resumption of cell division (Berrie and Drennan, 1971; Osborne et al., 2002) or, more frequently, DNA synthesis (Sargent et al., 1981; Dasgupta et al., 1982; Deltour, 1985; Osborne and Boubriak, 1994; Osborne, 2000; Boubriak et al.,

2000). It has been shown that cells in the G2 phase of the cell cycle, with duplicated DNA, are more sensitive to stress than cells that are still in the G1, pre-synthetic phase (Deltour, 1985; Sliwinska, 2003). It has been suggested, however, that although DNA replication is a suggestive developmental marker for the loss of DT in germinating seeds, it is not necessarily the reason (Sargent et al., 1981). Another component of the cell cycle that may be involved in the loss of DT during germination is the microtubular cytoskeleton, which is markedly sensitive to desiccation stress (Sargent et al., 1981). Microtubules (MTs) are elongated tubular structures, made of α and β-tubulin, which in plant cells play an important part in cell elongation, determination of the division site, chromosome separation and cytokinesis (Alberts et al., 2002; Wasteneys and Galway, 2003). In desiccation-sensitive seeds MTs can be irreversibly deranged by dehydration (Berjak and Pammenter, 2000). Microtubule reassembly is among the cellular repair processes that have been linked to DT (Oliver, 1996; Mycock et al., 2000).

44 DNA and microtubules during loss and re-establishment of DT in Medicago truncatula seeds

The maintenance of the genetic information carried by the DNA is essential to cell survival upon dehydration and rehydration (Osborne et al., 2002). The stability of the DNA on dehydration and the ability for its repair on rehydration is a prominent feature displayed by desiccation-tolerant seeds (Boubriak et al., 1997). Seeds that are allowed to germinate until (or beyond) the point where they become desiccation-sensitive, can experience irreversible DNA degradation when subjected to dehydration. When cells are confronted with environmental stresses, they can either die passively (accidental cell death) or can self- destruct (programmed cell death), depending on the stress type and intensity (Danon et al., 2000). Self-destruction is orchestrated by an active mechanism known as apoptosis, the biochemical hallmark of which is the cleavage of the DNA at internucleosomal sites by endonucleases, generating oligonucleosomal fragments. This DNA fragmentation can be detected by the formation of DNA ladders on agarose gels (Stein and Hansen, 1999). For cell death processes other then apoptosis a smear of broken DNA, instead of a clear fragmentation pattern is seen (Wang et al., 1998). The genetic information for DT is certainly present in the genome of plants that bear orthodox seeds. In those plants, the restriction of DT to specific stages of seed development is due to differences in the control of gene expression (Bartels and Salamini, 2001). The cellular protection system exhibited by those seeds may be induced in vegetative tissues by environmental clues related to drying (Oliver et al., 2000). The feasibility of the re- establishment of DT in seedlings originated from orthodox seeds by applying an osmotic stress, as shown by Bruggink and van der Toorn (1995) and Buitink et al. (2003) has appeared as an outstanding tool for studies on the mechanisms of desiccation tolerance and sensitivity in seeds. By using such an approach, the present study aimed to investigate the relationship of the DNA (relative content and integrity) and microtubule configurations with the loss and re-establishment of desiccation tolerance in germinating seeds of Medicago truncatula Gaertn. cv. Jemalong A17.

Material and Methods

Plant material

Medicago truncatula Gaertn. cv. Jemalong A17 plants were routinely grown in an environmentally controlled growth chamber (16/8 h photoperiod; 170 µmol/m2/s; 25oC; 60% RH). Mature pods were collected at shedding, around 30 d after flowering, stored at 20oC (Journet et al., 2001) and seeds were extracted manually when needed.

45 Chapter 3

Dormancy release and seed germination

Medicago truncatula seeds exhibit a combination of physical (coat-imposed) and physiological dormancy, with the latter lasting for three to four months following pod abscission (Journet et al., 2001). In order to overcome these dormancies, seeds were chemically scarified by immersing them in concentrated sulphuric acid for 5 to 10 min and subjected to cold imbibition (36 h at 4oC), in the dark, in Petri dishes (9 cm diameter; 250 seeds per dish) with two filter papers (No. 595, Schleicher & Schuell, Germany) moistened with 9 mL distilled water. During cold imbibition Petri dishes were kept shaking on a benchtop shaker (70 rpm). Seeds were transferred to new Petri dishes (9 cm diameter; 50 seeds/dish) with two filter papers moistened with 2.5 mL of distilled water and kept in the dark at 20oC (modified from Sieberer et al., 2002). Seeds that showed radicle protrusion were considered germinated. To characterise the dormancy release better, the germination of 2-month-old seeds treated with sulphuric acid plus cold imbibition was compared with seeds subjected to only one treatment (acid scarification or cold imbibition) as well as to untreated (control) seeds, in three replications of 50 seeds.

Moisture content (MC) determination

Moisture content was assessed in four replications of 10 seeds (or radicles), by oven drying at 103oC for 17 h, according to the International Seed Testing Association (ISTA, 1996). MC is expressed on a dry weight basis, i.e. in g H2O/g dry matter or simply g/g.

DNA content assessment

Relative DNA content assessment was done by flow cytometry, using suspensions of intact nuclei prepared from radicles excised from dry seeds or seedlings with increasing protruded radicle length (1, 2, 3 and 4 mm). Only the tip (1 mm) of the radicles, which includes the root cap and the meristematic region, was used in the analyses. Each treatment consisted of five replications of 10 radicle tips. Sample preparation was done according to Arumuganthan and Earle (1991) and analyses were performed with a flow cytometer (EPICS XL-MCL, Beckman-Coulter, Miami, FL, USA) equipped with an argon ion laser at 488 nm. Histograms were processed using ModFit LT (Verity Software House, Topsham, ME, USA) for data analysis and correction of the background noise. In each replication 10,000 nuclei were analysed. Statistical analyses were performed with the software SPSS 11.0.1.

46 DNA and microtubules during loss and re-establishment of DT in Medicago truncatula seeds

Assessment of the loss of desiccation tolerance during germination

Seeds were put to germinate as described previously, germination was scored at various times and then seeds/seedlings were dehydrated. Dehydration was done over a saturated o solution of K2CO3 (43% RH) in a closed box with circulating air at 23 C for 3 d, based on Buitink et al. (2003). After dehydration seeds were pre-humidified in humid air (100% RH) o for 24 h at 20 C to avoid imbibitional damage and then rehydrated (2.5 mL H2O/Petri dish 9 cm; two filter papers; 20oC; in the dark). Seeds that germinated and seedlings that continued their normal development were considered desiccation-tolerant. The data of DT (%) were normalized to the maximum germination attained (%) in the germination test. The experiment was replicated three times with 50 seeds being used each time that germination and DT were evaluated.

Re-establishment of desiccation tolerance

To assess the re-establishment of DT in seedlings, they were selected by their radicle length (1, 2, 3, 4 and 5 mm) using a dissection microscope and a metallic ruler with 0.5 mm scale divisions and either dried directly or after 3 d of incubation in a polyethylene glycol (PEG)

6000 solution (355 g PEG dissolved in 1.0 L H2O). Incubation was done in the dark, in 9 cm Petri dishes containing two filter papers wetted with 7 mL of PEG solution. The incubation temperature of 10oC which gives a water potential of -1.7 MPa, used by Buitink et al. (2003) for the re-establishment of DT in seedlings of M. truncatula cv. Paraggio, was not suitable for the cultivar Jemalong A17 used in the present study because it could not inhibit protruding radicles from continuing growth during incubation. The problem was solved by carrying out the incubation at 5oC. By decreasing the temperature, the water potential of the PEG solution was slightly lowered to -1.8 MPa. After incubation, seedlings were rinsed thoroughly in distilled water and then dehydrated, pre-humidified and rehydrated as described before. Seedlings that resumed normal growth after rehydration were considered desiccation- tolerant. Four independent experiments with 50 seedlings each were carried out.

Viability test (tetrazolium test)

Seedlings with a radicle length of 2 mm were dehydrated (with or without previous PEG treatment), pre-humidified, and incubated in a 1% (w/v) solution of 2,3,5-tryphenyl tetrazolium chloride (Merck, Darmstadt, Germany), at 20oC for 18 h in the dark. Stained tissues were considered viable, and unstained white tissues were considered dead (ISTA, 1996). The test was done using three replications of 50 seedlings per treatment.

47 Chapter 3

DNA isolation and electrophoresis to assess DNA fragmentation

Chromosomal DNA was extracted from 2 mm long radicles of seedlings (control and dehydrated with and without PEG treatment) and isolated following a protocol modified from Liu et al. (1995). Approximately 40 mg of radicles from hydrated (control) and 20 mg from dehydrated (PEG-treated and untreated) seedlings were ground to a fine powder with a mortar and pestle in liquid nitrogen and mixed with the extraction buffer (0.6 mL NaCl, 100 mM TRIS-HCl pH 7.5, 40 mM EDTA, 4% sarkosyl, and 1% SDS) previously diluted with urea and phenol. Phenol:chloroform was added, the mixture was centrifuged and the aqueous phase collected and mixed with isopropanol. DNA was precipitated by inverting the tubes a few times. After incubation for 10 min at room temperature and centrifugation, the pellet was washed with 80% ethanol and dissolved in TRIS-EDTA pH 8.0 containing RNase A. Samples (5 µg/lane) of DNA were loaded on a 1% agarose gel stained with ethidium bromide.

Immunohistochemical detection of the microtubular cytoskeleton

Radicles were excised from seeds/seedlings, both before imbibition (dry seeds) and after germination, with lengths of 1, 2, 3 and 4 mm. Radicles (2 mm long) were also excised from seedlings after each of the following steps, with or without PEG treatment: dehydration, pre- humidification, and rehydration. The type of fixation used depended on the moisture content of the tissue. Radicles excised from dry seeds/seedlings (both before imbibition and after germination followed by dehydration) were chemically fixed in water-free methanol + 0.1% glutaraldehyde for 4 h, at 20oC. Radicles excised from undried seedlings were plunged into liquid propane (cooled down in liquid nitrogen) and transferred to cryo-tubes containing frozen freeze-substitution medium (water-free methanol + 0.1% glutaraldehyde), also cooled down in liquid nitrogen. The cryo-tubes were put in a freeze substitution unit (FreasySub, Cryotech Benelux, Schagen, The Netherlands) for 78 h. The next steps were similar for both chemically- and cryo-fixed samples. The fixative medium was replaced by ethanol (series of increasing concentrations), followed by embedding in butylmethylmethacrylate (BMM) and UV polymerization at -20oC for 48 h, according to Baskin et al. (1992). Five roots were analysed for each treatment. Longitudinal sections (3 µm thick) were placed on slides and BMM was removed by washing in acetone. The slides were then rinsed in phosphate buffered saline (PBS) and sections were blocked in 100 mM hydroxyl tetra ammonium chloride (HAC) and in 26 mM bovine serum albumin (BSA). Next, they were incubated with the primary antibody (mouse anti-α-tubulin [Sigma, Zwijndrecht, The Netherlands], diluted 1:200 v/v), followed by the secondary antibody (goat anti-mouse IgG conjugated with fluorescein-5-isothiocyanate - FITC [Molecular Probes, Leiden, The Netherlands], diluted 1:200 v/v). As a control, slides without the first antibody were used for

48 DNA and microtubules during loss and re-establishment of DT in Medicago truncatula seeds

every treatment. A confocal laser scanning microscope (Biorad MRC-600) and an epifluorescence microscope (Nikon Labophot) were used for the visualisation of microtubules.

Results

Medicago truncatula seed germination: dormancy release and imbibition

Germination of seeds without pre-germination treatment (control) remained around 5% until 11 d, increasing afterwards and reaching its maximum (95%) at 17 d of imbibition. Seeds subjected to chemical scarification or cold imbibition germinated faster than the control, with a better performance by acid-treated seeds, suggesting that in the combined dormancy present in M. truncatula seeds, the physical (seed coat) dormancy is stronger than the physiological (embryo) dormancy. The most efficient treatment was achieved by the combination of acid scarification and cold imbibition that allowed the start of germination after a few hours of imbibition at 20oC, reaching the maximum (96%) within 1 d (Fig. 1). The imbibition curve of seeds chemically scarified and subjected to cold imbibition is shown in Fig. 2. Phase 1 of imbibition, characterized by a rapid increase in fresh weight, occurred in the first 9 h. Between 9 h and 40 h of imbibition, the gain in fresh weight was very small, characterizing the plateau or phase 2 of imbibition. Visible germination (radicle protrusion) started after 42 h of imbibition (or 6 h after transfer to 20oC). After this point, germinated seeds (seedlings) entered phase 3, resuming the increase in fresh weight (Fig. 2). Although seeds germinated only after transfer to 20oC, they were also able to do so at 4oC, when kept at this temperature for 4 d (not shown).

DNA content

Flow cytometric analyses of nuclear DNA contents in radicles of mature dry seeds of M. truncatula revealed a high 4C DNA content (45%), which remained unchanged during germination and radicle growth until 2 mm. With further growth the relative content of 4C nuclei increased significantly, reaching 63% in 3 mm long radicles, levelling off afterwards, with 65% of 4C DNA in 4 mm long radicles (Fig. 3). The 4C DNA content in 2 mm long radicles remained unchanged during incubation in PEG for 3 d (not shown).

49 Chapter 3

100

80

60

40 Acid + Cold Germination (%) Acid 20 Cold Control

0 024681012141618 Time (days)

Figure 1. Germination of Medicago truncatula seeds at 20oC after various treatments for dormancy release: chemical scarification with sulphuric acid; cold imbibition (4oC) for 36 h; chemical scarification plus cold imbibition; and control. Each data point is the mean of three replications of 50 seeds. Bars represent standard deviation.

The relation between the progress of germination and loss of desiccation tolerance

In order to relate the course of germination with the loss of DT, seeds were chemically scarified, imbibed (cold imbibition followed by imbibition at 20oC) and, at various times, germination and DT evaluated. Germination, as assessed by radicle protrusion, did not occur during the 36 h of cold imbibition, but was observed after 7 h of imbibition at 20oC (43 h of total imbibition time). From this point germination raised sharply until its maximum (96 %) by 56 h of total imbibition time (Fig. 2). Before the start of the germination, i.e. until 43 h of imbibition, DT remained unchanged, at 100%, dropping fast afterwards, inversely related to the progress of germination (Fig. 2).

50 DNA and microtubules during loss and re-establishment of DT in Medicago truncatula seeds

1.4 100 1.2 80 1.0 60 0.8 40 0.6

20 seeds) (g/100 weight Fresh Germination (%) Germination 0.4

Desiccation tolerance (%) 0 0.2 0 102030405060 Time (h)

Figure 2. Imbibition curve, germination, and loss of desiccation tolerance of chemically scarified Medicago truncatula seeds. Seeds were imbibed for 36 h at 4oC and then transferred to 20oC. Desiccation tolerance was determined after drying of the imbibed seeds/seedlings, followed by pre-humidification and rehydration. Seeds that germinated or seedlings that resumed radicle growth and normal development were considered desiccation-tolerant. Each data point is the average of three independent experiments of 50 seeds/seedlings. Bars represent standard deviation.

Typically, even for a homogeneous batch of seeds, germination did not occur uniformly. Consequently, at any given time-point of the course of germination after 43 h (Fig. 2), the population of seeds is comprised of germinated (at different stages) and non- germinated seeds. Thus, to characterise the loss of DT to the progress of the germination in a more accurate way, seeds were put to germinate, selected by their protruded radicle length and tested again for DT. The results show that right after visible germination, when the protruded radicle length was 1 mm, only 12% of the seedlings was still desiccation- tolerant, i.e. able to resume normal growth after being dehydrated and rehydrated. Seedlings with a radicle length of 2 mm or longer lost DT completely (Fig. 3). Seedlings that did not resume radicle growth, frequently showed growth of the cotyledons and, to a lesser extent, also of the hypocotyl. However, the longer the radicle before dehydration, the less frequent the growth of cotyledons and hypocotyl (not shown).

51 Chapter 3

70 100 PEG-treated

80 4C DNA content 60

60 Untreated

40 50

20 4C DNA content (%)

Desiccation tolerance (%) tolerance Desiccation 40 0

012345 Radicle length (mm)

Figure 3. Protruded radicle length and desiccation tolerance in Medicago truncatula seedlings with (closed symbols) and without (open symbols) previous incubation in PEG, and 4C DNA content of radicle cells from dry seeds and seedlings. Desiccation tolerance was determined after drying the seedlings with or without PEG treatment, followed by pre- humidification and rehydration. Seedlings that resumed radicle growth and normal development upon rehydration were considered desiccation-tolerant. Each data point is the average of four independent experiments of 50 seedlings. Bars represent standard deviation. For flow cytometry each data point is the average of five replications of 10 radicle tips. Bars represent standard deviation.

Re-establishment of desiccation tolerance in seedlings by incubation in PEG

In order to relate re-establishment of DT to the progress of germination, seedlings of M. truncatula with radicle lengths ranging from 1 mm to 5 mm were incubated in PEG solution, dehydrated, pre-humidified, and rehydrated. Seedlings that resumed radicle growth after dehydration and rehydration and showed normal development were considered desiccation- tolerant. The results in Fig. 3 show that DT could be substantially re-induced (84%) in seedlings with a radicle length of up to 2 mm. From 2 mm onwards there was an abrupt drop in DT, decreasing in value to 33% at 3 mm and to near zero at 4 mm (Fig. 3). As had already been observed for untreated seedlings, PEG-treated seedlings that did not resume radicle growth frequently showed elongation of the cotyledons and, to a lesser extent, also

52 DNA and microtubules during loss and re-establishment of DT in Medicago truncatula seeds

of the hypocotyl (up to 3 cm). This was mainly observed when the protruded radicle before dehydration was short, i.e. 1-2 mm (not shown). Thus, the radicle appeared to be the more desiccation-sensitive part of the seedling, followed by the hypocotyl and cotyledons.

Seedling viability

Although seedlings with a radicle of 2 mm did not resume radicle growth after dehydration (without PEG), pre-humidification, and rehydration (Fig. 3), their radicles were turgid and apparently healthy during the first days following rehydration. Therefore, a tetrazolium test was performed in order to assess biochemically the viability of both untreated and PEG- treated seedlings (2 mm long radicles) after dehydration and pre-humidification. All untreated seedlings showed dark-red stained cotyledons and unstained radicles (Fig. 4A), indicating that cotyledon cells survived dehydration whereas radicle cells did not. Nevertheless 10% of these seedlings showed a dark-red stained hypocotyl, indicating that cells in that region were still alive. In 100% of the PEG-treated seedlings the cotyledons were also dark-red stained, while different situations were observed in the radicle and hypocotyl. In 33% of them, both radicle and hypocotyl were dark-red stained (Fig. 4B); in 36% a dark-red staining of the hypocotyl occurred and a partial staining (dark- or light-red) of the radicle, normally in the tip (Fig. 4C); 12% showed dark-red stained hypocotyls and light-red stained radicles (Fig. 4D); 8% remained with white, unstained radicles and hypocotyls (Fig. 4E); and 11% showed unstained radicle and dark-red stained hypocotyl (Fig. 4F).

Changes in moisture content (MC) of the radicles during incubation in PEG, dehydration, pre-humidification, and rehydration

During incubation in PEG, MC of 2 mm long radicles decreased steadily in the first 9 h, from 3.74 g/g to 2.35 g/g, and then slowly until 72 h (2.24 g/g) (Fig. 5A). In terms of percentage, 37% of the water was removed in the first 9 h and, by the end (72 h), 40% of the water had been lost. During dehydration the rate and extent of the water loss were similar in radicles of both PEG-treated and untreated seedlings, which rapidly lost, respectively, 82% and 93% of the water in the first 2 h of drying (Fig. 5B, inset). By the end (72 h) the radicle MCs of PEG- treated and untreated seedlings were 0.20 g/g and 0.15 g/g, respectively (Fig. 5B), showing no statistical difference between them and with the original (dry seed) radicle MC (0.19 g/g). During pre-humidification (for 24 h) and the first 24 h of rehydration, changes in MC were again similar in radicles from both PEG-treated and untreated seedlings. During pre- humidification MC increased at a constant and low rate to around 1.3 g/g (Fig. 5C, 0-24 h) and rehydration quickly increased MC, with the highest rates occurring in the first hour. After

53 Chapter 3

24 h of rehydration, the radicle MCs of PEG-treated and untreated seeds were 6.21 and 5.72 g/g, respectively (Fig. 5C).

A

B

C D

E F

Figure 4. Tetrazolium test performed on seedlings (2 mm long radicles) of Medicago truncatula following dehydration (with or without previous PEG treatment) and pre- humidification. (A) Untreated seedlings showing unstained radicles and dark-red stained cotyledons. (B–F) PEG-treated seedlings. Cotyledons of all seedlings stained dark-red. (B) Radicle and hypocotyl totally dark-red stained; (C) radicle partially stained (arrow) and hypocotyls (arrowhead) stained; (D) radicle light-red stained and hypocotyl (arrowhead) stained; (E) radicle and hypocotyl unstained; (F) radicle unstained and hypocotyl (arrowhead) stained.

54 DNA and microtubules during loss and re-establishment of DT in Medicago truncatula seeds

A

4.0

3.6

3.2

2.8

2.4 (g/g) (g/g) content content Moisture

2.0 0 122436486072 Time of incubation in PEG (h) B

4.0 4.0 Untreated 3.5 PEG-treated 3.0 3.0 2.0 2.5 1.0

2.0 0.0 0.0 0.5 1.0 1.5 2.0 1.5

1.0

(g/g) content Moisture 0.5 0.0 0 122436486072

Time of dehydration (h)

C 7.0 Untreated 6.0 PEG-treated 5.0

4.0 3.0

2.0

(g/g) (g/g) content content Moisture 1.0 0.0 0 12243648 Time of pre-humidification and rehydration (h)

Figure 5. Changes in the moisture content (g H O/g dry matter) of the radicles (2 mm 2 long) of Medicago truncatula seedlings during (A) incubation in PEG; (B) dehydration, and

(C) pre-humidification (for 24 h) followed by rehydration.

55 Chapter 3

Detection of DNA fragmentation in seedlings subjected to dehydration

Analysis of DNA integrity in 2 mm long protruded radicles revealed DNA degradation in radicles excised from seedlings subjected to dehydration (Fig. 6; lanes 2 and 3), while control seedlings (not dehydrated) showed intact DNA. DNA fragmentation was much stronger in untreated than in PEG-treated seedlings, with the laddering pattern composed of multimers of about 200 bp.

kb M 1 2 3

10.0

2.0 1.5

1.0

0.5

Figure 6. Agarose gel of genomic DNA extracted from 2 mm long radicles of seedlings of Medicago truncatula. (M) Marker with the band lengths (kb) shown in the left; (1) control (not subjected to dehydration); (2) PEG-treated seedlings (dehydrated after incubation in PEG); and (3) untreated seedlings (dehydrated without previous incubation in PEG). DNA samples (5 µg) were loaded on a 1% agarose gel stained with ethidium bromide.

Microtubular cytoskeleton in radicles before and after germination

Radicles of dry seeds and of seedlings were analysed for microtubular cytoskeleton configurations in order to characterize their changes during radicle growth and to relate them to the loss of DT. In dry seeds a high level of fluorescence was detected in the form of granules, indicating that, at that stage, tubulin was present in granules in the cytoplasm, instead of being assembled into MTs, (Fig. 7A). In seedlings with a radicle length of 1 mm (Fig. 7B) and 2 mm (Fig. 7C) abundant cortical microtubular arrays were observed with MTs transversely oriented to the direction of cell elongation. In 3 mm long radicles cortical MTs

56 DNA and microtubules during loss and re-establishment of DT in Medicago truncatula seeds

Figure 7. Fluorescence micrographs of radicle cells of dry seeds and seedlings of Medicago truncatula, labelled with α-tubulin antibody and with a fluorescent secondary antibody. (A) Dry seed showing abundant fluorescent granules; (B) 1 mm protruded radicle with well- established cortical microtubular cytoskeleton arrays oriented perpendicularly to cell (and radicle) elongation; (C) 2 mm protruded radicle showing the same situation as the previous figure; (D, E) 3 mm protruded radicle: presence of cortical microtubules (D) and first appearance of mitotic configurations (E, arrows pointing to phragmoplast arrays), indicating the presence of cell division; and (F) 4 mm protruded radicle, with abundant cortical and mitotic microtubules. Bars (A–E) 25 µm, (F) 100 µm.

57 Chapter 3

(Fig. 7D) and the first mitotic MTs were detected (Fig. 7E), indicating the start of cell division. In 4 mm long radicles the number of cells entering the M phase of the cell cycle was much higher than in 3 mm long radicles, with all the mitotic MT configurations (preprophase band, spindle and phragmoplast) displayed. As for the shorter radicles (1-3 mm), cells with cortical microtubular cytoskeleton were also abundantly present (Fig. 7F).

