This is the accepted version of the article:
Zinsmeister, Julia; Leprince, Olivier; Buitink, Julia. «Molecular and environ-
mental factors regulating seed longevity». Biochemical Journal, Vol. 447, issue 2
(Jan. 2020), p. 305–323. DOI 10.1042/BCJ20190165


This version is avaible at https://ddd.uab.cat/record/230978

under the terms of the             license
   Molecular and environmental factors regulating seed longevity
Julia Zinsmeister1, Olivier Leprince2, Julia Buitink2
1 Centre
       for Research in Agricultural Genomics (CRAG), Bellaterra (Barcelona), Spain,
2 Institut
        de Recherche en Horticulture et Semences, Agrocampus Ouest, INRA, Université
d’Angers, SFR 4207 QUASAV, 49070 Beaucouzé, France

Correspondence: Julia Buitink (julia.buitink@inra.fr)

Abstract

Seed longevity is a central pivot of the preservation of biodiversity, being of main importance
to face the challenges linked to global climate change and population growth. This complex,
quantitative seed quality trait is acquired on the mother plant during the second part of seed
development. Understanding what factors contribute to lifespan is one of the oldest and most
challenging questions in plant biology. One of these challenges is to recognize that longevity
depends on the storage conditions that are experimentally used because they determine the type
and rate of deleterious conditions that lead to cell death and loss of viability. In this review, we
will briefly review the different storage methods that accelerate the deteriorative reactions
during storage and argue that a minimum amount of information is necessary to interpret the
longevity data. Next, we will give an update on recent discoveries on the hormonal factors
regulating longevity, both from the ABA signaling pathway but also other hormonal pathways.
In addition, we will review the effect of both maternal and abiotic factors that influence
longevity. In the last section of this review, we discuss the problems in unraveling cause-effect
relationship between the time of death during storage and deteriorative reactions leading to seed
ageing. We focus on the three major types of cellular damage, namely membrane permeability,
lipid peroxidation and RNA integrity for which germination data on seed stored in dedicated
seed banks for long period times are now available.

Introduction

Seeds of the vast majority of plant species terminate their development by losing most of their
cellular water content and reach a state of anhydrobiosis, where cellular activities and
metabolism are suspended and the cellular content is solidified. This dry state confers to the
seeds the extraordinary property to escape from time and space, necessary for the dissemination
of the species and underpinning global agriculture and food security as it secures the planting
value for the next sowing season. The capacity to remain alive during storage, or longevity, and
the benefits of keeping seeds in a dry state to ensure establishment of a new crop was recognized
as early as -300 BC [1]. It has been exploited as a vehicle for the long-term ex situ conservation
of plant germplasm in over 1750 dedicated gene banks holding altogether ca. 7.4 million
accessions [2-4]. Maintaining this genetic diversity alive is necessary for breeders to produce
new varieties for tomorrow’s needs and constitutes an insurance policy to preserve threatened
species [5]. For this purpose, seeds are stored at a constant low temperature (-20°C) and low
humidity (3-7% moisture). Unlike these dry cold conditions, in the soil seeds of wild species
will survive for decades while being exposed to rehydration-dehydration cycles and a range of
temperatures and to pathogens. This seedbank longevity is understood as an adaptive trait to
seasonality and aridity [6-9].
        Understanding what factors contribute to seed lifespan is one of the oldest and most
challenging questions in plant biology (reviewed by [10]). One of these challenges is to
recognize that longevity depends on the storage conditions that are experimentally used because
                                                                                                  1
they determine the type and rate of deleterious conditions that lead to cell death and loss of
viability. Here, we will briefly review the different storage methods that accelerate the
deteriorative reactions during storage and argue that a minimum amount of information needs
to be provided to correctly interpret the longevity data.
        Longevity is progressively acquired during seed maturation [9, 11, 12], after the
induction of desiccation tolerance. Depending on the species, the acquisition of longevity can
be can be spread out over several weeks after seed filling and can even still occur after harvest
[9, 13]. Longevity acquisition is regulated by hormonal and maternal factors, and is under
strong influence of the environment. In this review, an update of the current knowledge of these
regulatory mechanisms is presented that is complementary to recent excellent reviews
describing the protective factors involved in longevity [7, 9, 14, 15]. In the last section of this
review, we discuss problems in unravelling cause-effect relationship between the time of death
during storage and deteriorative reactions leading to seed ageing. We focus on the three major
types of cellular damage, namely membrane permeability, lipid peroxidation and RNA
integrity, for which germination data on seeds stored in dedicated seed banks for long period
times are now available. Next to longevity, developing seeds acquire dormancy, a physiological
state that blocks germination despite favorable conditions. Although genetic links have been
established between both longevity and dormancy [16], the latter trait will not be further
developed in this review.

Seed longevity in a nutshell

Over the past decades, it has become clear that there are numerous factors contributing to seed
longevity. To provide a background for the different sections in this review, Figure 1 gives a
summary of these factors. They include protective factors encompassing the seed coat, whose
compounds limit the diffusion of oxygen to the embryo, a range of lipophilic and water soluble
antioxidant compounds, stress proteins (Heat Shock Protein (HSP), Late Embryogenesis
Abundant (LEA) proteins) with moonlight functions, RNA binding proteins that conserve
mRNA in the dry seeds and non-reducing sugars (sucrose and oligosaccharides) that prevent
protein denaturation, protect membranes, and participate to the formation of a glassy state (Fig.
1). Longevity depends also on mechanisms that repair genome lesions, amino acid
modifications and aggregated proteins. For more detail we refer the reader to recent excellent
reviews describing the protective factors involved in longevity [7, 9, 14, 15]. Longevity is a
multifactorial trait, evident from the multiple genetic factors have been identified using
Quantitative trait locus (QTL) and Genome-Wide Association Study (GWAS) approaches
using historical data from long-term storage or conditions that accelerate seed ageing in a range
of crops such as wheat [17, 18], soybean [19, 20], oilseed rape [21-23], maize [24], rice [25-
27], barley [28], tobacco [29], lettuce [30] and Arabidopsis [16]. In these studies, different
storage / aging protocols have been used, giving rise do different QTL, which complicates a
comparative analysis of the data. In the next section, the importance of the storage protocol to
assess longevity will be discussed.

‘Wet’ vs ‘dry’ ageing

Experimentally, assessing seed longevity on a tractable scale requires to increase at least one
of the storage conditions that influences it, namely seed water content (or relative humidity
[RH] at which seeds were equilibrated), temperature and oxygen availability. Ageing protocols
used to assess the actual lifespan of the seeds and its extrapolation to colder or drier conditions
are important issues. There is a wide range of temperature / RH combinations during storage
conditions that are used by laboratories and a wide range of wordings to describe them, such as
                                                                                                 2
ambient or natural ageing, accelerated ageing or controlled deterioration (reviewed by Hay et
al., 2018 [31]). This makes it difficult to compare the molecular responses to seed ageing across
studies, or even within a species. Indeed, storage conditions will determine the type of ageing
reactions and the underlying damage that will lead to cell death. For example, storage over
water vapor at 100% relative humidity (RH) and 42-45°C will allow both the resumption of
respiration as seeds are progressively imbibing and a heat stress leading to protein denaturation,
whereas storage at 75% RH at 35°C does not allow respiration to occur because the cytoplasm
is too viscous to allow molecular motion necessary to produce ATP. In barley, the biochemical
changes occurring during loss of viability spanning over 4 years of storage in the laboratory
conditions (50% RH, 20°C) were not identical to those occurring during the loss of viability at
75% RH, 45°C [32]. Similar conclusions have been reported using biophysical approaches [33]
or genetic approaches [30, 34]. Therefore, great care must to be taken in extrapolating
mechanisms or genes identified under conditions that accelerate ageing to conditions that will
be found in the soil, seed banks or the laboratory bench.
        Recently, a new ageing method was introduced to accelerate ageing by increasing the
oxygen concentration during storage by elevating the pressure to 18 kPa using scuba diving
tanks [35]. Whether this accelerated ageing can be construed as a means to mimic deterioration
during storage in ambient conditions remains an open question. Variations in pressure during
the experiment is itself detrimental. Barley seeds exposed to high nitrogen pressure for 6 weeks
showed a significant number of abnormal seedlings with white shoots and no roots compared
to unaged controls [36]. It is likely that a high pressure disturbs the physico-chemical and
structural properties of the cytoplasmic glass as demonstrated in other glass systems. For
example, in a hydrous silicate glass, an increase in pressure induces the H2O molecules to act
as a network former rather than a modifier, making the glass elastically stiffer [37]. In meat, an
increased pressure decreases lipoxygenase activity [38], whereas this is known to be an enzyme
that contributes to seed ageing [39, 40].
        Hay and colleagues [31] recommended that seed ageing protocols should be designed
based on the potential downstream use of the findings and the biological significance of
longevity. In an effort to contribute towards this objective, Figure 2 presents the minimal
information necessary for an unambiguous interpretation of an ageing experiment. In soil
seedbanks, seeds are exposed to hydration-dehydration cycles, allowing some metabolism to
occur because the cytoplasm remains fluid [7]. This could be referred to as “wet ageing”,
corresponding to storage at a RH above 80% and temperatures ranging from 4 to 45°C. Also
classified in this category are the “accelerated ageing” and “controlled deterioration” tests.
These are two different and specific tests defined by the International Seed Testing Association
as a means to evaluate seed performance during germination rather than longevity per se (see
[31] for the detailed protocols). An intermediate storage condition that is close to ‘dry’ aging
or bench storage corresponds to 75% RH and 30-35°C. These conditions will bring the
cytoplasm in a highly viscous, rubbery state, with deteriorative reactions severely slowed down,
but still being slightly above the glass transition temperature of the cytoplasm. Finally, we
consider that storing seed at or below 70% RH corresponds to a “dry” ageing. This corresponds
to conditions used in 1) gene banks where seeds are stored under constant dry conditions (<35%
RH) and low temperature (4°C/-20°C) where the cytoplasm is solidified and in a glassy state
with restricted molecular mobility and 2) ambient storage or storage under laboratory conditions
(22-25°C, 50-70% RH). These latter conditions correspond to the limit where short scale
molecular mobility is allowed but not long range slow diffusive motion (or α relaxations) [33,
41]. Typically, the upper limit of RH/temperature combinations for a dry ageing that would
represent a good compromise between the above conditions and those leading to a manageable
loss of viability are 60-75% RH and 35-45°C. Good correlations were obtained for longevity
of seeds of the tree Astronium fraxinifolium that were stored at 60%, 65% and 70% RH [42].
The boundaries of what is “wet” or “dry” are difficult to set. They depend on the temperature
                                                                                                3
and are likely a continuum from short-range molecular mobility and long-range diffusive
motion within the cells [33, 41, 43].

