Final Report Summary - ECOSEED (Impacts of Environmental Conditions on Seed Quality)
Executive Summary:
PROJECT CONTEXT AND OBJECTIVES
Seed quality is of paramount importance to agriculture, food security and the conservation of wild plant species. Considerable economic losses result from sub-optimal seed performance, undermining food security and livelihoods. Seed quality is strongly influenced by the environmental stresses experienced by the mother plant. Climate change will further exacerbate economic losses and decrease the predictability of seed yield and quality for the farmer. The imminent challenges of climate change and food security require new knowledge of how stress impacts on seed quality, as well as a re-appraisal of optimal seed storage conditions. The EcoSeed project addressed these challenges by bringing together a group of distinguished European experts in seed science to characterise seed quality and resilience to environmental stress factors. State-of-the-art "omics", epigenetics, and post-"omics" approaches were brought together with the ambition to unravel the molecular switchboards that determine seed fate from development, through storage, germination and seedling development. Three crop plants were studied, Hordeum vulgare (barley), Brassica oleracea (B. oleracea) and Helianthus annuus (sunflower) as well as the model plant Arabidopsis thaliana (Arabidopsis), also with the aim to transfer knowledge gained from studying the model and crop plants to wild relatives in their genera, to the benefit of agriculture and conservation alike.
MAIN RESULTS ACHIEVED
Model and crop plants were subjected to environmental stresses such as those predicted in climate change scenarios, i.e. temperature stress and drought stress, and the impact of these stresses on seed quality were assessed. In the first year (2013), the consortium conducted pilot experiments to develop and optimize experimental procedures, and produced most of the biological material needed for molecular analyses. In 2014, key seed quality traits were assessed in the different seed lots and the first molecular analyses were conducted on selected samples. The year 2015 was mainly dedicated to the detailed molecular analysis of seeds selected after further in-depth seed phenotyping. Lab work was completed by the end of 2016. Importantly, it was found that drought and temperature stress applied to the mother plant impacted on seed yield, mean seed size and quality, including dormancy and vigour, in a species- and genotype-specific manner.
FINAL RESULTS AND THEIR POTENTIAL IMPACT AND USE, INCLUDING THE SOCIO-ECONOMIC IMPACT AND THE WIDER SOCIETAL IMPLICATIONS OF THE PROJECT
Novel markers for seed quality and information generated in this project will assist plant breeders, the seed trade and conservationists alike. In this way, the EcoSeed consortium intends to help solve problems related to seed quality and storability, and also play a leading role in enabling associated industries to better capture current and emerging markets. EcoSeed knowledge is transferred to the scientific community through peer-reviewed publications. As yet, 24 papers were published in scientific journals, including in Nature Communications, Nature Plants and PNAS, although the main results are yet to be published. Knowledge is also transferred to the public through the internet, the press and other activities such as YouTube videos, Podcasts and an EcoSeed Flyer. The consortium also recognized the importance of training and knowledge transfer to, and support of early career scientists. Nine PhD students and five Postdocs (on a head count basis) were employed by EcoSeed, and another 52 students (MSc, BSc, work experience and Erasmus exchange students) were trained by the consortium. Scientific exchange between partner labs was strongly encouraged, and 10 short scientific exchange visits of early career scientists were supported by the consortium.
Project Context and Objectives:
PROJECT CONTEXT
Seeds are pivotal to agricultural productivity and ecosystem conservation, yet there remain substantial gaps in our understanding of the critical role that the environment plays during seed development and storage, and its effect on seed quality. The EcoSeed project addressed this important knowledge gap by bringing together 11 EU Partners with well-established track records in seed biology and converging sciences. The consortium adopted an integrated approach with the aim to ultimately help to provide improved food security in times of increased population growth, greater pressure on cultivated land and changing climates. The EcoSeed partners aimed to deliver a step change in our understanding of seeds in relation to the environment, providing a basis for improving seed quality traits and seed handling protocols required for ensuring the availability of future plant genetic resources across Europe. Research focussed on the two major stresses, drought and elevated temperature, that are predicted to have a major impact on European crop production as a result of global climate change. Another ambition of the consortium was to provide young world class researchers with the skills needed to address the significant global challenges of food security for a population expected to reach 9.7 billion in 2050.
State-of-the-art scientific techniques were used to identify the important molecular regulatory pathways and key cellular switches that mediate seed environmental sensing and signalling, and their importance for downstream seed quality and seedling establishment. The intention was to provide a framework for the assessment of environmental effects on seed performance, urgently needed for improving agricultural production and for optimising conservation strategies. "Omics" techniques, epigenetic and post-"omics" approaches such as specific analysis of DNA repair and post-translational modifications (PTMs) of proteins, were used to define the signalling pathways that underpin the seed phenome. Special emphasis was put on redox regulation and signalling involving reactive oxygen species (ROS), and the interactions of the redox signalling hub that influence quality traits. The cellular redox hub processes information from metabolism and the environment to regulate plant growth and defence through integration with signalling pathways. Studies into redox regulation are necessary to provide a better understanding of the cross-talk between these signalling pathways that control dormancy, longevity, germination and vigour of model, crop and wild seeds. EcoSeed did not include research on the nutritional properties of seeds, which are the objectives of plant breeding programmes and have no direct relevance for conservation activities. The main focus was on understanding how seed stress incurred during development on the mother plant and during dry post-harvest storage impact on seed quality, germination and seedling establishment.
THE SPECIFIC OBJECTIVES OF ECOSEED WERE
1. to determine how environmental stresses on the mother plant during seed development affect key seed quality traits (WP1);
2. to determine how environmental stresses during seed development influence seed physiology during dry storage and define the molecular switches that trigger downstream biochemical pathways (WP2);
3. to determine the effects of the above-mentioned pre-harvest stresses and storage conditions on subsequent seed germination, and identify the signalling pathways involved (WP3);
4. to identify genes and technologies that contribute to the prediction of seed quality, storage and germination (WP4);
5. to translate the knowledge gained from the study of model and crop plants to wild species and provide a model for seed quality trait dependency on environmental factors as required by the seed trade and conservationists (WP5).