Effects of dehydration, pre-humidification, rehydration and PEG-treatment of M. truncatula seedlings on the microtubular cytoskeleton

PEG-treated and untreated seedlings with a 2 mm radicle were used to study the effect of dehydration, pre-humidification and rehydration on the microtubular cytoskeleton in radicle cells. Incubation in PEG for 3 d caused no changes in the microtubular cytoskeleton (compare Fig. 8A and B). Dehydration of both untreated and PEG-treated seedlings dismantled partially the well-established microtubular cytoskeleton and led to the appearance of tubulin granules (Fig. 8C, D). Pre-humidification worsened the situation, with no MTs and only few granules of tubulin being detected in both untreated (Fig. 8E) and PEG- treated (Fig. 8F) seedlings. Further, when pre-humidified untreated seedlings were rehydrated for 24 h, some tubulin granules could still be detected, but this was rather a rare event and the prevalent situation was that of a total absence of MTs and tubulin granules (Fig. 8G). However, a different situation occurred in PEG-treated seedlings, which exhibited, after 24 h of rehydration, a rebuilt, functional microtubular cytoskeleton in the radicle cells, with both interphase and mitotic configurations (Fig. 8H).

Figure 8 (right page). Fluorescence micrographs of radicle cells of seedlings of Medicago truncatula, with a radicle length of 2 mm, subjected to dehydration (with or without PEG- treatment), pre-humidification and rehydration. Sections were labelled with α-tubulin antibody and with fluorescent secondary antibody. (A) Control (before dehydration) showing abundant cortical MTs. (C, E, G) untreated seedlings. (C) Dehydration led to a decrease in the abundance of MTs and appearance of tubulin granules (arrow); (E) after dehydration and pre-humidification, MTs disappeared totally and tubulin granules could hardly be detected; (G) after dehydration, pre-humidification, and 24 h of rehydration, the total absence of MTs and tubulin granules. (B, D, F, H) PEG-treated seedlings. (B) After incubation in PEG, the situation of the microtubular cytoskeleton remained unchanged, compared with the control; (D) after dehydration, a decrease in the abundance of MTs and appearance of tubulin granules (arrow) was detected; (F) after dehydration and pre-humidification, although some cells still exhibited tubulin granules (inset, arrow), the general picture was of total absence of MTs and tubulin granules; (H) after dehydration, pre-humidification, and 24 h of rehydration, the normal situation was restored, with cortical MTs being seen throughout the radicle cells. In addition, MTs functioning in mitotic configurations were also observed. Bars (A, B, E, F, G, H) 50 µm; (C) (both pictures), (D) and (F, inset) 25 µm.

58 DNA and microtubules during loss and re-establishment of DT in Medicago truncatula seeds

59 Chapter 3

Discussion

The radicle length at which seedlings from orthodox seeds lose DT varies among species, with 2 mm being reported for tomato (Lin et al., 1998) and M. truncatula cv. Paraggio (Buitink et al., 2003), 1 mm for okra and mung bean (Lin et al., 1998) and 0.5 mm for snow pea and cucumber (Lin et al., 1998). Besides the expected variation among species, the different experimental procedures adopted, especially the drying rate, are certainly another cause for the differences found. Decrease in DT before radicle protrusion appears to be a relatively rare event. It has been reported for coffee seeds (Ellis et al., 1991) and was found in the present study. For instance, at 44 h of imbibition, 20% of the seeds had germinated (Fig. 2), suggesting that at least the 80% that had not germinated could still tolerate desiccation. However, only 40% were still desiccation tolerant. The drying condition used to assess DT (43% RH; corresponding to a water potential of -115 MPa) which led to a very fast dehydration of the seeds certainly diminished the chances of de novo synthesis of protective components. It has been shown that slow water loss may allow protective changes to occur, not only in germinating (Sun, 1999) but also in developing orthodox seeds (Kermode and Finch-Savage, 2002), in somatic embryos (Senaratna et al., 1989) and in the whole plant (Oliver et al., 1998), enabling them to withstand subsequent severe dehydration. In the present study such a slow (and limited) water loss was achieved by subjecting the seedlings to a mild osmotic stress through incubation in PEG solution (-1.8 MPa), resulting in re-establishment of DT at high rates in seedlings with a radicle length of up to 2 mm. In a study with M. truncatula cv. Paraggio, Buitink et al. (2003) showed that during incubation in PEG, synthesis of possibly protective substances such as sucrose and a dehydrin, occurred cumulatively until 24 h, although water was lost only in the first 6 h. The tetrazolium test revealed that radicles of untreated dried seedlings (2 mm long radicles) had lost their viability, confirming the absence of DT. In PEG-treated seedlings (also with 2 mm long radicles) 45% of the radicles were totally stained, and 36% only partially. The sum of these values (81%) is very close to the 84% of DT shown by these seedlings, suggesting that the partially stained radicles should also be considered viable. Radicles of dry mature M. truncatula seeds contained relatively high 4C DNA content (45%). This suggests that, by the end of seed maturation, there are two blocks acting in the cell cycle: one at the G2/M boundary, excluding those cells with 4C nuclei progress to mitosis, and another at G1/S keeping the cells with 2C nuclei at the pre-synthetic phase. However, this is not the prevalent situation in orthodox seeds, in which the quiescent embryo normally exhibits most (or all) cells with a 2C DNA content, reflecting a stringent arrest of the cell cycle at the pre-synthetic G1 phase (Deltour, 1985; Bino et al., 1993). Although high for an orthodox seed, the 4C value found in M. truncatula is lower than that reported for mature orthodox seeds of another legume species, Phaseolus vulgaris, in which 55% of the cells in the root tip showed 4C DNA content (Bino et al., 1993).

60 DNA and microtubules during loss and re-establishment of DT in Medicago truncatula seeds

The relationship between the progress of the cell cycle and stress sensitivity in plants has not yet been clarified, although it has been known for a long time that cells in G2 are more stress-sensitive than cells in G1 (Sybenga, 1972). Several studies relating cell cycle and stress resistance in seeds have shown that cells in G1 are more resistant to desiccation, cold- or heat-shock, storage, and radiation (reviewed by Deltour, 1985; Saracco et al., 1995; Sliwinska, 2003). Deltour (1985) hypothesised that nuclei with a 2C content might be more stress-resistant by offering a smaller target for mutation-inducing factors than those with a 4C content. However, very high 2C nuclei contents have also been found in mature embryos of intermediate and recalcitrant seeds, such as coffee (da Silva, 2002), neem (Azadirachta indica) (Sacandé, 2000), Castanea sativa (Bino et al., 1993) and Inga vera (Faria et al., 2004). Furthermore, seeds of the related tree species Acer platanoides (desiccation-tolerant) and A. pseudoplatanus (desiccation-sensitive) are shed with a similar 4C DNA content in the radicles (38 and 37%, respectively) (Finch-Savage et al., 1998). Thus, it appears that, in mature seeds, DT is not correlated with the arrest of the cell cycle at any particular DNA content. In seedlings from orthodox seeds, DNA content and DT normally show a high correlation (Sargent et al., 1981; Dasgupta et al., 1982; Deltour, 1985; Osborne and Boubriak, 1994; Osborne, 2000; Boubriak et al., 2000), although the resumption of DNA synthesis is unlikely to be the only effective agent in inducing the change from the tolerant to the intolerant state (Dasgupta et al., 1982). As DNA replication is, in general, a late event during germination, other processes may be more tightly linked to the loss of DT, with DNA content playing only an additive role in the increasing stress sensitivity upon germination (Saracco et al., 1995; Boubriak et al., 1997). In PEG-treated seedlings the greatest drop in DT (from 84% to 33%) occurred simultaneously with the increase in 4C DNA content. A second significant decrease in DT of PEG-treated seedlings (from 33% to 5%) was observed between the radicle lengths of 3 mm and 4 mm. In this interval, the DNA content remained unaltered, but a great number of cells had entered the M phase of the cell cycle. Dividing cells are less tolerant to desiccation than those that are elongating (Dasgupta et al., 1982). Programmed cell death has been shown to occur in different plant organs and tissues during normal development (e.g. senescence of leaves and post-germinative megagametophyte cell death) or induced by pathogens and stress (Danon et al., 2000; He and Kermode, 2003). Dehydration of desiccation-sensitive seeds (both recalcitrant and germinating orthodox seeds) may lead to the fractionation of DNA (Osborne and Boubriak, 1994; Boubriak et al., 2000). In the present study, directly dried seedlings with 2 mm long radicles displayed degradation of nuclear DNA, which was visualised by the formation of DNA ladders. The fragment lengths were multiples of approximately 200 bp. These multimers, with lengths of 170 to 200 bp, are generated by the cleavage of the chromatin by endonucleases at internucleosomal sites (Stein and Hansen, 1999). In desiccation-tolerant seeds some DNA damage that occurs during dehydration or dry storage may be repaired when water is again available (Osborne, 2000). However, DNA laddering is an indicator of

61 Chapter 3

the endpoint of the apoptotic process and cannot be reversed (Boubriak et al., 2000). It appears thus that the weak signal of laddering shown by PEG-treated seedlings possibly comes from the 16% that did not survive dehydration. To our knowledge this is the first time that DNA laddering is shown to occur during drying of intolerant plant tissue. It has been suggested that compounds such as sugars and LEA proteins may act as protecting factors, stabilizing cellular structures during drying (Crowe and Crowe, 1986; Dure, 1997). It is also known that mild stresses can trigger the synthesis of protective substances in plant tissues (Farnsworth, 2000) and seedlings (Buitink et al., 2003). It can thus be speculated that in our study PEG incubation induced the synthesis of nuclear proteins with a protective role of the DNA. It is thought that nuclear desiccation-induced proteins, such as QP47 isolated from Pisum sativum seeds, protect DNA during desiccation (Chiatante et al., 1995). Besides the synthesis of protectants, loss of water may also cause reversible conformational changes in the DNA, altering the recognition of specific base- sequence domains by enzymes (Osborne and Boubriak, 1994; Osborne et al., 2002), thereby hindering the action of the nucleases, although this is yet to be proven in plant cells. There were no MTs in radicle cells of dry M. truncatula seeds. Only granules of tubulin were detected, as in seeds of tomato (de Castro, 1998) and coffee (da Silva, 2002). Upon germination free tubulin assembled into a cortical microtubular cytoskeleton in cells of protruded radicles with a length of 1 and 2 mm and, afterwards, together with mitotic MTs. Hence, cell elongation alone was sufficient for radicle protrusion and early radicle growth, with cell division being additionally required later. When desiccation-sensitive tissues are exposed to water potentials below -2 MPa, dehydration may lead to the loss of membrane organization, cellular integrity and degradation of macromolecules (Osborne and Boubriak, 1994). Around -5 MPa there is a general trend towards contraction or dismantling of organelles (Walters et al., 2002). In the present study, seedlings were exposed to much more severe dehydration conditions (43% RH; -115 MPa) and the consequence in both untreated and PEG-treated seedlings was a decrease in abundance of MTs and appearance of tubulin granules. The dismantling of the cytoskeleton in seeds caused by dehydration has also been reported for recalcitrant (desiccation-sensitive) seeds, such as Quercus robur (Mycock et al., 2000), Trichilia dregeana (Gumede et al., 2003) and Inga vera (Faria et al., 2004). The subsequent pre- humidification of the dried seedlings resulted in the total disappearance of the MTs and a great reduction of tubulin granules. Again, the decay of the microtubular cytoskeleton was comparable in both untreated and PEG-treated seedlings. The difference between untreated and PEG-treated seedlings only appeared when the pre-humidified seedlings were rehydrated: PEG-treated seedlings were able to reconstruct a functional microtubular cytoskeleton and continue normal development, while untreated seedlings showed a total absence of MTs and tubulin granules, and, consequently, the ability to resume normal growth. It is clear that the ability of PEG-treated seedlings to survive dehydration, as far as

62 DNA and microtubules during loss and re-establishment of DT in Medicago truncatula seeds

the microtubular cytoskeleton is concerned, did not rest on its protection during dehydration, but on its reconstruction upon rehydration. Desiccation-tolerant organisms must rely on one or both of the following strategies: avoidance of the accumulation of desiccation-induced damage, and the activation of repair mechanisms upon rehydration (Buitink et al., 2002). The present study showed that both strategies were distinctly applied by seedlings in which DT was re-established by PEG treatment. Nuclear DNA was kept intact during dehydration, whilst MTs were dismantled and later rebuilt upon rehydration.

Acknowledgements

We thank Olivier Leprince (Anjou Recherche Semences, Angers, France) for showing us how to re-establish DT in M. truncatula seedlings. Jan Bergervoet (Plant Research International, Wageningen) is acknowledged for his assistance with the flow cytometry analysis. We thank CNPq (National Council for Scientific and Technological Development, from the Ministry of Science and Technology, Brazil) for financial support of the studies of J.M.R. Faria.

References

Adams, C.A. and Rinne, R.W. (1980) Moisture content as a controlling factor in seed development and germination. International Review of Cytology 68, 1-8. Alberts, B., Johnson, A., Lewis, J., Raff, M., Roberts, K. and Walter, P. (1998) Essential cell biology: an introduction to the molecular biology of the cell. New York, USA: Garland Publishing, Inc. Arumuganthan, K. and Earle, E.D. (1991) Estimation of nuclear DNA content of plants by flow cytometry. Plant Molecular Biology Report 9, 229-233. Bartels, D. and Salamini, F. (2001) Desiccation tolerance in the resurrection plant Craterostigma plantagineum. A contribution to the study of drought tolerance at the molecular level. Plant Physiology 127, 1346-1353. Baskin, T.I., Busby, C.H., Fowke, L.C., Sammut, M. and Gubler, F. (1992) Improvements on immunostaining samples embedded in methacrylate: localization of microtubules and other antigens throughout developing organs in plants of diverse taxa. Planta 187, 405-413. Berjak, P. and Pammenter, N.W. (2000) What ultrastructure has told us about recalcitrant seeds. Revista Brasileira de Fisiologia Vegetal 12, 22-55. Berrie, A.M.M. and Drennan, D.S.H. (1971) The effect of hydration-dehydration on seed germination. New Phytologist 70, 135-142. Bewley, J.D. (1979) Physiological aspects of desiccation tolerance. Annual Review of Plant Physiology 30, 195-238.

63 Chapter 3

Bino, R.J., Lanteri, S., Verhoeven, H.A. and Kraak, H.L. (1993) Flow cytometric determination of nuclear replication stage in seed tissues. Annals of Botany 72, 181-187. Boubriak, I., Kargiolaki, H., Lyne, L. and Osborne, D.J. (1997) The requirement for DNA repair in desiccation tolerance of germinating embryos. Seed Science Research 7, 97-105. Boubriak, I., Dini, M., Berjak, P. and Osborne, D.J. (2000) Desiccation and survival in the recalcitrant seeds of Avicennia marina: DNA replication, DNA repair and protein synthesis. Seed Science Research 10, 307-315. Bruggink, T. and van der Toorn, P. (1995) Induction of desiccation tolerance in germinated seeds. Seed Science Research 5, 1-4. Buitink, J., Hoekstra, F.A. and Leprince, O. (2002) Biochemistry and biophysics of tolerance systems, in Desiccation and Survival in Plants: Drying without Dying (eds. M. Black and H.W. Pritchard), Cabi Publishing, Wallingford, pp. 293-318. Buitink, J., Vu, B.L., Satour, P. and Leprince, O. (2003) The re-establishment of desiccation tolerance in germinated radicles of Medicago truncatula Gaertn. seeds. Seed Science Research 13, 273- 286. Chiatante, D., Onelli, E., Patrignani, G. and Scippa, G. (1995) Localization of a nuclear protein (QP47) in embryonic meristems during seed maturation and germination and its distribution among crop plants. Journal of Experimental Botany 46, 815-821. Crowe, J.H. and Crowe, L.M. (1986) Stabilization of membranes in anhydrobiotic organisms, in Membranes, Metabolism and Dry Organisms (ed A.C. Leopold), Comstock Publ. Assoc., Ithaca, USA, pp. 188-209. da Silva, E.A.A. (2002) Coffee seed (Coffea arabica cv. Rubi) germination: mechanism and regulation. PhD thesis, Wageningen University, Wageningen, The Netherlands. Danon, A., Delorme, V., Mailhac, N. and Gallois, P. (2000) Plant programmed cell death: a common way to die. Plant Physiology and Biochemistry 38, 647-655. Dasgupta, J., Bewley, J.D. and Yeung, E.C. (1982) Desiccation-tolerant and desiccation-intolerant stages during the development and germination of Phaseolus vulgaris seeds. Journal of Experimental Botany 33, 1045-1057. de Castro, R.D. (1998) A functional analysis of cell cycle events in developing and germinating tomato seeds. PhD thesis, Wageningen Agricultural University, Wageningen, The Netherlands. Deltour, R. (1985) Nuclear activiation during early germination of the higher plant embryo. Journal of Cell Science 75, 43-83. Dure, L. (1997) LEA proteins and the desiccation tolerance of seeds, in Cellular and Molecular Biology of Plant Seed Development (eds B.A. Larkins and I.K. Vasil), Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 525-543. Ellis, R.H., Hong, T.D. and Roberts, E.H. (1991) An intermediate category of seed storage behaviour? II. Effects of provenance, immaturity and imbibition on desiccation-tolerance in coffee. Journal of Experimental Botany 42, 653-667. Faria, J.M.R., van Lammeren, A.A.M. and Hilhorst, H.W.M. (2004) Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis. Seed Science Research 14, 165-178. Farnsworth, E. (2000) The ecology and physiology of viviparous and recalcitrant seeds. Annual Review of Ecology and Systematics 31, 107-138.

64 DNA and microtubules during loss and re-establishment of DT in Medicago truncatula seeds

Finch-Savage, W.E., Bergervoet, J.H.W., Bino, R.J., Clay, H.A. and Groot, S.P.C. (1998) Nuclear replication activity during seed development, dormancy breakage and germination in three tree species: Norway Maple (Acer platanoides L.), Sycamore (Acer pseudoplatanus L.) and Cherry (Prunus avium L.). Annals of Botany 81, 519-526. Gumede, Z., Merhar, V. and Berjak, P. (2003) Effect of desiccation on the microfilament component of the cytoskeleton in zygotic embryonic axes of Trichilia dregeana Sond, in Proceedings of the 4th International Workshop on Desiccation Tolerance and Sensitivity of Seeds and Vegetative Plant Tissues. Blouwaterbaai, South Africa, p. 22. He, X. and Kermode, A.R. (2003) Nuclease activities and DNA fragmentation during programmed cell death of megagametophyte cells of white spruce (Picea glauca) seeds. Plant Molecular Biology 51, 509-521. Hoekstra, F.A., Golovina, E.A. and Buitink, J. (2001) Mechanisms of plant desiccation tolerance. Trends in Plant Science 6, 431-438. Hong, T.D. and Ellis, R.H. (1992) The survival of germinating orthodox seeds after desiccation and hermetic storage. Journal of Experimental Botany 43, 239-247. ISTA (International Seed Testing Association) (1996) International rules for seed testing. Seed Science and Technology 24 (suppl.). Journet, E.P., Barker, D., Harrison, M. and Kondorosi, E. (2001) Medicago truncatula as biological material (Module 1), in EMBO Practical Course on the New Plant Model System Medicago truncatula, pp. 1-29. Kermode, A.R. and Finch-Savage, B.E. (2002) Desiccation sensitivity in orthodox and recalcitrant seeds in relation to development, in Desiccation and Survival in Plants: Drying without Dying (eds M. Black and H.W. Pritchard), Cabi Publishing, Wallingford, Oxon, UK, pp. 149-184. Koster, K.L., Reisdorph, N. and Ramsay, J.L. (2003) Changing desiccation tolerance of pea embryo protoplasts during germination. Journal of Experimental Botany 54, 1607-1614. Leprince, O., Harren, F.J.M., Buitink, J., Alberda, M. and Hoekstra, F.A. (2000) Metabolic dysfunction and unabated respiration precede the loss of membrane integrity during rehydration of germinating radicles. Plant Physiology 122, 597-608. Lin, T.P., Yen, W.L. and Chien, C.T. (1998) Disappearance of desiccation tolerance of imbibed crop seeds is not associated with the decline of oligosaccharides. Journal of Experimental Botany 49, 1203-1212. Liu, Y.G., Norihiro, M., Teruko, O. and Whittier, R.F. (1995) Efficient isolation and mapping of Arabidopsis thaliana T-DNA insert junctions by thermal asymmetric interlaced PCR. The Plant Journal 8, 457-463. Mycock, D.J., Berjak, P. and Finch-Savage, W.E. (2000) Effects of desiccation on the subcellular matrix of the embryonic axes of Quercus robur, in Seed biology: Advances and Applications (eds M. Black, K.J. Bradford and J. Vazquez-Ramos), CABI Publishing, Wallingford, pp. 197-203. Oliver, M.J. (1996) Desiccation tolerance in vegetative plant cells. Physiologia Plantarum 97, 779-787. Oliver, M.J., Wood, A.J. and O'Mahony, P. (1998) "To dryness and beyond" - preparation for the dried state and rehydration in vegetative desiccation-tolerant plants. Plant Growth Regulation 24, 193– 201. Oliver, M.J., Tuba, Z. and Mishler, B.D. (2000) The evolution of vegetative desiccation tolerance in land plants. Plant Ecology 151, 85–100.

65 Chapter 3

Osborne, D.J. (2000) Hazards of a germinating seed: available water and the maintenance of genomic integrity. Israel Journal of Plant Sciences 48, 173-179. Osborne, D.J. and Boubriak, I. (1994) DNA and desiccation tolerance. Seed Science Research 4, 175- 185. Osborne, D.J., Boubriak, I. and Leprince, O. (2002) Rehydration of dried systems: membranes and the nuclear genome, in Desiccation and Survival in Plants: Drying without Dying (eds M. Black and H.W. Pritchard), Cabi Publishing, Wallingford, Oxon, UK, pp. 343-364. Pukacka, S. (2001) Loss of tolerance to desiccation in germinated Norway maple (Acer platanoides L.) seeds. Changes in carbohydrate content. Dendrobiology 46, 43-48. Reisdorph, N.A. and Koster, K.L. (1999) Progressive loss of desiccation tolerance in germinating pea (Pisum sativum) seeds. Physiologia Plantarum 105, 266-271. Ren, J. and Tao, L. (2003) Effect of hydration–dehydration cycles on germination of seven Calligonum species. Journal of Arid Environments 55, 111-122. Sacandé, M. (2000) Stress, storage and survival of neem seed. PhD thesis, Wageningen Agricultural University, Wageningen, The Netherlands. Saracco, F., Bino, R.J., Bergervoet, J.H.W. and Lanteri, S. (1995) Influence of priming-induced nuclear replication activity on storability of pepper (Capsicum annuum L.) seed. Seed Science Research 5, 25-29. Sargent, J.A., Mandi, S.S. and Osborne, D.J. (1981) The loss of desiccation tolerance during germination: an ultrastructural and biochemical approach. Protoplasma 105, 225-239. Senaratna, T. and McKersie, B.D. (1983) Dehydration injury in germinating soybean (Glycine max L. Merr.) seeds. Plant Physiology 72, 620-624. Senaratna, T. and McKersie, B.D. (1986) Loss of desiccation tolerance during seed germination: a free radical mechanism of injury. In: Leopold AC, ed. Membranes, metabolism and dry organisms. Ithaca, USA: Comstock Publ. Assoc., 85-101. Senaratna, T., McKersie, B.D. and Bowley, S.R. (1989) Desiccation tolerance of alfalfa (Medicago sativa L.) somatic embryos. Influence of abscisic acid, stress pretreatments and drying rates. Plant Science 65, 253-259. Sieberer, B.J., Timmers, A.C.J., Lhuissier, F.G.P. and Emons, A.M.C. (2002) Endoplasmic microtubules configure the subapical cytoplasm and are required for fast growth of Medicago truncatula root hairs. Plant Physiology 130, 977-988. Sliwinska, E. (2003) Cell cycle and germination of fresh, dried and deteriorated sugarbeet seeds as indicators of optimal harvest time. Seed Science Research 13, 131-138. Stein, J.C. and Hansen, G. (1999) Mannose induces an endonuclease responsible for DNA laddering in plant cells. Plant Physiology 121, 71–79. Sun, W.Q. (1999) Desiccation sensitivity of recalcitrant seeds and germinated orthodox seeds: Can germinated orthodox seeds serve as a model system for studies of recalcitrance?, in IUFRO Seed Symposium 1998. Recalcitrant Seeds. Proceedings of the Conference (eds M. Marzalina et al.), Forestry Research Institute Malaysia, Kuala Lumpur, pp. 29-42. Sybenga, J. (1972) General cyogenetics. Elsevier, New York, USA. Walters, C., Farrant, J.M., Pammenter, N.W. and Berjak, P. (2002) Desiccation stress and damage, in Desiccation and Survival in Plants: Drying without Dying (eds M. Black and H.W. Pritchard), Cabi Publishing, Wallingford, Oxon, UK, pp. 263-291.

66 DNA and microtubules during loss and re-establishment of DT in Medicago truncatula seeds

Wang, M., Oppedijk, B.J., Caspers, M.P.M., Lamers, G.E.M., Boot, M.J., Geerlings, D.N.G., Bakhuizen, B., Meijer, A.H. and van Duijn, B. (1998) Spatial and temporal regulation of DNA fragmentation in the aleurone of germinating barley. Journal of Experimental Botany 49, 1293–1301. Wasteneys, G.O. and Galway, M.E. (2003. Remodeling the cytoskeleton for growth and form. Annual Review of Plant Biology 54, 691-722.