Effects of maternal tissues on seed longevity

The contribution of the maternal-derived outer tissues of the seed to its longevity can be
substantial. These tissues include the seed coat that differentiated from the ovule integuments
during development and, if present, the pericarp, which is derived from the ovule. The
endosperm is a triploid tissue that develops into a storage organ in endospermic seeds and to a
thin layer surrounding the embryo, controlling germination, in non-endospermic seeds. The
endosperm has an increased dosage of the maternal genome. The importance of the endosperm
in seed longevity has been shown indirectly by meta-analyses of seed traits. Within 197 species
across the globe [44] and 172 Australian species [8], life span of endospermic seeds stored at
45°C and 60% RH were on average 3.3-fold lower than non-endospermic seed. The positive
effect of reduced endosperm on seed longevity was recently confirmed on the panel of 39
grassland species from a single habitat [45]. Why the presence of endosperm would reduce
longevity remains unknown. Floris and Anguillesi (1974) [46] concluded that in wheat, the
aged endosperm had a “debilitating” effect on the aging embryo by leaking some unknown
toxic substance. Intriguingly, when freshly harvested embryos were imbibed in the presence of
an isolated endosperm that had first been aged for 3 years in ambient conditions, they carried
more chromatin damage during imbibition than embryos imbibed in the presence of non-aged
endosperm [46]. It is noteworthy that in cereals, endosperm dies during maturation drying or
very fast in the dry state, probably due to insufficient protective compounds. An exhaustive -
omics comparison of embryo and endosperm in rice revealed that the endosperm has no or
decreased amount of polypeptides with an important role in seed longevity such as protective
proteins (several LEA proteins, small HSPs, lipocalin, an antioxidant) and repair proteins
(PIMT, protein-L-isoaspartate O-methyltransferase and methylsulfoxide reductase, Figure 1,
[47]). Therefore, it will be an important future research goal to better understand the influence
of endosperm on seed longevity.
        Pioneering work by Debeaujon and colleagues [48] established that mutants of
Arabidopsis seeds defective in genes regulating seed coat composition and properties exhibited
decreased life span compared to wild type. The worst impact on longevity is the absence of the
two integument layers in the ats mutant [48]. The TT (TRANSAPRENT TESTA) genes encode
maternal regulators of seed coat development and enzymes involved in flavonoid synthesis.
Seeds of mutants defective in these genes also have a reduced longevity when stored at ambient
conditions or at high RH / temperature combinations [12, 48, 49]. Next to a decreased longevity,
these mutants also exhibited an increased permeability to tetrazolium salts, suggesting that the
primary role of the seed coat is to isolate the embryo from the storage environment such as
oxygen and water. The origin of the increased permeability has been recently elucidated using
Arabidopsis mutants that are defective in genes involved in the biosynthesis of very long chain
fatty acids. It was found that seeds in these biosynthesis mutants were defective in a cutin layer
between the endosperm and the seed coat [49]. In Arabidopsis, this cutin layer is only 300 nm
thick and is made of abundant deposits of aliphatic polymers of glycerol esterified to long-chain
fatty acids and hydroxyacids along with proanthocyanidins [49, 50]. In these mutants, wet
ageing at 82% RH, 42°C induced an increase in peroxidized lipids and decreased tocopherols,
suggesting that the cutin layer prevents oxygen diffusion through the embryo [49]. Whether this
protection occurs during dry storage remain to be assessed. Originally thought to be of
endosperm origin [49], the deposition of a cutin actually originates from the maternal inner seed
coat and is deposited to the surface of the endosperm under the control of TT1 and TT16 [50,
51]. In Arabidopsis, the formation of cutin occurs very early after fertilization. It is probably
triggered by a signal originating from the fertilization of the central cell under the relief of the
                                                                                                  4
FERTILIZATION INDEPENDENT SEED Polycomb group repressive mechanism. Its original
function is to prevent the fusion of tissues so that the seed coat is not attached to the embryo
and also ensure proper embryonic growth [51]. The presence of this cutin layer may also
influence the communication between the embryo and maternal tissues during development.
Possibly, cutin does not act alone to provide a barrier separating the embryo from the
surrounding environment. Indeed, suberin is another hydrophobic polyester located in the outer
layer of the seed coat that provides impermeability to oxygen and water during seed imbibition
[52, 53]. Its role in seed longevity has not been investigated yet.

Hormonal regulation and signaling pathways involved in longevity

Seed maturation and the underlying developmental programs are under the control of the so
called LAFL transcriptional regulators which include the plant-specific family of B3 domain
transcription factors ABI3 (ABSCISIC ACID INSENSITIVE3), FUS3 (FUSCA3), and LEC2
(LEAFY COTYLEDON2) and the LEC1 (LEAFY COTYLEDON1), a member of the NFYB
protein family (see [54] for a review). In Arabidopsis, loss of function mutants of ABI3 and
LEC1 produce seeds that lose their viability during the first few weeks after harvest [55],
implying that these transcription factors regulate the acquisition of longevity. However, these
transcription factors do not act alone in regulating seed longevity. They form complexes
between them and with other proteins yet to be identified and act redundantly [54, 55], making
it difficult to unravel their precise role in seed longevity.
         Downstream of ABI3 are several genes that are implicated in seed longevity (Figure 1),
determined mostly by wet ageing. They include the seed specific HEAT SHOCK FACTOR A9
(HSFA9), controlling the developmentally regulated expression of sHSP [56]. This
transcription factor is part of the genetic program controlling resistance again a deterioration
induced by incubating wet seeds (28% moisture) for 2 days at 50°C [57, 58]. Likewise,
induction of Mt-HSFA9 expression was also correlated with the acquisition of longevity both
in M. truncatula [59] and in soybean [60]. In Arabidopsis, ABI3 also regulates seed degreening
via the induction of STAYGREEN [61], a process also associated with the acquisition of
longevity. By introgressing loci from different Arabidopsis ecotypes into abi3-5 and lec1-3
lines and restoring longevity, Sugliani et al. 2009 [55] showed that ABI3 and LEC1 regulate
directly or indirectly and in a redundant manner a set of protective proteins such as small HSP
and LEA proteins as well as storage proteins that are associated with longevity [11, 62].
         An additional layer of complexity in the regulation of longevity involves alternative
splicing of ABI3 mRNA. In developing Arabidopsis seeds, the relative abundance of transcripts
encoding the full-length ABI3 decreases between 14 and 20 days after pollination (i.e. dry
mature seeds), concomitantly with the acquisition of seed longevity. Also, during that period,
transcripts of a spliced form, ABI3-β, increases and becomes preponderant in dry seeds [65].
ABI3-β encodes a truncated protein that contains two of the four functional domains (A1 and
B1), but whose function in seed maturation is unknown. This splicing event is controlled by
SUPRESSOR OF ABI3-5 (SUA1 to SUA4), a RNA binding protein that is involved in seed
longevity. Arabidopsis seeds of double mutant sua1 abi3-5 exhibited a longevity during dry
storage that was similar to wild type seeds, and restored the reduced longevity phenotype found
for abi3-5 seeds [63].
         During maturation, ABI3 is not the only gene to be alternatively spliced. Twenty percent
of the genes expressed between 14 and 20 Days after pollination (DAP) are alternatively spliced
and enriched in Gene Ontology (GO) functions associated with mRNA catabolic processes [64].
Alternative splicing affects a higher percentage of genes at 20 DAP compared to 14 DAP,
possibly representing a mechanism that allows the rapid end of seed maturation once
desiccation is achieved. Considering that spliced forms of ABI3 have been detected in seeds of
many species (see references in [64]), alternative splicing might represent an important and
                                                                                               5
conserved form of longevity regulation that deserves further attention. Taking advantage of the
slow acquisition of longevity in the developing legume seed M. truncatula, a conditional-
dependent network of global transcription interactions revealed distinct co-expression modules
related to the successive acquisition of desiccation tolerance and longevity [59]. The desiccation
tolerance module was associated with abiotic stress responsive genes (typically LEA proteins),
whereas the longevity module was enriched in genes involved in RNA processing and
translation [59], in concordance with an increased splicing activity at the end of maturation
[64]. This approach was taken a step further by incorporating into the network the
environmental effects on the acquisition of longevity in Medicago and genes co-regulated with
the acquisition of longevity in Arabidopsis [12] (Figure 3). While 85% of the genes in the
module associated with the acquisition of desiccation tolerance were deregulated in Mt-abi3
mutants, less than 50% of the genes present in the longevity module were found to be
downstream targets of Mt-ABI3, suggesting that these genes are regulated via alternative
pathways that remain to be elucidated.
        Another player of the ABA signaling pathway, ABI5, has also been shown to be an
important regulator of longevity in legumes such as M. truncatula and pea seeds [65]. In both
species, seeds of abi5 mutants were desiccation tolerant but exhibited a 40 to 60% reduction in
longevity during dry storage (75% RH) compared to wild-type. ABI5 is a b-Zip transcription
factor of the ABF (ABRE‐BINDING FACTOR) family that in response to ABA regulates many
genes involved in conferring stability in the dry state such as genes encoding LEA proteins and
Raffinose Family Oligosaccharide (RFO) synthesis [65-67]. It also regulates the seed
antioxidant status because compared to wild type seeds, Mt-abi5 seeds were more sensitive to
oxygen during storage and exhibited an altered tocopherol profile [65]. Interestingly, both in
legumes and in Arabidopsis, ABI5 regulates photosynthesis and chlorophyll metabolism genes
[65, 68]. As a result, in abi5 mutants of pea and Medicago, mature seeds retained significant
amounts of chlorophylls [65, 68]. In oily chlorophyllous seeds, the presence of chlorophyll in
mature seeds is an undesirable trait leading to reduced seed vigor and longevity [61, 69, 70]. In
Arabidopsis, abi5 seeds did not exhibit a longevity phenotype or a green seed phenotype [65],
probably because of the action of redundant bZIP transcription factors homologous to ABI5,
altogether forming the ABF family [67, 68].
        While in Arabidopsis, ABI5 is acting downstream of ABI3 during seed germination to
regulate growth arrest upon drought stress [71], it remains to be established whether this is the
case during seed development or whether ABI5 acts independently of ABI3 [66]. In M.
truncatula, Mt-ABI5 expression was not deregulated in Mt-abi3 mutants [59]. Furthermore, in
this species, ABI3 and ABI5 have complementary roles in inducing desiccation tolerance [72].
In Arabidopsis, ABI5 appears to be regulated by DOG1 (DELAY OF GERMINATION 1) [67].
Seed transcriptomes of these mutants showed an intriguing overlap, with genes involved in
longevity being deregulated such as the molecular chaperone genes encoding LEA proteins and
HSP. Consistent with this, dog1 mutants exhibited a reduced life span when stored under dry
laboratory conditions [73]. Recently, major progress has been made in elucidating the
regulatory role of DOG1. DOG1 is a heme binding protein that interacts with ABA
HYPERSENSITIVE GERMINATION1 (AHG1, a group A type 2C protein phosphatases
(PP2C)) to trigger an ABA response and dormancy [74]. This signaling pathway seems to act
in parallel with the core ABA signaling pathway upstream of ABI3. This scenario was
demonstrated in the modulation of seed dormancy in response to temperature. Whether a similar
scenario can be envisaged for longevity remains to be assessed. A genetic link with DOG1 and
ABI3 in the regulation of seed longevity is supported by the observations that in Arabidopsis,
the dog1-1 mutant enhances abi3-1 phenotypes and phenocopies a severe reduction in longevity
and chlorophylls retention in seeds of stronger abi3 alleles [67]. Therefore, DOG1, ABI3, ABI5
and homologous ABF genes form an important module that regulates the acquisition of