Two additional workpackages were dedicated to project management (WP6) and to the dissemination and exploitation of project results, and training of early career scientists (WP7).
SPECIES SELECTION
The consortium took advantage of the fully sequenced model plant Arabidopsis thaliana (Arabidopsis), representative of an endospermic seed type within the dicotyledonous plants. Several partners provided various Arabidopsis lines, including several mutants suitable for studying the molecular pathways implicated in plant and seed resilience to perturbation.
Responses to the environment gained in Arabidopsis were compared with those of its close relative B. oleracea, representative of vegetable crops.
Hordeum vulgare (barley) is a starch-accumulating, endospermic seed type within the monocotyledonous plants. Barley is one of the four most important cereal crops worldwide, with about one third of the global production of barley (62 million tons) produced in the EU. Due to its importance to agriculture and the brewing industry, barley germination has been extensively investigated and it is therefore a suitable crop model. Partner 3 provided selected varieties chosen from over 23,245 barley accessions.
Helianthus annuus (sunflower) is representative of a non-endospermic, dicotyledonous seed type in which the oily reserves are stored in the cotyledons. Sunflower is the second most important oil crop in the EU economy after Brassica napus (canola). Over seven million tons of sunflower seeds are harvested globally every year (http://faostat.fao.org). Partner 2 provided selected varieties chosen from over 700 wild sunflower accessions. Commercial sunflower seeds were provided by Partner 11.
Trait improvement that enables crops to withstand harsh climatic conditions with altered growing seasons, and the growth of crops on more marginal land is fundamentally dependent on gene mining from wild species. Translational research in WP5 aimed at gaining an in-depth understanding of the underlying physiology of seeds from wild species in the three crop genera, and validating seed quality markers developed for crops in WP4 for use on wild species. Using the unique germplasm collections available through Partners 2 and 3, it was tested if the mechanisms of stress resistance identified in WPs 1-4 are evolutionarily conserved and can be found in crop wild relatives that have adapted to stressful environments. New markers of seed quality could help with the regular monitoring of the quality of seeds in storage and the results could be valuable to provide a better basis for the use of genetic material of wild plants.
TO THE BENEFIT OF LAY READERS, A LIST OF THE MOST IMPORTANT DEFINITIONS IS PROVIDED BELOW:
After-ripening: a naturally occurring process, and also a commonly used method, to alleviate dormancy and to promote germination by dry storage of freshly harvested, mature seeds.
Ageing: the process of seed deterioration during seed storage, whereby vigour and then total germination gradually decrease over time, finally resulting in the death of all seeds. All seeds age and eventually die, whether stored in agricultural seed banks or in gene banks. However, gene bank storage at low temperature (e.g. 20 °C or lower) and low seed moisture content (e.g. 5% moisture content) increases seed longevity, maintaining viability in the range from decades (in short-lived species) to millennia (in long-lived species).
Controlled Deterioration: a method to accelerate seed deterioration, in this report also termed "ageing" for simplicity, by subjecting seeds with defined, relatively low moisture contents (e.g. 10 % seed moisture content) to high temperatures (e.g. 40 °C), resulting in relatively rapid decline in total germination (in the range of weeks to months).
Cultivar: A crop genotype produced by modern plant breeding techniques for selecting plants with desirable characteristics.
Dormancy: the inability of an imbibed seed to germinate under favourable conditions before certain environmental cues (e.g. low temperatures for an extended time) have been received. Seed dormancy controls the timing of seedling establishment by preventing germination of a viable, imbibed seed during (temporary) favourable conditions in an unfavourable season.
Germination: the process by which a quiescent seed develops into a seedling, starting with water uptake and ending with the protrusion of the radicle (the embryonic root) through the surrounding covers in conjunction with cell elongation in the region immediately behind the radicle.
ISSS: International Society for Seed Science.
ISTA: International Seed Testing Association.
Landrace: a local variety of a crop genotype that has adapted to tolerate biotic and abiotic stress factors typical for the natural environment in which it grows, resulting in a high yield stability and an intermediate yield level under a low-input agricultural system.
Non-viable: dead.
Priming: a technique used by the seed trade to enhance seed vigour, involving a transient increase of seed moisture content followed by re-drying and seed storage.
PGR: plant genetic resources.
ROS: reactive oxygen species.
Seed: the product of the double fertilization in angiosperm plants, which results in the formation of a small embryonic plant enclosed by the seed coat. For simplicity, in this report the term "seed" is also used for the caryopses of the Poaceae family (botanically, these are fruits in which the fruit coat, or pericarp, is fused with the thin seed coat) and for de-coated seeds (particularly for sunflower achenes without pericarp).
Seed quality: here defined as "the sum of all seed traits that are acquired from the time of seed development on the mother plant to seed germination". Seed quality determines storability and subsequent performance during germination and also has significant impacts on seedling, and hence crop, establishment.
Viability (adjective viable): capability of a seed to live and germinate (alive).
Vigour: the sum total of those properties of a seed that determine the potential level of activity and performance of the seed during germination and seedling emergence, often best observed under stress conditions.
WP: Workpackage.
Project Results:
WORKPACKAGE 1:
ENVIRONMENTAL AND GENETIC FACTORS INFLUENCING SEED QUALITY
The purpose of WP1 was
1) to characterise the impact of environmental stresses applied during grain filling and maturation on seed physiology and quality;
2) to use high-throughput "-omics" technologies (i.e. transcriptomics, proteomics, metabolomics) and study nuclear size and chromatin compaction, to explore the effects of theses stresses on seed metabolism;
3) to produce sufficient seeds for use in other WPs.
Regarding point 2) we want to point out that vast amounts of –omics data were produced for seeds grown under different environmental conditions (WP1), then subjected to storage and ageing conditions (WP2), followed by imbibition (WP3), but these cannot be described in this final public report that is limited to 25 pages, and only brief summaries are given below.