67

Chapter 4

Changes in gene expression during loss and re-establishment of desiccation tolerance in germinated Medicago truncatula seeds

José Marcio Rocha Faria, Michiel Lammers, Wilco Ligterink, Jan Kodde and Henk W. M. Hilhorst

Abstract

This study investigated the expression of various genes related to seed development, desiccation tolerance (DT), cell cycle and cytoskeleton during loss and re-establishment of DT in germinated seeds of Medicago truncatula Gaertn. cv. Jemalong A17. Gene expression was monitored by quantitative real time PCR before, during and after germination, after osmotic treatment and after dehydration of PEG-treated germinated seeds, using specific primers for M. truncatula. Clear changes in transcript abundance were detected during and after germination and in response to osmotic treatment and dehydration. DT-related genes (EM6, PER1 and sHSP18.2) were down regulated during germination and up regulated by osmotic treatment, which correlated with the loss and reacquisition of DT in the radicles. The expression pattern of ABI3 after germination was similar to that of the stress related genes, corresponding with a possible control of the stress response by this gene. Abundance of LEC1 transcript correlated more with the germination process than with osmotic stress. The cytoskeleton genes (ACT and TUB) were up-regulated during germination, not affected substantially by osmotic treatment and down regulated by subsequent dehydration, which was related with the massive breakdown of the cytoskeleton upon germination. Expression of CDC2a, one of the key regulators of the G1-to-S transition, was clearly associated with the occurrence of the first cell cycle in the growing radicle. The negative response to dehydration diminished with increasing radicle length. Our data shows that radicles that have gone through the first cell cycle may respond to osmoticum in a similar fashion as desiccation tolerant radicles. However, the occurrence of the first cell cycle appears to be an overriding factor that abolishes re-establishment of DT. Chapter 4

Introduction

Orthodox seeds acquire desiccation tolerance (DT) during their development and are shed in a metabolically quiescent state. DT is a complex feature, and its acquisition has been related to the expression of developmental genes, such as ABI3, VP1, FUS3, LEC1 and LEC2 (Koornneef et al., 1984; Parcy et al., 1994, 1997; Harada, 2001; Zeng et al., 2003). The ABI3 (abscisic acid-insensitive) gene is one of the most studied genes in relation to seed development. It is specifically expressed in seeds (Parcy et al., 1994) and its protein is a member of a large group of transcription factors that act as intermediates in regulating ABA- responsive genes during seed development. In the abi3 Arabidopsis mutant, several aspects of seed maturation, such as DT, degradation of chlorophyll and accumulation of storage compounds are abolished. Seeds show no dormancy and display a poor longevity (Koornneef et al., 1984; Léon-Kloosterziel et al., 1996; Zeng et al., 2003). LEC1, one of the Arabidopsis LEAFY COTYLEDON (LEC) genes also affects embryo development. LEC1 mRNA accumulates specifically during seed development (Lotan et al., 1998), in the embryo, suspensor and endosperm, and could not be detected in vegetative organs or in flower buds prior to anthesis (Harada, 2001). Several characteristics of Arabidopsis lec mutants show that the genes play critical roles also during the maturation phase. In lec mutants, embryos exhibit defective synthesis and accumulation of storage reserve, as well as higher desiccation sensitivity. Embryos of lec1 display characteristics of seedlings, such as the activation of the shoot and root apical meristems and the expression of genes that are normally expressed post germination. The cotyledons are in general not well defined and revert partially to a more leaf-like organ, with trichomes. Vivipary occurs occasionally (West et al., 1994; Harada, 2001). Another developmental gene, not reported to be related to acquisition of DT, but to the switch between the opposing developmental programs of dormancy and germination in Arabidopsis is COMATOSE (CTS). CTS acts, at least in part, by profoundly affecting the metabolism of stored lipids. It regulates germination potential by enhancing after-ripening, sensitivity to gibberellins and pre-chilling, and by repressing the activities of loci that activate embryo maturation (Holdsworth et al., 2001). CTS expression transiently increases shortly after imbibition during germination, but not in imbibed dormant seeds (Footitt et al., 2002). In Arabidopsis, the cts mutation specifically blocks germination potential. Such a mutation requires the prior action of ABA1, ABI3, FUS3 and LEC1 to induce embryo dormancy, suggesting that these loci are suppressed by CTS (Russell et al., 2000). Several stress-related genes have been associated with DT in orthodox seeds, such as those that code for late embryogenesis abundant (LEA) proteins, heat shock proteins (HSPs), peroxiredoxins (Prx) and ABA-responsive protein kinases (PKABA). It is generally assumed that LEA genes play a role in the establishment of DT during seed development (Delseny et al., 2001). LEA proteins have physical properties consistent with a role in DT, e.g. they are extremely hydrophilic and resistant to denaturation (Oliver et al., 2000). The

70 Gene expression during loss and re-establishment of DT in germinated Medicago truncatula seeds

Em genes correspond to class I of the LEA genes and have been found in a large number of species (monocots, dicots and gymnosperms) (Delseny et al., 2001). HSPs are thought to be involved in the correct refolding of proteins that are partially denatured by heat shock (Delseny et al., 1994). However, their importance extends beyond the protection from high- temperature stress (Vierling, 1991), since their role is also seen as either preserving or repairing macromolecular structures during dehydration or rehydration, respectively (Helm and Abernethy, 1990). HSPs are actually members of multi-gene superfamilies in which not all members are regulated by heat (Vierling, 1991). Peroxiredoxins are enzymatic antioxidants, and are part of complex protective mechanisms that have been evolved by plants to mitigate and repair the damage initiated by free radicals (Mowla et al., 2002). The 1-Cys Prx genes are expressed solely in seeds, and only in the parts of the seeds surviving desiccation (Aalen, 1999). The expression level increases late in seed development and is maintained in mature seeds during storage (Mowla et al., 2002). Protein kinases are important components in signal transduction pathways leading to cellular adjustments in response to changes in extracellular conditions (Holappa and Walker- Simmons, 1995). In embryos, the accumulation of transcripts of ABA-responsive protein kinase 1 (PKABA1) correlates with the ABA levels as the seeds mature, implying a role of PKABA1 in the control of dehydration tolerance or seed dormancy (Xiong and Zhu, 2001). Upon imbibition of non-dormant orthodox seeds, metabolism is quickly resumed with elementary cellular activities such as respiration, enzyme and organelle activity, and RNA and protein synthesis (Bewley and Black, 1994). With this shift from the quiescent state to a growing seedling, it is likely that thousands of genes are turned on or off (Bradford et al., 2000). The cell cycle, which is regulated by the activity of cyclin-dependent kinases (CDKs) (Huntley and Murray, 1999), is also resumed during germination or early seedling growth when DNA synthesis and cell division are required. Microtubules and actin filaments, the two key components of the cytoskeleton, are reconstituted during seed germination and post- germinative growth, and are involved in organelle positioning, cell elongation and cell division (Jing et al., 1999; Kost et al., 2002; Faria et al., 2005). Desiccation tolerance is lost as germination progresses (Reisdorph and Koster, 1999), and several events are accredited to be involved with this transition from a tolerant to a sensitive state. These events include the resumption of DNA synthesis (Sargent et al., 1981; Deltour, 1985; Boubriak et al., 2000) and cell division (Berrie and Drennan, 1971; Osborne et al., 2002), cell elongation (Dasgupta et al., 1982) and a decrease in the content of sucrose and larger oligosaccharides (Koster and Leopold, 1988). It has been shown that DT can be re-established in young seedlings of cucumber and M. truncatula by osmotic treatment through incubation in polyethylene glycol – PEG (Bruggink and van der Toorn, 1995; Buitink et al., 2003). The slight desiccation experienced by plant tissues exposed to water potentials of about -1.5 MPa may induce production of protectants in the cells (Vertucci and Farrant, 1995).

71 Chapter 4

The present work aimed to identify putative control points in the re-establishment of DT. For this we investigated the changes in the expression of various key genes (Table 1), associated with the aforementioned events, at the transcript level during loss and re- acquisition of DT in imbibing and germinated seeds of M. truncatula. We hypothesize that re- establishment of DT depends on the repression of genes related to the vegetative stage, and re-activation of genes related to seed development and maturation.

Materials and Methods

Seed material

Medicago truncatula Gaertn. cv. Jemalong A17 plants were routinely grown in an environmentally-controlled growth chamber (16/8 h photoperiod; 170 µmol/m2/s; 25oC; 60% RH). Mature pods were collected at shedding, around 30 d after flowering, stored at 20oC (Journet et al., 2001) and seeds were extracted manually when needed.

Seed germination and assessment of desiccation tolerance (DT)

To overcome the combined coat-imposed and embryo dormancy, seeds of M. truncatula were chemically scarified in concentrated sulphuric acid for 5 min and cold imbibed (4oC) for 36 h, before set to germinate at 20oC in the dark (modified from Sieberer et al., 2002). To characterize loss of DT after germination, germinated seeds with radicle lengths of 1, 2 and

3 mm were dehydrated over a saturated solution of K2CO3 (43% RH) in a closed box with circulating air at 23oC for 3 d. After dehydration, germinated seeds were pre-humidified in humid air (100% RH) for 24 h at 20oC to avoid imbibitional damage and then rehydrated in Petri dishes over moistened filter papers at 20oC in the dark (Faria et al., 2005). Seeds that resumed radicle growth and developed normally were considered desiccation tolerant. Four independent experiments of 50 germinated seeds each were done.

Re-establishment of desiccation tolerance in germinated seeds

The technique to re-establish DT in germinated seeds of M. truncatula Gaertn. cv. Jemalong A17 is described elsewhere (Faria et al., 2005). In short, germinated seeds with radicle lengths of 1, 2 and 3 mm were incubated for 3 d in a -1.8 MPa solution of polyethylene

72 Gene expression during loss and re-establishment of DT in germinated Medicago truncatula seeds

glycol (PEG) 6000 at 5oC in the dark. After incubation, they were rinsed thoroughly in distilled water and then dehydrated as described before. After pre-humidification and rehydration (as described before), seeds that resumed radicle growth and developed normally were considered desiccation tolerant.

RNA isolation and cDNA synthesis

Preliminary tests had shown that the radicle was the most desiccation sensitive part of the germinated seed of M. truncatula. For this reason, and to avoid mixing embryo parts with different levels of DT, RNA was isolated from radicles only. Radicles were excised from the seeds before (dry seed), during and after germination, following PEG treatment of germinated seeds and following dehydration of PEG-treated germinated seeds, totalizing 12 treatments (Table 2). For each treatment, approximately 100 radicles (20 to 50 mg) were excised, frozen in liquid nitrogen and ground with a mortar and pestle, also in liquid nitrogen. Total RNA was extracted using the RNeasy Plant Mini Kit (Qiagen Benelux, Venlo, The Netherlands), according to the manufacturer’s directions. RNA was further treated with DNase for 30 min at 37°C (20 µL total RNA, 3 µL RNase-free DNase, 10 µL buffer and 67 µL diethyl pyrocarbonate (DEPC) water). DNase was removed by phenol–chloroform–isoamyl alcohol (25:24:1), RNA was precipitated for 2 h at -20°C (0.1 volume 3 M sodium acetate, 2.5 volume 100% ethanol) and then resuspended in 20 µL DEPC water. RNA quality was analysed in 1.5% agarose gel electrophoresis, stained with ethidium bromide and quantified with a spectrophotometer (Eppendorf BioPhotometer, Eppendorf, Hamburg, Germany). cDNA synthesis was performed using the iScript cDNA synthesis kit (Bio Rad, Hercules, CA, USA) according to the manufacturer's protocol. Reactions were prepared by mixing the total RNA with iScript reaction mix and iScript reverse transcriptase and performed in a thermal cycler (iCycler, Bio Rad, Hercules, CA, USA), as follows: 5 min at 25oC; 30 min at 42oC and 5 min at 85oC. A negative control for cDNA synthesis was also included, by omitting reverse transcriptase to the reaction.

Quantitative real-time PCR (qRT-PCR) reaction

Quantitative real-time PCR was performed using an iCycler iQ instrument (Bio Rad, Hercules, CA, USA) with gene specific primers, cDNA and iQ SYBR green supermix (Bio Rad, Hercules, CA, USA). The amplification protocol consisted of 3 min at 95oC; then 40 cycles of 15 s at 95oC and 1 min at 60oC, followed by 1 min at 95oC and 1 min at 55oC. Melting curves of the products formed were determined in order to assure their specificity. Specific primers for M. truncatula (Table 3) were designed based on gene sequences available at the TIGR database (MtGI Release 8.0; www.tigr.org/tigr-scripts/tgi/T_index.cgi?species=medicago). Primer

73 Chapter 4

design was performed with Jellyfish software version 3.2 (LabVelocity, Inc., Los Angeles, USA; www.jellyfish.labvelocity.com). A specific primer for the M. truncatula 18S gene was used as internal control to normalize the other products. A negative control (negative cDNA) and a water control were used in every PCR plate. Real-time PCR evaluations were replicated three times for each treatment. Values of fold change in gene expression (mRNA) in relation to the control (dry seed) were calculated using the 2-∆∆Ct method and plotted in graphs for comparison purposes. Data from real-time PCR evaluations of gene expression were statistically analyzed using Genestat (VSN International Ltd., Herts, UK).

Table 1. Genes (mRNAs) monitored during loss and re-establishment of desiccation tolerance in imbibing and germinated seeds of Medicago truncatula.

Gene symbol Gene name Related to ABI3 Abscisic acid (ABA)-insensitive Seed development1 ACT Actin Cytoskeleton2 CDC2a Cyclin-dependent kinase Cell cycle3 CTS Comatose Seed development4 EM6 Early methionine 6 Desiccation stress5 LEC1 Leafy cotyledon Seed development6 PKABA1 ABA-induced protein kinase Desiccation stress7 PER1 1-Cys peroxiredoxin Desiccation stress8 sHSP18.2 Small heat-shock protein Desiccation stress9 TUB Tubulin Cytoskeleton2 1Koornneef et al. (1984); 2Kost et al. (2002); 3Mironov et al. (1999); 4Holdsworth et al. (2001); 5Delseny et al. (2001); 6Lotan et al. (1998); 7Anderberg and Walker- Simmons (1992); 8Aalen (1999); 9Wehmeyer and Vierling (2000).

Table 2. Treatments applied to Medicago truncatula seeds and seedlings in order to study changes in gene expression during loss and re-acquisition of desiccation tolerance.

Treatment Sample None dry seed (control) 36h at 4 °C radicle before protrusion 36h at 4 °C + 6h at 20 °C radicle before protrusion 36h at 4 °C + xh at 20 °C 1-2-3 mm radicle protrusion 3d PEG, -1.8 MPa, 5 °C 1-2-3 mm radicle protrusion 3d PEG, -1.8 MPa, 5 °C + 3d dehydration 1-2-3 mm radicle protrusion

74 Gene expression during loss and re-establishment of DT in germinated Medicago truncatula seeds

Table 3. Specific primers for Medicago truncatula used for real-time PCR.

Name Forward (5’Æ3’) Reverse (5’Æ3’) ABI3 ACCGGTGATTTTGTGAAAGC ATGGCGTGCCATTATTATCC ACT TCCATCATGAAGTGCGATGT AACCTCCGATCCAGACACTG CDC2a ACCCCAGTTGATGTTTGGTC CCACGGTTGCTAGGTCCTTA CTS GGTCTTCCACATGGCAAGTAA AAACCCAAGGCCAAGTAACA EM6 GGCAAAGCAAGGAGAGACTG ACCTTCCTCTTCAGCACGTT LEC1 TGGGTTTTGGAAAATGGAAG CAGCTCCGAATGAAAAGACC PKABA1 GGCGCTTATCCTTTTGAAGA CAGCTTCCACCTTCCCATTA PER1 GCACATACTCCAGGTGCAAA ACCACCCTCAGCACTTCATC sHSP18.2 CACGTGTGGACTGGAAAGAA TCTCTCAACACGATGCCACT TUB GGATAACGAGGCGATCTACG CGAGGATACGGCACAAGATT 18S TGACGGAGAATTAGGGTTCG CCTCCAATGGATCCTCGTTA

Results

Protruded radicles can be rescued from desiccation damage by PEG-treatment

DT was quickly lost after germination of M. truncatula seeds as assessed by the survival of directly dehydrated (without previous PEG treatment) protruded radicles followed by pre- humidification and rehydration. As soon as the protruded radicle reached a length of 1 mm, DT dropped from 100% (non germinated, dry seed) to 12%. Germinated seeds with protruded radicle length of 2 mm or greater showed no DT any more. However, when incubated in PEG before dehydration seeds with a protruded radicle length of 1 and 2 mm were still highly desiccation tolerant (95% and 84%, respectively). From 2 mm onwards there was an abrupt drop in DT to 33% at 3 mm, 5% at 4 mm and 0% at 5 mm (Faria et al., 2005).

Developmental genes (ABI3, LEC1 and CTS)

ABI3 expression increased 13-fold from the dry seed level after 36h of imbibition at 4 °C (Fig. 1A). After an additional 6h at 20 °C, which is approximately 2 hours before radicle protrusion, ABI3 expression had increased 16-fold. However, within two hours after radicle protrusion (1 mm germinated seed) ABI3 expression level had dropped close to the dry seed level and decreased further to 14-fold lower than the dry seed level. The PEG treatment resulted in higher ABI3 expression in the growing radicle which did not drop below dry seed

75 Chapter 4

levels, both before and after dehydration. Dry seed LEC1 expression was not affected by imbibition for 36h at 4 °C but increased approximately 10-fold after an additional 6h at 20 °C (Fig. 1B). Expression of this gene did not clearly drop below this level at all stages of radicle growth with or without the PEG treatment, with or without dehydration. CTS expression increased 3-fold upon imbibition and did not change much until radicles became longer than 2 mm, after which expression decreased (Fig. 1C). As compared to the untreated seeds, PEG treatment resulted in higher expression level in 1-mm radicles and similar expression in 2- mm radicles which was maintained in the 3-mm radicles. The most profound effect of dehydration after the PEG treatment was observed in the 1-mm radicles, which displayed a strong decrease compared to the non-dehydrated seeds, whereas expression in the 2- and 3-mm radicles was similar to the non-dehydrated seeds.

Stress-related genes (EM6, PER1, sHSP18.2 and PKABA1)

Transcript abundance of EM6 gene decreased slightly below dry seed levels during imbibition and decreased from 20-fold lower in 1-mm protruded radicles to more than 250-fold lower in 3-mm radicles (Fig. 2A). The PEG-treatment re-induced EM6 expression to approximately 5 to 10-fold lower expression than in the dry seeds in 1- and 2-mm radicles, but to significantly lower (20-fold) expression in the 3-mm radicles. However, this was still substantially higher than in the untreated seeds. Dehydration had only minor effects on gene expression in the PEG treated seeds. Expression patterns of both PER1 and sHSP18.2 were generally similar to that of EM6 over the different stages of imbibition and germination (Fig. 2B, C). A notable difference was displayed by sHSP18.2 which was significantly down regulated during the first 42h of imbibtion (Fig. 2C). PEG treatment (with and without dehydration) re-induced expression of these genes up to, or slightly below, dry seed expression levels from as low as 275-fold lower than dry seeds (in 3-mm radicles). Expression of PKABA1 increased 8-fold upon imbibition and gradually decreased to the original dry seed levels after germination (Fig. 2D). PEG treatment substantially increased (20-30-fold) transcript levels whereas dehydration caused a decrease in PKABA1 expression in all radicle lengths. However, expression levels remained 5- to 15-fold higher than in the dry seeds.

76 Gene expression during loss and re-establishment of DT in germinated Medicago truncatula seeds

20 20 A. ABI3 15 15

10 10 a 5 a 5 b b b 0 0 b b

Fold change -5 c -5

-10 -10

-15 a -15 20 20 B. LEC1 a

15 15 a

a b a 10 a 10 a a c 5 5

Fold change change 0 0

-5 -5

7 C. CTS a 7

6 6

5 5 b 4 4 a a a a

3 a 3 changeFold 2 2 c b 1 1

0 0

36h36h 36+6h 36+6h 1 mm 2 mm 3 mm DT (%) = 12 95 0 84 0 33

imbibition time protruded radicle length Figure 1. Relative fold change in transcript abundance of (A) ABI3, (B) LEC1 and (C) CTS in radicles of Medicago truncatula, following imbibition and subsequent radicle growth, as compared to the dry seed. Treatments: imbibed and germinated seeds (black bars), PEG- treated germinated seeds (light gray bars) and PEG-treated + dehydrated germinated seeds (grey bars). Data show the mean of three replicates. Lower case letters compare means within every group of radicle length. Different letters indicate significance of difference based on least significant difference (l.s.d.) of measured Ct values (p ≤ 0.05). Desiccation tolerance (DT) of untreated (black bars) and PEG-treated + dehydrated (grey bars) germinated seeds is indicated below each radicle length. 77 Chapter 4

0 0 c c c c 0 0 -10 b -10 b c b -10 b b -10 -20 -20 a c -20 a -20 -30 -30 b -30 -30

-40 -40 -40 -40 changeFold a a -50 -50 -50 -50 -260 -260 -265 A. EM6 a -265 -270 C. sHSP18.2 a -270 -300 -300 20 c 20 D. PKABA1 a 30 30 10 10 b b c a 0 0 a a b b -10 -10 20 a 20 -20 -20 a -30 -30 changeFold 10 b 10 -40 a -40 b -200 -200 b B. PER1 b -2400 a -240 0 0 36h 36+6h 1 mm 2 mm 3 mm 36h 36+6h 1 mm 2 mm 3 mm DT (%) = 12 95 0 84 0 33 DT (%) = 12 95 0 84 0 33

imbibition time protruded radicle length imbibition time protruded radicle length

Figure 2. Relative fold change in transcript abundance of (A) EM6, (B) PER1, (C) sHSP18.2 and (D) PKABA1 in radicles of Medicago truncatula, following imbibition and subsequent radicle growth, as compared to the dry seed. Treatments: imbibed and germinated seeds (black bars), PEG-treated germinated seeds (light gray bars) and PEG-treated + dehydrated germinated seeds (grey bars). Data show the mean of three replicates. Lower case letters compare means within every group of radicle length. Different letters indicate significance of difference based on l.s.d. of measured Ct values (p ≤ 0.05). Desiccation tolerance (DT) of untreated (black bars) and PEG-treated + dehydrated (grey bars) germinated seeds is indicated below each radicle length.

Cytoskeleton (ACT and TUB) and cell cycle (CDC2a) genes

Expression of ACT, TUB and CDC2a was 50-, 220- and 75-fold up regulated, respectively, during imbibition (Fig. 3). Expression of both ACT and TUB did not show large changes at the 1- and 2-mm stages but decreased to 20- and 100-fold at the 3-mm stage, respectively. Transcript abundance of CDC2a increased further to 140-fold as compared to the dry seed and gradually decreased during radicle growth. PEG treatment had no significant effect on gene expression levels of all 3 genes in 1- and 2-mm radicles, except for ACT expression, which doubled by PEG in 1-mm radicles. The effect of PEG was most profound in 3-mm radicles in which gene expression was maintained at the 2-mm level, resulting in levels that

78 Gene expression during loss and re-establishment of DT in germinated Medicago truncatula seeds

80 80 A. ACT a

60 60 a a a 40 b 40

change Fold b 20 20 c c b

0 0 250 250 B. TUB a a 200 a a 200 a 150 150 b 100 100

Fold change

50 50 c b b 0 0 160 160 C. CDC2a a 140 140 120 a 120

100 100 a 80 a 80

60 a 60

Fold change ab 40 40 b b 20 b 20 0 0 36h 36+6h 1 mm 2 mm 3 mm DT (%) = 12 95 0 84 0 33

imbibition time protruded radicle length

Figure 3. Relative fold change in transcript abundance of (A) ACT, (B) TUB and (C) CDC2a in radicles of Medicago truncatula, following imbibition and subsequent radicle growth, as compared to the dry seed. Treatments: imbibed and germinated seeds (black bars), PEG- treated germinated seeds (light gray bars) and PEG-treated + dehydrated germinated seeds (grey bars). Data show the mean of three replicates. Lower case letters compare means within every group of radicle length. Different letters indicate significance of difference based on l.s.d. of measured Ct values (p ≤ 0.05). Desiccation tolerance (DT) of untreated (black bars) and PEG-treated + dehydrated (grey bars) germinated seeds is indicated below each radicle length.

79 Chapter 4

were twice that of the non-treated radicles. Dehydration after PEG treatment caused a substantial lowering of expression to levels at or below initial expression during imbibition, except CDC2a, for which the decrease was less profound.