                                                                                                6
longevity in Arabidopsis. A DOG1 domain-containing protein exists in the genome of M.
truncatula and it would be interesting to investigate its function in longevity.
         While evidence is secured that longevity is regulated by the components of the ABA
signaling cascade, a direct role of ABA metabolism in the acquisition of longevity is still
unclear. Studies on hormone synthesis and degradation in Arabidopsis show that effects on seed
longevity were only reported for aba1-5 seeds, with a significant loss in viability observed after
4 years of storage [75]. However, two Brassica oleracea genotypes with allelic differences at
two QTLs that result in differences in ABA content did not differ in their longevity when stored
at 75% RH and 40°C [76]. Likewise, Arabidopsis mutants of ABA degradation (cyp707a1,
cyp707a2 and cyp707a3) were not affected in seed longevity as determined by a wet ageing at
80-85% RH, 37 -40°C [77, 78]. However, the role of ABA metabolism in longevity needs
probably to be assessed by taking into account the environmental conditions during seed
maturation. When stored at approximately the same conditions (85% RH, 40°C), seeds of the
double mutants nced6 nced9 affected in ABA biosynthesis died faster than wild type if seeds
were grown at 15°C but no difference was found between then genotypes at 25°C [78]. Also, a
recent exhaustive study on ABA metabolism in developing seeds of aba-2 mutants defective in
ABA synthesis demonstrated the accumulation of ABA glucosyl ester, suggesting an alternative
route of ABA synthesis [79].
         Auxin is another hormone that is involved in the regulation of the acquisition of
longevity [12, 58, 80]. A pioneering work on identifying the regulator of HaHSA9, a seed
specific heat shock factor involved in the resistance against deterioration during wet storage,
led to the isolation of the sunflower AUXIN-RESPONSIVE PROTEIN 27 (HaIAA27), a short-
lived transcriptional repressor of early auxin response genes. HaIAA27 repressed HaHSFA9
and the subsequent accumulation of sHSP via direct interaction [58]. An auxin resistant form
of the sunflower HaIAA27 was able to repress the activation of the sHSP promoter by
HaHSFA9 in mature bombarded tobacco embryos and its ectopic overexpression resulted in
poor germination after 2 days of storage at 28% moisture, 50°C, comparable to HSFA9 loss of
function lines [58]. Our own recent work also demonstrated a direct link between auxin
signaling and the acquisition of longevity [80]. This study was based on the discovery of a
conserved gene module related to seed longevity (Figure 3) that contains genes that are
significantly enriched in the cis-regulatory element ARFAT, an auxin response factor (ARF)
binding site [12]. In addition, this module contains CYP79B2, a gene involved in a minor auxin
biosynthesis pathway (i.e. indole-3-acetaldoxime pathway, IAOx) [81, 82]. During maturation
of Arabidopsis seeds, auxin signaling input and output increased and spread progressively
throughout the embryo, concomitant with the acquisition of longevity. When auxin was
supplied exogenously during maturation, seed longevity increased. Consistent with this,
longevity of seeds of single mutants defective in the different auxin biosynthesis pathways was
altered in a dose-response manner depending on the level of auxin signaling activity, with lower
signaling activity coinciding with decreased longevity [80]. The downstream effect of this auxin
signaling was associated with the ABA signaling pathway in the embryo. The expression of
ABI3 and its LEA protein target EARLY METHIONE1 (EM1) were induced by auxin and
deregulated in the auxin biosynthesis mutants. Furthermore, the beneficial effect of exogenous
auxin during seed development on longevity was abolished in abi3-1 mutants. Besides the ABA
signaling pathway, auxin might also directly regulate genes involved in longevity. Indeed, other
longevity-associated genes that are not regulated by ABI3 were also affected in their expression
downstream of the auxin signaling [80].
         The role of gibberellins in seed longevity has also been suggested but needs to be further
secured. Mutants of the transcription factor HOMEOBOX25 (ATHB25) are more sensitive to
wet ageing at high temperature [85]. ATHB25 positively regulates GA synthesis and reinforces
the seed coat. The positive role of gibberellin in resistance against deterioration was further
demonstrated by the increased resistance of seeds from GA3-treated plants or the quintuple
                                                                                                 7
DELLA mutant [83]. However, like for the studies on HSFA9, a wet ageing was used in this
study, corresponding to an imbibition at 42°C at 100% RH. In addition, seeds were incubated
for 4 days at 4°C before testing seed viability. Considering that such cold treatment releases
dormancy, decreases ABA content and induces GA synthesis, it remains to be assessed whether
a similar conclusion would have been reached if athb25 seeds were aged under dry conditions
without a 4°C imbibition. In contrast to the above observations, ga1-3 mutant defective in
gibberellin synthesis and the gibberellin insensitive gai mutant did not exhibit a decreased
germination after 4 years of dry storage compared to wild type [75].
        Recently, it emerged that brassinosteroids (BR) could modulate negatively seed
longevity during priming, a controlled imbibition and drying treatment that improves
germination performance [77]. After priming, seeds of the BR-deficient mutants cyp85a1/a2
and det2 showed significantly longer longevity than the wild type. It was suggested that the loss
of longevity in primed seed is due to increased seed coat permeability, which is positively
regulated, at least partly, via BR signaling. Whether this hormone is also related to maturation-
induced acquisition of longevity needs investigation. There is no report mentioning the
involvement of other hormones in longevity, apart from the observation that the ethylene-
resistant1 (etr1) and jasmonic acid-resistant1 (jar1-1) mutant seeds did not show any
significant viability loss after storage for 4 years [75].

Environmental regulation of longevity during seed maturation

Longevity, like many other seed traits such as seed filling and dormancy, is influenced by the
abiotic environment that is experienced both by the mother plant and the zygotic tissues.
Conditions that influence longevity include light, temperature, drought, and salinity (Table 1),
with temperature and water availability seeming to be the dominant factors. For crops, most of
our knowledge is inferred from works based on wet ageing conditions that were used to assess
the planting value of the seed, whereas for wild species a wider of range of dry storage
conditions were employed. Our current understanding of the environmental impact on longevity
and how it affects the molecular factors contributing to longevity is fragmentary but known to
strongly involve genotype x environment interactions. Environmental conditions can be
experienced by the seed or the mother plant as a constraint that limits or slows down the
developmental programs leading to the acquisition of longevity [6, 12, 84]. There are also
significant variations in seed longevity in relation to the climate or environmental conditions
for which the species is adapted. A longevity survey of 197 wild species stored at 60% RH,
45°C showed that seeds originating from species collected in hot, dry environments tend to
have a longer lifespan from those found in cool, wet conditions, regardless of the phylogenetic
relatedness, suggesting that species have adapted to dry climates by increasing seed longevity
[44]. Concordant with this, variations in longevity were found in Silene species originating
from differing climates, with a longer life span for Mediterranean accessions found in dry, hot
habitats compared to alpine accessions [85]. A similar variation of longevity of Arabidopsis
seeds stored under wet ageing conditions (85% RH, 35°C) was observed using DOG near
isogenic lines that are the result of adaptations to local environmental conditions [78]. At a local
geographic scale, environmental characteristics can also influence longevity. Plants of Silene
vulgaris growing in warm environments produced longer-lived seeds from those from colder
environments [6]. All these studies led to the conclusion that longevity might be construed as a
long-term adaptation to the dry, arid environment, with the consequence that it would exert a
cascading selective pressure on future generations. Whether a similar hypothesis can be posited
for crops remains to be assessed. Limited information suggests that it might not be the case, at
least for barley. In a survey of longevity in 175 accessions of barley including landraces and
cultivars from Europe, Africa, Asia and Australia, genotypic variation in seed longevity was
independent of the geographical origin of the accession or its breeding status [34].
                                                                                                  8
        Alternatively, intraspecific variations in longevity might represent phenotypic plasticity
mediated by the mother plant as the response to the environment. This interpretation is
supported by experiments showing that it is possible to decrease or increase longevity by
manipulating the environment, such as temperature and drought in greenhouse conditions for
M. truncatula [12] or growing Silene mother plants in garden outside their natural habitat [6].
Another strong argument in favor of plasticity is the observation made by [86] on the parental
influence on longevity of Wahlenbergia tumidifructa seeds, a semi-arid species. When mother
plants were submitted to cold and wet conditions before returning to optimal conditions after
fecundation, longevity doubled compared to seeds obtained from mother plants grown under
warm dry conditions as prezygotic conditions. Understanding the causal genes implicated in
long term adaptation and/or plasticity of longevity and whether transgenerational effects are at
play represent an exciting new field of investigation.

Temperature
Temperature can affect longevity positively or negatively in a species- and genetic-dependent
manner. Warm temperatures during seed development of alpine species and Arabidopsis tend
to increase longevity during dry storage [87-89]. In contrast, warm temperatures negatively
affected seed longevity in rice [84] and M. truncatula [12]. Low temperature during seed
development led to a reduction of seed longevity in Arabidopsis [78] but had no effect on M.
truncatula [12], probably because this species is adapted to such conditions.
        How a change in temperature imparts changes in seed longevity is not understood but
likely implies mechanisms both from the mother plant and the zygote. MacGregor and
colleagues [90] showed that cold temperature influences the phenylpropanoid composition and
permeability of the seed coat in Arabidopsis, a maternal tissue for which defective mutants
showed decreased longevity [48]. We mentioned earlier that DOG1 affects longevity in
Arabidopsis [67]. In fact, DOG1 is mostly known for its function as a temperature sensor that
modulates seed dormancy (references in [91]). DOG1 protein levels vary according to seed-
maturation temperature, and mutations in this gene compromise the ability of the seeds to enter
dormancy in response to low maturation temperatures. An exciting possibility would be that a
similar mechanism could be acting to regulate the longevity response to temperature.

Water stress
Drought can be beneficial or detrimental to seed longevity according to the species and intensity
of the water deficit. In soybean, a water stress during seed maturation led to mature green seeds,
a trait leading to decreased longevity [70, 92]. A similar effect of drought was found on seeds
of M. truncatula, although seeds did not show a retained chlorophylls phenotype [12]. In
contrast, in Brassica rapa, arresting irrigation during early seed filling fastens the acquisition
of seed longevity, resulting in mature seeds with a longer lifespan during dry storage [93, 94].
Similarly, in peanut, a drought stress during seed development enhanced seed longevity [95]. It
is likely that water availability will have a different impact on the longevity depending on the
stage of seed development because developing seeds can modulate their maturation to regain
to some extent the longevity that was lost due to the water availability. This is suggested by
field experiments on wheat for which rainfall was simulated by wetting the ears at different
stages of development or decreasing rainfall by providing shelter [96]. Wetting ears for 30 min
at all stages of seed development increased the seed water content by 5-10% only but led to
reduced longevity after harvest (tested under dry storage). However, these seeds could regain
considerable longevity when wetting was suspended and seeds were able to re-dry in planta for
several days [96], highlighting the plasticity of the development program leading to longevity.
The physiological causes for this behavior remain to be understood, but since ABA signaling
is involved in chloroplast dismantling during seed maturation that in turn is affected by water

                                                                                                9
stress in soybean [70], ABA signaling might be important for longevity in response to water
stress.

Light
During seed filling and despite the presence of surrounding layers of tissues, embryos are green
and can perceive 20 to 30% of light, which is enriched in green and far red wavelengths [97].
Photosynthesis is necessary to provide NADPH and ATP for seed reserve deposition. For this
purpose, chloroplasts are adapted to these shade conditions. They display a high Chla/Chlb ratio
and photosystem I activity is enhanced at the expense of photosystem II [97]. During the later
phase of maturation, and concomitant with the acquisition of longevity, chloroplasts are
dismantled and chlorophyll molecules degraded via mechanisms different from those in
senescing leaves [98, 99]. During this phase, light may actually be detrimental for these
chlorophyllous embryos. Seeds of many species in which traces of chlorophylls remain at
harvest or due to gene defects in chlorophylls degradation exhibit a decreased shelf life during
wet [98] and dry storage [65, 67, 69]. While Li et al. [98] provide some evidence for
photodamage, the role of light and chlorophylls during storage are not yet elucidated because
these longevity phenotypes are also observed on seeds that are stored in the dark [65]. Perhaps,
the detrimental effects of light on seeds with retained chlorophylls occur during seed maturation
rather than during storage or imbibition. Despite the shade conditions provided by the seed
layers, control of photochemical reactions during chloroplast dismantling induced by
maturation drying might be necessary to avoid the buildup of toxic compounds that could
jeopardize longevity. In support of this hypothesis, mutants defective in the assembly of the
cytochrome b6f complex or wild type seeds treated with DCMU (3-(3,4-dichlorophenyl)-1,1-
dimethylurea), a specific inhibitor of photosystem II activity, display a reduced lifespan during
storage at 75% RH, 40°C [97].
         Another clue that light might be important for longevity acquisition comes from a study
on the effect of photoperiod on longevity of lettuce seeds, a species displaying photosensitivity
during germination [100, 101]. Seeds produced under red-rich light conditions exhibited
reduced longevity during storage at 30°C and 74% RH compared to seeds produced under far
red light [100]. In Arabidopsis, light intensity but not photoperiod had a positive effect on
longevity, and it was suggested that this might regulate RFO accumulation [102]. This suggests
that developing seeds integrate the light signal to modulate longevity but how they do so remain
unknown. In Arabidopsis, light perception involves the underlying genes of DOG3 and DOG6
loci, since seeds of the near-isogenic lines NILDOG3, NILDOG6 produced under different light
conditions exhibited an altered longevity compared to the Landsberg erecta (Ler) ecotype [78].
In the future, cloning the genes underlying DOG3 and DOG6 loci should reveal the regulatory
function of light intensity on longevity.