The plant species studied in WP1 were Arabidopsis thaliana (Arabidopsis), Brassica oleracea (B. oleracea), Hordeum vulgare (barley) and Helianthus annuus (sunflower). In the initial experiments the effects of the maternal environment ("drought stress" and "temperature stress") on plant and seed performance were tested. The stress conditions for further experimentation were chosen according to the following criteria: 1) changes in water availability and temperature should be close to those predicted in climate change scenarios, and 2) the stresses applied to the mother plants should not be too detrimental to the plant or to seed production, so that the requirements to produce sufficient seed for other project tasks could be met. The stress conditions were evaluated for each species. To test the stress factor "temperature", a "control temperature" for good plant growth and seed quality was chosen and compared to a regime with "lowered temperature" (3-5 °C lower than the control temperature) and one with "elevated temperature" (3-5 °C higher than control temperature). "Drought" was simulated by maintaining a defined level of plant water potential in growth cabinets (Arabidopsis, B. oleracea) and by a controlled water deficit in glasshouses (barley) or by withholding irrigation in field experiments (sunflower).
Both stresses applied individually strongly influenced seed yield, so that it was not always possible to produce enough seeds when drought and temperature stress were combined. Therefore, the consortium focused on the effects of the individual stresses and produced seeds of selected genotypes for each species in two or three environmental regimes that provided a good environmental contrast, and also allowed a sufficient yield of seeds for analyses in WPs 1, 2 and 3. Following harvest, the seeds produced from all four species were dried and stored under defined conditions to minimise the impacts of non-experimental variables and changes to physiological behaviour during storage. A detailed analysis of yield characteristics was conducted and seed quality was assessed using a range of species-specific measures. These included speed of germination at appropriate temperature, dormancy characteristics, and germination following controlled deterioration. In general, stress factors in the maternal environment had a negative effect on seed yield, mean seed size and quality.
ARABIDOPSIS
Seeds of the model plant Arabidopsis (Col-0 accession) and a range of mutant genotypes were produced in growth cabinets under three selected temperature regimes, a "control temperature" regime, a "lowered temperature" regime and an "elevated temperature" regime. Measurements of stomatal conductance and photosynthesis were taken to evaluate the stress effects under the lowered and elevated temperature regimes. Plants grown under lower temperatures produced more seeds. However, seeds produced in the lowest temperature regime exhibited a high degree of thermodormancy and seed quality was poorer. Seed performance was similar in those produced in the two other temperature regimes, but the highest temperature regime significantly reduced the numbers of seeds produced.
Furthermore, seeds from a large panel of Arabidopsis mutants were produced and screened for quality traits and seed longevity. These included a range of hormone and dormancy mutants and mutants affected in reactive oxygen species (ROS) metabolism with an assumed role in dormancy regulation. Dormancy assays were performed on freshly harvested dry seeds of all genotypes. From this, 19 mutant genotypes were selected based on their dormancy index and their sensitivity to the growth temperature of the mother plant. These were used to investigate maternal environmental effects with a focus on the influence of plant hormones on primary dormancy. All seeds were more dormant when produced in colder conditions than in warmer conditions. Seed production temperatures impacted on both seed yield and weight. In general, the higher the growth temperature the greater the reduction in seed yield and the higher the seed weight. Nonetheless, seeds from a subset of the mutants showed the opposite behaviour indicating that sensitivity to the maternal environment during seed maturation was modified in these genotypes. Furthermore, growth temperatures affected seed sensitivity to controlled deterioration (CD).
Transcriptomic analyses of the seeds produced in WP1 revealed that the maternal environment affected transcript regulation. The identified differentially expressed transcripts were further analysed by a Gene Ontology analysis. Transcripts differentially expressed in seeds matured under low temperatures corresponded to a lower numbers of Gene Ontology categories than those found in seeds produced under control or elevated temperatures. Differential proteomic analyses using a gel-free shotgun proteomic approach showed that more than 200 proteins were differentially accumulated in seeds produced under the different temperature regimes in at least one pairwise comparison. A Gene Ontology analysis revealed that proteins that were up-accumulated under elevated or lowered temperatures corresponded to relatively distinct biological and molecular functions. Changes in primary metabolites were studied using a gas chromatography-mass spectrometry (GC-MS)-based metabolomics approach, and flavonoid and phytohormone composition were studied using ultra-high performance liquid chromatography-mass spectrometry (UHPLC-MS). Over 100 primary metabolites showed differential accumulation in seeds produced in the different temperature regimes in at least one pairwise comparison. The effects of temperature and drought in the maternal environment on nuclear size and chromatin compaction was investigated to determine whether these characteristics reflected differences in seed quality.
B. OLERACEA
In paper 1, using a trait to gene analysis Morris, Barker, Walley, Lynn and Finch-Savage (New Phytologist, 2016; http://onlinelibrary.wiley.com/doi/10.1111/nph.14102/abstract) reported that allelic variation in three genes determines the key vigour trait of rapid germination in B. oleracea. A mechanism was proposed in which both seed ABA content and ABA sensitivity determines speed of germination.
Paper 2 by Finch-Savage and Bassel (Journal of Experimental Botany, 2016; https://academic.oup.com/jxb/article/67/3/567/2893341/) considers seed vigour at an ecophysiological, molecular, and biomechanical level. This review discusses how some seed characteristics that serve as adaptive responses to the natural environment are not suitable for agriculture. Past domestication has provided incremental improvements, but further actively directed change is required to produce seeds with the characteristics required now and in the future. Furthermore, ways in which basic plant science could be applied to enhance seed performance in crop production are discussed.
Two B. oleracea genotypes,"A12DHd" (Chinese Kale) and "AGSL101", a substitution line originating from a cross between A12DHd and "GDDH33" (var. italica; broccoli calabrese), were used to assess the impact of drought and temperature stress on seed quality. Seeds from both genotypes are non-dormant under benign production conditions, but A12DHd has inherently lower vigour than AGSL101. Seeds were produced in two temperature regimes, a "control" regime and an "elevated temperature" regime in the absence of water stress. Plants grown under control temperature were also subjected to a "drought" regime following seed set.