Discussion

During germination seeds will switch from a developmental to a germination program (Bewley, 1997). They display high metabolic activity and will start breaking down reserves for energy supply and de novo synthesis of components required for growth. During this process DT will be lost as it is part of the seed developmental program. By definition, seed germination is completed when the radicle protrudes the surrounding tissues. From this moment on the germinated seed is regarded as a vegetatively growing seedling. LEC1 and ABI3 encode for transcription factors that are active during late embryogenesis and seed maturation, respectively, in Arabidopsis (Raz et al., 2001). LEC1 has been implicated in the arrest of embryonic growth and ABI3 in the acquisition of DT and dormancy. Studies with mutants have clearly demonstrated that ABI3, LEC1, LEC2 and FUS3 genes are required for normal seed maturation (Léon-Kloosterziel et al., 1996), with “normal” referring to the processes occurring in wild type orthodox seeds. In general, LEC1 and ABI3 (with FUS3) are thought to activate embryo development and to repress seedling development and are considered seed specific (Parcy et al., 1997; Harada, 2001; Holdsworth et al., 2001). However, ectopic expression of ABI3 in vegetative tissues of Arabidopsis resulted in increased freezing tolerance through enhanced responsiveness to ABA (Tamminen et al., 2001). In the present study we show that ABI3 expression increases during imbibition but diminishes during radicle growth, as compared to the dry seed. However, PEG treatment of germinated seeds maintained ABI3 expression in the growing radicle above or at the dry seed level. This suggests that ABI3 expression may occur in vegetative tissues as a response to osmotic stress and may be associated with the re- establishment of DT in the radicle tissue. However, despite a still 15-fold higher expression in the 3-mm PEG treated radicles, as compared to the untreated seed, dehydration led to a loss of viability. LEC1 expression was maintained at approximately 10-fold higher expression during radicle growth in both treated and untreated seeds, demonstrating that this gene is not exclusively expressed in seed tissue but also in the growing seedling. CTS is required for the generation of germination potential in Arabidopsis. It has been demonstrated that it does so by repressing embryo dormancy through negative interaction with ABI3, ABA1, FUS3 and LEC1 (Russell et al., 2000). However, the present study shows that the higher expression levels of ABI3 and LEC1 are not resulting in suppression of CTS expression. For example, CTS expression in 1-mm radicles was significantly higher after PEG treatment than in untreated seeds but also ABI3 expression was significantly higher in the treated seeds.

80 Gene expression during loss and re-establishment of DT in germinated Medicago truncatula seeds

Synthesis of small heat shock proteins, LEAs and peroxiredoxins is part of the developmental program of seed maturation. Concomitantly with the progress of germination the expression of these stress-related genes diminishes, reflecting the loss of DT. This general picture has been shown for a number of stress related proteins and genes in different species, including dehydrins in pea seeds (Baker et al., 1995), peroxiredoxin in Arabidopsis (Haslekas et al., 1998), group-1 Lea genes in soybean (Calvo et al., 1997) and small heat shock protein (HSP17.9) in sunflower (Coca et al., 1994). However, osmotic stress may also induce or enhance expression of this group of genes in mature seeds. For example, in somatic embryogenic cultures of white spruce (Picea glauca), the inclusion of PEG in the maturation medium increased the transcript levels of various stress-related genes, such as LEAs and HSPs (Stasolla et al., 2003). Furthermore, Em expression can be induced in young seedlings by water and salt stress (Delseny et al., 1994). This is in accordance with our observations in M. truncatula in which expression of sHSP18.2, EM6 and PER1 was down regulated to levels approximately 250-fold lower in 3-mm radicles than in dry seeds. PEG treatment of the germinated seeds restored gene expression, often to dry seed levels, even after dehydration. This pattern was similar to that of ABI3. Expression of EM6 in Arabidopsis is under control of ABI3 and LEC1 as it was absent in the abi3-4 and lec1-1 mutants (Parcy et al., 1994; Vicient et al., 2000). Also the synthesis of small heat shock proteins appears to be controlled by ABI3. Wehmeyer et al. (1996) have shown that seeds of abi3 mutants, as well as the lec1-2 mutant had reduced levels of class 1 sHSPs. However, application of heat stress to mature Arabidopsis embryos induced expression of HSP17.4 in the mutants, indicating independence of the control of the HSP17.4 promoter during heat stress and seed development (Wehmeyer and Vierling, 2000). The expression of AtPER1 has also been shown to be reduced in the ABA-insensitive abi3-1 mutant, suggesting control by ABI3 (Haslekas et al., 1998). The ABA-responsive protein kinase gene PKABA1 is up-regulated in wheat seedlings by ABA, dehydration, cold and osmotic stress and was suggested to be part of the initial response to abiotic stresses, resulting in the accumulation of osmoprotectants (Anderberg and Walker-Simmons, 1992; Holappa and Walker-Simmons, 1995). The growing radicle of germinated M. truncatula seeds responded to the PEG osmotic treatment by a 20- to 30-fold up-regulation of the PKABA1 transcript, which was reduced after dehydration but still well above dry seed levels. Also the untreated seeds showed a moderate response, likely caused by the 36-h cold imbibition. These results suggest that the protruded radicles are competent to generate the PKABA1 response irrespective of their tolerance to desiccation. Our results show a correlation between expression patterns of the developmental ABI3 and the stress related sHSP18.2, EM6 and PER1 genes. This might indicate a partial reversion of the radicle tissue to the developmental state, thereby re-inducing DT. It seems obvious that ABA signalling plays a role in this process. For example, AtEM mRNAs are normally not detected in young germinating Arabidopsis seedlings, but they can be induced by an ABA treatment during the first five days of germination, indicating that the genes are

81 Chapter 4

ABA responsive (Delseny et al., 2001). However, application of ABA to 2.7-mm-protruded radicles of M. truncatula could not mimic the effect of PEG on the re-establishment of DT, whereas fluridone, an inhibitor of ABA-synthesis, reduced the effect of PEG, indicating involvement of ABA signaling (Buitink et al., 2003). These results suggest that both ABA- signaling and signaling pathways related to osmotic stress are required to re-establish DT. Radicle growth of germinated M. truncatula seeds was effectively inhibited by PEG but not by ABA (Buitink et al., 2003). Prevention of radicle (expansion) growth may be necessary to avoid transition from the G1 to S phase of the cell cycle. In M. truncatula this transition occurs between radicle lengths of 2 and 3 mm when the number of cells with 4C DNA nuclei increases from 43 to 63%. In 2-mm radicles 4C DNA content did not increase during prolonged incubation in PEG, which indicates that PEG inhibited DNA replication (Faria et al., 2005). ACT and TUB expression did not clearly respond to the PEG treatment as compared to the non-treated seeds. However, dehydration after PEG-treatment resulted in a massive drop in expression of both genes. This corroborates the observation of a complete loss of the microtubular cytoskeleton after dehydration of PEG treated germinated M. truncatula seeds (Faria et al., 2005). CDC2a is an A-type CDK and is a key regulator in the G1-to-S and G2-to- M transitions in the cell cycle (Mironov et al., 1999; Vázquez-Ramos and Sánchez, 2003). Expression of CDC2a in growing radicles is likely confined to the root apical meristem and vascular tissues (Burssens et al., 2000). Therefore, expression levels of CDC2a in the radicle of M. truncatula are increasingly underrated with the length of the protruding radicle (Fig. 3C). Compensating for this dilution effect results in maximal expression in 2-mm radicles, which is just prior to the G1-to-S transition in the majority of the seed population (Faria et al., 2005). PEG treatment maintained transcript abundance in the 1- and 2-mm radicles but doubled expression in the 3-mm radicle. The extent of reduction of transcript levels by dehydration of the treated seeds diminished with increasing radicle length. These results suggest that after the first G1-to-S transition osmotic stress or dehydration no longer repressed CDC2a transcription and, hence, a ‘point-of-no-return’ may have been crossed to allow subsequent cell cycles. Since ABA may inhibit CDC2a activity through induction of ICK1, an inhibitor of CDK activity (Wang et al., 1998), this could be due to a loss of sensitivity to ABA with increasing radicle length. During development of M. sativa seeds, sensitivity of germination to both ABA and osmoticum decreased during the maturation stage (Xu and Bewley, 1991). It is possible that this decrease continues during early seedling growth. A loss of sensitivity to ABA is also reflected by decreasing transcript abundance during radicle growth of ABI3 and the stress-related genes, which are under ABA-control. In summary, our results suggest that growing radicles of M. truncatula that are sensitive to desiccation may become desiccation tolerant when they are subjected to osmotic stress, but only prior to the occurrence of the first cell cycle. Osmotic stress induces both ABA-dependent and –independent signaling pathways that lead to the accumulation of protectants. These signaling pathways are similar to pathways occurring during seed

82 Gene expression during loss and re-establishment of DT in germinated Medicago truncatula seeds

maturation and maturation drying and involve, among others, seed specific developmental genes. Our data also shows that gene expression in radicles that have gone through the first cell cycle and have lost DT may still respond in a similar fashion, in terms of gene expression, as desiccation tolerant radicles. Apparently, the occurrence of the first cell cycle is an overriding factor that abolishes re-establishment of DT.

References

Aalen, R. B. (1999) Peroxiredoxin antioxidants in seed physiology. Seed Science Research 9, 285–295. Anderberg, R.J. and Walker-Simmons, M.K. (1992) Isolation of a wheat cDNA clone for an abscisic acid-inducible transcript with homology to protein kinases. Proceedings of the National Academy of Sciences USA 89, 10183-10187. Baker, E.H., Bradford, K.J., Bryant, J.A. and Rost, T.L. (1995) A comparison of desiccation-related proteins (dehydrin and QP47) in peas (Pisum sativum). Seed Science Research 5, 185-193. Berrie, A.M.M. and Drennan, D.S.H. (1971) The effect of hydration-dehydration on seed germination. New Phytologist 70, 135-142. Bewley, J.D. (1997) Seed germination and dormancy. The Plant Cell 9, 1055-1066. Bewley, J.D. and Black, M. (1994) Seeds: Physiology of Development and Germination, 2nd edn, Plenum Press, New York. Boubriak, I., Dini, M., Berjak, P. and Osborne, D.J. (2000) Desiccation and survival in the recalcitrant seeds of Avicennia marina: DNA replication, DNA repair and protein synthesis. Seed Science Research 10, 307-315. Bradford, K.J., Chen, F., Cooley, M.B., Dahal, P., Downie, B., Fukunaga, K.K., Gee, O.H., Gurusinghe, S., Mella, R.A., Nonogaki, H., Wu, C.T. and Yim, K.O. (2000) Gene expression prior to radicle emergence in imbibed tomato seeds, in Seed Biology: Advances and Applications (eds M. Black, K.J. Bradford and J. Vazquez-Ramos), CAB International, Wallingford, U.K., pp. 231-251. Bruggink, T. and van der Toorn, P. (1995) Induction of desiccation tolerance in germinated seeds. Seed Science Research 5, 1-4. Buitink, J., Vu, B.L., Satour, P. and Leprince, O. (2003) The re-establishment of desiccation tolerance in germinated radicles of Medicago truncatula Gaertn. seeds. Seed Science Research 13, 273- 286. Burssens, S., Himanen, K., van de Cotte, B., Beeckman, T., Van Montagu, M., Inzé, D. and Verbruggen, N. (2000) Expression of cell cycle regulatory genes and morphological alterations in response to salt stress in Arabidopsis thaliana. Planta 211, 632-640. Calvo, E.S., Wurtle, E.S. and Shoemaker, R.C. (1997) Cloning, mapping and analyses of expression of the Em-like gene family in soybean [Glycine max (L). Merr.]. Theoretical and Applied Genetics 94, 957-967. Coca, M.A., Almoguera, C. and Jordano, J. (1994) Expression of sunflower low-molecular-weight heat- shock proteins during embryogenesis and persistence after germination: localization and possible functional implications. Plant Molecular Biology 25, 479-492.

83 Chapter 4

Dasgupta, J., Bewley, J.D. and Yeung, E.C. (1982) Desiccation-tolerant and desiccation-intolerant stages during the development and germination of Phaseolus vulgaris seeds. Journal of Experimental Botany 33, 1045-1057. Delseny, M., Gaubier, P., Hull, G., Saez-Vasquez, J., Gallois, P., Raynal, M., Cooke, R. and Grellet, F. (1994) Nuclear genes expressed during seed desiccation: relationship with responses to stress, in Stress-Induced Gene Expression in Plants (ed A.S. Basra), Harwood Academic Publishers, Switzerland, pp. 25-59. Delseny, M., Bies-Etheve, N., Carles, C., Hull, G., Vicient, C., Raynal, M., Grellet, F. and Aspart, L. (2001) Late Embryogenesis Abundant (LEA) protein gene regulation during Arabidopsis seed maturation. Journal of Plant Physiology 158, 419-427. Deltour, R. (1985) Nuclear activiation during early germination of the higher plant embryo. Journal of Cell Science 75, 43-83. Faria, J.M.R., Buitink, J., van Lammeren, A.A.M. and Hilhorst, H.W.M. (2005) Changes in DNA and microtubules during loss and re-establishment of desiccation-tolerance in germinating Medicago truncatula seeds. Journal of Experimental Botany 56, 2119-2130. Footitt, S., Slocombe, S.P., Larner, V., Kurup, S., Wu, Y., Larson, T., Graham, I., Baker, A. and Holdsworth, M. (2002) Control of germination and lipid mobilization by COMATOSE, the Arabidopsis homologue of human ALDP. EMBO Journal 21, 2912-2922. Harada, J.J. (2001) Role of Arabidopsis LEAFY COTYLEDON genes in seed development. Journal of Plant Physiology 158, 405-409. Haslekas, C., Stacy, R.A.P., Nygaard, V., Culianez-Macia, F.A. and Aalen, R.B. (1998) The expression of a peroxiredoxin antioxidant gene, AtPer1, in Arabidopsis thaliana is seed specific and related to dormancy. Plant Molecular Biology 36, 833-845. Helm, K.W. and R.H. Abernethy (1990) Heat shock proteins and their mRNAs in dry and early imbibing embryos of wheat. Plant Physiology 93, 1626-1633. Holappa, L.D. and Walker-Simmons, M.K. (1995) The wheat abscisic acid-responsive protein kinase mRNA, PKABA1, is up-regulated by dehydration, cold temperature, and osmotic stress. Plant Physiology 108, 1203-1210. Holdsworth, M., Lenton, J., Flintham, J., Gale, M., Kurup, S., McKibbin, R., Bailey, P., Larner, V. and Russell, L. (2001) Genetic control mechanisms regulating the initiation of germination. Journal of Plant Physiology 158, 439-445. Huntley, R.P. and Murray, J.A.H. (1999) The plant cell cycle. Current Opinion in Plant Biology 2, 440- 446. Jing, H.C., van Lammeren, A.A.M., de Castro, R.D., Bino, R.J., Hilhorst, H.W.M. and Groot, S.P.C. (1999) ß-tubulin accumulation and DNA synthesis are sequentially resumed in embryo organs of cucumber (Cucumis sativus L.) seeds during germination. Protoplasma 208, 230-239. Journet, E.P., Barker, D., Harrison, M. and Kondorosi, E. (2001) Medicago truncatula as biological material (Module 1), in EMBO Practical Course on the New Plant Model System Medicago truncatula, pp. 1-29. (www.isv.cnrs-gif.fr/embo01/manuels/index.html; accessed on November 14, 2005). Koornneef, M., Reuling, G. and Karssen, C.M. (1984) The isolation and characterization of abscisic acid-insensitive mutants of Arabidopsis thaliana. Physiologia Plantarum 61, 377-383. Kost, B., Bao, Y.Q. and Chua, N.H. (2002) Cytoskeleton and plant organogenesis. Philosophical Transactions of the Royal Society London 357, 777–789.

84 Gene expression during loss and re-establishment of DT in germinated Medicago truncatula seeds

Koster, K.L. and Leopold, A.C. (1988) Sugars and desiccation tolerance in seeds. Plant Physiology 88, 829-832. Léon-Kloosterziel, K.M., van de Bunt, G.A., Zeevaart, J.A.D. and Koornneef, M. (1996) Arabidopsis mutants with a reduced seed dormancy. Plant Physiology 110, 233-240. Lotan, T., Ohto, M., Yee, K.M., West, M.A.L., Lo, R., Kwong, R.W., Yamagishi, K., Fischer, R.L., Goldberg, R.B. and Harada, J.J. (1998) Arabidopsis LEAFY COTYLEDON1 is sufficient to induce embryo development in vegetative cells. Cell 93, 1195–1205. Mironov, V., de Veylder, L., van Montagu, M. and Inzé, D. (1999) Cyclin-dependent kinases and cell division in plants - The nexus. The Plant Cell 11, 509-522. Mowla, S.B., Thomson, J.A., Farrant, J.M. and Mundree, S.G. (2002) A novel stress-inducible antioxidant enzyme identified from the resurrection plant Xerophyta viscosa Baker. Planta 215, 716–726. Oliver, M.J., Tuba, Z. and Mishler, B.D. (2000) The evolution of vegetative desiccation tolerance in land plants. Plant Ecology 151, 85–100. Osborne, D.J., Boubriak, I. and Leprince, O. (2002) Rehydration of dried systems: membranes and the nuclear genome, in Desiccation and Survival in Plants: Drying Without Dying (eds M. Black and H.W. Pritchard), CAB International, Wallingford, Oxon, UK, pp 343-364. Parcy, F., Valon, C., Kohara, A., Misera, S. and Giraudat, J. (1997) The ABSCISIC ACID- INSENSITIVE3, FUSCA3, and LEAFY COTYLEDON1 loci act in concert to control multiple aspects of Arabidopsis seed development. The Plant Cell 9, 1265-1277. Parcy, F., Valon, C., Raynal, M., Gaubier-Comella, P., Delseny, M. and Giraudat, J. (1994) Regulation of gene expression programs during Arabidopsis seed development: roles of the ABI3 locus and of endogenous abscisic acid. The Plant Cell 6, 1567-1582. Raz, V., Bergervoet, J.H.W. and Koornneef, M. (2001) Sequential steps for developmental arrest in Arabidopsis seeds. Development 128, 243-252. Reisdorph, N.A. and Koster, K.L. (1999) Progressive loss of desiccation tolerance in germinating pea (Pisum sativum) seeds. Physiologia Plantarum 105, 266-271. Russell, L., Larner, V., Kurup, S., Bougourd, S. and Holdsworth, M. (2000) The Arabidopsis COMATOSE locus regulates germination potential. Development 127, 3759-3767. Sargent, J.A., Mandi, S.S. and Osborne, D.J. (1981) The loss of desiccation tolerance during germination: an ultrastructural and biochemical approach. Protoplasma 105, 225-239. Sieberer, B.J., Timmers, A.C.J., Lhuissier, F.G.P. and Emons, A.M.C. (2002) Endoplasmic microtubules configure the subapical cytoplasm and are required for fast growth of Medicago truncatula root hairs. Plant Physiology 130, 977-988. Stasolla, C., van Zyl, L., Egertsdotter, U., Craig, D., Liu, W. and Sederoff, R.R. (2003) Transcript profiles of stress-related genes in developing white spruce (Picea glauca) somatic embryos cultured with polyethylene glycol. Plant Science 165, 719-729. Tamminen, I., Mäkelä, P., Heino, P. and Palva, E.T. (2001) Ectopic expression of ABI3 gene enhances freezing tolerance in response to abscisic acid and low temperature in Arabidopsis thaliana. The Plant Journal 25, 1-8. Vazquez-Ramos, J.M. and Sanchez, M. de la Paz (2003) The cell cycle and seed germination. Seed Science Research 13, 113-130.

85 Chapter 4

Vertucci, C.W. and Farrant, J.M. (1995) Acquisition and loss of desiccation tolerance, in Seed Development and Germination (eds J. Kigel and G. Galili) Marcel Dekker Inc., New York, pp. 237- 271. Vicient, C.M., Bies-Etheve, N. and Delseny, M. (2000) Changes in gene expression in the leafy cotyledon1 (lec1) and fusca3 (fus3) mutants of Arabidopsis thaliana L. Journal of Experimental Botany 51, 995-1003. Vierling, E. (1991) The roles of heat shock proteins in plants. Annual Review of Plant Physiology and Plant Molecular Biology 42, 579-620. Wang, H., Qi, Q., Schorr, P., Cutler, A.J., Crosby, W.L. and Fowke, L.C. (1998) ICK1, a cyclin- dependent protein kinase inhibitor from Arabidopsis thaliana interacts with both Cdc2a and CycD3, and its expression is induced by abscisic acid. The Plant Journal 15, 501-510. Wehmeyer, N., and Vierling, E. (2000) The expression of small heat shock proteins in seeds responds to discrete developmental signals and suggests a general protective role in desiccation tolerance. Plant Physiology 122, 1099-1108. Wehmeyer, N., Hernandez, L.D., Finkelstein, R.R. and Vierling, E. (1996) Synthesis of small heat- shock proteins is part of the developmental program of late seed maturation. Plant Physiology 112, 747-757. West, M.A.L., Yee, K.M., Danao, J., Zimmerman, J.L., Fischer, R.L., Goldberg, R.B. and Harada, J.J. (1994) LEAFY COTYLEDON1 is an essential regulator of late embryogenesis and cotyledon identity in Arabidopsis. The Plant Cell 6, 1731-1745. Xiong, L. and Zhu, J.K. (2001) Abiotic stress signal transduction in plants: Molecular and genetic perspectives. Physiologia Plantarum 112, 152–166. Xu, N. and Bewley, J.D. (1991) Sensitivity to abscisic acid and osmoticum changes during embryogenesis of alfalfa (Medicago sativa). Journal of Experimental Botany 42, 821-826. Zeng, Y., Raimondi, N. and Kermode, A.R. (2003) Role of an ABI3 homologue in dormancy maintenance of yellow-cedar seeds and in the activation of storage protein and Em gene promoters. Plant Molecular Biology 51, 39-49.

86 Chapter 5

Improvement of storability of Inga vera subsp. affinis embryos

José Marcio Rocha Faria, Lisete Chamma Davide, Edvaldo A. Amaral da Silva, Antonio Claudio Davide, Roselaine C. Pereira, André A.M. van Lammeren and Henk W.M. Hilhorst

Abstract

The remarkably short storability of recalcitrant seeds has prevented their inclusion in programs of ex situ conservation. The causes of their desiccation sensitivity and rapid decline in viability during storage are not fully elucidated yet. In the present study the highly recalcitrant but fully viable seeds of Inga vera subsp. affinis were stored under various conditions and analyzed physiologically and cytologically at intervals, in order to obtain more insight in their loss of viability during storage. Seeds stored fully hydrated at 5oC totally lost viability in 18 days. Sealed storage of partially dehydrated seeds slightly prolonged their viability, with 95% germination being attained after 14 days of storage. However, after 30 days of storage, viability was completely lost. Storage of hydrated seeds in a solution of polyethylene glycol (PEG) at -1.7 MPa was capable of maintaining high germinability until 30 days of storage. When added to the PEG solution, abscisic acid (ABA) showed a strong temperature-dependent interaction, with a positive effect on the longevity of seeds stored at 20oC and a negative effect on seeds stored at 5oC. However, in any case, seed viability could not be maintained longer than 50 days. Seeds collected from the same trees in the following year showed better storability and still attained 45% germination after 62 days of storage. Starch content seemed to decline with time during storage but no clear correlation was found with viability. Analysis of cellular alterations during storage and viability loss of the seeds showed disappearance of starch granules and various damages to the cells, such as cell wall folding and cytoplasm fragmentation. Chapter 5

Introduction

In recent years, awareness concerning the loss of plant diversity has captured worldwide attention and germplasm conservation has become necessary as a means of maintaining species diversity to prevent genetic erosion (Marzalina and Krishnapillay, 1999). While it is relatively easy to store orthodox and intermediate seeds in the long- and medium-term, respectively (Roberts, 1973; Ellis et al.; 1990), one of the greatest challenges for those who deal with seed banks is the conservation of recalcitrant seeds. The difficulty of conserving recalcitrant seeds is well recognized (Marzalina and Krishnapillay, 1999) and has prevented their inclusion in programs of ex situ conservation. Generally, recalcitrant seeds are shed with high moisture content and high metabolic activity. Their viability is lowered or even totally lost when moisture content is reduced to a certain relatively high value and, in general, those of tropical origin are also chilling sensitive and cannot be stored at temperatures below about 15oC (Dussert et al., 1999; Pammenter and Berjak, 1999). Storage of recalcitrant seeds using the conventional facilities of seed banks is virtually unfeasible. Therefore the Genebank Standards (1994) published by FAO and IPGRI with the guidelines for seed storage, deal solely with orthodox seeds. For the long-term storage of recalcitrant seeds, cryo-storage is the only viable option to date, even though not achievable for every species (Berjak and Pammenter, 2003) and often with erratic results (Krishnapillay, 2000). The refinement of storage techniques for intermediate and recalcitrant seeds is among the key storage research subjects for the future (Smith et al., 2003). Although long- term storage of recalcitrant seeds under common conditions is not yet possible, any improvement of short-term storage would be valuable (Chin, 1989) in order to provide a suitable seed supply for forestry operations. The bad storability of recalcitrant seeds has been widely reported, but the mechanisms involved in seed death during storage are not fully elucidated yet. In general, germinability has been the only parameter used to characterize the loss of viability (Goldbach, 1979; Tompsett, 1984 and 1985; Probert and Longley, 1989; Farrant et al., 1993; Fu et al., 1993; Hong and Ellis, 1998; Tompsett and Pritchard, 1998; Carvalho, 2000; Marshall et al., 2000). Although essential for characterizing the recalcitrant behavior, decrease in germination rates does not reveal per se the causes of the poor storability, since it is the end point of several processes occurring within the stored seed, which ultimately lead to seed death. Because recalcitrant seeds are not desiccation-tolerant they must be stored moist (Greggains et al., 2000). However, even in moist conditions, their longevity is short, varying from a few weeks to a few months, depending on the species (Roberts and King, 1980a). One of the possible causes of viability loss of stored recalcitrant seeds is that they remain metabolically active during storage (Pammenter et al., 1994), thereby exhausting the seed’s food reserves. Seed death is the result of a gradual and cumulative process in which an increasing number of cells die until certain parts of the seed fail to perform essential

88 Improvement of storability of Inga vera subsp. affinis embryos

functions (Woodstock, 1973). Intracellular damages that have been reported for hydrated recalcitrant seeds during storage include degeneration of plastids, disruption of nuclear morphology, spatial disorientation of organelles and collapse of cell wall (Drew et al., 2000). Furthermore, due to their high moisture content, recalcitrant seeds can germinate or be damaged by microbial contamination during storage (King and Roberts, 1980). If metabolism is an underlying cause of the short storage life span, then reducing metabolic rate should extend longevity (Berjak and Pammenter, 2003). One possible approach is to store the seeds in an osmotic medium, such as polyethylene glycol (PEG) with or without the addition of germination inhibitors, such as abscisic acid (ABA) (Tompsett, 1985). Studies with ABA in desiccation-sensitive seeds are scarce and so far have shown that reduced ABA levels are associated with recalcitrance in embryos. Furthermore, it has been shown that embryos of recalcitrant seeds of five mangrove species were able to produce ABA in response to artificial drying (Farrant et al., 1996; Farnsworth and Farrant, 1998). Besides slowing down the metabolism, ABA can also induce desiccation tolerance in sensitive seeds and tissues, such as desiccation sensitive abi3 mutant seeds of Arabidopsis thaliana (Ooms et al., 1994), somatic embryos of Medicago sativa (Senaratna et al., 1989; Anandarajah and McKersie, 1990) and germinated seeds of cucumber (Bruggink and van der Toorn, 1995). Many of the physiological and biochemical changes caused by ABA in developing embryos can also be induced by low osmotic potentials (Bewley and Black, 1994). Either partial dehydration or ABA increased desiccation tolerance in the cyanobacterial lichen Peltigera polydactylon (Beckett et al., 2005). ABA was successfully used in watermelon seedlings in order to reduce their chilling sensitivity (Korkmaz, 2002). In cucumber seedlings, chilling tolerance was enhanced by exposing them to osmotic or heat stress, which induced the synthesis of dehydrin-like and heat-shock proteins respectively (Kang et al., 2005). Since these two protein families have been associated with desiccation tolerance (Delseny et al., 1994; Baker et al., 1995; Black et al., 1999; Wehmeyer and Vierling, 2000; Ramanjulu and Bartels, 2002), it is possible that both osmotic and heat stress can also enhance desiccation tolerance in sensitive seeds. Osmotic stress (PEG) is also capable of re-establishing desiccation tolerance in germinated seeds of cucumber (Cucumis sativus) (Bruggink and van der Toorn, 1995; Leprince et al., 2000) and Medicago truncatula (Buitink et al., 2003; Faria et al., 2005). Intermittent heat-shock treatment has been used to improve low-temperature fruit storage of recalcitrant species, such as mango and avocado (Marcellin, 1992). Incubation in PEG, slow drying and heat-shock, independently, were able to increase longevity in primed seeds (Bruggink et al., 1999). It seems therefore that these techniques have the potential to decrease the desiccation sensitivity shown by recalcitrant seeds and/or increase their longevity. Inga vera subsp. affinis, seeds of which were used in this study, is a leguminosae tree species typical of the riparian forests in south-eastern Brazil, and one of the most important species for ecological restoration of those ecosystems. Its seeds are recalcitrant and have one of the worst storabilities known (Faria et al., 2004). In order to develop

89 Chapter 5

storage protocols for recalcitrant seeds a better knowledge of the events associated with storage is mandatory. In the present study we attempted to decrease metabolic activity and/or desiccation sensitivity by low osmotic potential, ABA, heat shock and low temperature to improve current storability protocols. In parallel we characterized physiologically and cytologically the loss of viability that occurs in Inga vera embryos during storage.