Nutrient availability
Nutrient availability in the soil and the status of macro- and microelements in the mother plant
can affect seed yield components. Likewise, nitrate, phosphate and sulfate availability also
impact various traits associated with germination performance such as dormancy in a complex
interaction between genotype and other environmental conditions such as temperature and light,
as shown for tomato [103], Arabidopsis [78] and oilseed rape [102-104]. Therefore, it is
expected that longevity might be affected as well, although this aspect has not received much
attention. In barley, longevity assessed by intermediate ageing conditions (75% RH) was higher
in seeds produced from plants cultivated with an adequate nutrient supply (P, K, N, Mg and
organic matter) compared to those from plants grown with a lower nutrient supply [34]. In the
Arabidopsis Ler ecotype, longevity also depended on the N supply to the seeds, higher nitrate
leading to longer life span when aged at 82% RH. The effect was also observed with Col for
high N supply, but to a lesser extent [78]. As a result of a low N supply, contents in amino acids
                                                                                               10
and glucoronate were affected in mature dry seeds together with gene transcripts impacted in
cell wall metabolism, which led the authors to suggest that the C/N balance was impacted.
Interestingly, the transcriptome and metabolome of seeds grown without nitrate overlapped
with those of seeds produced at 15°C [102], leaving the possibility open that temperature might
directly impact the C/N balance.
        Another evidence suggesting that the C/N balance is important for longevity is provided
by data showing that altering metabolic sensors during seed development impacts longevity. So
far, two metabolic sensors have been identified for which seeds from mutants showed decreased
resistance against ageing. The first one is the Target of Rapamycin (TOR) complex [105]. In
Arabidopsis, mutants of the REGULATORYASSOCIATED PROTEIN OF TOR 1B
(RAPTOR1B), a conserved TOR interactor, showed a decrease in germination after one year
in ambient, cold storage as well as decreased resistance against wet ageing [105]. Decreased
longevity was associated with changes in hormone profiles including ABA and auxins, a
reduction in sugar contents, changes in secondary metabolite profiles and seed coat composition
[105]. A second sensor is the Sucrose non fermenting Related Kinase 1 (SnRK1) complex
[106]. In developing M. truncatula embryos, silencing MtSNF4b, a seed-specific regulatory γ‐
subunit of the SnRK1 complex, led to reduced longevity during dry storage concomitant with
an 80% decrease in RFO and a 60% increase in sucrose content [106]. Both sensors have crucial
roles in nutrient and energy sensing and translate this information into metabolic and
developmental adaptations.

The continuing challenge of assessing causes of seed ageing and future
perspectives

During the past decade, our knowledge of the fate of lipids, proteins and nucleic acids during
seed ageing has greatly improved. However, a causal relationship between accumulation of
damage with time of death during storage remains difficult to assess, making it complicated to
pin-point the origin of the deterioration during storage and to trace back the overall effect to a
particular set of genes. First, when cell death occurs at some point during seed storage, this does
not mean that deteriorative reactions stop. Secondly, the relative contribution of viable and non-
viable individuals (seeds) or cells (within a seed) to the amount of damage incurred by the seed
population remains also difficult to establish. Thirdly, it is still unknown whether deteriorative
reactions occur randomly during storage and are governed by stochastic processes, or whether
there are specific compounds that are degraded during storage, leading to a reproducible
deteriorative path [107]. To illustrate these issues, we review three major types of cellular
damage, namely membrane permeability, lipid peroxidation and RNA integrity, for which
germination data on seed stored in dedicated seed banks for long period times are now available.

Membrane permeability
Using EPR spectroscopy, Golovina and colleagues [108] were able to assess membrane damage
in vivo on a seed to seed basis in wheat seeds stored over an 18-years period. In this experiment,
membrane damage was assessed by measuring the plasma membrane permeability after storage
in individual embryonic axis using a method that is based on the differential permeability of the
plasma membrane to a nitroxide spin probe and a polar broadening agent. The amphipatic spin
probe freely crosses membranes and partitions between the hydrophobic environment of the
phospholipid bilayer and hydrophilic cytoplasm, whereas the broadening agent ferricyanide,
which cannot penetrate membranes, erases the cytosolic signal of the probes, leaving only the
lipid signal. Model membrane systems demonstrated that the ratio between lipid and cytosol
signals allowed for a quantitative assessment of membrane damage, with a high R value
indicative of a nonpermeable and therefore intact membrane. Figure 4A shows that in a

                                                                                                11
population of embryonic axes from seeds germinating at 100%, R values are distributed around
8, probably reflecting the variability of cell types [108]. Concomitant with time of storage and
decreased viability, the frequency distribution shifted towards lower values, indicative of
increased membrane damage. However, in the dead seeds that were stored for another 9 years,
the frequency distribution continued to shift to lower values, suggesting that membrane damage
had continued to progress (Figure 4A).
         A frequency distribution was also found in histone/DNA ratios of wheat at two storage
intervals [109]. A decrease in the ratio was found in the seed lot that was retrieved after 6 years
compared to 4 years of storage (Figure 4B). These authors attributed the change in ratio to
histones, since nuclei were arrested in the pre-synthetic phase of the mitotic cycle (G1).
However, the distribution of the ratio from both seed batches overlapped, despite the contrasting
germination capacity of the seed lots (Figure 3B). It is worth investigating whether changes in
this frequency distribution of damage could be described quantitatively using the population-
based threshold model [110] that considers the diversity of individuals that have a determinate
shelf-life during storage.

Lipid peroxidation
Our current view is that seed deterioration during storage is a result of a stochastic oxidation
cascade that produces peroxides, hydroperoxides, carbonyl and nitrosyl groups that further react
with neighboring molecules leading to further oxidation, fragmentation and formation of
adducts. Propanal, butanal, and hexanal, which are downstream products of lipid peroxidation
have been shown to be progressively produced in legume seeds during storage [111, 112]. In
pea seeds, the amount of propanal and hexanal first increased up to 12 days of storage at 50°C
prior to the loss of viability, then decreased by half after 25 days while viability decreased to
80% [112]. This observation gives rise to two possible interpretations: 1) complex oxidative
reactions within lipids leading to accumulation of alkoxy radicals do not lead to death or 2) only
a limited number of critical cells need to be damaged by lipid oxidation to cause loss of viability.
In contrast, in Lathyrus pratensis, no apparent significant change in hexanal content was found
whereas half of the seed population died during storage. Upon further aging, when 90% of the
seeds had died, hexanal content increased 20-fold. Similar observations were made in seeds of
lettuce, argulla and caraway [111, 113]. No obvious correlations were found between rate of
ageing and apparent increase in byproducts of lipid peroxidation. Furthermore, seed ageing was
not affected in a mixture of seeds emitting different amounts of these compounds, suggesting
these compounds are a consequence of deterioration rather than the cause [113].
        The role of lipid oxidation in seed deterioration has gained momentum from detailed
analyses of the changes of the lipidome during long term dry storage of a range of genotypes
of oilseed rape wheat, barley and Arabidopsis [3, 39, 40]. These authors documented complex
changes in the lipidome and oxylipidome after storage, involving lipids both from oil reserves
and membranes and originating both from enzyme activities and non-enzymatic degradation.
In general, during long-term storage, hundreds of singly and multiply oxidized lipids of various
classes increased with decreased germination after storage, following first- or second-order
kinetics, reinforcing the role of oxidation and oxygen as a key factor governing seed longevity
[40, 42, 114]. However, [42] pointed out that some of these oxidized lipids were already highly
abundant before the onset of loss of viability, making it impossible to untangle a cause-effect
relationship with the primary events triggering loss of viability. Such relationship might also be
obscured by the impact that these degraded lipids might have during seed imbibition and
germination. Mobilization of storage oil such as triacylglycerols appears to be more important
for seedling establishment than germination [115]. Therefore, the level of oxidized
triacylglycerols might not be that critical for seed germination after storage. In contrast, subtle
changes in membrane phospholipid composition during storage might strongly influence its
physico-chemical properties upon rehydration, leading for example to increased susceptibility
                                                                                                 12
to imbibitional injury and cell death [116-118]. Further work is therefore needed to understand
what triggers the pathways leading to lipid oxidation and hydrolysis and how it affects seed
germination after ageing.

RNA integrity
Seed ageing during dry storage is also associated with RNA degradation, for which a complex
picture starts to emerge. Correlation of ribosomal RNA integrity with storage time or storage
temperature was significant for a range of species over a 50 years period, making it an
interesting tool to quickly assess seed ageing [119]. However, like for lipid degradation and
membrane damage, loss of rRNA integrity occurred both before and during the loss of
germination, and there was not a single threshold RIN value for all species for either high
viability or death. Likewise, RIN values could not distinguish germinable from nongerminable
soybean seeds [119]. Perhaps, the translational machinery is very well protected in dry seeds
and very low RIN values are symptoms of post-mortem decay.
Dry seeds store large amount of mRNAs as they will rely on them to synthesize proteins early
during imbibition to ensure germination [120-122]. Their number range from ca. 12.000 in
Arabidopsis to > 17.000 in rice, and include gene products from all different categories, but
with an over-representation of genes involved in metabolism and protein synthesis (ribosomal
proteins, translation initiation and elongation factors). To further understand the fate of RNA
in dry seeds, Flemming et al. [119] assessed the degradation of stored mRNA in soybean seeds
during dry storage using Minion technology that allows for sequencing both full-length and
fragmented transcripts. They compared cohorts of seeds that were 2 years old and germinated
to 100% with those that had been stored for 23 years and for which the viability declined from
61% to 80% one year prior to RNA sequencing. Four patterns of degradation were detected
during long-term storage. Two categories consisted of transcripts of various lengths that
degraded progressively over the 23 years storage, whereas a third category was represented by
transcripts of different lengths that were equally degraded before and after storage, suggesting
a fraction of degradation occurred during drying or soon after drying [119]. The fourth category
was constituted by the shortest transcripts that remained intact even when seeds died.
Interestingly, these transcripts might offer internal standards to normalize changes in the
transcript population during seed ageing. Thus, this work offers a rich and unique set of markers
that will be helpful to understand the link between seed ageing, particularly during the so-called
asymptomatic phase when storage does not yet induce the loss of viability. In parallel, we need
to further explore the translational efficiency of these degraded transcripts during seed
imbibition, which will be a challenging task. In healthy Arabidopsis seeds, the degree and
dynamics of translational regulation exhibited two temporal phases associated with seed
rehydration and preparation for germination during imbibition [122]. This regulation was
associated with unique mRNA features such as transcript length, secondary structure and
specific motifs. It is noteworthy that in stored soybean seeds, the transcripts that remained intact
during seed ageing were related to ribosomal functions, highlighting the importance of the
translational machinery during germination [120, 121]. Yet the translational machinery remains
sensitive to seed ageing [120], which might have causes unrelated to RNA. Dry seeds
Arabidopsis knockout lines for the elongation factor eIF(iso)4G1 and eIF(iso)4G2 showed a
severe reduced longevity when stored over one year in the laboratory [123]. This suggests that
transcription elongating factors contribute to the survival in the dry state. Currently, the cause
of transcript fragmentation and how it leads to loss of seed germination is unknown. Oxidation
is strongly suspected but the chemistry remains to be elucidated. In sunflower, oxidized mRNA
in the form of 8-hydroxyguanosine was found to increase by 50% during a storage period that
did not kill the seeds but instead released dormancy [124]. This oxidation was not detected
using total RNA, reinforcing the idea a differential sensitivity to ageing according to the RNA
species. In contrast, in soybean there was no correlation between the amount of 8’OH mRNA
                                                                                                 13
and with loss of viability [119]. It should be noted that like RNA, DNA integrity is also modified
during storage, especially during oxidative and UV stress exposure. These damages must be
overcome during germination to avoid the mutation to be spread to the offspring [15, 125]. To
do so, many DNA repair proteins have been found to play a role in seed longevity in the past
decades, and has been extensively reviewed by Waterworth and colleagues [15].
        Being able to discriminate non-destructively in a cohort of stored seeds those that are
alive and already dead would constitute a powerful tool to further understand how deterioration
during aging leads to seed death during storage. The recent developments in imaging analysis,
machine learning and statistics look promising developments. Multivariate analysis of spectral
signatures of single cowpea seeds extracted from multispectral images after storage was able to
classify seed cohorts with reasonable accuracy (68 to 98%) into different categories as aged vs.
non aged, viable vs. nonviable and fast vs slowly germinating seeds [126]. However, the ageing
conditions that were used were 98% RH and 45°C. These conditions are not only far away from
those found in seed bank storage conditions but are permissive for respiration and protein
denaturation. It remains to be assessed whether a similar predictability can be obtained from
seeds stored in more appropriate storage conditions, close to the glassy state, where metabolism
has stopped and molecular mobility is severely restricted.