The genotype effects were generally consistent, with AGSL101 seeds providing higher quality seeds overall and being less affected by adverse seed production environments. Thus, the difference in quality between the genotypes was greater when seeds were produced under stress. Elevated temperature and drought significantly reduced stomatal conductance and the rate of photosynthesis during seed development. A likely consequence of this was that seeds produced from both genotypes in the elevated temperature regime were generally smaller and had poorer quality. However, the effect can differ between genotypes, e.g. there was a positive impact on germination speed of the elevated temperature regime on seeds of the lower vigour genotype. Furthermore, despite a similarly reduced rate of photosynthesis, seed size was hardly affected by water stress compared to that of seeds produced in the elevated temperature regime. This suggests that the adverse effects do not simply result from reduced resources production in the leaves, but requires a more complex explanation. Transcriptome, proteome and metabolome data were also generated for B. oleracea seeds and statistical analyses were started, revealing more significant differences between seeds produced under the control and elevated temperature regimes than between seeds produced under control and drought conditions.
BARLEY
More than 230 spring barley accessions from the collection of the Federal ex situ genebank for agricultural and horticultural plants at the IPK in Gatersleben, Germany, were screened following storage since 1974. From these, two genotypes with contrasting longevity phenotypes, the long-lived genotype "HOR 4710" and the short-lived genotype "HOR 6489", were chosen for seed production and analyses. Seeds were produced in a "control temperature" regime, and compared to an "elevated temperature" regime and a "drought stress" regime (the latter at control temperature), applied 7 days after flowering.
Barley is a rather stress tolerant crop grown in wide ranges of latitude and altitude, and the maternal environment had no strong effects on seed yield and germination performance, but in both genotypes the "thousand seed weight" was strongly affected by drought, and to a lesser extent by elevated temperature. However, elevating the temperature in the maternal environment reduced thermodormancy. Furthermore, no strong effects of the maternal environment were found on seed performance after CD. Isolated embryos of the produced barley seeds were submitted to transcriptomic, proteomic and metabolomic analyses. Compared to seeds produced under control conditions, those produced under drought showed pronounced molecular and biochemical rearrangements in both genotypes, whereas the elevated temperature regime affected genotype HOR4710 more than HOR2110.
SUNFLOWER
Sunflower seeds from five genotypes differing in oil content, dormancy status and earliness were produced in the field under commercial conditions in Marchena, Andalusia, Spain. At this location, there is almost no rain between June and September. "Drought" was applied during grain filling according to the following protocol: plants were irrigated five times before grain filling, and then irrigation was stopped. For comparison, seeds were produced under "control" conditions by applying two further irrigations during grain filling. The five genotypes tested did not show consistent responses to drought stress. The most pronounced effect was found in the seeds of genotype B (a semi-early linolenic hybrid poor in oil), whereby drought stress during seed maturation led to increased germination at 15 °C. Interestingly, this variety also had greater resistance to ageing, and this genotype was chosen for further investigations.
WORKPACKAGE 2:
BIOCHEMICAL AND BIOPHYSICAL CHANGES IN STORED SEEDS THAT AFFECT SEED QUALITY
In WP2 the seeds of Arabidopsis, B. oleracea, barley and sunflower produced in WP1 were used to evaluate the effects of the maternal environment on seed after-ripening and storability, and to study the associated molecular mechanisms. Seeds produced in WP1 grown in the control, the lowered, the elevated temperature and the drought regime were aged using controlled deterioration (CD) for different time intervals. The seeds were also tested for dormancy. In addition, seeds produced in WP1 were stored under agricultural conditions (by Advisory Board member Syngenta) and seed bank conditions (by Partner 3) so that they are available for future studies into the effects of the maternal environment on long-term storage.
THE EFFECTS OF MATERNAL AND STORAGE ENVIRONMENTS ON SEED RESPONSE TO CD
Seeds of Arabidopsis and B. oleracea, but not barley, produced under elevated temperatures were more susceptible to CD (i.e. seed viability and vigour were more negatively affected) compared to seeds grown in the control regime. Arabidopsis seeds grown in the lowered temperature regime had higher levels of thermodormancy, which was released upon CD. Seeds of sunflower and B. oleracea produced under drought stress responded to CD similarly as the seeds grown in the control regime. In sunflower, seed dormancy was not affected by drought stress. To assess the impact of oxygen upon seed storage, seeds were aged under ambient oxygen, termed “normoxia” (around 21% of oxygen), under elevated oxygen (>70% of oxygen) and under low oxygen, termed “hypoxia” (less than 3% of oxygen). For hypoxic conditions seeds were kept in airtight boxes flushed daily with gaseous nitrogen to maintain an oxygen level below 3%.
OXIDATIVE PROCESSES IN SEED AFTER-RIPENING AND LONGEVITY
The combined effects of the maternal and storage conditions on ROS production and antioxidant response were assessed in selected samples of interest prior to and during CD. In the seeds of all four plant species, seed ageing was generally associated with an accumulation of ROS, such as hydrogen peroxide and superoxide, and a breakdown of antioxidant compounds. Different organic radicals were found in the species investigated (reported in WP4), but no significant changes in either organic free radical chemistry or abundance were found in response to the maternal environment upon seed maturation or CD at 75% relative humidity and 40 °C. Changes in the maternal environment affected antioxidant and pro-oxidant status in the seeds of all species. However, the relationship between ROS metabolism, cellular redox state and seed germinability appeared to be, in part, species-dependent. The most consistent results across the species studied were found for the ROS hydrogen peroxide and the water-soluble antioxidant glutathione, which were significantly altered in response to the maternal environment and ageing. For example, the drought regime resulted in a higher concentration of glutathione in seeds of both barley and sunflower, and both species did not appear to display any signs of oxidative stress as a consequence of the maternal seed production environment. The composition of the lipid-soluble tocochromanols (the “Vitamin E family”) differed greatly between species. Barley seeds produced in the elevated temperature regime contained higher concentrations of tocochromanols. In Arabidopsis and B. oleracea seeds, the maternal environment affected the concentrations and redox state of glutathione. In B. oleracea, the adverse effects of the elevated temperature regime on seed vigour were associated with an imbalance between ROS and antioxidants, which can cause oxidative stress. In summary, the maternal environment during seed maturation affected redox status, ROS and antioxidants levels of the seeds.