Materials and Methods

Plant material

Mature seeds of Inga vera subsp. affinis were obtained from fruits hand-harvested in January 2004 and January 2005 from 20 adult, healthy trees, growing in riparian forests along the Rio Grande at the surroundings of the city of Lavras, Minas Gerais State, Brazil. The coordinates of the collection site are 21o09’S and 44o53’W, and the altitude is 808 m. The climate is classified as a transition between Cwb and Cwa, characterized by hot and humid summers with mild and dry winters, according to Köppen’s classification (Köppen, 1936). On the same day of fruit collection seeds were manually extracted, processed and sampled for the experiments described below. During processing of the seeds, the sarcotesta was removed, which caused removal of the inner layers of the seed coat as well, ending up with the naked embryo. Removal of the sarcotesta was an absolute prerequisite since it contains high amounts of sugars which results in excessive growth of micro-organisms during storage. All embryos were surface sterilized by washing in a solution of sodium hypochlorite (1%) for 10 minutes, followed by thorough rinsing in distilled water.

Assessment of moisture content

Moisture contents (MCs) of the embryos were determined gravimetrically on four replications of 5 embryos by oven drying them for 17 h at 103oC (ISTA, 1996). MCs were calculated as g

H2O/g dry matter (hereafter g/g). Embryos stored in PEG were opened by separating the cotyledons and washed rapidly before MC determination, in order to remove PEG residues.

Germination tests

Germination tests of fresh and stored embryos were carried out on moistened sand in plastic boxes (11 x 11 x 3.5cm), at 30oC and constant light. Four replications of 20 embryos were

90 Improvement of storability of Inga vera subsp. affinis embryos

used. Germination was scored daily, with visible growth of the radicle as the criterion for germination.

Embryo storage

The storability of I. vera embryos was assessed in 15 treatments, distributed over two experiments.

Storage experiment 1

Embryos from seeds collected in January 2004 were placed in a bench top humidity chamber (Hotpack model 435314, Warminster, PA, USA), with running air (43% RH and 25oC) and were partially dried, from 1.35 to 1.02 g/g. The embryos were then packed in aluminized plastic bags (light- and air-protected; 150 embryos/bag). Air was removed from the packages by using a vacuum pump until the plastic pressed moderately hard against the embryos. Viability, as assessed by a germination test immediately after packing and sealing, was not affected by this method of air removal (data not shown). The bags were sealed and stored at 5 and 20oC, in a refrigerator and an incubator, respectively. This protocol is further referred to as ‘sealed storage’. After one week a heat-shock was applied to half of the embryos stored at 5oC, by transferring the bags to an incubator at 30oC and keeping them at this temperature for 24 hours, after which they were transferred back to 5oC. Experiment 1 was thus composed of three treatments (Table 1). Non-stored embryos (control, kept at 20°C) and embryos stored for 14 and 30 days were sampled for germination test, moisture content assessment and cytological studies (scanning electron microscopy).

Storage experiment 2

In the second experiment, embryos from seeds collected in January 2004 were stored with their original moisture content (1.35 g/g), in a single layer in plastic trays (40 x 25 x 8 cm; 350 embryos/tray) containing a solution of polyethylene glycol (PEG 8000; Sigma-Aldrich, Zwijndrecht, The Netherlands) at -1.7 MPa, enough to slightly cover the embryos. PEG solutions were prepared with (100 µM) or without abscisic acid (ABA; Sigma-Aldrich, Zwijndrecht, The Netherlands). The trays were covered with a plastic sheet to decrease the evaporation and stored at 5 and 20oC, in a refrigerator and an incubator, respectively, in the dark. After one week, a heat-shock was applied to half of the embryos stored at 5oC, by transferring the trays to an incubator at 30oC, in the dark for 24 hours, after which they were transferred back to 5oC. Two weeks after the beginning of storage, half of the stored embryos was removed from the PEG solution, washed, blot-dried, air dried (method as in experiment 1) to circa 0.64 g/g, to test for a possible decrease in desiccation sensitivity, and put into aluminized plastic bags (sealed conditions, as in experiment 1). The packages were

91 Chapter 5

then transferred back to their original temperatures (5 or 20oC). The other half of the stored embryos remained in PEG. Experiment 2 was therefore composed of 12 treatments (Table 1). Non-stored embryos (control) and embryos stored for 14, 30 and 50 days were sampled for germination test, moisture content assessment and cytological studies (light and scanning electron microscopy). Based on the results of 2004, more seeds were collected in 2005 and treatments 7, 11, 13 and 15 (Table 1) were repeated. Non-stored embryos (control) and embryos stored for 30 days (treatments 7, 11, 13 and 15) and 62 days (treatment 15) were sampled for germination test, moisture content assessment and transmission electron microscopy.

Table 1. Storage of I. vera embryos after partial dehydration (Experiment 1) or in PEG solution (Experiment 2), with or without ABA, heat-shock or dehydration in combination with sealed storage, as indicated by the plus and minus signs, respectively.

Treatment Temperature of ABA Heat- Air-dried to 0.64g/g storage (oC) shock after 14 d of storage + sealed storage Experiment 1 (partial dehydration: 1.35 Æ 1.02 g/g + sealed storage) 1 5 - + - 2 5 - - - 3 20 - - - Experiment 2 (moist storage in PEG) 4 5 - + + 5 5 - + - 6 5 - - + 7 5 - - - 8 5 + + + 9 5 + + - 10 5 + - + 11 5 + - - 12 20 - - + 13 20 - - - 14 20 + - + 15 20 + - -

Cytological analysis

For all cytological studies, 3-5 embryonic axes per treatment, were excised from the embryos and fixed by immersion for 24 hours in a modified Karnovsky solution (glutaraldehyde 2.5%, formaldehyde 2% in sodium cacodylate buffer 0.05M, CaCl2 0.001 M, pH 7.2). Thereafter

92 Improvement of storability of Inga vera subsp. affinis embryos

different protocols were followed, depending on the microscopy technique, as described below.

Scanning Electron Microscopy (SEM)

Samples were transferred to a cryo-protectant solution (glycerol 30% v/v in water) for 30 min and sectioned longitudinally in liquid nitrogen using a scalpel blade. Sections were transferred to a 1% aqueous solution of osmium tetroxide for 1 hour at room temperature and subsequently dehydrated in a series of acetone solutions (25, 50, 75, 90 and 100% twice), for 10 minutes each. Dehydrated samples were critical-point dried. Specimens were mounted on aluminum stubs, sputter-coated with gold and observed with a scanning electron microscope (LEO Evo40 XVP – Leo Electron Microscopy, Cambridge, UK). Images were generated and recorded with a working distance of 9 mm using an accelerating voltage of 20 kV. Observations were always made in the meristematic region of the radicle.

Transmission Electron Microscopy (TEM)

Samples were rinsed three times (10 min each) with cacodylate buffer 0.05M and post-fixed in a 1% aqueous solution of osmium tetroxide for 1 hour at room temperature, rinsed three times with MQ H2O for 15 min each, stained in uranyl acetate (0.5%) overnight at 4°C, and subsequently dehydrated in a series of acetone solutions (25, 50, 75, 90 and 100% twice). The dehydrated tissue was gradually infiltrated with Spurr resin/acetone (33 and 67%) and embedded in pure Spurr resin and polymerized at 68°C for 48 h. Thin sections (<100 nm) were cut in a Reichert-Jung (ultracut) ultramicrotome equipped with a diamond knife, picked up on copper slot grids, and allowed to dry onto Formvar-coated aluminum racks (Rowley and Moran, 1975). Sections were post-stained with uranyl acetate followed by lead citrate (3 min each) and examined using a transmission electron microscope (Zeiss EM 109, Carl Zeiss, Jena, Germany) operating at 80 kV. Observations were made of cells in the radicle meristeme.

Starch

Starch contents were determined enzymatically, using a commercial assay kit (Sigma, Saint Louis, MO, USA). The assay is based on enzymatic hydrolysis of starch to glucose and glucose to glucose-6-phosphate, which is subsequently assayed by NAD dependent conversion to 6-phosphogluconate. Thus, the value found for starch content includes the amount of free glucose in the sample. Samples were prepared as follows: approximately 150 mg of tissue (15 embryonic axes) from each storage treatment were ground with a mortar and pestle and transferred to a flask with 20 mL of DMSO and 5 mL of 8 M HCl. The flask

93 Chapter 5

was covered and incubated for 30 min at 60oC in a shaking water bath. Next, 50 mL of deionized water was added and the pH adjusted to 4-5 with 5 N NaOH. The solution was cooled to room temperature and diluted to 100 mL with deionized water. Starch assays were subsequently performed according to the supplier’s instructions.

Results

Moisture content

Experiment 1 – Embryos stored partially dried

Just after seed processing, MC of the embryos was 1.35 g/g, decreasing to 1.02 g/g by the partial dehydration prior to sealed storage. During 30 days of storage, MC showed a slight increase, if any, in the 3 treatments, to about 1.05 to 1.25 g/g (Fig. 1 A).

Experiment 2 – Embryos stored moist in PEG

The MC of the embryos at the beginning of the storage experiment performed in 2004 was 1.35 g/g. Embryos removed from PEG at 14 days of storage, dehydrated to 0.64 g/g and stored again in plastic bags did not show subsequent significant changes of MC (data not shown). In embryos that were not dried and remained in PEG, MC varied differently among treatments, increasing or decreasing, depending on the treatment (Fig. 2 A) and sometimes on the seed batch (data not shown). The MCs of seeds stored at 5°C (+ ABA, with or without HS) were consistently and significantly lower than the MCs of seeds stored in the absence of ABA. In contrast, ABA led to significantly higher MC at 20°C. Both the ABA effects at 5°C and at 20°C were visible after 14, 30 and 50 d of storage (Fig. 2A). In 2005 no significant effects of ABA on MC were observed (data not shown).

Germination

Experiment 1 – Embryos stored partially dried (at sealed conditions)

After 14 days of storage of partially dehydrated embryos at sealed conditions, germination dropped from 100% to less than 20% for those stored at 20oC and remained high (over 90%) for those stored at 5oC, with or without heat-shock treatment. However, after 30 days of storage, all seeds failed to germinate, irrespective of the storage temperature (Fig. 1B).

94 Improvement of storability of Inga vera subsp. affinis embryos

Embryos did not germinate during storage, neither did they show any sign of degradation upon macroscopic inspection.

AB 1.5 5oC 100 5oC 5oC (HS) 5oC (HS) 1.4 o o 20 C 80 20 C

1.3 60

1.2 40 O/g dry matter) O/g dry matter) 2 2 Germination (%) Germination content content Moisture 1.1 20 (g H (g H 1.0 0 0 0 1 14 30 001430 14 30 Before After drying drying Storage time (days) Storage time (days) Figure 1. Moisture content (A) and germination (B) of partially dried Inga vera embryos after sealed storage (experiment 1). Experiment executed in 2004. Averages are from four replications of 20 embryos. Bars represent standard deviation.

Experiment 2 – Embryos stored moist in PEG

After 14 days of storage in PEG, germination in all treatments was still high (≥ 95%; Fig. 2B). In all treatments in which embryos were removed from PEG, dehydrated to 0.64 g/g after 14 days of storage and stored again in plastic bags at sealed conditions, there was an almost total loss of viability (less than 10% of germination) in the assessment done at 30 days of storage (data not shown). Apparently, storage under sealed conditions leads to rapid viability loss, especially at higher temperatures, possibly by depletion of oxygen necessary for metabolic activity. The only treatment not subjected to dehydration and sealed storage (Fig. 2B) that also showed an almost complete loss of viability at 30 days, was that of embryos stored at 20oC: under these conditions metabolic activity is also supposed to be high. In the same evaluation, embryos stored at 20oC in the presence of ABA (supposed to slow down metabolism), attained maximum germination (100%), like fresh, non-stored embryos (Fig. 2B). High germination (over 94%) was also observed in embryos stored at 5oC (without ABA, with or without heat-shock). Embryos stored at 5oC in the presence of ABA attained 53% (with heat-shock) and 40% (without heat-shock) of germination in the evaluation done 30 days after the beginning of the storage. There was a clear relationship between MC (Fig. 2A) and germination after 30 d of storage (Fig. 2B). In the final

95 Chapter 5

evaluation, at 50 days, the only treatments still showing some germination, albeit at low levels, ranging from 8 to 16%, were storage at 5oC without ABA and at 20oC with ABA (Fig. 2B). Again, there were similarities with the MC at 50 d of storage. When part of the experiment was carried out again in 2005 to generate samples for TEM visualization, the results of storability were in general better than the previous year. The four treatments that were repeated (5°C, 5°C + ABA, 20°C and 20°C +ABA) attained over 80% germination after 30 days of storage in PEG (data not shown). Embryos stored at 20oC in the presence of ABA reached 45% germination after 62 days of storage. Because the storage experiment done in 2005 was based on the results of the previous year, only a limited number of embryos was stored enough for evaluation (germination, moisture content and TEM visualization) at 30 days (5°C, 5°C + ABA and 20°C) and 62 days (20°C + ABA) of storage. For this reason the experiment could not be extended to determine when embryos from 2005 would completely lose viability.

Starch contents

SEM observations suggested that the abundance of starch granules in radicle cells declined as storage proceeded (Fig. 3). For example, most cells from embryonic axes of non-stored fresh embryos contained abundant starch granules (Fig. 3A) whereas cells after 50 d of storage at 20°C in PEG + ABA generally appeared empty (Fig. 3B). Starch was quantified in a limited number of embryonic axes (Table 3). Embryos stored in PEG at 5°C both with and without ABA, showed a minor decrease in starch content after 30 d of storage as compared with the fresh control. In the low temperature treatment without ABA germination was maintained at or close to 100%, while ABA had a negative effect on germination (40%). Storage for 30 d at 20°C in PEG resulted in loss of viability and significant degradation of starch. In the presence of ABA, a treatment which retained full viability to 30 days, starch degradation was similar as in the embryos stored without ABA (Table 3). After 50 d of PEG- storage with ABA both germination and starch content were close to zero (cf. Fig. 3B). Partial dehydration resulted in an increase of starch content as compared with the fresh control (Table 3). Subsequent storage at 5°C for 14 d hardly lowered the starch content and the viability of the partially hydrated embryos. The heat shock treatment of partially dehydrated seeds during storage at 5°C also resulted in a low starch content after 14 d but retained a high viability. After 14 d at 20°C both starch content and germination had dropped to low levels.

96 Improvement of storability of Inga vera subsp. affinis embryos

A 2.0 5oC 5oC (HS) o 1.8 5 C (ABA) 5oC (ABA+HS) 20oC 1.6 20oC (ABA)

1.4

O/g dry matter) 1.2 2

Moisture content

(g H 1.0

0.8 0143050 B 100

80

60

40

Germination (%) 20

0 0143050

Storage time (days)

Figure 2. Moisture content (A) and germination (B) of Inga vera embryos (2004) after storage in PEG solution (experiment 2) at different temperatures, with or without ABA or heat-shock (HS). Averages are from four replications of 20 embryos. Bars represent standard deviation.

Cytological observations in stored embryos

TEM micrographs of axis cells of the fresh control embryos generally showed that full-grown Inga seeds are still metabolically active, as can be expected for such a very recalcitrant species. The typical picture emerges of a metabolically active and viable cell, with polysomes and a round nucleus with dense nucleolus and peripherally arranged heterochromatin (Fig. 4A). Mitochondria and amyloplasts (Fig. 4B) were present in large numbers. Axis cells of

97 Chapter 5

embryos stored moist in PEG showed various changes, including confluence of lipid droplets resulting in large lipid masses in the cytoplasm and loss of cellular structure (Fig. 4C). More severely damaged cells appeared totally devastated with cell wall folding, fragmented cytoplasm and amorphous nuclei (Fig. 4D). In the same tissues and at the same storage conditions both healthy appearing cells and cells damaged to various degrees were observed next to each other. Because of the limited number of observations we could not correlate the extent of the damage or the number of damaged cells to the viability of the whole embryo, as assessed by the germination tests.

A

B

Figure 3. Scanning electron micrographs of embryonic axes from stored Inga vera embryos. (A) Control, non-stored, fresh embryos (100% germination); (B) embryos stored in PEG + ABA at 20°C for 50 days (7.5% germination). Bars = 10 µm.

98 Improvement of storability of Inga vera subsp. affinis embryos

Table 3. Percentage germination and starch content of embryonic axes after different storage conditions and durations.

Germination Starch content Treatment (%) (mg/g dry matter)* Fresh control (cf. Fig. 3A) 100 44 5°C - PEG - 30d 94 37 5°C - PEG – ABA - 30d 40 37 20°C - PEG - 30 d 2.5 16 (± 3) 20°C - PEG - ABA - 30 d 100 23 (± 1) 20°C - PEG - ABA - 50 d (cf. Fig. 3B) 7.5 8 (± 1)

Partially dehydrated (control) 100 76 5°C - 14 d 95 69 5°C + heat shock - 14 d 89 24 20°C - 14 d 16 21 (± 4)

* Standard deviations are shown between brackets. Where missing, there was insufficient material for replications.

Discussion

Just after processing, MC of the embryos of I. vera was 1.35 g/g, which corresponds to a water potential close to 0 MPa (Faria et al., 2004). During storage in PEG, with a water potential of -1.7 MPa, it was expected that in all treatments embryos would lose water to the osmotic medium, thus becoming slightly dehydrated. This was indeed the case in embryos stored in PEG at 5oC with ABA, at 5oC with ABA and HS and at 20oC without ABA or HS. However, a significant increase in MC was observed with storage in PEG + ABA at 20°C after 20 and 50 days of storage. Intense respiration of stored recalcitrant seeds, which generates water (Bilia et al., 2003), can increase MC during storage, as reported for Inga uruguensis (Bilia, 1997) and Trichilia dregeana (Drew et al., 2000). However, this is unlikely to have occurred in this study since the increases in MC were not clearly associated with treatments that would enhance respiration, such as a temperature of 20oC (as compared to 5oC) or the presence or absence of ABA. The variation in MC in the different storage conditions never fell below the critical moisture content of approximately 0.7 g/g for I. vera embryos (Faria et al., 2004). In orthodox seeds, decreasing MCs increase the period of viability in storage (Roberts, 1973). It is thus reasonable to consider that partial dehydration of recalcitrant seeds also has the potential to increase their longevity, as long as MC does not drop beyond the point where dehydration injuries start to occur. Specifically for mature I. vera embryos, it has already been shown that a partial dehydration, from 1.43 to 1.00 g/g, has no negative effect on germination (Faria et al., 2004). However, in the present study, all combinations of

99 Chapter 5

sealed storage of partially dried embryos in aluminized plastic bags lead to a complete loss of viability after 30 days of storage. Low temperature storage under these conditions only slightly delayed loss of viability. These results suggest that oxygen is depleted rapidly under sealed conditions, especially at higher temperatures. The quick reduction of oxygen likely leads to fermentation and ethanol production. However, upon opening of the bags there was no smell of ethanol. This may indicate that the seeds lost their viability before oxygen levels became critically low and would induce fermentation. It is also possible that the potential for

A B

Am

Nuc

Ncl

C D

Cw Nuc

Figure 4. Transmission electron micrographs of embryonic axes from non-stored (A, B) and stored (C, D) Inga vera embryos. (A) Round nucleus (Nuc) with dense nucleolus (Ncl) and heterochromatin arranged peripherally (arrowheads); (B) amyloplasts (Am) and mitochondria (arrowheads) with normal appearance; (C) confluence of lipid droplets resulting in large masses in the cytoplasm (white arrowheads) and in the extraprotoplasmic space (black arrowhead); (D) more severely damaged cell, showing cell wall (Cw) folding, cytoplasm degradation and amorphous nucleus (Nuc). Bars: A = 5 µm; B-D = 10 µm.