Concluding remarks

Here, we have synthetized our knowledge on the environmental effects and signaling pathway
regulating seed longevity. The main conclusion is that longevity is a highly plastic seed trait
that has served species to be adapted to their environment. This is clearly established in wild
species. Yet, there is a dearth of studies aiming at identifying the molecular factors that
modulate longevity in response to the environment. Such knowledge would be pivotal to further
understand its role as an adaptive trait. Both the mother plant and the zygote are able to regulate
longevity during seed development in response to the environment, probably by interfering with
ABA and auxin signaling pathways as well as metabolic regulation. Next to efforts in
unravelling the molecular factors contributing to seed lifespan, the future challenge is now to
understand the gene x environment interactions that govern the acquisition of longevity in
crops. For this, there is a need to better describe in the environmental conditions used to grow
mother plants and reconcile the storage protocols used to determine seed longevity and bring
them to consensus.

Abbreviations
ABA, abscisic acid; ATP, adenosine triphosphate; BR, brassinosteroids; DAP, days after
pollination; DCMU, 3-(3,4-dichlorophenyl)-1,1-dimethylurea; DNA, GA, gibberellic acid;
EMB, Embryogenesis; GWAS, Genome-Wide Association Study; HSP, Heat Shock Protein;
LEA, Late Embryogenesis Abundant; NADPH, Nicotinamide Adenine Dinucleotide
Phosphate; QTL, Quantitaive Trait Locus; RH, relative humidity; RFO, Raffinose Family
Oligosaccharides; sHSP, Small Heat Shock Protein; TT, transparent testa; WC, Water content.

Funding
OL and JB are supported from an ANR grant REGULEG (ANR-15-CE20-0001) and the
“Objectif Végétal” project funded by the Région-des-Pays-de-la-Loire. JZ is funded by grants
BIO2017-90877-REDT and BIO2017-84041-P from the Spanish Ministry of Science,
Innovation, and Universities (MCIU), State Research Agency (AEI) and European Regional
Development Fund (FEDER) and grant 2017SGR-710 from Generalitat de Catalunya to MRC.
We also thank the financial support of the MINECO Severo Ochoa Programme for Centres of
Excellence in R&D 2016-2019 (SEV‐2015‐0533) and the Generalitat de Catalunya CERCA

                                                                                                14
Programme to CRAG. JZ is supported by a CRAG Severo Ochoa International Postdoc
fellowship.

Author contribution
JZ, OL and JB conceived and wrote the manuscript.

Competing interest
The Authors declare that there are no competing interests associated with the manuscript.

Figure Legends

Figure 1. Seed longevity in a nutshell.
Summary of the intrinsic and external factors contributing to seed lifespan during storage.
ATM, ATAXIA TELANGIECTASIA MUTATED, HSP, heat shock protein, LEA late
embryogenesis abundant protein, LIG6, LIGASE 6, MSR: methylsulfoxide reductase, PIMT,
protein-L-isoaspartate O-methyltransferase, PM25: a LEA protein from the SMP family. RFO:
Raffinose Family oligosaccharides. Pictograms corresponding to know environmental effects
on longevity are respectively (light, water availability, temperature, nutrients). ABA, abscisic
acid; AIA, auxins; DOG1, DELAY OF GERMINATION1; GA, gibberellins.

Figure 2: Information necessary for an unambiguous interpretation and reproduction of
an ageing experiment to test longevity.
Knowledge of any drying treatment after harvest or pre-equilibration treatments before storing
the seed lots is necessary to determine whether longevity is assessed during water sorption (i.e.
mature seeds are imbibed to higher water content then stored) or desorption (i.e. mature seeds
are dried to lower water content then stored [127]). Wet ageing corresponds to storage
conditions above 80% relative humidity (RH) whereas dry storage corresponds to RH below
75% RH.

Figure 3. Desiccation tolerance and longevity specific gene coexpression modules
identified by a gene regulatory network analysis during seed development of M.
truncatula.
Nodes represent genes that are either upregulated during maturation in parallel to the acquisition
of desiccation tolerance (DT, green) or with the acquisition of longevity, measured as the P50,
the time for 50 percent of the seed population to die during storage (blue). Within each module,
large squares represent transcription factors and circles, other hub genes. The upper network
represents genes that are confirmed targets of MtABI3 whereas the bottom network corresponds
to genes that are not regulated by MtABI3. Data are from Righetti et al. (2015) [12] and
reproduced with permission ([128]. Copyright 2018, Taylor and Francis). EMB,
embryogenesis. WC, water content.

Figure 4. Distribution of the progress of damage symptoms during wheat seed ageing in
dry storage.
A) Distribution of membrane permeability values (R) calculated by electron paramagnetic
resonance (EPR) spectroscopy in individual embryonic axes within seed lots retrieved at
indicated time of storage (Y, years) and exhibiting different percentage of germination (G). For
each individual axis, the R value represents the ratio between the EPR signals of a guest probe
present respectively in the hydrophilic and the hydrophobic compartments of the cell in the
presence of a broadening agent. The broadening agent which is water soluble cancels the signal
emitted by the probe molecules that are localized in the hydrophilic compartment. Since this
agent does not permeate through intact membranes, high R value indicates a partitioning of the
                                                                                               15
EPR probe between the cytosol and lipids whereas a low (or nil) R value is due the loss of the
membrane integrity, allowing the broadening agent to exert its cancelling effect on the EPR
hydrophilic signal. Data are from [109]. B) Distribution of histone/DNA ratios measured by
histochemistry on the individual 2C nuclei from the root meristem of squashed individual axis
from seed lots stored for 4 and 6 years (Y). DNA content was assessed by microscope
photometry at 535 nm after staining with the Feulgen method. The same nuclei were thereafter
stained with fast-green FCF and scanned for histones at 635 nm. Data are from [109].