Paper 9 by Morscher, Kranner, Arc, Bailly and Roach (Annals of Botany, 2015) reported on glutathione redox state, tocochromanols, fatty acids, antioxidant enzymes and protein carbonylation associated with after-ripening and ageing in dormant and after-ripened (non-dormant) sunflower embryos subjected to CD under normoxia (21 % oxygen) or elevated oxygen levels (75 % oxygen). After-ripening was accompanied by a shift in the thiol-based cellular redox environment towards more oxidizing conditions. Controlled deterioration under elevated oxygen led to a faster loss of seed dormancy and significant decreases in glutathione reductase and glutathione peroxidase activities, but viability was lost at the same rate as under normoxia. Irrespective of oxygen concentration, the overall thiol-based cellular redox state increased significantly over 21 d of CD to strongly oxidizing conditions and then plateaued, while viability continued to decrease. Viability loss was accompanied by a rapid decrease in glucose-6-phosphate-dehydrogenase, an enzyme that provides NADPH for reductive processes such as required by glutathione reductase. Protein carbonylation, a marker of protein oxidation, increased strongly in deteriorating seeds. The lipid-soluble tocochromanols, dominated by α-tocopherol, and fatty acid profiles remained stable. In conclusion, after-ripening, dormancy-breaking during ageing, and viability loss appeared to be associated with oxidative changes of the cytosolic environment and proteins rather than the lipid environment. High oxygen concentrations accelerated dormancy alleviation but, surprisingly, did not accelerate the rate of viability loss.
Furthermore, paper 17 by Arc, Galland, Godin, Cueff and Rajjou (Frontiers in Plant Science, 2013; http://journal.frontiersin.org/article/10.3389/fpls.2013.00346/full) gives an overview of the role of nitric oxide in the control of seed dormancy and germination in Arabidopsis with a focus on the implications of this radical in the control of seed dormancy and germination, proposing that nitric oxide-dependent post-translational modifications of proteins are key to nitric oxide signalling during early seed germination (WP3).
OXIDATIVE MODIFICATIONS OF MACROMOLECULES DURING DRY STORAGE
Paper 18 by Meimoun, Mordret, Langlade, Balzergue, Arribat, Bailly, El-Maarouf-Bouteau (PLoS One, 2014; http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0086442) showed that the alleviation of seed dormancy upon dry after-ripening is not related to regulated changes in gene expression. However, oxidative modifications to nucleic acids, proteins and lipids may occur in dry seeds, and are thought to affect seed fate upon seed imbibition (studied in WP3). In WP2, molecular targets of oxidative modification in dry seeds produced in the different maternal environments were identified.
Proteins are rather sensitive to oxidation of certain amino-acid residues, which can be oxidized reversibly or irreversibly. Methionine can be oxidized to methionine sulfoxide which, in turn can be reduced to methionine via the action of methionine sulfoxide reductase. Oxidation of methionine was detected in all protein samples extracted from dry seeds, but was not affected by the production or storage environment. Carbonylation, an irreversible oxidation of proteins, was found at altered levels when conditions in the maternal environment were altered, and upon CD in Arabidopsis and B. oleracea.
The accumulation of DNA single strand breaks (SSB) and double strand breaks (DSB) was quantified by electrophoretic methods and image analysis. For example, in B. oleracea seeds decreasing viability upon CD was associated with significant DNA damage. Moreover, using lipidomics it was demonstrated for Arabidopsis that the temperature during seed maturation had a marked effect on membrane and reserve lipid composition. In conclusion, seed ageing resulted in a more oxidizing cellular environment, although the detection of oxidatively-modified macromolecules depended on the extent of seed ageing and plant species.
MOLECULAR MOBILITY AND WATER SORPTION PROPERTIES DURING DRY STORAGE
It was also determined whether conditions of seed production and storage modulate cellular mobility in the dry state, in order to establish whether a relationship exists between oxidative modifications of biomolecules and molecular mobility. Seed water status (sorption isotherms) were studied using IGAsorp and thermodynamic analyses of water sorption isotherms. Biophysical properties of dry seeds were assessed by DMA (dynamic mechanical analysis) and DSC (Differential Scanning Calorimetry), and oil content was measured using TD-NMR.
Differences in water relations between aged and non-aged seeds of the four species could be explained mainly by variation in seed oil content. Although some isotherm shifts as a result of the maternal environment or ageing could not be fully explained, the consistent fits for the isotherms suggests a general robustness of the IGAsorp and LiCl methods. No obvious impacts of vigour or maternal environment on the glass forming properties of sunflower, B. oleracea or barley seeds were found by DMA. Using DSC, it was shown that the maternal environment had no effects on the glass properties of sunflower and barley seeds. In B. oleracea seeds, the elevated temperature regime slightly modified glass properties but this was not associated with an altered storability upon CD. In sunflower seeds, there was a clear effect of ageing on glass formation, as glass transition temperature appeared to be lower in aged seeds, suggesting that the glass weakened upon CD. No glass transitions were detected in Arabidopsis seeds. In summary, the maternal environment affected some characteristics of molecular mobility and water properties, mainly due to modified seed oil content, and significant effects were found upon CD, which induced changes in water sorption isotherms (B. oleracea, Arabidopsis) and glass properties (sunflower).
In conclusion, novel and original data were produced related to the downstream effects of the maternal environment upon seed ageing in the dry state. The major outcomes of WP2 were that the temperature regime in the maternal environment can modify dormancy alleviation and longevity, whereas drought stress had less clear effects. Furthermore, the maternal environment affected ROS homeostasis and related oxidative processes. In all species, seed ageing was associated with oxidative damage to proteins and DNA.
WORK PACKAGE 3:
MOLECULAR CHANGES DURING IMBIBITION AND GERMINATION
WP3 was dedicated to characterizing how the maternal environment (WP1) followed by post-harvest dry storage (WP2) affects seed germination. High-throughput technologies were used to capture the intricate molecular processes that control seed quality, including dormancy, vigour and longevity. Dry seeds (from WP1 and WP2) were compared with imbibed seeds.