100 Improvement of storability of Inga vera subsp. affinis embryos

fermentation is limited in I. vera embryos. The effect of partial dehydration on recalcitrant seed storage appears to be species-specific, since it has led to contrasting results. Positive results have been reported for Eugenia involucrata (stored in perforated plastic bags; Maluf et al., 2003), Inga uruguensis (sealed polythene bags; Bilia, 1997) and cocoa and rubber seeds (perforated polythene bags; Chin, 1989), whereas negative effects were monitored in Trichilia dregeana (sealed plastic bags; Drew et al., 2000) and Myrciaria dubia (sealed plastic bags; Ferreira and Gentil, 2003). No significant effect was found in Aesculus hippocastanum (inflated polythene bags, ventilated weekly; Tompsett and Pritchard, 1998). In our experiments the unfavourable effect of oxygen depletion during sealed storage may have masked a possible positive effect of partial dehydration on the storability. For the embryos stored moist in PEG, effects of the treatments were apparent after 30 days of storage. ABA completely prevented the loss of viability of embryos stored at 20oC but displayed a moderately negative effect on the viability of embryos stored at 5oC. According to Berjak and Pammenter (2003) the efficacy of ABA in extending the storage life span of recalcitrant seeds depends on maturity status and whether or not the seeds are responsive to it. The results shown here suggest that the efficacy of ABA is also dependent on the storage temperature. The use of exogenous ABA in recalcitrant seeds to improve storability has shown inconsistent results, with reports of positive results for some species (Melicoccus bijugatus and Eugenia brasiliensis [Goldbach, 1979]; Avicennia marina [Pammenter et al., 1997] and Acer saccharinum [Marshall et al., 2000]), as well as small or no effect for other species (Virola guatemalensis [Gonzalez and Fisher, 1997] and Inga uruguensis [Barbedo and Cicero, 2000]). The effect of ABA on recalcitrant seed storage seems to be species-specific, although the conditions of the experiments (ABA concentration and storage temperature) may obviously influence the response of the seeds. The heat shock treatment had no significant effect on the storability at 5°C of moist- stored I. vera embryos. Incubation in PEG, slow drying and heat-shock, independently increased longevity of primed seeds of Impatiens walleriana (Bruggink et al., 1999). The heat-shock treatment was most effective at the highest MCs of these seeds. Intermittent heat-shock treatment has also been used to improve fruit storability of (recalcitrant) species, such as mango and avocado (Marcellin, 1992). This treatment was applied to reduce or prevent cold damage to the fruits by allowing repair processes to proceed at the intermittent elevated temperature. Although these treatments suggest having the potential to decrease the desiccation sensitivity shown by recalcitrant seeds and/or increase their longevity, their effectiveness remains to be shown. Re-establishment of desiccation tolerance in germinated Impatiens walleriana and Medicago truncatula seeds was accompanied by an increase in sucrose content (Bruggink et al., 1999; Buitink et al., 2003). The likely source of sucrose synthesis is starch, which is a common food reserve in seeds. Thus, apart from a role as energy source in the metabolically active embryos during moist storage, starch hydrolysis could also produce the sucrose necessary for the reduction of desiccation sensitivity. We found no direct relationship

101 Chapter 5

between starch content and viability/germination: low germination was always accompanied by (relatively) low starch, but the reverse was not true: high germination also occurred together with (relatively) low starch content. In many species starch is the last reserve material to be degraded during germination (Werker, 1997). Because recalcitrant embryos are in a germination mode of development it seems plausible that they utilize starch to maintain their energy metabolism during moist storage and that low temperatures slow down metabolism. Starch mobilization has also been reported in the integument of early developing Brassica napus seeds, when it undergoes programmed cell death (Wan et al., 2002). On the contrary, seeds of four (recalcitrant) Quercus species stored at 2oC in plastic bags with their original MCs (0.35 to 0.67 g/g), showed increasing contents of insoluble carbohydrates, probably starches, during the first months of storage. Starch content attained a maximum in the fourth month, prior to the radicle protrusion, and decreased as the radicle elongated (Clatterbuck and Bonner, 1985). Apparently, seeds from these species completed their maturation program during storage, thereby accumulating food reserves which were utilized after the switch to the germination mode. In the present study we found an increase in starch content after partial dehydration. Recalcitrant seeds do not dry during their maturation period. Thus, an increase in starch content as a reflection of extended maturation is very unlikely. Alternatively, transitory starch accumulation may occur in germinated seeds (young seedlings) as a response to sucrose production from hemi-cellulose or lipid degradation, to maintain osmotic status of the cells, as has been shown in fenugreek and soybean (Dirk et al., 1999; Bewley and Black, 1994). As I. vera seeds do not contain an obvious endosperm, cell wall-located hemi-cellulose is not a likely source for such starch accumulation. It is not known if lipids are a major reserve food in I. vera embryos but our microscopic observations clearly show the presence of lipid vesicles in axis cells (Fig. 4D). However, in the present study the availability of starch does not appear to be the primary cause of cell death during storage of I. vera embryos. The presence of ABA did not affect starch degradation but it affected viability negatively at 5°C and positively at 20°C when stored in PEG. Thus, low temperature and ABA may affect viability by different mechanisms. Various features observed in embryonic cells from the control treatment, such as numerous mitochondria and polysomes and well organized rER are indicators of ongoing metabolism and progress to germination as might be expected in newly-shed recalcitrant seeds (Berjak et al., 1993; Kioko et al., 1998). Most of these characteristics were also observed in stored embryos, but with the simultaneous occurrence of sub-cellular damages that have been associated with viability loss during storage of recalcitrant seeds (Berjak et al., 1993). A correlation between the extent of cellular damages and viability seems obvious but is difficult to demonstrate. This correlation may depend on cell type, the number of cells within the axis with non-repairable damage, as well as the number of non-viable embryos within the population that is used for the viability assessment. Seeds collected in 2005 and stored in PEG performed better than those of 2004 as they were still able to germinate for 45% after 62 days of storage. A similar storability (43%

102 Improvement of storability of Inga vera subsp. affinis embryos

germination after 60 days) was found by Bilia et al. (1999), with partially dehydrated (from 1.34 to 1.05 g/g) embryos of Inga uruguensis stored in impermeable plastic bags at 10oC. Together, these are the longest effective storage periods reported for Inga embryos. In one single species the characteristics of mature seeds may vary from year to year within provenances, due to environmental factors and plant reproductive strategies. The best known examples are related to seed production (Kelly and Sork, 2002; Greene and Johnson, 2004), seed mass (Leishman et al., 2000) and seed dormancy (Baskin and Baskin, 1998). However there is a scarcity of such studies concerning desiccation sensitivity and storability. Tompsett and Pritchard (1993) showed that germination of mature Aesculus hippocastanum seeds collected from the same trees in 2 successive years decreased from 77 to 17% when dried to a MC of approx. 30% (fwb). The year-to-year variation was also shown in relation to the storability of recalcitrant seeds of Machilus kusanoii, a tree species from Taiwan. After 4 months of wet storage, germination of mature seeds decreased from around 100 to below 10%, while in the next year, seeds collected from the same tree and stored for the same period under the same conditions, maintained germinability close to 100% (Chien and Lin, 1997). It is likely that in the present study the different results of storability of I. vera embryos attained with seeds collected in 2005, compared to 2004, were due to this natural variation. For germplasm conservation purposes, seeds need to be stored for periods greater than that from sowing to sexual maturity of the plant (Roberts and King, 1980 a,b). Inga vera trees start to bear fruits three to four years after planting, which is much longer than the storability achieved in the present study. Therefore, there is still an urgent need to develop techniques that lengthen substantially the viability of stored Inga vera and recalcitrant seeds in general. Since recalcitrant seeds usually lack developmental arrest after maturation, storage protocols should aim at suppressing metabolic activity during moist storage. From our results it appeared that this will not easily be achieved by partial dehydration in combination with sealed storage. Treatments with osmotic agents, such as PEG, or heat shock, which have proven successful in re-induction of DT or extension of longevity of germinated orthodox seeds (Bruggink and van der Toorn, 1995; Bruggink et al., 1999; Faria et al., 2005) were not effective. PEG may be used as a means to slow down metabolism rather than inducing DT. Application of growth regulators, such as ABA requires further study since it seems that ABA is able to extend storability of I. vera embryos, depending on the storage temperature. In addition, it is possible that mature recalcitrant seeds have completely lost unique seed traits such as DT and should therefore be regarded as vegetatively growing plants rather than seeds. Storage under controlled atmosphere may prove successful. This type of storage is common in long term conservation of fresh fruits.

103 Chapter 5

Acknowledgements

We thank CNPq (National Council for Scientific and Technological Development, from the Ministry of Science and Technology, Brazil) for financial support of the studies of J.M.R. Faria. The technical staff from the Laboratory of Forest Tree Seeds (Lavras Federal University - UFLA) is acknowledged for performing seed collection and handling. We also thank professor Eduardo Alves (Department of Plant Pathology, UFLA) for his assistance with the microscopy studies.

References

Anandarajah, K. and McKersie, B.D. (1990) Manipulating the desiccation tolerance and vigor of dry somatic embryos of Medicago sativa L. with sucrose, heat shock and abscisic acid. Plant Cell Reports 9, 451-455. Baker, E.H., Bradford, K.J., Bryant, J.A. and Rost, T.L. (1995) A comparison of desiccation-related proteins (dehydrin and QP47) in peas (Pisum sativum). Seed Science Research 5, 185-193. Barbedo, C.J. and Cicero, S.M. (2000) Effects of initial quality, low temperature and ABA on the storage of seeds of Inga uruguensis, a tropical species with recalcitrant seeds. Seed Science and Technology 28, 793-808 Baskin, C.C. and Baskin, J.M. (1998) Seeds: Ecology, biogeography, and evolution of dormancy and germination. Academic Press, San Diego, CA, USA. Beckett, R.P., Mayaba, N., Minibayeva, F.V. and Alyabyev, A.J. (2005) Hardening by partial dehydration and ABA increase desiccation tolerance in the cyanobacterial lichen Peltigera polydactylon. Annals of Botany 96, 109-115. Berjak, P., Vertucci, C.W. and Pammenter, N.W. (1993) Effects of developmental status and dehydration rate on characteristics of water and desiccation-sensitivity in recalcitrant seeds of Camellia sinensis. Seed Science Research 3, 155-166. Berjak, P. and Pammenter, N.W. (2003) Understanding and handling desiccation-sensitive seeds, in Seed Conservation - Turning Science into Practice (eds R.D. Smith, J.B. Dickie, S.H. Linington, H.W. Pritchard and R.J. Probert), Kew: Royal Botanic Gardens, UK, pp. 417-430. Bewley, J.D. and Black, M. (1994) Seeds. Physiology of Development and Germination, 2nd edn, Plenum Press, New York. Bilia, D.A.C. (1997) Tolerância à dessecação e armazenamento de sementes de Inga uruguensis (Hook. et Arn.). Phd. Thesis, ESALQ, Piracicaba, SP, Brazil. Bilia, D.A.C., Marcos Filho, J. and Novembre, A.D.C.L. (1999) Desiccation tolerance and seed storability of Inga uruguensis (Hook. et Arn.). Seed Science and Technology 27, 77-89. Bilia, D.A.C., Barbedo, C.J., Cicero, S.M. and Marcos Filho, J. (2003) Ingá: uma espécie importante para recomposição vegetal em florestas ripárias, com sementes interessantes para a ciência. Informativo Abrates 13, 26-30.

104 Improvement of storability of Inga vera subsp. affinis embryos

Black, M., Corbineau, F., Gee, H. and Come, D. (1999) Water content, raffinose, and dehydrins in the induction of desiccation tolerance in immature wheat embryos. Plant Physiology 120, 463-471. Bruggink, T. and van der Toorn, P. (1995) Induction of desiccation tolerance in germinated seeds. Seed Science Research 5, 1-4. Bruggink, G.T., Ooms, J.J.J. and van der Toorn, P. (1999) Induction of longevity in primed seeds. Seed Science Research 9, 49-53. Buitink, J., Vu, B.L., Satour, P. and Leprince, O. (2003) The re-establishment of desiccation tolerance in germinated radicles of Medicago truncatula Gaertn. seeds. Seed Science Research 13, 273- 286. Carvalho, L.R. (2000) Classificação fisiológica de sementes de espécies florestais quanto à capacidade de armazenamento. MSc. Thesis, Universidade Federal de Lavras, Lavras, Brazil. Chien, C.T. and Lin, T.P. (1997) Effect of harvest date on the storability of desiccation-sensitive seeds of Machilus kusanoi Hay. Seed Science and Technology 25, 361–371. Chin, H.F. (1989) Recalcitrant seeds. Extension Bulletin no. 288. ASPAC: Food and Fertilizer Technology Center, Taipei City, Taiwan. Clatterbuck, W.K. and Bonner, F.T. (1985) Utilisation of food reserves in Quercus seed during storage. Seed Science and Technology 13, 121-128. Delseny, M., Gaubier, P., Hull, G., Saez-Vasquez, J., Gallois, P., Raynal, M., Cooke, R. and Grellet, F. (1994) Nuclear genes expressed during seed desiccation: relationship with responses to stress, in Stress-Induced Gene Expression in Plants (ed A.S. Basra), Harwood Academic Publishers, Switzerland, pp. 25-59. Dirk, L.M.A., VanDerKrol, S, Vreugdenhil, D., Hilhorst, H.W.M. and Bewley, J.D. 1999. Galactomannan, soluble sugar, and starch mobilization following germination of fenugreek (Trigonella foenum- graecum) seeds. Plant Physiology and Biochemistry 37, 41-50 Drew, P.J., Pammenter, N.W. and Berjak, P. (2000) “Sub-imbibed” storage is not an option for extending longevity of recalcitrant seeds of the tropical species Trichilia dregeana Sond. Seed Science Research 10, 355-363. Dussert, S., Chabrillange, N., Engelmann, F. and Hamon, S. (1999) Quantitative estimation of seed desiccation sensitivity using a quantal response model: application to nine species of the genus Coffea L. Seed Science Research 9, 135-144. Ellis, R.H., Hong, T.D. and Roberts. E.H. (1990) An intermediate category of seed storage behaviour? I. Coffee. Journal of Experimental Botany 41, 1167-1174. Faria, J.M.R., van Lammeren, A.A.M. and Hilhorst, H.W.M. (2004) Desiccation sensitivity and cell cycle aspects in seeds of Inga vera subsp. affinis. Seed Science Research 14, 165-178. Faria, J.M.R., Buitink, J., van Lammeren, A.A.M. and Hilhorst, H.W.M. (2005) Changes in DNA and microtubules during loss and re-establishment of desiccation-tolerance in germinating Medicago truncatula seeds. Journal of Experimental Botany 56, 2119-2130. Farnsworth, E.J. and Farrant, J.M. (1998) Reductions in abscisic acid are linked with viviparous reproduction in mangroves. American Journal of Botany 85, 760-769. Farrant, J.M., Berjak, P. and Pammenter, N.W. (1993) Studies on the development of the desiccation- sensitive (recalcitrant) seeds of Avicennia marina (Forssk.) Vierh.: the acquisition of germinability and response to storage and dehydration. Annals of Botany 71, 405-410.

105 Chapter 5

Farrant, J.M., Pammenter, N.W., Berjak, P., Farnsworth, E.J. and Vertucci, C.W. (1996). Presence of dehydrin-like proteins in recalcitrant (desiccation-sensitive) seeds may be related to habitat. Seed Science Research 6, 175-182. Ferreira, S.A.N. and Gentil, D.F.O. (2003) Armazenamento de sementes de camu-camu (Myrciaria dubia) com diferentes graus de umidade e temperaturas. Revista Brasileira de Fruticultura 25, 440-442. Fu, J.R., Xia, Q.H. and Tang, L.F. (1993) Effects of desiccation on excised embryonic axes of three recalcitrant seeds and studies on cryopreservation. Seed Science and Technology 21, 85-95. Genebank Standards (1994) Food and Agriculture Organization of the United Nations, Rome, International Plant Genetic Resources Institute, Rome. Goldbach, H. (1979) Imbibed storage of Melicoccus bijugatus and Eugenia brasiliensis (E. dombeyi) using abscisic acid as a germination inhibitor. Seed Science and Technology 7, 403-406. González, J.E. and Fisher, R.F. (1997) Effect of desiccation, temperature, and moisture content on seed storage of three tropical tree species. Forest Science 43, 595-601. Greene, D.F. and Johnson, E.A. (2004) Modelling the temporal variation in the seed production of North American trees. Canadian Journal of Forest Research 34: 65–75. Greggains, V., Finch-Savage, W.E., Quick, W.P. and Atherton, N.M. (2000) Metabolism-induced free radical activity does not contribute significantly to loss of viability in moist-stored recalcitrant seeds of contrasting species. New Phytologist 148, 267-276. Hong, T.D. and Ellis, R.H. (1998) Contrasting seed storage behaviour among different species of Meliaceae. Seed Science and Technology 26, 77-95. ISTA (International Seed Testing Association) (1996) International rules for seed testing. Seed Science and Technology 24 (suppl.). Kang, H.M., Park, K.W. and Saltveit, M.E. (2005) Chilling tolerance of cucumber (Cucumis sativus) seedling radicles is affected by radicle length, seedling vigor, and induced osmotic- and heat- shock proteins. Physiologia Plantarum 124, 485-492. Kelly, D. and Sork, V.L. (2002) Mast seeding in perennial plants: Why, how, where? Annual Review of Ecology and Systematics 33, 427-447 King, M.W. and Roberts, E.H. (1980) Maintenance of recalcitrant seeds in storage, in Recalcitrant Crop Seeds (eds H.F. Chin, E.H. Roberts), Tropical Press, Kuala Lumpur, pp. 53-89. Kioko J., Berjak, P., Pammenter, N.W., Watt, P.M. and Wesley-Smith, J. (1998). Desiccation and cryopreservation of embryonic axes of Trichilia dregeana Sond. Cryo-Letters 19, 15-26. Köppen, W. (1936) Das geographische system der climate, in Handbuch der Klimatologie, Vol. 1, part C. (eds W. Köppen and R. Geiger), Gebrüder Borntraeger, Berlin, pp. 1-44. Korkmaz, A. (2002) Amelioration of chilling injuries in watermelon seedlings by abscisic acid. Turkish Journal of Agriculture and Forestry 26, 17-20. Krishnapillay, D.B. (2000) Attempts at conservation of recalcitrant seeds in Malaysia. Forest Genetic Resources. FAO, Rome, Italy. 28, 34-37. Leishman, M.R., Wright, I.J., Moles, A.T. and Westoby, M. (2000) The evolutionary ecology of seed size, in Seeds: The Ecology of Regeneration in Plant Communities, 2nd edn (ed M. Fenner), Cabi Publishing, Wallingford, pp. 31-57.

106 Improvement of storability of Inga vera subsp. affinis embryos

Leprince, O., Harren, F.J.M., Buitink, J., Alberda, M. and Hoekstra, F.A. (2000) Metabolic dysfunction and unabated respiration precede the loss of membrane integrity during rehydration of germinating radicles. Plant Physiology 122, 597-608. Maluf, A.M., Bilia, D.A.C. and Barbedo, C.J. (2003) Drying and storage of Eugenia involucrata DC. seeds. Scientia Agricola 60, 471-475. Marcellin, P. (1992) Les maladies physiologiques du froid, in Les Végétaux et le Froid (ed D. Côme), Hermann, Paris, pp. 53-105. Marshall, J., Beardmore, T., Whittle, C.A., Wang, B., Rutledge, R.G. and Blumwald, E. (2000) The effects of paclobutrazol, abscisic acid, and gibberellin on germination and early growth in silver, red, and hybrid maple. Canadian Journal of Forest Research 30, 557–565. Marzalina, M. and Krishnapillay, B. (1999) Recalcitrant seed biotechnology applications to rain forest conservation, in Plant Conservation Biotechnology (ed. E.E. Benson), Taylor and Francis, London, pp. 265-276. Ooms, J.J.J., van der Veen, R. and Karssen, C.M. (1994) Abscisic acid and osmotic stress or slow drying independently induce desiccation tolerance in mutant seeds of Arabidopsis thaliana. Physiologia Plantarum 92, 506-510. Pammenter, N.W., Motete, N. and Berjak, P. (1997) The response of hydrated recalcitrant seeds to long-term storage, in Basic and Applied Aspects of Seed Biology (eds R.H. Ellis, M. Black, A.J. Murdoch and T.D. Hong), Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 673-687. Pammenter, N.W. and Berjak, P. (1999) A review of recalcitrant seed physiology in relation to desiccation-tolerance mechanisms. Seed Science Research 9, 13-37. Pammenter, N.W., Berjak, P., Farrant, J.M., Smith, M.T. and Ross, G. (1994) Why do stored hydrated recalcitrant seeds die? Seed Science Research 4, 187-191. Probert, R.J. and Longley, P.L. (1989) Recalcitrant seed storage physiology in three aquatic grasses (Zizania palustris, Spartina anglica and Porteresia coarctata). Annals of Botany 63, 53-63. Ramanjulu, S. and Bartels, D. (2002) Drought- and desiccation-induced modulation of gene expression in plants. Plant, Cell & Environment 25, 141-151. Roberts, E.H. (1973) Predicting the storage life of seeds. Seed Science and Technology 1, 499-514. Roberts, E.H. and King, M.W. (1980a) The characteristics of recalcitrant seeds, in Recalcitrant Crop Seeds (eds H.F. Chin, E.H. Roberts), Tropical Press, Kuala Lumpur, pp. 1-5. Roberts, E.H. and King, M.W. (1980b) Storage of recalcitrant seeds, in Crop Genetic Resources: the Conservation of Difficult Material (eds L.A. Withers and J.T. Williams), Reading, UK, pp. 39-48. Rowley, C.R. and Moran, D.T. (1975) A simple procedure for mounting wrinkle-free sections on formvar-coated slot grids. Ultramicrotomy 1, 151-155. Senaratna, T., McKersie, B.D. and Bowley, S.R. (1989) Desiccation tolerance of alfalfa (Medicago sativa L.) somatic embryos. Influence of abscisic acid, stress pretreatments and drying rates. Plant Science 65, 253-259. Smith, R.D., Dickie, J.B., Linington, S.H., Pritchard, H.W. and Probert, R.J. (2003) An editorial perspective on seed conservation, in Seed Conservation - Turning Science into Practice (eds R.D. Smith, J.B. Dickie, S.H. Linington, H.W. Pritchard and R.J. Probert), Kew: Royal Botanic Gardens, UK, pp. 969-980. Tompsett, P.B. (1984) The effect of moisture content and temperature on the seed storage life of Araucaria columnaris. Seed Science and Technology 12, 801-816.

107 Chapter 5

Tompsett, P.B. (1985) The influence of moisture content and storage temperature on the viability of Shorea almon, Shorea robusta, and Shorea roxburghii seed. Canadian Journal of Forest Research 15, 1074-1079. Tompsett, P.B. and Pritchard, H.W. (1993) Water status changes during development in relation to the germination and desiccation tolerance of Aesculus hippocastanum L. seeds. Annals of Botany 71, 107-116. Tompsett, P.B. and Pritchard, H.W. (1998) The effect of chilling and moisture status on the germination, desiccation tolerance and longevity of Aesculus hippocastanum L. seed. Annals of Botany 82, 249-261. Wan, L., Xia, Q., Qiu, X. and Selvaraj, G. (2002) Early stages of seed development in Brassica napus: a seed coat-specific cysteine proteinase associated with programmed cell death of the inner integument. The Plant Journal 30, 1-10. Wehmeyer, N. and Vierling, E. (2000) The expression of sHSPs in seeds responds to discrete developmental signals and suggests a general protective role in desiccation tolerance. Plant Physiology 122, 1099-1108. Werker, E. (1997) Seed Anatomy. Gebruder Borntraeger, Berlin. Woodstock, L.W. (1973) Physiological and biochemical tests for seed vigour. Seed Science and Technology 1, 127-157.

108 Chapter 6

General discussion

Introduction

This thesis deals with desiccation tolerance (DT) and sensitivity (DS) in seeds, in order to better understand some of the phenomena that underlie these antagonistic traits. The ultimate aim of this study was to generate fundamental knowledge to underpin future protocols for diminishing the DS inherent to the recalcitrant seeds, which would allow their dehydration to levels low enough to prolong their remarkably short storability. It has been proposed that the study of DS can be done in orthodox seeds during/after germination, when they lose DT (Chapter 1, Fig. 1), that is, when they become comparable to their recalcitrant counterparts (Farrant et al., 1997; Sun, 1999). The feasibility of the re- establishment of DT in germinated orthodox seeds, as shown by Bruggink and van der Toorn (1995) and Buitink et al., (2003), enables the comparison of different levels of DT in seeds of the same species. Thus, in the present study, we used the truly recalcitrant seeds of the tree species Inga vera, as well as germinating seeds and seedlings of the orthodox-seeded model species Medicago truncatula. Physiological, cytological and molecular aspects of DT and DS were investigated. The results may shed more light on these two seed traits.

Desiccation sensitivity and the cell cycle

It has been suggested that cells in the G2 phase of the cell cycle, that is, those whose nucleus contains 4C DNA, are more desiccation sensitive than cells that are in the pre- synthetic phase (G1), with 2C DNA (Deltour, 1985), although the reason for this is not yet fully understood. One possible explanation is that cells in G2 have more target (DNA) to be damaged by stress (Bino, 1993). It seems therefore that the cell cycle is associated with DT in seeds. Surprisingly little attention has been paid to this subject and, consequently, very few data are available on the possible link between the stress response and cell cycle progression (Reichheld et al., 1999). In this thesis, I. vera seeds showed a slight decrease in desiccation sensitivity during development (6 to 14 weeks after flowering), while the percentage of cells with 4C DNA remained practically constant, around 10% in the shoot and 15% in the root (Chapter 2). At shedding, seeds of M. truncatula, which are desiccation Chapter 6

tolerant, showed a high percentage of radicle cells (45%) with 4C DNA content (Chapter 3). These results, together with others (Bino et al., 1993), suggest that in mature seeds the relative number of cells with 4C DNA content does not dictate per se the level of DT. Other factors must be involved and DNA content might have only a secondary effect. However, the expected inverse relation between DT and the relative number of cells with 4C DNA content was clearly detected in growing radicles of M. truncatula seedlings subjected to osmotic treatment with PEG in order to re-establish DT (Chapter 3). In this case the interval of radicle length (from 2 to 3 mm) in which the extent of re-induction of DT fell substantially (from 84 to 33%) coincided with the resumption of the cell cycle, as shown by the increase in the percentage of radicle cells with 4C DNA (from 44 to 63%). However, a causal relationship remains to be established.

Desiccation sensitivity, microtubules and DNA integrity

Microtubules (MTs) showed to be sensitive to dehydration, both in embryonic cells of mature seeds of I. vera (Chapter 2) and in seedlings of M. truncatula with a 2 mm long radicle (Chapter 3). As cells lost water, MTs were dismantled by depolymerization and the tubulin units reverted to the cytoplasmic pool. With the continuation of dehydration, tubulin was degraded and no MTs neither tubulin granules could be observed anymore. Upon rehydration, cells were not able to reconstitute their microtubular cytoskeleton. Seedlings of M. truncatula with 2 mm long radicles that were treated with PEG before dehydration showed a similar behavior regarding disappearance of MTs and tubulin granules caused by dehydration. However, upon rehydration, the microtubular cytoskeleton was restructured and the cells followed their normal course, leading to the development of normal seedlings. The ability of rebuilding the microtubular cytoskeleton after disappearance of MTs and tubulin, shown by PEG-treated seedlings is in contrast with the untreated seedlings. This can be explained by the maintenance of DNA integrity during dehydration of PEG-treated seedlings. Contrarily, in untreated seedlings DNA was degraded, generating fragments of approximately 200 bp, suggesting that an apoptotic-like process took place during dehydration. To our knowledge this is the first time that DNA laddering is shown to occur during drying of desiccation intolerant plant tissues. Desiccation tolerant organisms must rely on one or both of the following strategies: avoidance of the accumulation of desiccation- induced damage and activation of repair mechanisms upon rehydration (Buitink et al., 2002). The present study shows that both strategies were distinctly applied by M. truncatula seedlings in which DT was re-established by PEG treatment. Nuclear DNA was kept intact during dehydration, whilst MTs were dismantled and later rebuilt upon rehydration.