References

1. Hort, A. (1900) Theophrastus. Enquiry into plants. Harvard University Press.
2. Walters, C., Wheeler, L.M. and Grotenhuis, J.M. (2005) Longevity of seeds stored in a
genebank: species characteristics. Seed Sci. Res. 15, 1-20
3. Nagel, M. and Börner, A. (2010) The longevity of crop seeds stored under ambient
conditions. Seed Sci. Res. 20, 1-12
4. Colville, L. and Pritchard, H.W. (2019) Seed life span and food security. New Phytol. 224,
557-562
5. Li, D.Z. and Pritchard, H.W. (2009). The science and economics of ex situ plant conservation.
Trends Plant Sci. 14, 614-621.
6. Mondoni, A., Orsenigo, S., Dona, M., Balestrazzi, A., Probert, R.J., Hay, F.R., Petraglia, A.,
Abeli, T. (2014). Environmentally induced transgenerational changes in seed longevity:
maternal and genetic influence. Ann. Bot. 113, 1257-1263
7. Long, R.L., Gorecki, M.J., Renton, M., Scott, J.K., Colville, L., Goggin, D.E., Commander,
L. E., Westcott, D. A., Cherry, H., Finch-Savage, W. E. (2015). The ecophysiology of seed
persistence: a mechanistic view of the journey to germination or demise. Biological Rev. 90,
31-59
8. Merritt, D.J., Martyn, A.J., Ainsley, P., Young, RE, Seed, L.U., Thorpe, M., Hay, F. R.,
Commander, L. E., Shackelford, N., Offord, C. A., Dixon, K. W., Probert, R. J. (2014). A
continental-scale study of seed lifespan in experimental storage examining seed, plant, and
environmental traits associated with longevity Biodivers. Conserv. 23, 1081–1104
9. Leprince, O., Pellizzaro, A., Berriri, S., Buitink, J. (2017) Late seed maturation: drying
without dying. J. Exp. Bot. 68, 827-841
10. Priestley, D.A. (1986) Seed aging: implications for seed storage and persistence in the soil.
Comstock Associates. 304pp
11. Chatelain, E., Hundertmark, M., Leprince, O., Gall, S. L., Satour, P., Deligny-Pennick,
Rogniaux, H., Buitink, J. (2012). Temporal profiling of the heat‐stable proteome during late
maturation of Medicago truncatula seeds identifies a restricted subset of late embryogenesis
abundant proteins associated with longevity. Plant Cell Environ. 35, 1440-1455.
12. Righetti, K., Vu, J. L., Pelletier, S., Vu, B. L., Glaab, E., Lalanne, D., Pasha, A., Patel, R.
V., Provart, N. J., Verdier, J., Leprince, O., Buitink, J. (2015). Inference of longevity related
genes from a robust coexpression network of seed maturation identifies regulators linking seed
storability to biotic defense-related pathways. Plant Cell 27, 2692-2708.
13. Probert, R., Adams, J., Coneybeer, J., Crawford, A., Hay, F. (2007). Seed quality for
conservation is critically affected by pre-storage factors. Aust. J. Bot. 55, 326-335.
14. Sano, N., Rajjou, L., North, H. M., Debeaujon, I., Marion-Poll, A., Seo, M. (2015). Staying
alive: molecular aspects of seed longevity. Plant Cell Phys. 57, 660-674.
15. West, C. E., Waterworth, W. M., Bray, C. M. (2019). Seeds and the art of genome
maintenance. Front. Plant Sci. 10, 706.
16. Nguyen, T.P., Keizer, P., van Eeuwijk, F., Smeekens, S., Bentsink L. (2012) Natural
variation for seed longevity and seed dormancy are negatively correlated in Arabidopsis. Plant
Physiol. 160, 2083–2092
                                                                                                16
17. Nagel, M., Behrens, A.K. and Börner, A. (2013) Effects of Rht dwarfing alleles on wheat
seed vigour after controlled deterioration. Crop Pasture Sci. 64, 857-864
18. Börner, A., Nagel, M., Agacka-Mołdoch, M., Gierke, P. U., Oberforster, M., Albrecht, T.,
Mohler, V. (2018). QTL analysis of falling number and seed longevity in wheat (Triticum
aestivum L.). J. Appl. Genet. 59, 35-42.
19. Dargahi, H., Tanya, P., Srinives, P. (2014). Mapping of the genomic regions controlling
seed storability in soybean (Glycine max L.). J.Genet. 93, 365-370.
20. Zhang, X., Hina, A., Song, S., Kong, J., Bhat, J. A., Zhao, T. (2019). Whole-genome
mapping identified novel “QTL hotspots regions” for seed storability in soybean (Glycine max
L.). BMC Genomics 20, 499.
21. Nagel, M., Rosenhauer, M., Willner, E., Snowdon, R. J., Friedt, W., Börner, A. (2011).
Seed longevity in oilseed rape (Brassica napus L.)–genetic variation and QTL mapping. Plant
Genet. Resour. 9, 260-263.
22. Schatzki, J., Allam, M., Klöppel, C., Nagel, M., Börner, A., Möllers, C. (2013). Genetic
variation for secondary seed dormancy and seed longevity in a set of black‐seeded European
winter oilseed rape cultivars. Plant Breed. 132, 174-179.
23. Wang, T., Hou, L., Jian, H., Di, F., Li, J., Liu, L. (2018). Combined QTL mapping,
physiological and transcriptomic analyses to identify candidate genes involved in Brassica
napus seed aging. Mol. Genet. Genomics 293, 1421-1435.
24. Liu, Y., Zhang, H., Li, X., Wang, F., Lyle, D., Sun, L., Wang, G., Wang, J., Gu, R. (2019).
Quantitative trait locus mapping for seed artificial aging traits using an F2: 3 population and a
recombinant inbred line population crossed from two highly related maize inbreds. Plant Breed.
138, 29-37.
25. Miura, K., Lin, S., Yano, M., Nagamine, T. (2002). Mapping quantitative trait loci
controlling seed longevity in rice (Oryza sativa L.). Theor. Appl. Genet. 104, 981-986.
26. Li, L., Lin, Q., Liu, S., Liu, X., Wang, W., Hang, N. T., Liu, F., Zhao, Z., Jiang, L., Wan, J.
(2012). Identification of quantitative trait loci for seed storability in rice (Oryza sativa L.). Plant
Breed. 131, 739-743.
27. Liu, W., Pan, X., Li, Y., Duan, Y., Min, J., Liu, S., Sheng, X., Li, X. (2018). Detection and
validation of QTL s associated with seed longevity in rice (Oryza sativa L.). Plant Breed. 137,
546-552.
28. Wozny, D., Kramer, K., Finkemeier, I., Acosta, I. F., Koornneef, M. (2018). Genes for seed
longevity in barley identified by genomic analysis on near isogenic lines. Plant Cell Environ.
41, 1895-1911.
29. Agacka-Mołdoch, M., Nagel, M., Doroszewska, T., Lewis, R. S., Börner, A. (2015).
Mapping quantitative trait loci determining seed longevity in tobacco (Nicotiana tabacum L.).
Euphytica 202, 479-486.
30. Schwember, A. R., Bradford, K. J. (2010). Quantitative trait loci associated with longevity
of lettuce seeds under conventional and controlled deterioration storage conditions. J. Exp. Bot.
61, 4423-4436.
31. Hay, F. R., Valdez, R., Lee, J. S., Sta. Cruz, P. C. (2018). Seed longevity phenotyping:
recommendations on research methodology. J. Exp. Bot. 70, 425-434.
32. Roach, T., Nagel, M., Börner, A., Eberle, C., Kranner, I. (2018). Changes in tocochromanols
and glutathione reveal differences in the mechanisms of seed ageing under seedbank conditions
and controlled deterioration in barley. Environ. Exp. Bot. 156, 8-15.
33. Ballesteros, D., Walters, C. (2011). Detailed characterization of mechanical properties and
molecular mobility within dry seed glasses: relevance to the physiology of dry biological
systems. Plant J. 68, 607-619.
34. Nagel, M., Kranner, I., Neumann, K., Rolletschek, H., Seal, C. E., Colville, L., Fernandez-
Marin, B., Börner, A. (2015). Genome‐wide association mapping and biochemical markers

                                                                                                    17
reveal that seed ageing and longevity are intricately affected by genetic background and
developmental and environmental conditions in barley. Plant
Cell Environ. 38, 1011-1022.
35. Groot, S.P.C., Surki, A.A., de Vos, R.C.H., Kodde, J. (2012). Seed storage at elevated
partial pressure of oxygen, a fast method for analysing seed ageing under dry conditions. Ann.
Bot. 110, 1149-1159.
36. Nagel, M., Kodde, J., Pistrick, S., Mascher, M., Börner, A., Groot, S. P. (2016). Barley seed
aging: genetics behind the dry elevated pressure of oxygen aging and moist controlled
deterioration. Front. Plant Sci. 7, 388.
37. Murakami, M. (2018). Water makes glass elastically stiffer under high-pressure. Sci. Rep.
8, 11890.
38. Huang, Y., Wu, Z., Wang, Y., Li, F. (2015). Examination of the effects of temperature and
pressure on lipoxygenase activities in pork using response surface methodology. Food Sci.
Biotechnol. 24, 1257-1263.
39. Boca, S., Koestler, F., Ksas, B., Chevalier, A., Leymarie, J., Fekete, A., Mueller, M. J,
Havaux, M. (2014). Arabidopsis lipocalins AtCHL and AtTIL have distinct but overlapping
functions essential for lipid protection and seed longevity. Plant Cell Environ. 37, 368-381.
40. Wiebach, J., Nagel, M., Börner, A., Altmann, T., Riewe, D. (2019). Age‐dependent loss of
seed viability is associated with increased lipid oxidation and hydrolysis. Plant Cell Environ.
1-12
41. Ballesteros, D., Walters, C. (2019). Solid-state biology and seed longevity: a mechanical
analysis of glasses in pea and soybean embryonic axes. Front. Plant Sci. 10, 920.
42. Pereira Neto L. G., Pereira Sartori, M. M., Toorop, P. E., da Silva, E. A. A. (2019) Seed
longevity differs in Astronium fraxinifolium Scott from two geographical regions in Brazil.
Revist. Bras.Cien. Agr. 14, e6776.
43. Buitink, J., Leprince, O., Hemminga, M. A., Hoekstra, F. A. (2000). Molecular mobility in
the cytoplasm: an approach to describe and predict lifespan of dry germplasm. Proc. Nat. Acad.
Sci. 97, 2385-2390.
44. Probert, R. J., Daws, M. I., Hay, F. R. (2009). Ecological correlates of ex situ seed longevity:
a comparative study on 195 species. Ann. Bot. 104, 57-69.
45. Tausch, S., Leipold, M., Reisch, C. and Poschlod Peter (2019) Dormancy and endosperm
presence influence the ex situ conservation potential in central European calcareous grassland
plants. AoB PLANTS 11, 1-12
46. Floris, C. and Anguillesi, M.C. (1974) Ageing of isolated embryos and endosperms of
durum wheat: an analysis of chromosome damage. Mutation Res. 22, 133-138.
47. Galland, M., He, D., Lounifi, I., Arc, E., Clément, G., Balzergue, S. (2017) An integrated
“Multi-Omics” comparison of embryo and endosperm tissue-specific features and their impact
on rice seed quality. Front. Plant Sci. 8, 1984
48. Debeaujon, I., Léon-Kloosterziel, K. M., Koornneef, M. (2000). Influence of the testa on
seed dormancy, germination, and longevity in Arabidopsis. Plant Physiol. 122, 403-414.
49. De Giorgi J Piskurewicz U Loubery S Utz-Pugin A Bailly C, Mène-Saffrané L Lopez-
Molina L. (2015). An endosperm-associated cuticle is required for Arabidopsis seed viability,
dormancy and early control of germination. PLOS Genet. 11, 1-32
50. Loubéry, S., De Giorgi, J., Utz-Pugin, A., Demonsais, L., Lopez-Molina, L. (2018). A
maternally deposited endosperm cuticle contributes to the physiological defects of transparent
testa Seeds. Plant Physiol. 177, 1218-1233.
51. Coen, O., Lu, J., Xu, W., De Vos, D., Péchoux, C., Domergue, F., Grain, D., Lepiniec, L.,
Magnani, E. (2019). Deposition of a cutin apoplastic barrier separating seed maternal and
zygotic tissues. BMC Plant Biol. 19, 304.