FUNCTIONAL GENOMICS AND METABOLOMICS ASSESSED BY HIGH-THROUGHPUT TECHNOLOGIES
ARABIDOPSIS
Changes in gene expression products (mRNAs and proteins) and low-molecular-weight metabolites of seeds produced under the three temperature regimes were analysed in dry seeds and imbibed seeds. Transcriptomic analyses revealed gene sets displaying different tendencies of transcript regulation in response to the temperature regimes upon seed maturation. The differentially expressed transcripts identified were further analysed by Gene Ontology analysis. Although several Gene Ontology categories were shared between treatments, the elevated temperature regime seemed to enhance stress responses. To characterize the impact of the maternal environment on the seed proteome during germination, a shotgun proteomics approach was conducted allowing us to quantify approximately 1000 individual proteins. For each temperature regime, statistical analyses revealed proteins that were significantly up- or down-accumulated upon imbibition. Changes in the maternal environment also resulted in the differential accumulation of metabolites in dry and imbibed seeds, in accordance with transcriptome and proteome results. Integrative statistics were used to study the complex interactions within transcriptome, proteome and metabolome datasets. The selected variables (transcripts, proteins and metabolites) revealed connexions (correlations) within and between datasets based on the accumulation levels for each variable, highlighting predictive biomarkers of seed quality in Arabidopsis.
Paper 10 by Galland, Huguet, Arc, Cueff, Job, Rajjou (Molecular & Cellular Proteomics, 2014; http://www.mcponline.org/content/13/1/252.short) provides a detailed description of the dynamics of protein synthesis during the time course of germination, demonstrating that mRNA translation is both sequential and selective during this process. The data confirms that during early imbibition, the Arabidopsis translatome keeps reflecting an embryonic maturation program until a certain developmental checkpoint.
Paper 13, an opinion article by Galland and Rajjou (Frontiers in Plant Science, 2015; http://journal.frontiersin.org/article/10.3389/fpls.2015.00284/full) discusses the regulation of mRNA translation, which controls seed germination and is critical for seedling vigour.
Paper 23 by Basbouss-Serhal, Pateyron, Cochet, Leymarie and Bailly (Plant Physiology, 2017; http://dx.doi.org/10.1104/pp.16.01933) shows that the regulation of germination by dormancy in Arabidopsis seeds depends on the specific decay of positive and negative regulators of this complex process. It was demonstrated that 5’ to 3’ mRNA degradation, mRNA secondary structure and specific motifs are implicated in the selectivity of the decay. This work also allowed to reveal several new regulators of seed dormancy and germination.
B. OLERACEA
As described for Arabidopsis, high-throughput omics analyses were also conducted for the seeds of the two B. oleracea genotypes, A12DHd and AGSL101, produced in the control, the elevated temperature and the drought regimes. Again, dry seeds (from WP1) were compared to imbibed seeds. Correlation networks were established between proteins and metabolites revealing major correlations between proteome and metabolome levels as a fingerprint of seed quality in B. oleracea.
BARLEY
As for Arabidopsis and B. oleracea, high-throughput omics analyses were conducted for barley seeds produced in the control, the elevated temperature and the drought regimes. Again, dry seeds (WP1) were compared to imbibed seeds. Metabolomics experiments were conducted on the dry seeds of both barley genotypes initially selected (HOR4710 and HOR2110) and produced under different conditions. The metabolome of HOR2110 seeds was less affected by stress than the metabolome of HOR4710 seeds. Due to the low sensibility of HOR2110 to both drought and heat stress, it was decided to conduct detailed transcriptomic, proteomic and metabolomic analyses only for the HOR4710 genotype. Quantitative and qualitative RNAseq and mass spectrometry data were generated and the statistical analyses are in progress.
SUNFLOWER
Sunflower seeds were produced in the field under control and drought conditions during grain filling, and analysed by high-throughput methods as described for the other three species.
Furthermore, paper 6 by Layat, Leymarie, El-Maarouf-Bouteau, Caius, Langlade and Bailly (New Phytologist, 2014; http://onlinelibrary.wiley.com/doi/10.1111/nph.13002/full) highlighted the importance of post-transcriptional regulation of germination. The data suggests that seed germination requires timely regulated and selective recruitment of mRNAs into polysomes. The polysome-associated mRNAs, that is, the translatome, were fractionated and characterized with microarrays in dormant and nondormant sunflower embryos during imbibition at 10°C. The proteins corresponding to the polysomal mRNAs in nondormant embryos were involved mainly in transport, regulation of transcription and cell wall modifications.
ROLE OF QUINONE REDUCTASES IN THE CONTROL OF ROS LEVELS DURING GERMINATION
Paper 21 by Biniek, Heyno, Kruk, Sparla, Trost and Krieger-Liszkay (Planta, 2016; http://link.springer.com/article/10.1007%2Fs00425-016-2643-y) showed that a NAD(P)H quinone oxidoreductase, NQR, and the cytochrome b AIR12 are involved in controlling superoxide generation at the plasma membrane. Two proteins which are important to control ROS production at the level of the apoplast were identified. NAD(P)H:quinone oxidoreductase (NQR) and cytochrome b (called AIR12 for Auxin Induced in Root culture) are two proteins attached to the plant plasma membrane which may be important for generating and controlling ROS levels in the apoplast. AIR12 is a single gene in Arabidopsis that codes for a mono-heme cytochrome b. The NQR is a two electron-transferring flavoenzyme that contributes to the generation of superoxide in isolated plasma membranes. Arabidopsis double knockout plants of both NQR and AIR12 generated more superoxide and germinated faster than the single mutants affected in AIR12 or in NQR. An electron transfer catalyzed by NQR from cytosolic NAD(P)H to a quinone located in the plasma membrane was demonstrated. The quinol (reduced state of the quinone) is then oxidized by AIR12. In addition, the nature of the quinones in plant plasma membranes was identified and quantified by HPLC. Biochemical analysis showed that purified plasma membranes contained phylloquinone and menaquinone-4 as redox carriers. This is the first report on the occurrence of menaquinone-4 in eukaryotic photosynthetic organisms. It can be concluded from the mutant studies and the in vitro experiments that both NQR and AIR12 play an important role in controlling ROS production at the plasma membrane. Absence of both proteins increased the superoxide level, which may be responsible for the observed stimulation of germination.