110 General discussion

Changes in gene expression during loss and re-establishment of desiccation tolerance (DT) in germinating seeds and in seedlings of Medicago truncatula

Upon imbibition, non-dormant, orthodox seeds switch from a developmental to a germination program, resuming the metabolic activity that leads to the formation of a vegetatively growing seedling. With this shift from the quiescent state to a growing seedling, it is likely that thousands of genes are turned on or off (Bradford et al., 2000). Acquisition of desiccation tolerance is part of the developmental program and is lost upon, or right after, radicle protrusion. Since PEG-treatment re-establishes DT in M. truncatula seedlings, we hypothesized that it does so by driving cells back to the developmental program, with the reactivation of genes that had been turned off during and after germination. The genes LEC1 and ABI3 are considered seed specific and have been implicated as key-genes in the arrest of embryonic growth and in the acquisition of DT and dormancy (Parcy et al., 1997; Harada, 2001; Holdsworth et al., 2001). In the present study, transcript levels of ABI3 in radicles increased during cold imbibition and diminished during radicle growth, as compared to the dry seed. PEG treatment of seedlings maintained ABI3 expression in the growing radicle above or at the dry seed level. LEC1 expression showed an approximately 10-fold higher expression during radicle growth in both PEG-treated and untreated seeds, demonstrating that this gene may also be expressed in vegetative tissue. Thus, both LEC1 and ABI3 expression may occur in seedling tissues as a response to cold- and/or osmotic stress although only ABI3 expression was correlated with the re- establishment of DT in the radicle tissue. In addition, expression of (ABA-dependent) PKABA1 which plays a role in the initial response to abiotic stresses (Anderberg and Walker- Simmons, 1992; Holappa and Walker-Simmons, 1995) responded to the PEG treatment by a 20- to 30-fold up-regulation of the PKABA1 transcript. Heat shock proteins (sHSP), peroxiredoxins (PER) and late embryogenesis abundant (LEA) proteins accumulate during the maturation of orthodox seeds. As germination progresses, the expression of these stress-related genes diminishes, concomitantly with the loss of DT, as shown in this thesis (Chapter 4), when the expression of sHSP18.2, EM6 (a LEA gene) and PER1 in 3-mm radicles of M. truncatula seedlings reached levels approximately 250-fold lower than in dry seeds. PEG treatment of the seedlings restored transcript levels, often to dry seed levels, even after dehydration. This pattern was similar to that of ABI3. In Arabidopsis, the expression of EM6 (Parcy et al., 1994; Vicient et al., 2000), sHSPs (Wehmeyer et al., 1996) and PER1 (Haslekas et al., 1998) appears to be controlled by ABI3. Our results show that this may also be the case in protruded radicles of Medicago truncatula. The correlation between expression patterns of the developmental ABI3 gene and the stress-related sHSP18.2, EM6 and PER1 genes might indicate a (partial) reversion of the radicle tissue to the developmental state, thereby re-inducing DT.

111 Chapter 6

ABI3 and LEC1 suppress CTS expression, related to germination potential, in Arabidopsis seeds, thereby maintaining the developmental state (Russell et al., 2000). However, the present study shows that the higher expression levels of ABI3 and LEC1 did not result in suppression of CTS expression. Thus, it seems that the processes responsible for the re-establishment of DT in seedlings may not be exactly the same as during the acquisition of DT in developing orthodox seeds. We hypothesize that the protruded radicle possesses both seed and seedling characteristics. However, despite still substantially higher expression in the 3-mm PEG-treated radicles of developmental and/or stress related genes, as compared to the untreated seed, dehydration led to a loss of viability. Hence, these genes may be a pre-requisite for re- induction of DT but are not sufficient. The transcript levels of the cytoskeleton genes ACT and TUB either increased or did not change in response to PEG treatment. However, dehydration after PEG treatment led to a substantial drop in the expression of both genes, which is in accordance with the complete loss of the microtubular cytoskeleton observed after dehydration of PEG treated M. truncatula seedlings (Chapter 3). CDC2a is an A-type CDK (cyclin-dependent kinase) and acts in the G1-to-S and G2-to-M transitions in the cell cycle (Mironov et al., 1999; Vázquez-Ramos and Sánchez, 2003). Highest CDC2a expression was found in 2-mm radicles, which is just prior to the G1-to-S transition in the majority of the seed population (Chapter 3). In summary, our results show that, in general, gene expression in radicles that have gone through the first cell cycle and have lost DT may still respond, in terms of gene expression, in a similar fashion as desiccation tolerant radicles. Apparently, the occurrence of the first cell cycle is an overriding factor that abolishes reversion to the developmental state and, hence, re-establishment of DT.

Improvement of storability of Inga vera embryos

Desiccation sensitivity and short storability are the most important (and unwanted) characteristics of the recalcitrant seeds, for those who want to conserve them. Germinability of embryos of Inga vera fell from 100 to 50% after drying them from 1.43 to 0.69 gH2O/g dry matter. They also presented one of the shortest longevities known, living no much longer than one or two weeks if stored at 20 or 5oC, respectively (Chapter 2). Because recalcitrant seeds are metabolically active in storage (Pammenter et al., 1994), an attempt was made to slow down the metabolism of I. vera embryos and consequently prolong their short storability (Chapter 5). For this an approach was used, proposed by Tompsett (1985), which consists of storage in an osmotic medium (in the present study, PEG at -1.7 MPa) with or without the addition of germination inhibitors (in the present study, ABA). Besides the slowing down of metabolism, osmotic stress and ABA, alone or together, can lead to the expression of genes involved in DT in plants or seeds (Skriver and Mundy, 1990). Thus, it

112 General discussion

was expected that after storage in PEG (with or without ABA), desiccation sensitivity of I. vera embryos would be reduced. Storage in PEG was capable of keeping the germination rate of I. vera embryos at 100% until 30 days, either at 5 or 20oC. Obviously, a storage duration of one month is insignificant in terms of germplasm conservation, but considering that, when stored at 20oC without PEG, embryos completely lost viability within 10 days, storage in PEG caused a three-fold increase in storability. The effect of ABA showed to be temperature-dependent, being positive at 20oC and negative at 5oC. Storage of the embryos in PEG for 14 days did not render them more tolerant to desiccation, since after being dried to 0.64 g H2O/ g dry matter and stored in plastic bags for 14 more days, they totally lost their viability. This may have been caused by the following: the desiccation was too severe, the storage conditions were unsuitable to reduce the desiccation sensitivity of the embryos (by stimulating the expression of desiccation-related genes), the sealed storage conditions resulted in a quick depletion of oxygen, or I. vera does not have the required genes. Since recalcitrant seeds do not acquire DT during their development this means that they may not have the genes required for the induction of DT, or that they are not active. To date very few molecular studies with recalcitrant seeds have been carried out, thus it is still an enigma what events occur at the molecular level. For this reason one can not rule out the possibility that desiccation-related genes are present in the genome of recalcitrant-seeded species, but that the plant does not sense the environmental signals required for expression of those genes because of absent or defective signal-transduction pathways. Effectively, the genes may have been “turned off” during evolution. In genomic DNA from I. vera leaves we have found sequences with high homology to sequences associated with sHSP (unpublished data). Many attempts have been made to isolate RNA from I. vera embryos, using different protocols, but all failed. Thus, it could not be investigated if (and which) genes are being differentially regulated during storage under different conditions.

Final conclusions

The use of germinating and germinated seeds of the model species M. truncatula has proven very useful to study the capacity of protruded radicles to regain desiccation tolerance by the application of a mild osmotic stress. This approach was not successful in mature (germinating) embryos of the recalcitrant species I. vera. We therefore conclude that germinated Medicago seeds are not just growing seedlings but they retain, latently, orthodox seed characteristics. When radicle protrusion has exceeded a length of 3 mm they must be considered actively growing seedlings with increasing relative numbers of nuclei with 4C DNA content, which cannot return to the developmental mode anymore and, hence, cannot be made desiccation tolerant. Mature recalcitrant embryos resemble growing seedlings rather than seeds. DT cannot be established, not because they have passed the ‘seed-to-plant’

113 Chapter 6

threshold but because they do not have the genetic program to induce DT during seed development. To bridge the gap between sowing and the attainment of sexual maturity, which is 3- 4 years for Inga vera, the storability of its seeds needs to be extended substantially. Thus, the urgency for suitable seed storage protocols for I. vera and other recalcitrant species remains high. Since mature recalcitrant seeds resemble actively growing seedlings rather than proper seeds, storage protocols should aim at suppressing metabolic activity during moist storage thereby delaying the moment of viability loss by depletion of reserve food. Our study has shown that partial dehydration in combination with sealed storage is not successful. Treatments with osmotic agents, such as PEG, or heat shock appear equally ineffective. However, PEG may be used as a means to suppress metabolic activity. Application of growth regulators, such as ABA, to (further) suppress metabolism, holds some promise but requires further study. ABA is able to extend storability of I. vera embryos at high but not at low temperature. It seems likely that mature recalcitrant seeds have completely lost their unique seed traits such as DT and should therefore be regarded as vegetatively growing plants rather than seeds. Storage under controlled atmosphere may prove successful. This type of storage is common in long term conservation of fresh fruits and vegetables. Examples include hypobaric (or low oxygen) and cold storage with intermittent ventilation or heating, respectively, or storage in elevated CO2 atmosphere.

References

Anderberg, R.J. and Walker-Simmons, M.K. (1992) Isolation of a wheat cDNA clone for an abscisic acid-inducible transcript with homology to protein kinases. Proceedings of the National Academy of Sciences USA 89, 10183-10187. Bino, R.J., Lanteri, S., Verhoeven, H.A. and Kraak, H.L. (1993) Flow cytometric determination of nuclear replication stage in seed tissues. Annals of Botany 72, 181-187. Bradford, K.J., Chen, F., Cooley, M.B., Dahal, P., Downie, B., Fukunaga, K.K., Gee, O.H., Gurusinghe, S., Mella, R.A., Nonogaki, H., Wu, C.T. and Yim, K.O. (2000) Gene expression prior to radicle emergence in imbibed tomato seeds, in Seed Biology: Advances and Applications (eds M. Black, K.J. Bradford and J. Vazquez-Ramos), CAB Publishing, Wallingford, pp. 231-251. Bruggink, T. and van der Toorn, P. (1995) Induction of desiccation tolerance in germinated seeds. Seed Science Research 5, 1-4. Buitink, J., Hoekstra, F.A. and Leprince, O. (2002) Biochemistry and biophysics of tolerance systems. In M Black, HW Pritchard, eds, Desiccation and Survival in Plants: Drying Without Dying. Cabi Publishing, Wallingford, pp 293-318. Buitink, J., Vu, B.L., Satour, P. and Leprince, O. (2003) The re-establishment of desiccation tolerance in germinated radicles of Medicago truncatula Gaertn. seeds. Seed Science Research 13, 273- 286.

114 General discussion

Burssens, S., Himanen, K., van de Cotte, B., Beeckman, T., van Montagu, M., Inzé, D. and Verbruggen, N. (2000) Expression of cell cycle regulatory genes and morphological alterations in response to salt stress in Arabidopsis thaliana. Planta 211, 632-640. Deltour, R. (1985) Nuclear activiation during early germination of the higher plant embryo. Journal of Cell Science 75, 43-83. Farrant, J.M., Pammenter, N.W., Berjak, P. and Walters, C. (1997) Subcellular organization and metabolic activity during the development of seeds that attain different levels of desiccation tolerance. Seed Science Research 7, 135-144. Harada, J.J. (2001) Role of Arabidopsis LEAFY COTYLEDON genes in seed development. Journal of Plant Physiology 158, 405-409. Haslekas, C., Stacy, R.A.P., Nygaard, V., Culianez-Macia, F.A. and Aalen, R.B. (1998) The expression of a peroxiredoxin antioxidant gene, AtPer1, in Arabidopsis thaliana is seed specific and related to dormancy. Plant Molecular Biology 36, 833-845. Holappa, L.D. and Walker-Simmons, M.K. (1995) The wheat abscisic acid-responsive protein kinase mRNA, PKABA1, is up-regulated by dehydration, cold temperature, and osmotic stress. Plant Physiology 108, 1203-1210. Holdsworth, M., Lenton, J., Flintham, J., Gale, M., Kurup, S., McKibbin, R., Bailey, P., Larner, V. and Russell, L. (2001) Genetic control mechanisms regulating the initiation of germination. Journal of Plant Physiology 158, 439-445. Leprince, O., Harren, F.J.M., Buitink, J., Alberda, M. and Hoekstra, F.A. (2000) Metabolic dysfunction and unabated respiration precede the loss of membrane integrity during rehydration of germinating radicles. Plant Physiology 122, 597-608. Mironov, V., de Veylder, L., van Montagu, M. and Inzé, D. (1999) Cyclin-dependent kinases and cell division in plants - The nexus. The Plant Cell 11, 509-522. Pammenter, N.W., Berjak, P., Farrant, J.M., Smith, M.T. and Ross, G. (1994) Why do stored hydrated recalcitrant seeds die? Seed Science Research 4, 187-191. Parcy, F., Valon, C., Kohara, A., Miséra, S. and Giraudat, J. (1997) The ABSCISIC ACID- INSENSITIVE3, FUSCA3, and LEAFY COTYLEDON1 loci act in concert to control multiple aspects of Arabidopsis seed development. The Plant Cell 9, 1265-1277. Parcy, F., Valon, C., Raynal, M., Gaubier-Comella, P., Delseny, M. and Giraudat, J. (1994) Regulation of gene expression programs during Arabidopsis seed development: roles of the ABI3 locus and of endogenous abscisic acid. The Plant Cell 6, 1567-1582. Reichheld, J.P., Vernoux, T., Lardon, F., van Montagu, M. and Inzé, D. (1999) Specific checkpoints regulate plant cell cycle progression in response to oxidative stress. The Plant Journal 17, 647- 656. Russell, L., Larner, V., Kurup, S., Bougourd, S. and Holdsworth, M. (2000) The Arabidopsis COMATOSE locus regulates germination potential. Development 127, 3759-3767. Skriver, K. and Mundy, J. (1990) Gene expression in response to abscisic acid and osmotic stress. The Plant Cell 2, 503-512. Sun, W.Q. (1999) Desiccation sensitivity of recalcitrant seeds and germinated orthodox seeds: can germinated orthodox seeds serve as a model system for studies of recalcitrance?, in IUFRO Seed Symposium 1998. Recalcitrant Seeds. Proceedings of the Conference (eds M. Marzalina et al.), Forestry Research Institute Malaysia, Kuala Lumpur, pp. 29-42.

115 Chapter 6

Tompsett, P.B. (1985) The influence of moisture content and storage temperature on the viability of Shorea almon, Shorea robusta, and Shorea roxburghii seed. Canadian Journal of Forest Research 15, 1074-1079. Vazquez-Ramos, J.M. and Sánchez, M. de la P. (2003) The cell cycle and seed germination. Seed Science Research 13, 113-130. Vicient, C.M., Bies-Etheve, N. and Delseny, M. (2000) Changes in gene expression in the leafy cotyledon1 (lec1) and fusca3 (fus3) mutants of Arabidopsis thaliana L, Journal of Experimental Botany 51, 995-1003. Wehmeyer, N., Hernandez, L.D., Finkelstein, R.R. and Vierling, E. (1996) Synthesis of small heat- shock proteins is part of the developmental program of late seed maturation. Plant Physiology 112, 747-757.

116 Summary

Orthodox seeds acquire desiccation tolerance (DT) during their development which enables them to pass through the phase of maturation drying by the end of their development and enter a state of quiescence. After harvesting, these seeds can be dried further and stored for the long-term without significant loss of viability. On the other hand, there are many species that produce recalcitrant seeds, which have a developmental history and post-harvest behaviour rather opposite from the orthodox types. These are shed with high moisture contents (MCs), are metabolically active and can not be dried to MCs low enough to allow safe storage. Recalcitrant seeds thus represent a big challenge to those who need to store them in seed banks for germplasm conservation purposes. In order to gain insight into the recalcitrance phenomenon, this thesis addresses questions concerning the physiological, cytological and molecular aspects of desiccation sensitivity in developing and mature recalcitrant seeds of the tree species Inga vera Willd. subsp. affinis and in mature and germinating orthodox seeds of the model species Medicago truncatula Gaertn. cv. Jemalong A17. The desiccation sensitivity of I. vera seeds was analyzed in terms of water relations, DNA content and the microtubular cytoskeleton. Developing (6, 9 and 10 weeks after flowering - WAF) and mature seeds (14 WAF) were collected and processed, ending up with the naked embryos. Slight desiccation of immature embryos increased germination, but further drying resulted in a quick decline of germinability. Total loss of viability occurred while the embryos still showed high MCs (around 0.4 g H2O/g dry matter) and high water activity (aw), around 0.88, irrespective of the developmental stage. During development the desiccation sensitivity of the embryos decreased slightly, while the percentage of cells with 4C DNA remained constant, around 15% in the root and 10% in the shoot, suggesting no relation between DT and DNA content. Immuno-histochemical detection of microtubules (MTs) in embryonic axes cells of mature embryos showed abundant cortical microtubule arrays, which were not affected by mild desiccation, but were totally dismantled by further drying. Upon rehydration, damaged cells were not able to reconstitute the microtubular cytoskeleton and this might be related to the viability loss during dehydration of the embryo. The desiccation sensitivity was also studied in seeds of M. truncatula during and after germination. When seedlings with a radicle length as short as 1 mm were dried back to the original MC found in dry seeds and rehydrated, only 15% survived. Seedlings with a radicle length equal or longer than 2 mm did not survive dehydration at all. By subjecting seedlings to an osmotic treatment with polyethylene glycol - PEG (-1.8 MPa) before drying, DT could be re-established in seedlings with a radicle length up to 2 mm. Flow cytometric analyses and MTs visualisation in radicle cells of growing M. truncatula seedlings showed that up to a radicle length of 2 mm, the cell cycle had not been resumed, as shown by the absence of DNA synthesis and cell division, which were first detected in 3 mm long radicles. Therefore

117 DT could be re-established only before the resumption of the cell cycle in the radicles. Dehydration of seedlings with a 2 mm protruded radicle, with or without previous PEG treatment, caused disassembly of MTs. Upon rehydration MTs were not reassembled in radicle cells of untreated seedlings, while PEG-treated seedlings were able to reconstitute the microtubular cytoskeleton and develop into normal seedlings. Dehydration of untreated seedlings with a 2 mm protruded radicle also led to an apoptotic-like DNA fragmentation in radicle cells, while in PEG-treated seedlings DNA integrity was maintained. The results showed that for different cellular components, desiccation-tolerant seedlings may apply distinct strategies to survive dehydration, either by further repair or avoidance of the damages. This thesis also investigated the changes in the expression of various genes related to seed development, DT, cell cycle and cytoskeleton during loss and re-establishment of DT in germinating seeds and in seedlings of M. truncatula. The transcript levels of the studied genes in radicle cells were relatively quantified by real time PCR, using specific primers for M. truncatula. Clear changes in transcript abundance were detected during and after germination and in response to osmotic treatment and dehydration. DT-related genes (EM6, PER1 and sHSP18.2) were down regulated during germination and up regulated by osmotic treatment, which correlated with the loss and re-acquisition of DT in the radicles. The expression pattern of the developmental gene ABI3 was similar to that of the stress related genes, corresponding with a possible control of the stress response by this gene. Abundance of LEC1 transcript correlated more with the germination process than with osmotic stress. The cytoskeleton genes (ACT and TUB) were up-regulated during germination, not affected substantially by osmotic treatment and down-regulated by subsequent dehydration, which was related with the massive breakdown of the cytoskeleton upon dehydration of seedlings.

Expression of CDC2a, one of the key regulators of the G1-to-S transition, was clearly associated with the occurrence of the first cell cycle in the growing radicle. Radicles that have gone through the first cell cycle (3 mm long) may respond to osmoticum in a similar fashion as desiccation tolerant radicles, in terms of gene regulation. However, the resumption of the cell cycle appears to be an overriding factor that abolishes re- establishment of DT. Recalcitrant seeds are metabolically active in storage and this leads to a short longevity. I. vera embryos are among the worst storable species known, retaining viability for not much longer than one or two weeks if stored in semi-permeable bags at 20 or 5oC, respectively. Thus, an attempt was made to slow down the metabolism of I. vera embryos and, consequently, prolong their short storability. For this, embryos were stored in an osmotic medium (PEG at -1.7 MPa) with or without the addition of ABA, a well known germination inhibitor. Besides slowing down the metabolism, osmotic stress and ABA can lead to the expression of genes involved in DT in plants or seeds. Storage in PEG was capable of keeping the germination rate of I. vera embryos at 100% until 30 days, either at 5 or 20oC, thus causing a three-fold increase in storability, when compared to the embryos stored in semi-permeable bags. Starch content, as major food reserve, generally decreased

118 with increasing storage duration. However, we found no direct relationship between starch content and viability/germination. The effect of ABA showed to be temperature-dependent, being positive at 20oC and negative at 5oC. The permanence of the embryos in PEG for 14 days did not render them more tolerant to desiccation. Another technique applied in order to prolong the storability was sealed storage of partially dried embryos. In this case, storability was better than the control, but not as long as in the PEG storage. Anyway, both approaches used in this thesis seem to be promising to prolong the naturally short storability of recalcitrant seeds. We hypothesize that mature recalcitrant seeds have completely lost unique seed traits such as DT and should therefore be regarded as vegetatively growing seedlings/plants rather than seeds and, hence, storage under controlled atmosphere may prove successful. This type of storage is common in long term conservation of fresh fruits and vegetables.