                                                                                                 18
52. Fedi, F., O’Neill, C. M., Menard, G., Trick, M., Dechirico, S., Corbineau, F., Bailly, C.,
Eastmond, P. J., Penfield, S. (2017). Awake1, an ABC-type transporter, reveals an essential
role for suberin in the control of seed dormancy. Plant Physiol. 174, 276-283.
53. Vishwanath, S. J., Kosma, D. K., Pulsifer, I. P., Scandola, S., Pascal, S., Joubès, J., Dittrich-
Domergue, F., Lessire, R., Rowland, O., Domergue, F. (2013). Suberin associated fatty alcohols
in Arabidopsis: distributions in roots and contributions to seed coat barrier properties. Plant
Physiol. 163, 1118-1132.
54. Lepiniec, L., Devic, M., Roscoe, T. J., Bouyer, D., Zhou, D. X., Boulard, C., Baud, S.,
Dubreucq, B. (2018). Molecular and epigenetic regulations and functions of the LAFL
transcriptional regulators that control seed development. Plant Reprod. 31, 291-307.
55. Sugliani, M., Rajjou, L., Clerkx, E. J., Koornneef, M., Soppe, W. J. (2009). Natural
modifiers of seed longevity in the Arabidopsis mutants abscisic acid insensitive3‐5 (abi3‐ 5)
and leafy cotyledon1‐3 (lec1‐3). New Phytol. 184, 898-908.
56. Kotak, S., Vierling, E., Bäumlein, H., von Koskull-Döring, P. (2007). A novel
transcriptional cascade regulating expression of heat stress proteins during seed development
of Arabidopsis. Plant Cell 19, 182-195.
57. Prieto-Dapena, P., Castaño, R., Almoguera, C., Jordano, J. (2006). Improved resistance to
controlled deterioration in transgenic seeds. Plant Physiol. 142, 1102-1112.
58. Carranco, R., Espinosa, J. M., Prieto-Dapena, P., Almoguera, C., Jordano, J. (2010).
Repression by an auxin/indole acetic acid protein connects auxin signaling with heat shock
factor-mediated seed longevity. Proc. Nat. Acad. Sci. 107, 21908-21913.
59. Verdier, J., Lalanne, D., Pelletier, S., Torres-Jerez, I., Righetti, K., Bandyopadhyay, K.,
Leprince, O., Chatelain, O., Ly Vu, B., Gouzy, J., Gamas, P., Udvardi, M. K., Buitink, J. (2013).
A regulatory network-based approach dissects late maturation processes related to the
acquisition of desiccation tolerance and longevity of Medicago truncatula seeds. Plant Physiol.
163, 757-774.
60. Pereira, J. L., Buitink, J., Lalanne, D., Rossi, R. F., Pelletier, S., Leprince, O. (2017).
Molecular characterization of the acquisition of longevity during seed maturation in soybean.
PloS one 12, e0180282.
61. Delmas, F., Sankaranarayanan, S., Deb, S., Widdup, E., Bournonville, C., Bollier, N.,
Samuel, M. A. (2013). ABI3 controls embryo degreening through Mendel's I locus. 110,
Proc.Nat. Acad. Sci. 110, e3888-e3894
62. Hundertmark, M., Buitink, J., Leprince, O., Hincha, D. K. (2011). The reduction of seed
specific dehydrins reduces seed longevity in Arabidopsis thaliana. Seed Sci. Res. 21, 165- 173.
63. Sugliani, M., Brambilla, V., Clerkx, E. J., Koornneef, M., Soppe, W. J. (2010). The
conserved splicing factor SUA controls alternative splicing of the developmental regulator
ABI3 in Arabidopsis. Plant Cell 22, 1936-1946.
64. Srinivasan, A., Jiménez‐Gómez, J. M., Fornara, F., Soppe, W. J., Brambilla, V. (2016).
Alternative splicing enhances transcriptome complexity in desiccating seeds. J. Integr. Plant
Biol. 58, 947-958.
65. Zinsmeister, J., Lalanne, D., Terrasson, E., Chatelain, E., Vandecasteele, C., Vu, B. L.,
Dubois-Laurent, C., Geoffriau, E., Le Signor, C., Dalmais, M., Gutbrod, K., Dörmann, P.,
Gallardo, K., Bendahmane, A., Buitnk, J., Leprince, O. (2016). ABI5 is a regulator of seed
maturation and longevity in legumes. Plant Cell 28, 2735-2754.
66. Reeves, W. M., Lynch, T. J., Mobin, R., Finkelstein, R. R. (2011). Direct targets of the
transcription factors ABA-Insensitive (ABI) 4 and ABI5 reveal synergistic action by ABI4 and
several bZIP ABA response factors. Plant Mol. Biol. 75, 347-363.
67. Dekkers, B. J., He, H., Hanson, J., Willems, L. A., Jamar, D. C., Cueff, G., Rajjou, L.,
Hilhorst, H. W. M., Bentsink, L. (2016). The Arabidopsis DELAY OF GERMINATION 1 gene
affects ABSCISIC ACID INSENSITIVE 5 (ABI 5) expression and genetically interacts with
ABI 3 during Arabidopsis seed development. Plant J. 85, 451-465
                                                                                                  19
68. Nakabayashi K, Okamoto M, Koshiba T, Kamiya Y, Nambara E (2005) Genome-wide
profiling of stored mRNA in Arabidopsis thaliana seed germination: epigenetic and genetic
regulation of transcription in seed: Molecular profiling in Arabidopsis seed. Plant J. 41, 697–
709.
69. Nakajima, S., Ito, H., Tanaka, R., Tanaka, A. (2012). Chlorophyll b reductase plays an
essential role in maturation and storability of Arabidopsis seeds. Plant Physiol. 160, 261-273.
70. Teixeira, R. N., Ligterink, W., França-Neto, J. D. B., Hilhorst, H. W., da Silva, E. A. (2016).
Gene expression profiling of the green seed problem in soybean. BMC Plant Biol. 16, 37.
71. Lopez‐Molina, L., Mongrand, S., McLachlin, D. T., Chait, B. T., Chua, N. H. (2002). ABI5
acts downstream of ABI3 to execute an ABA‐dependent growth arrest during germination.
Plant J. 32, 317-328.
72. Terrasson, E., Buitink, J., Righetti, K., Ly Vu, B., Pelletier, S., Zinsmeister, J., Lalanne, D.,
Leprince, O. (2013). An emerging picture of the seed desiccome: confirmed regulators and
newcomers identified using transcriptome comparison. Front. Plant Sci. 4, 497.
73. Bentsink, L., Jowett, J., Hanhart, C. J., Koornneef, M. (2006). Cloning of DOG1, a
quantitative trait locus controlling seed dormancy in Arabidopsis. Proc. Nat. Acad. Sci. 103,
17042-17047.
74. Nishimura, N., Tsuchiya, W., Moresco, J. J., Hayashi, Y., Satoh, K., Kaiwa, Irisa, T.,
Kinoshita, T., Schroeder, J. I., Yates III, J. R., N., Hirayama, T., Yamazaki, T. (2018). Control
of seed dormancy and germination by DOG1-AHG1 PP2C phosphatase complex via binding
to heme. Nat. Comm. 9(1), 2132.
75. Clerkx, E. J., Blankestijn‐De Vries, H., Ruys, G. J., Groot, S. P., Koornneef, M. (2004).
Genetic differences in seed longevity of various Arabidopsis mutants. Physiol. Plant. 121, 448-
461.
76. Schausberger C, Roach T, Stöggl W, Arc E, Finch-Savage WE, Kranner I (2019) Abscisic
acid-determined seed vigour differences do not influence redox regulation during ageing.
Biochem. J. 476, 965–974.
77. Sano N., Kim J.-S., Onda Y., Nomura T., Mochida K., Okamoto M. et al. (2017) RNASeq
using bulked recombinant inbred line populations uncovers the importance of brassinosteroid
for seed longevity after priming treatments. Sci. Rep. 7, 8075
78. He, H., de Souza Vidigal, D., Snoek, L. B., Schnabel, S., Nijveen, H., Hilhorst, H., Bentsink,
L. (2014). Interaction between parental environment and genotype affects plant and seed
performance in Arabidopsis. J. Exp. Bot. 65, 6603-6615.
79. Chauffour F, Bailly M, Perreau F, Cueff G, Suzuki H, Collet B et al. (2019). Multi-omics
analysis reveals sequential roles for ABA during seed maturation. Plant Physiol. 180, 1198-
1218.
80. Pellizzaro, A., Neveu, M., Lalanne, D., Ly Vu, B., Kanno, Y., Seo, M., Leprince, O.,
Buitink, J. (2019). A role for auxin signaling in the acquisition of longevity during seed
maturation. New Phytol. 225: 284–296
81. Hull, A. K., Vij, R., Celenza, J. L. (2000). Arabidopsis cytochrome P450s that catalyze the
first step of tryptophan-dependent indole-3-acetic acid biosynthesis. Proc.Nat. Acad. Sci. 97,
2379-2384.
82. Zhao, Y., Hull, A. K., Gupta, N. R., Goss, K. A., Alonso, J., Ecker, J. R., Normanly, J.,
Chory, J., Celenza, J. L. (2002). Trp-dependent auxin biosynthesis in Arabidopsis: involvement
of cytochrome P450s CYP79B2 and CYP79B3. Genes Dev. 16, 3100-3112.
83. Bueso E, Muñoz-Bertomeu J, Campos F, Brunaud V, Martínez L, Sayas E, Ballester P,
Yenush L, Serrano R. 2014. ARABIDOPSIS THALIANA HOMEOBOX25 uncovers a role for
gibberellins in seed longevity. Plant Physiol. 164, 999–1010.
84. Ellis, R. H., Hong, T. D., Jackson, M. T. (1993). Seed production environment, time of
harvest, and the potential longevity of seeds of three cultivars of rice (Oryza sativa L.). Ann.
Bot. 72, 583-590.
                                                                                                  20
85. Zani, D., Müller, J. V. (2017). Climatic control of seed longevity of Silene during the post‐
zygotic phase: do seeds from warm, dry climates possess higher maturity and desiccation
tolerance than seeds from cold, wet climates?. Ecol. Res. 32, 983-994.
86. Kochanek, J., Buckley, Y. M., Probert, R. J., Adkins, S. W., Steadman, K. J. (2010).
Prezygotic parental environment modulates seed longevity. Austral Ecol. 35, 837-848.
87. Mondoni, A., Probert, R. J., Rossi, G., Vegini, E., Hay, F. R. (2010). Seeds of alpine plants
are short lived: implications for long-term conservation. Ann. Bot. 107, 171-179.
88. Bernareggi, G., Carbognani, M., Petraglia, A., Mondoni, A. (2015). Climate warming could
increase seed longevity of alpine snowbed plants. Alp. Bot. 125, 69-78.
89. Mondoni, A., Pedrini, S., Bernareggi, G., Rossi, G., Abeli, T., Probert, R. J., Ghitti, M.,
Bonomi, C., Orsenigo, S. (2015). Climate warming could increase recruitment success in
glacier foreland plants. Ann. Bot. 116, 907-916.
90. MacGregor, D. R., Kendall, S. L., Florance, H., Fedi, F., Moore, K., Paszkiewicz, K.,
Smirnoff, N., Penfield, S. (2015). Seed production temperature regulation of primary dormancy
occurs through control of seed coat phenylpropanoid metabolism. New Phytol. 205, 642-652.
91. Penfield, S., MacGregor, D. R. (2017). Effects of environmental variation during seed
production on seed dormancy and germination. J. Exp. Bot. 68, 819-825.
92. Smolikova, G. N., Medvedev, S. S. (2016). Photosynthesis in the seeds of
chloroembryophytes. Russ. J. Plant Physiol. 63, 1-12.
93. Bettey, M., Sinniah, U. R., Finch-Savage, W. E., Ellis, R. H. (1998). Irrigation and seed
quality development in rapid-cycling Brassica: accumulation of stress proteins. Ann. Bot. 82,
657-663.
94. Sinniah, U. R., Ellis, R. H., John, P. (1998). Irrigation and seed quality development in
rapid-cycling Brassica: seed germination and longevity. Ann. Bot. 82, 309-314.
95. Ramamoorthy, K., Basu, R. N. (1996). Studies on the effect of moisture stress at different
growth phases on seed vigour, viability and storability in peanut (Arachis hypogaea L.). J.
Agron. Crop Sci. 177, 33-37.
96. Ellis, R. H., Yadav, G. (2016). Effect of simulated rainfall during wheat seed development
and maturation on subsequent seed longevity is reversible. Seed Sci.Res. 26, 67-76.
97. Allorent, G., Osorio, S., Ly Vu, J., Falconet, D., Jouhet, J., Kuntz, M., Fernie, A. R, Lerbs-
Mache, S., Macherel, D., Courtois, F. (2015). Adjustments of embryonic photosynthetic
activity modulate seed fitness in Arabidopsis thaliana. New Phytol. 205, 707-719.
98. Li, Z., Wu, S., Chen, J., Wang, X., Gao, J., Ren, G., Kuai, B. (2017). NYEs/SGRs mediated
chlorophyll degradation is critical for detoxification during seed maturation in Arabidopsis.
Plant J. 92, 650-661.
99. Smolikova, G., Kreslavski, V., Shiroglazova, O., Bilova, T., Sharova, E., Frolov, A.,
Medvedev, S. (2018). Photochemical activity changes accompanying the embryogenesis of pea
(Pisum sativum) with yellow and green cotyledons. Funct. Plant Biol. 45, 228-235.
100. Contreras, S., Bennett, M. A., Metzger, J. D., Tay, D. (2008). Maternal light environment
during seed development affects lettuce seed weight, germinability, and storability.
HortScience 43, 845-852.
101. Contreras, S., Bennett, M. A., Metzger, J. D., Tay, D., Nerson, H. (2009). Red to far-red
ratio during seed development affects lettuce seed germinability and longevity. HortScience 44,
130-134.
102. He, H., Willems, L. A., Batushansky, A., Fait, A., Hanson, J., Nijveen, H., Hilhorst, H. W.
M., Bentsink, L. (2016). Effects of parental temperature and nitrate on seed performance are
reflected by partly overlapping genetic and metabolic pathways. Plant Cell Physiol. 57, 473-
487.
103. Geshnizjani, N., Sarikhani Khorami, S., Willems, L. A., Snoek, B. L., Hilhorst, H. W.,
Ligterink, W. (2019). The interaction between genotype and maternal nutritional environments
affects tomato seed and seedling quality. J. Exp. Bot. 70, 2905-2918.
                                                                                               21
104. Brunel-Muguet, S., d'Hooghe, P., Bataillé, M. P., Larré, C., Kim, T. H., Trouverie, J.,
Avice, J. C., Etienne, P., Dürr, C. (2015). Heat stress during seed filling interferes with sulfur
restriction on grain composition and seed germination in oilseed rape (Brassica napus L.).
Front. Plant Sci. 6, 213.
105. Salem, M. A., Li, Y., Wiszniewski, A., Giavalisco, P. (2017). Regulatory‐associated
protein of TOR (RAPTOR) alters the hormonal and metabolic composition of Arabidopsis
seeds, controlling seed morphology, viability and germination potential. Plant J. 92, 525-545.
106. Rosnoblet, C., Aubry, C., Leprince, O., Vu, B. L., Rogniaux, H., Buitink, J. (2007). The
regulatory gamma subunit SNF4b of the sucrose non‐fermenting‐related kinase complex is
involved in longevity and stachyose accumulation during maturation of Medicago truncatula
seeds. Plant J. 51, 47-59.
107. Walters, C., Ballesteros, D., Vertucci, V. A. (2010). Structural mechanics of seed
deterioration: standing the test of time. Plant Sci. 179, 565-573.
108. Golovina, E. A., Wolkers, W. F., Hoekstra, F. A. (1997). Long-term stability of protein
secondary structure in dry seeds. Comp. Biochem. Physiol. A Physiol. 117, 343-348.
109. Innocenti, A. M., Bitonti, M. B., Bennici, A. (1983). Cytophotometric analyses and in vitro
culture test in the embryo first node of old Triticum durum caryopses. Caryologia 36, 83-87.
110. Bradford, K. J. (2018). Interpreting biological variation: seeds, populations and sensitivity
thresholds. Seed Sci. Res. 28, 158-167.
111. Mira, S., González-Benito, M. E., Hill, L. M., Walters, C. (2010). Characterization of
volatile production during storage of lettuce (Lactuca sativa) seed. J. Exp. Bot. 61, 3915-3924.
112. Colville, L., Bradley, E.L., Lloyd, A.S., Pritchard, H.W., Castle, L. and Kranner, I. (2012)
Volatile fingerprints of seeds of four species indicate the involvement of alcoholic fermentation,
lipid peroxidation and Maillard reactions in seed deterioration during ageing and desiccation
stress. J. Exp. Bot. 63, 6519-6530.
113. Mira, S., Hill, L.M., Gonzàlez-Benito, M.E., Ibáñ, M.A. and Walters, C. (2016) Volatile
emission in dry seeds as a way to probe chemical reactions during asymptomatic deterioration.
J. Exp. Bot. 67, 1783-1793.
114. Riewe, D., Wiebach, J., Altmann, T. (2017). Structure annotation and quantification of
wheat seed oxidized lipids by High-Resolution LC-MS/MS. Plant Physiol. 175, 600-618.
115. Kelly, A. A., Quettier, A. L., Shaw, E., Eastmond, P. J. (2011). Seed storage oil
mobilization is important but not essential for germination or seedling establishment in
Arabidopsis. Plant Physiol. 157, 866-875.
116. Hoekstra, F. A., Golovina, E. A., van Aelst, A. C., Hemminga, M. A. (1999) Imbibitional
leakage from anhydrobiotes revisited. Plant, Cell Environ. 22, 1121-1131.
117. Scherber, C. M., Schottel, J. L., Aksan, A. (2009). Membrane phase behavior of
Escherichia coli during desiccation, rehydration, and growth recovery. BBA Biomebranes 1788,
2427-2435.
118. Yu, X., Li, A.,Li, W. (2015). How membranes organize during seed germination: three
patterns of dynamic lipid remodelling define chilling resistance and affect plastid biogenesis.
Plant Cell Environ. 38, 1391-1403.
119. Fleming, M. B., Richards, C. M., Walters, C. (2017). Decline in RNA integrity of drystored
soybean seeds correlates with loss of germination potential. J. Exp. Bot. 68, 2219-2230.
120. Rajjou, L., Debeaujon, I. (2008). Seed longevity: survival and maintenance of high
germination ability of dry seeds. C. R. Biol. 331, 796-805.
121. Galland, M., Huguet, R., Arc, E., Cueff, G., Job, D., Rajjou, L. (2014). Dynamic
proteomics emphasizes the importance of selective mRNA translation and protein turnover
during Arabidopsis seed germination. Mol. Cell. Proteom. 13, 252-268.
122. Bai, B., Peviani, A., van der Horst, S., Gamm, M., Snel, B., Bentsink, L., Hanson, J. (2017).
Extensive translational regulation during seed germination revealed by polysomal profiling.
New Phytol. 214, 233-244.
                                                                                               22
123. Lellis, A.D., Allen, M.L., Aertker, A.W., Tran, J.K., Hillis, D.M., Harbin, C.R. et al. (2010)
Deletion of the eIFiso4G subunit of the Arabidopsis eIFiso4F translation initiation complex
impairs health and viability. Plant Mol.Biol., 74, 249-263
124. Bazin, J., Langlade, N., Vincourt, P., Arribat, S., Balzergue, S., El-Maarouf-Bouteau, H.,
Bailly, C. (2011). Targeted mRNA oxidation regulates sunflower seed dormancy alleviation
during dry after-ripening. Plant Cell 23, 2196-2208.
125. Waterworth, W. M., Footitt, S., Bray, C. M., Finch-Savage, W. E., West, C. E. (2016).
DNA damage checkpoint kinase ATM regulates germination and maintains genome stability in
seeds. Proc. Nati. Acad. Sci. 113, 9647-9652.
126. ElMasry, G., Mandour, N., Wagner, M. H., Demilly, D., Verdier, J., Belin, E., Rousseau,
D. (2019). Utilization of computer vision and multispectral imaging techniques for
classification of cowpea (Vigna unguiculata) seeds. Plant Methods 15, 24.
127. Hay, F.R. and Temple S. (2016) The longevity of desorbing and adsorbing rice seeds. Seed
Sci. Res. 26, 306–316
128. Buitink, J., Leprince, O. (2018). Letters to the twenty-first century botanist. Second series:
“What is a seed?–2. Regulation of desiccation tolerance and longevity in developing seeds: two
faces of the same coin. Botany Letters, 165, 181-185.