ROLE OF ASCORBATE AND NUCLEAR GLUTATHIONE SALICIL ACID-TRANSFERASES (GSTS) AND GLUTAREDOXINS (GRXS) IN CONTROLLING ROS-LEVELS DURING GERMINATION
Redox metabolism and signalling in the emerging radicles of germinating Arabidopsis seeds were studied.
Paper 24 by De Simone, Dong, Vivancos and Foyer (Molecular Physiology and Ecophysiology of Sulfur, 2015; http://link.springer.com/chapter/10.1007%2F978-3-319-20137-5_4) reported on the distribution of the antioxidant glutathione between the nucleus and the cytosol. The glutathione redox potential was similar in the nucleus and cytosol of developing radicles of Arabidopsis seeds after germination. However, in the arrested embryonic root meristem of the root meristemless 1 (rml1) mutant that have less than 5 % GSH of the wild-type, glutathione was predominantly localised in the nuclei. This was also the case in wild-type roots treated with the auxin transport inhibitor, N-1-napthylphthalamic acid, which have decreased root glutathione levels.
Furthermore, paper 20 by Diaz-Vivancos, de Simone, Kiddle and Foyer (Free Radical Biology and Medicine, 2015; http://www.sciencedirect.com/science/article/pii/S0891584915006085) considered the respective roles of ROS and glutathione in the regulation of plant growth, with a particular focus on the regulation of the plant cell cycle. Glutathione, in addition to its roles as a crucial redox buffer, appears to be a flexible regulator of genetic and epigenetic functions. It is proposed that regulated ROS production linked to glutathione-mediated signalling events are the hallmark of viable cells within a changing and challenging environment.
REPAIR MECHANISMS
Paper 12 by Sano, Rajjou, North, Debeaujon, Marion-Poll and Seo (Plant and Cell Physiology, 2016; https://academic.oup.com/pcp/article/57/4/660/2460931/Staying-Alive-Molecular-Aspects-of-Seed-Longevity) reviews molecular aspects of seed longevity. Reduction of seed longevity is often associated with oxidation of cellular macromolecules such as nucleic acids, proteins and lipids. Seeds possess two main strategies to combat these stressful conditions: protection and repair. The protective mechanism includes the formation of glassy cytoplasm to reduce cellular metabolic activities and the production of antioxidants that prevent accumulation of oxidized macromolecules during seed storage (see WP2). The repair system removes damage accumulated in DNA, RNA and proteins upon seed imbibition through enzymes such as DNA glycosylase and methionine sulfoxide reductase.
Paper 7 by Waterworth, Bray and West (Journal of Experimental Botany, 2015; https://academic.oup.com/jxb/article/66/12/3549/524229/The-importance-of-safeguarding-genome-integrity-in) on the importance of safeguarding genome integrity in germination and seed longevity reviewed recent advances that have established molecular links between genome integrity and seed quality. These studies identified that maintenance of genome integrity is particularly important to the seed stage of the plant lifecycle, revealing new insight into the physiological roles of plant DNA repair and recombination mechanisms.
Paper 4 by Waterworth, Footitt, Bray, Finch-Savage and West (PNAS, 2016; http://www.pnas.org/content/113/34/9647.abstract) demonstrated physiological functions for DNA damage checkpoint kinases (ATM and ATR) in linking genome integrity to germination, thereby influencing seed quality, crucial for plant survival in the natural environment and sustainable crop production.
WORKPACKAGE 4:
BIOTECHNOLOGY AND INNOVATIVE TOOLS
The goals of WP4 were to identify genes and technologies that enable a prediction of seed quality in relation to stress factors experienced in the maternal environment. Some of these genes were identified based on the -omics studies from WP1 and WP3. Additional genes were obtained from studies of natural variation, using natural modifiers, Quantitative Trait Loci (QTL) analysis and Genome Wide Association Study (GWAS) in Arabidopsis, B. oleracea and barley. In addition, seed quality genes were identified by studying the transcriptional network in seeds. Several known genes involved in seed quality were analysed for their tissue-specific roles. Finally, novel techniques like Activity-Based Protein Profiling (ABPP) and High-Field Electron Paramagnetic Resonance (HF-EPR) were investigated for their potential to provide easily scorable markers for seed quality.
IDENTIFICATION OF NOVEL GENES DETERMINING SEED QUALITY
Genes that showed a response to changed temperatures during seed development in Arabidopsis and that have a predicted function indicating potential involvement in seed quality were selected from WP1 and WP3 after combining the datasets produced from transcriptomic and proteomic analyses. These genes participate in different biological processes, such as defence mechanisms, redox regulation or protein and lipid metabolism. Insertion mutants for all selected candidates were obtained from the Arabidopsis Stock Centre. An initial analysis of the seed quality phenotypes of these insertion mutants revealed that most of them showed similar dormancy levels and resistance to CD as wild-type seeds. This suggests that investigated genes could have redundant roles regarding seed quality. However, for a few mutants reduced germination was observed indicating an important role in seed quality. Further analyses will indicate the potential of these genes as markers to predict and as tools to improve seed quality.