119

Samenvatting

Orthodoxe zaden worden uitdroogtolerant tijdens hun ontwikkeling en kunnen daardoor de uitdroging aan het einde van de rijpingsfase doorstaan en een staat van rust bereiken. Na de oogst kunnen deze zaden nog verder gedroogd worden en langdurig bewaard blijven zonder noemenswaardig verlies van vitaliteit. Er bestaan echter ook veel plantensoorten die zg. recalcitrante zaden produceren. Zaadontwikkeling en gedrag na de oogst van deze zaden zijn tegengesteld aan die van de orthodoxe types. Rijpe recalcitrante zaden hebben een hoog vochtgehalte, zijn metabolisch actief en kunnen niet worden gedroogd tot een vochtgehalte dat laag genoeg is voor langdurige bewaring. Recalcitrante zaden vormen dus een grote uitdaging voor alle betrokkenen bij de langdurige bewaring van zaden in zaadbanken voor de conservering van genetisch materiaal. Om meer inzicht te verkrijgen in het gedrag van recalcitrante zaden behandelt dit proefschrift vragen ten aanzien van de fysiologische, cellulaire en moleculaire aspecten van uitdrooggevoeligheid in zich ontwikkelende en rijpe (recalcitrante) zaden van de boomsoort Inga vera Willd. subsp. affinis en in rijpe en kiemende orthodoxe zaden van de modelsoort Medicago truncatula Gaertn. cv. Jemalong A17. De uitdrooggevoeligheid van de zaden van I. vera werd geanalyseerd op waterrelaties, DNA gehalte en het microtubulaire cytoskelet. Zich ontwikkelende (6, 9 and 10 weken na de bloei) en rijpe zaden (14 weken na de bloei) werden verzameld en bewerkt waardoor de naakte embryo’s overbleven. Een zeer geringe uitdroging van onrijpe embryo’s verhoogde de kieming enigszins maar verdere droging resulteerde in een snelle afname van de kiemkracht. Volledig verlies van vitaliteit werd al waargenomen bij relatief hoog watergehalte (ongeveer 0.4 g H2O/g droge stof) en wateractiviteit (aw = 0.88), ongeacht de ontwikkelingsfase van het embryo. Tijdens de ontwikkeling nam de uitdrooggevoeligheid van de embryo’s enigszins af terwijl het percentage cellen met 4C DNA constant bleef: +/- 10% in het hypocotyl en +/- 15% in het worteltje. Een relatie tussen DNA gehalte en uitdroogtolerantie ligt daarmee niet voor de hand. Immuno-histochemische detectie van microtubuli in de cellen van de embryonale as van rijpe embryo’s liet vele corticale bundels van microtubuli zien die niet waren aangetast door de milde uitdroging maar die volledig desintegreerden tijdens verdere droging. Na rehydratatie konden de beschadigde cellen het microtubulaire cytoskelet niet meer herstellen. Dit kan gerelateerd zijn aan het verlies van vitaliteit door dehydratatie van het embryo. Uitdrooggevoeligheid werd ook bestudeerd in zaden van M. truncatula tijdens en na kieming. Wanneer zaailingen met een wortellengte van minder dan 1 mm werden teruggedroogd tot het oorspronkelijke vochtgehalte van de droge zaden en vervolgens gerehydrateerd, was het overlevingspercentage slechts ongeveer 15%. Zaailingen met een wortellengte van 2 mm of meer overleefden dehydratatie in het geheel niet. Behandeling van

121 de zaden in een osmoticum, zoals polyethyleenglycol (PEG) van -1.8 MPa, voorafgaande aan de droging, resulteerde in een herwinning van de uitdroogtolerantie in zaailingen met een wortellengte tot 2 mm. Flowcytometrische analyse en visualisatie van microtubuli in wortelcellen van groeiende M. truncatula zaailingen liet zien dat tot een wortellengte van 2mm de celcyclus nog niet was hervat omdat DNA synthese en celdeling niet kon worden aangetoond totdat het worteltje minstens 3 mm lang was. De uitdrooggevoeligheid kon dus alleen hersteld worden vóór hervatting van de celcyclus in de wortelcellen. Dehydratatie van zaailingen met een doorgebroken worteltje van 2 mm, met of zonder voorafgaande PEG behandeling, veroorzaakte ontmanteling van de microtubuli. Na rehydratatie vond geen herstel plaats van microtubuli in de wortelcellen van onbehandelde zaailingen, terwijl dit wel het geval was in de behandelde. Deze groeiden uit tot normale plantjes. Dehydratatie van onbehandelde zaailingen met een worteltje van 2 mm leidde ook tot een op apoptose gelijkende DNA fragmentatie in de wortelcellen, terwijl in de met PEG behandelde zaailingen de DNA integriteit bewaard bleef. De resultaten lieten zien dat uitdroogtolerante zaailingen verschillende strategieën gebruiken om dehydratatie te overleven: via reparatie van de schade of door het vermijden van schade. Dit proefschrift heeft ook de veranderingen bestudeerd in de expressie van een aantal genen die gerelateerd zijn aan zaadontwikkeling, uitdroogtolerantie, cell cyclus, en het cytoskelet tijdens het verlies en herstel van uitdroogtolerantie in kiemende zaden en zaailingen van M. truncatula. De transcripthoeveelheid van de bestudeerde genen in de wortelcellen werd gekwantificeerd met behulp van QRT-PCR en specifieke primers voor M. truncatula. Duidelijke veranderingen in transcripthoeveelheid werden waargenomen tijdens en na keiming en als reactie op de osmotische behandeling en dehydratatie. De expressie van genen betrokken bij uitdroogtolerantie (EM6, PER1, en sHSP18.2) nam af tijdens de kieming maar nam toe door de osmotische behandeling. Dit correleerde met het verlies en herstel van uitdroogtolerantie van het kiemworteltje. Het expressiepatroon van het zaadontwikkelingsgen ABI3 gelijk aan dat van de stress gerelateerde genen wat een mogelijke regulatie van de stress response door dit gen suggereert. De hoevvelheid LEC1 transcript correleerde meer met het kiemingsproces dan met osmotische stress. De cytoskelet genen ACT en TUB lieten een hogere expressie zien tijdens kieming die nauwelijks werd beïnvloed door de osmotische behandeling. Echter, de daaropvolgende dehydratatie resulteerde in een verlaging van de expressie die was gerelateerd aan de massale afbraak van het cytoskelet tijdens kieming. Expressie van CDC2a, één van de belangrijkste regulerende genen in de G1-tot-S transitie, was duidelijk geassocieerd met het voorkomen van de eerste celcyclus in het groeiende worteltje. Worteltjes die reeds door de eerste celcyclus zijn gegaan, reageren op dezelfde manier als uitdroogtolerante wortels op osmotische behandeling in termen van genregulatie. Echter, kennelijk is de celcyclus een overheersende factor die het herstel van uitdroogtolerantie onmogelijk maakt. Recalcitrante zaden zijn tijdens bewaring metabolisch actief en dit leidt tot een korte levensduur. I. vera embryo’s behoren tot de slechtst bewaarbare soorten die bekend zijn. Bewaring in semi-permeabele zakken bij 5 of 20°C houdt de kiemkracht slechts 1-2 weken

122 op aanvaardbaar niveau. Daarom is in dit proefschrift een poging gedaan om de metabolische activiteit in I. vera embryo’s te onderdrukken en daarmee de bewaarduur te verlengen. Hiertoe werden de embryo’s bewaard in een osmotisch medium (PEG, -1.7 MPa) met of zonder toevoeging van abscisinezuur (ABA) een bekende remmer van de kieming. Naast het onderdrukken van metabolische activiteit kunnen osmotische stress en ABA de expressie van genen induceren die betrokken zijn bij de uitdroogtolerantie van planten of zaden. Bewaring in PEG resulteerde in handhaving van de kiemkracht op 100% gedurende 30 dagen, zowel bij 5 als 20°C, waarmee de maximale bewaarduur bijna 3 maal langer werd, vergeleken met bewaring in semi-permeabele zakken. Het zetmeelgehalte als belangrijkste reservevoedsel, nam in het algemeen af tijdens de bewaring. Er werd echter geen duidelijk verband gevonden tussen zetmeelgehalte en vitaliteit. Het effect van ABA bleek temperatuursafhankelijk en was positief bij 20°C maar negatief bij 5°C. Het verblijf van de embryo’s in PEG gedurende 14 dagen resulteerde niet in een verbeterde uitdroogtolerantie. Een andere techniek die werd gebruikt om de bewaarbaarheid te verbeteren was hermetische bewaring van gedeeltelijk gedroogde embryo’s. Ook in dit geval was de bewaarbaarheid beter dan die van de controle maar niet zo goed als de PEG bewaring. Niettemin lijken beide benaderingen een goed uitgangspunt voor het verbeteren van de bewaarbaarheid van recalcitrante zaden. De hypothese uit dit werk is dat rijpe recalcitrante zaden hun typische zaadkenmerken, zoals uitdroogtolerantie, volledig hebben verloren en daarom meer beschouwd moeten worden als vegetatief groeiende zaailingen/planten dan als zaden. Bewaring onder gecontroleerde atmosfeer zou daarom succesvol kunnen zijn. Dit type bewaring is gebruikelijk bij de langdurige conservering van verse groente en vruchten.

123

Resumo

Sementes ortodoxas adquirem tolerância à dessecação durante seu desenvolvimento, preparando-as para passar pela fase de dessecação e entrar em um estado de quiescência. Após a dispersão, essas sementes podem ser secas ainda mais e armazenadas a longo prazo sem significante perda de viabilidade. Por outro lado, muitas espécies produzem sementes recalcitrantes, as quais apresentam um desenvolvimento e comportamento pós-dispersão diferentes das ortodoxas, sendo dispersas com alto grau de umidade, metabolicamente ativas e intolerantes à dessecação a graus de umidade baixos o suficiente para permitir um armazenamento seguro. Sementes recalcitrantes representam portanto um grande desafio àqueles que necessitam armazená-las em banco de sementes com o propósito de conservação de germoplasma. Visando aumentar o conhecimento do fenômeno da recalcitrância, esta tese abrange aspectos fisiológicos, citológicos e moleculares da sensibilidade à dessecação em sementes em desenvolvimento e maduras da espécie arbórea Inga vera Willd. subsp. affinis (cujas sementes são recalcitrantes), e em sementes maduras, em germinação e germinadas da espécie modelo Medicago truncatula Gaertn. cv. Jemalong A17 (cujas sementes são ortodoxas). A sensibilidade à dessecação de sementes de I. vera foi analisada em termos de relações hídricas, conteúdo de DNA e o citoesqueleto microtubular. Sementes em desenvolvimento (6, 9 e 10 semanas após o florescimento - SAF) e maduras (14 SAF) foram coletadas e processadas, resultando nos embriões, já que o tegumento é removido durante o beneficiamento. Dessecação moderada dos embriões aumentou a germinação, mas sua continuidade resultou em um rápido declínio da germinação. A perda total da viabilidade ocorreu enquanto os embriões ainda apresentavam graus de umidade elevados (cerca de 0,4 g H2O/g matéria seca) e alta atividade de água (aw), em torno de 0,88, independente do estágio de desenvolvimento. Durante o desenvolvimento, a sensibilidade à dessecação dos embriões diminuiu levemente, enquanto que a porcentagem de células com conteúdo de DNA 4C permaneceu constante, em torno de 15% no eixo hipocótilo-radícula e 10% na parte aérea (epicótilo mais plúmula), sugerindo não haver uma relação entre tolerância à dessecação e conteúdo de DNA. A detecção imuno-histoquímica dos microtúbulos (MTs) nas células do eixo embrionário de embriões maduros revelou a existência de abundantes MTs corticais, os quais não foram afetados pela dessecação moderada, sendo porém totalmente destruídos com a continuação da dessecação. Após a reidratação, células danificadas não foram capaz de reconstituir o citoesqueleto microtubular e essa inabilidade pode estar relacionada à perda da viabilidade durante a dessecação do embrião. A sensibilidade à dessecação também foi estudada em sementes de M. truncatula durante e após a germinação. Plântulas com radículas de 1 mm de comprimento foram dessecadas até o grau de umidade apresentado pelas sementes antes da embebição e em seguida reidratadas, alcançando apenas 15% de sobrevivência. Plântulas com radículas

125 iguais ou maiores que 2 mm de comprimento não sobreviveram à dessecação. Através de um tratamento osmótico das plântulas com polietileno glicol (PEG, -1,8 MPa) antes da dessecação, foi possível restabelecer tolerância à dessecação em plântulas com radículas de até 2 mm de comprimento. Análise de citometria de fluxo e visualização de MTs mostraram que em radículas de até 2 mm de comprimento, o ciclo celular encontra-se paralisado, não havendo síntese de DNA e divisão celular, as quais foram detectadas pela primeira vez em radículas de 3 mm de comprimento. Portanto, a tolerância à dessecação pode ser restabelecida apenas antes do reinício do ciclo celular na radícula. A desidratação de plântulas com radículas de 2 mm, com ou sem tratamento prévio com PEG, causou o desmonte e desaparecimento dos MTs. Com a reidratação, não houve formação de novos MTs nas células de radículas de plântulas não tratadas com PEG, enquanto que plântulas tratadas com PEG foram capazes de reconstruir o citoesqueleto microtubular e continuar o desenvolvimento como plântulas normais. Desidratação de plântulas com radículas de 2 mm, não tratadas com PEG, causou ainda fragmentação do DNA, com aspecto de apoptose, nas células da radícula, enquanto que em plântulas tratadas com PEG, a integridade do DNA foi mantida. Os resultados mostraram que, para diferentes componentes celulares, plântulas tolerantes à dessecação (tratadas com PEG) podem aplicar estratégias distintas para sobreviver à desidratação, seja evitando os danos durante a dessecação, ou reparando-os durante a reidratação. Também foram estudadas as mudanças na expressão de vários genes relacionados ao desenvolvimento da semente, tolerância à dessecação, ciclo celular e citoesqueleto, durante a perda e o restabelecimento da tolerância à dessecação em células radiculares de sementes em processo de germinação e em plântulas de M. truncatula. Os níveis de transcritos dos genes estudados foram quantificados relativamente por PCR em tempo real, usando primers específicos para M. truncatula. Mudanças claras na abundância dos transcritos foram detectadas durante e após a germinação e em resposta ao tratamento osmótico e à desidratação. Genes relacionados à tolerância à dessecação (EM6, PER1 e sHSP18.2) foram regulados negativamente durante a germinação e positivamente pelo tratamento osmótico, correlacionando com a perda e o restabelecimento da tolerância à dessecação nas radículas. O padrão de expressão do gene ABI3 (ligado ao desenvolvimento da semente) foi similar ao dos genes relacionados a estresses, correspondendo a um possível controle da resposta a estresse por este gene. A abundância de transcritos de LEC1 correlacionou mais com o processo germinativo do que com o estresse osmótico. Genes do citoesqueleto (ACT e TUB) foram regulados positivamente durante a germinação, não foram afetados substancialmente pelo tratamento osmótico e foram regulados negativamente pela desidratação subsequente, relacionando-se com a intensa degradação do citoesqueleto em plântulas desidratadas. A expressão de CDC2a, um dos principais reguladores da transição de G1 para S, foi claramente associada com a ocorrência do primeiro ciclo celular na radícula em crescimento. Radículas que passaram pelo primeiro ciclo celular (3 mm de comprimento) responderam ao tratamento osmótico de maneira similar àquelas ainda tolerantes à

126 dessecação, em termos de regulação gênica. Entretanto, o reinício do ciclo celular parece ser um fator decisivo, que evita o restabelecimento da tolerância à dessecação. Sementes recalcitrantes permanecem metabolicamente ativas durante o armazenamento, levando a uma rápida perda de viabilidade. I. vera é uma das espécies cujas sementes apresentam as menores capacidades de armazenamento, retendo a viabilidade por não mais que uma ou duas semanas se armazenadas em embalagens semi- permeáveis a 20 ou 5oC, respectivamente. Assim, uma tentativa foi feita na presente tese, visando reduzir o metabolismo de embriões de I. vera, consequentemente prolongando seu período de armazenamento. Para tanto, os embriões foram armazenados em um meio osmótico (PEG a -1,7 MPa) com ou sem ácido abscísico (ABA), um inibidor de germinação. Estresse osmótico e ABA, além de reduzirem o metabolismo, podem levar à expressão de genes envolvidos na tolerância à dessecação em plantas ou sementes. O armazenamento em PEG, a 5 ou 20oC, foi capaz de manter a germinação dos embriões de I. vera em 100% até 30 dias, causando portanto um aumento de até três vezes no período de armazenamento, quando comparado aos embriões armazenados em embalagens semi-permeáveis. O conteúdo de amido, principal fonte de reserva, em geral diminuiu com o aumento do período de armazenamento. Entretanto, não houve uma relação direta entre o conteúdo de amido e a viabilidade dos embriões. O efeito do ABA foi dependente da temperatura, sendo positivo a 20oC e negativo a 5oC. A permanência dos embriões no PEG por 14 dias não os tornou mais tolerantes à dessecação. Outra técnica usada para prolongar a viabilidade foi o armazenamento de embriões parcialmente secos em sacos plásticos aluminizados selados. Neste caso, os resultados foram melhores que no controle, mas inferiores aos obtidos com o armazenamento em PEG. De qualquer maneira, as duas técnicas usadas parecem ser promissoras no sentido de prolongar a curta capacidade de armazenamento das sementes recalcitrantes. Nossa hipótese é que sementes recalcitrantes perderam completamente características únicas das sementes, como a tolerância à dessecação, devendo então ser consideradas como plântulas em crescimento, em vez de sementes. Portanto, o armazenamento em atmosfera controlada, comum na conservação a longo prazo de frutas e legumes, pode dar bons resultados.

127

Acknowledgements

It’s not a matter of luck. I believe that everything in life happens to us according to our worthiness. If you attract good things, it is because you deserve them. The same for the bad things. So, I thank God for being worthy of having far more good than bad moments in my life. I thank the National Council for Scientific and Technological Development (CNPq, Brazil) for financial support of my PhD study at Wageningen University and Research Centre (WUR). I also thank the Department of Forest Sciences (DCF) from the Lavras Federal University (UFLA), Brazil, where I work, for having allowed me to go abroad for my PhD. My warmest thanks go to my supervisor Henk Hilhorst for having accepted me as his student and for having taught me about seed science. You’re a great scientist and very kind person, and this kind of combination is rare. Words are not enough to express how grateful I am for your wise guidance and friendship. I am waiting for your next visit to Lavras, so we can have some beers with ‘bean soup’ again. I also thank André van Lammeren, my co-promotor, for having taught me about cytology, especially the microtubules. I’ll never forget the many hours you spent with me at the microscope, always patient and gentle. I thank my promotor Linus van der Plas for being helpful whenever I needed. Thank you for your critical reading of the manuscripts of this thesis. I also thank my promotor Anne Mie Emons for everything she taught me about microtubules and for the discussion of the results during the weekly meetings of the Laboratory of Plant Cell Biology (WUR). Thank you also for your critical reading of the manuscripts of this thesis. I thank the staff and PhD students at the Laboratory of Plant Physiology (WUR) for the enjoyable friendship: Oblé Neya (have you already bought your notebook and mobile phone? You must look like a busy man!), Jaap Nijsse (thanks for translating the letters in Dutch that I received), Folkert Hoekstra, Sheila Adimargono, Wytske Nijenhuis, Trees van Nieuwamerongen, Baukje Lobregt, Said Husaineid, Marielle Schreuder, Jan van Kreel, Rob van der Vossen and Ben van der Swaluw. I’ll never forget the excursions organized by the Activities Committee of the Botanisch Centrum, always by bike, to see mushrooms, birds, or just to enjoy the picturesque surroundings of Wageningen. I also thank the people at the Laboratory of Plant Cell Biology (WUR) for the friendship and for having helped me with my thesis. From small things (like finding a chemical in the fridge) to discussions of our research. Thank you Bjorn Sieberer, Franck Lhuissier, John Esseling, Agnieszka Ozdoba, Carolina Cifuentes, Miriam Akkerman and Tiny Franssen-Verheijen. A special thanks goes to Henk Kieft, Norbert de Ruijter and Adriaan van Aelst, for being helpful every time I needed. Doing the flow cytometric analysis can be joyful if you have the assistance of Jan Bergervoet, Jeroen Peters and Wim Dirkse from the GreenFlow Lab (Plant Research

129 International, PRI), the lab with the best soundtrack ever. Pure rock, thanks to Jeroen. Thanks guys, you are great! I greatly thank Michiel Lammers, my paranymph and a great Feyenoord, sorry, Ajax, supporter. Thank you for your crucial help during my stay at PRI. If today I know something about molecular biology and lab techniques, I greatly owe this to you. Thanks also for trying to teach me how to spell ‘Woningstichting’. I also thank Wilco Ligterink and Jan Kodde (PRI), for always solving my (many) doubts about gene expression. I had a wonderful time working with you in lab number 3. A lab where the biggest threat to my health was not the mutagen chemicals that I handled, but the damn dance music (or whatever that horrible stuff was) that played loud on the radio all day long. I only had peace when, bravely, I turned to Arrow Rock Radio, with Hans Dassen’s encouragement. Thanks also to Ronny Joosen and Martijn Fiers (from the lab next door) for having kindly helped me every time I needed. I also had a great time in Angers, France, in 2003, when I spent two pleasant weeks in the lab of Olivier Leprince and Julia Buitink, learning the technique to re-establish desiccation-tolerance in Medicago truncatula seedlings. Besides of being great scientists, you are a very kind couple. Thanks also to all researchers, technicians and students from your lab. Thanks Olivier also for having lent me your Tea Party CD. Sailing down...down the Styx again... That band drove me crazy! I feel honoured by having in my promotion committee Prof. Dr. Maarten Koornneef, Prof. Dr. Rens Voesenek, Prof. Dr. Olivier Leprince and Dr. Bill Finch-Savage. Thank you all for being present in this important moment of my life. I thank people in Brazil who helped me, in one way or another, with this present achievement. Very special thanks go to Prof. Antonio Claudio Davide (UFLA, Brazil). If today I reached this point, I greatly owe this to you. In 1990, when you invited me to work with you in the ‘Riparian Forest Project’, my professional life started to change (in fact, my professional life started). During almost eight years I worked gladly, dealing with seeds, seedlings and ecological restoration of riparian forests. Working for that Project was essential in preparing me for further achievements, including the PhD that I am finishing now. The building is still being built up and it is solid, because its groundwork is the ‘Riparian Forest Project’. Claudio, I will never forget the opportunity you gave me and I will always be grateful. I thank my assignee Soraya Botelho (DCF, UFLA) for solving my problems in Lavras while I was out. Not many problems, I think. That’s the advantage of being assignee of a moneyless guy. I also thank the people responsible for collection and processing of Inga seeds in the Laboratory of Tree Seeds (DCF, UFLA): José Carlos (corno), José Pedro (Padre Marcelo), Olívia and Francisca. Amaral (DCF, UFLA) is acknowledged for his crucial help with the experiments with Inga seeds at UFLA. For the same reason, I thank Lisete Davide and Roselaine Pereira (Department of Biology, UFLA). Eduardo Alves and Heloísa (Department of Plant Pathology, UFLA) are acknowledged for their help with the microscopy studies. I thank Renato Paiva (Department of Biology, UFLA) for giving me some clues on how to live abroad and for having shown me that plant physiology and molecular biology can also be fun. I thank Dulcinéia de Carvalho (DCF, UFLA) for providing the protocols of DNA extraction and

130 for substituting me in my teaching activities while I was out. I thank Maria Laene de Carvalho and Renato Guimarães (Department of Agriculture, UFLA) for having allowed me to carry out part of the experiments with desiccation of Inga seeds in their laboratory. I thank Renato Castro (UCSAL) for helping in my endless attempts to isolate RNA from Inga axes. And while doing the PhD there was also time for sports and fun! I’ll always have good memories of our weekly soccer (I like it!!!) at the Sportcentrum de Bongerd, where I had the pleasure to meet people from every corner of the world. I miss you all, my friends! I enjoyed every match we played, and of course, every beautiful goal I scored (thousands, I think…), thanks to my unparalleled skill and modesty. I’ll never forget our soccer team ‘Ranca-Toco Futebol e Regatas’ (champion of the soccer tournament of the We-Day 2003). By the way, who ate the trophy? Very special thanks go to Michel Uiterwijk, my paranymph and organizer of our weekly soccer. God bless you, your wife and your children. Changing from sports to music, I also had the invaluable chance to attend three KISS expos: Zaandam 2002 and Helmond 2003 (NL) and Witten 2004 (Germany). Rock and roll all nite… I also had the pleasure to share good moments (and gallons of Grolsch) with the Brazilian (and connected) people in Wageningen. Too numerous to be cited, but I’ll try: Amaral (are your hands clean?), Cláudia, Chalfun (more ‘atleticano’ than me!), Cristiane, Fausto (just like in Brazil!), Tarsis, Tarcísio (never drink and drive! Unless you have a blue passport…), Franciene, Paulinho (the Google man), Veridiana, Luiz Marcelo ‘Passarinho’ (wake up, the lights are green!), Simone, Alan (friend of Nelsinho Caliper Mouth. You’ll always be remembered - and never forgiven - for missing the penalty in the final match of the We-Day 2002), Ivete, André Fuleragem (Kezman’s twin brother), Raquel, Mário (what?? Coffee with salami?) Ana Cláudia, Celso, Isabella, Stefano, Alessandro (the only guy in the Netherlands (maybe Europe) besides me who knows - and enjoys - Bartô Galeno’s songs), Jaqueline, Irene (more ‘petista’ than me!), Arne, Nilvanira, Tânia, Eduardo, Isabela, Rômulo, Flávia, Vagner, Luiz and Gilma. Sorry if I forgot someone. And last, but not least, I thank my family. Lucília, my mother, thank you for the education (in all senses) you gave me and for teaching how important the simplicity is for the spiritual elevation. I also thank my brothers Paulo Tadeu, Carlos Rogério and Mário Sérgio for being part of my life. Regiane, my wife, I greatly thank you for your love and fun during the last 19 years. This PhD wouldn’t have been possible without your permanent support. I love you! Alissa, my beloved daughter, my inspiration, thank you for making my life even more happy. And I couldn’t finish without saying: Netherlands, I can’t get enough! A wonderful country with the things just in the way they should be. Words cannot describe how much I liked living in Wageningen. I enjoyed every single day I lived here and I’ll miss Netherlands forever. There are countless things that I'll never forget like the windmills, tulips, stroopwafel, krentenbollen and Grolsch. And I hope nobody will kill me, but I do like the Dutch weather! Uhh… Henk… are you still my friend?

131

Curriculum vitae

José Marcio Rocha Faria was born in Lavras, Minas Gerais State, Brazil, on July 11th, 1964. In 1982 he started the undergraduate program in Forest Engineering at the Lavras Federal University (UFLA), finishing in 1986. In 1987 and 1988 he worked for a project on rubber- tree plantation in the southeast region of Brazil. From 1990 to 1997 he worked for the “Riparian Forest Project”, carrying out research on tree seeds, seedling production, land reclamation and ecological restoration of riparian forests. While working for this project, he pursued his Master’s degree (1993-1996) in Forest Engineering, also at UFLA. In his Master’s thesis he dealt with the use of various tree species for reclamation of degraded land. In 1997, having been successful in a public selection process, he became a faculty member in the Department of Forest Sciences at UFLA. Then, besides continuing with the research, he started to teach silviculture. After all these years of working in forestry, he realized that the seed is the bottleneck of the process of ecological restoration. When a seed can not be easily stored or germinated, the whole process is hampered or impeded. Thus, he decided to carry out his Ph.D. on seeds, more specifically on desiccation sensitivity, in an internationally recognized education and research organization. He was then granted a scholarship by CNPq (National Council for Scientific and Technological Development, Brazil) to pursue his Ph.D. at Wageningen University and Research Centre, the Netherlands, from 2001 to 2005. The resulting thesis comprises the molecular, cytological and physiological aspects of the desiccation tolerance and sensitivity in seeds of Medicago truncatula and Inga vera. Back in Brazil, the author resumed his job at UFLA, with desiccation sensitivity in seeds as his main research subject.

E-mail address: [email protected]

133

This work was financially supported by The National Council for Scientific and Technological Development (CNPq), Brazil Process 200193/99-6