                                                                                                23
           Table 1. Summary of studies reporting on the effect of environment on seed longevity.

Environment Species                Aging method     Effect on seed longevity                             Reference
                                   75%RH, 35°C and Cold temperature during seed maturation
Temperature      Arabidopsis                                                                             12, 78
                                   85% RH, 40°C     negatively impacts seed longevity
                 Orysa japonica,                    In three rice cultivars, potential longevity is
Temperature                        15% RH, 40°C                                                          84
                 Oryza indica                       higher in cool compared to warm temperatures
                                                    Higher seed longevity for plants exposed to
                 Alpine snowbed
Temperature                        60% RH, 45°C     warmer temperatures compared to natural              88, 89
                 species
                                                    conditions
                                   60% RH, 45°C and Seeds from the field of lower nutrients supply
Nutrient       ordeum vulgare                                                                            34
                                   100% RH, 43°C    display both lower resistance to AA and CD tests
                                                    A long day treatment increases seed longevity
Light            Lactuca sativa    74% RH, 30°C                                                          100, 101
                                                    compared to short day treatment
                                                    Potential longevity is higher when optimal
Wet/dry          Brassica rapa     15% RH, 40°C                                                          93, 94
                                                    irrigation stops before the end of seed maturation
                 Arachis                            Drought stress during seed development
Wet/dry                            50-71% RH, 30°C                                                       95
                 hypogaea L.                        enhances seed longevity
                 Triticum                           Dry environment during seed development
Wet/dry                            15% RH, 40°C                                                          96
                 aestivum                           enhances seed longevity
                                                    Heat and drought during development decreases
Wet/dry          Glycine max                                                                             70
                                                    vigour and leads to stay green phenotype
                 Silene suecica
Wet/dry x        alpine and                           Seeds grown in dry and warm environment
                                   60% RH, 45°C                                                          85
temperature      subartic                             display higher seed longevity
                 populations




                                                                                                          24
Figure 1. Seed longevity in a nutshell.
Summary of the intrinsic and external factors contributing to seed lifespan during
storage. ATM, ATAXIA TELANGIECTASIA MUTATED, HSP, heat shock protein,
LEA late embryogenesis abundant protein, LIG6, LIGASE 6, MSR: methylsulfoxide
reductase, PIMT, protein-L-isoaspartate O-methyltransferase, PM25: a LEA protein
from the SMP family. RFO: Raffinose Family oligosaccharides. Pictograms
corresponding to know environmental effects on longevity are respectively light,
water availability, temperature, nutrients. ABA, abscisic acid;; AIA, auxins; DOG1,
DELAY OF GERMINATION1; GA, gibberellins.
Figure 2: Information necessary for an unambiguous interpretation and
reproduction of an ageing experiment to test longevity.
Knowledge of any drying treatment after harvest or pre-equilibration treatments
before storing the seed lots is necessary to determine whether longevity is assessed
during water sorption (i.e. mature seeds are imbibed to higher water content then
stored) or desorption (i.e. mature seeds are dried to lower water content then stored)
[130]). Wet ageing corresponds to storage conditions above 80% relative humidity
(RH) whereas dry storage correspond to RH below 75% RH.
                                DT gene              longevity gene
                                module                  module


Figure 3. Desiccation tolerance and longevity specific gene coexpression modules identified
by a gene regulatory network analysis during seed development of M. truncatula.
Nodes represent genes that are either upregulated during maturation in parallel to the acquisition
of desiccation tolerance (DT, green) or with the acquisition of longevity, measured as the P50, the
time for 50 percent of the seed population to die during storage (blue). Within each module, large
squares represent transcription factors and circles, other hub genes. The upper network represents
genes that are confirmed targets of MtABI3 whereas the bottom network corresponds to genes that
are not regulated by MtABI3. Data are from Righetti et al. (2015) [12] and reproduced with
permission ([128]. Copyright 2018, Taylor and Francis). EMB, embryogenesis. WC, water content.
            A                                                        B

                                                 1 Y, 100% G                60
            50
                                                 4 Y, 17% G                                                          4 Y, 100 % G
                                                                            50
                                                 9 Y, 0% G                                                           6 Y, 3% G
            40
                                                 18 Y, 0% G




                                                                Frequency
Frequency




                                                                            40
            30
                                                                            30
            20
                                                                            20

            10                                                              10

             0                                                               0
                 0   2    4    6     8    10    12   14                          0,0   0,2   0,4   0,6   0,8   1,0     1,2    1,4
             Membrane permeability (R value, arbitrary units)                                      Histone/DNA




            Figure 4. Distribution of the progress of damage symptoms during wheat seed ageing
            in dry storage.
            A) Distribution of membrane permeability values (R) calculated by electron paramagnetic
            resonance (EPR) spectroscopy in individual embryonic axes within seed lots retrieved at
            indicated time of storage (Y, years) and exhibiting different percentage of germination (G).
            For each individual axis, the R value represents the ratio between the EPR signals of a guest
            probe present respectively in the hydrophilic and the hydrophobic compartments of the cell
            in the presence of a broadening agent. The broadening agent which is water soluble cancels
            the signal emitted by the probe molecules that are localized in the hydrophilic compartment.
            Since this agent does not permeate through intact membranes, high R value indicates a
            partitioning of the EPR probe between the cytosol and lipids whereas a low (or nil) R value
            is due the loss of the membrane integrity, allowing the broadening agent to exert its
            cancelling effect on the EPR hydrophilic signal. Data are from [110]. B) Distribution of
            histone/DNA ratios measured by histochemistry on the individual 2C nuclei from the root
            meristem of squashed individual axis from seed lots stored for 4 and 6 years (Y). DNA
            content was assessed by microscope photometry at 535 nm after staining with the Feulgen
            method. The same nuclei were thereafter stained with fast-green FCF and scanned for
            histones at 635 nm. Data are from [109].