The transcriptional network implicated in seed quality was characterized by implementing a high-throughput in-plant expression screening of transcription factors. Key transcription factors involved in abiotic stress tolerance during seed development were identified. Ten independent lines were generated for each transcription factor-open reading frame (ORF) and the transgenic lines were phenotypically characterised in relation to a large number of seed quality-related traits. This screening allowed to identify lines affected in the responses to the plant hormone abscisic acid, salt stress or osmotic stress during seed germination. The results obtained after the analysis of approximately 634 transcription factors, corresponding to 1636 homozygous independent transgenic lines indicated that 13 lines showed insensitive phenotypes to low abscisic acid concentration and 5 lines showed insensitive phenotypes to abscisic acid and abiotic stresses (salt and osmotic stress) compared with the wild-type. Those five transcription factors that showed higher insensitivity to abscisic acid, salt and/or osmotic stress during seed germination and seedling development belonged to different families. The genes selected from the screening are likely to be involved in the mechanism by which different abiotic stresses (i.e. salt and osmotic stress) during seed development influence germination and seedling establishment. Their roles and functions will be further evaluated by the phenotypic analysis of simple and double mutants in these genes.
Furthermore, the impact of the maternal environment on subsequent seed quality was studied in B. oleracea by comparing two genotypes with inherent differences in seed vigour, A12DHd (low vigour) and AGSL101 (high vigour). This study concluded a long-term investigation on the genetic basis of the differences in vigour in these lines by developing a genetic model for the potential regulation of germination. It was demonstrated that allelic variation at three loci influence this key trait. Functional analysis in both B. oleracea and Arabidopsis identified and demonstrated activity of genes at these loci. Two candidate genes were identified at the principle Speed of Germination QTL (SOG1) in B. oleracea. One gene BoLCVIG2 is a homologue of the alternative-splicing (AS) regulator (AtPTB1). The other gene BoLCVIG1 is unknown, but different alleles had different splice forms that were coincident with altered abscisic acid sensitivity. A further QTL, Reduced ABscisic Acid 1 (RABA1), was identified that influenced abscisic acid content and provided evidence that this results from the activity of a homologue of the abscisic acid catabolic gene AtCYP707A2 at this locus (for further details see Paper 1, Morris et al. New Phytologist, 2016). Lines containing beneficial alleles of these three genes had greater seed vigour. This work provides a way forward to understand and manipulate the genetic pathways regulating this key agricultural trait.
Paper 22 by Sechet, Roux, Plessis, Effroy, Frey, Perreau, Biniek, Krieger-Liszkay, Macherel, North, Mireau and Marion-Poll (Molecular Plant, 2015; http://www.sciencedirect.com/science/article/pii/S1674205214000331) showed that the hot ABA-deficiency suppressor2 (has2) mutation increases drought tolerance and the ABA sensitivity of seed germination, in agreement with the above findings.
IDENTIFICATION OF GENES THAT DETERMINE SEED QUALITY IN GENETIC MATERIAL WITH ALTERED LONGEVITY
Previous work identified three natural modifiers from the Arabidopsis Shahdara accession, which were able to enhance seed longevity in a longevity-defective mutant (lec1-3) in the Ler background. Fine-mapping of the strongest modifier led to the identification of a candidate gene explaining the observed enhanced seed longevity phenotype. This candidate gene is involved in the regulation of flowering time, which complicates proper synchronisation of seed maturation periods and harvesting. In order to study seed longevity without the influence of altered flowering time, transgenic plants (which are mutants for the candidate genes) were constructed expressing the coding sequence of a candidate gene under the control of a seed-specific promoter. These transgenic plants showed altered seed longevity without altered flowering time. Further characterisation of these transgenic lines showed that the observed phenotypes perfectly matched transcript levels of the transgene. This work identified an interesting connection between the regulation of flowering time and seed quality which is relevant for further studies and can have practical implications.
Seed longevity loci/genes were also identified and characterised in barley by combining a quantitative genetics approach with -omics technologies. This work is based on a previous Quantitative Trait Loci (QTL) study in recombinant inbred lines (RILs), derived from the short lived Ethiopian spring barley landrace L94 and the long lived Argentinian spring barley landrace Cebada Capa (CC). In total, four putative QTLs for seed longevity were identified and near isogenic lines (NILs) were generated. Using these NILs, the four putative QTLs could be confirmed and further mapping of Cebada Capa introgressions was achieved in the so called L94 NILs using RNA seq. For candidate and downstream target gene identification, a combined transcriptome and proteome analysis was conducted. The analysis of mature, non aged seeds of the two parental lines and the L94 NILs by RNA seq and total seed proteomic profiling identified two putative candidate genes and one possible downstream target gene. To validate these, putative candidate and downstream target genes, T DNA knock out lines of homologous Arabidopsis genes, recently shown to have a seed longevity phenotype, were complemented with the respective Cebada Capa and L94 alleles. Both a candidate and downstream target gene were able to rescue their respective Arabidopsis mutant seed longevity phenotypes when expressed under the control of the 35S or the Ubi promoter. Since one of the identified candidates may affect the redox status in normal and deteriorating seeds, the corresponding Arabidopsis mutant was combined with the redox reporter roGFP2 to visualize the redox potential in seeds using confocal microscopy. First transformant generation (T1) lines, showing strong fluorescence in the cytosol, could be generated in this study and will allow the measurement of the NADP+/NADPH redox potential in future studies. Metabolite profiling, performed on the same genotypes utilized for the transcriptomic and proteomic analyses, identified one monosaccharide sugar and one flavonoid showing a similar trend in abundance in Cebada Capa and the L94 NILs. Sugars have been proposed in seeds to act as signals that regulate and influence seed development. Furthermore, several studies showed that sugars help maintain the structural integrity of membranes and proteins under dry conditions due to the formation of a glassy state to limit deteriorative reactions.
Furthermore, paper 19 by Nagel, Kodde, Pistrick, Mascher, Börner and Groot (Plant Science, 2016; http://journal.frontiersin.org/article/10.3389/fpls.2016.00388/full) analysed the genetic background of seed longevity in barley, revealing a tolerance QTL on Chromosome 5H.
Paper 8 by Nagel, Kranner, Neumann, Rolletschek, Seal, Colville, Fernández-Marín and Börner (Plant Cell and Environment, 2015; http://onlinelibrary.wiley.com/doi/10.1111/pce.12474/full) showed that seed ageing and longevity are intricately affected by genetic background and developmental and environmental conditions in barley. Seeds from 175 barley genotypes from four continents grown in field plots with different nutrient supply were subjected to ageing conditions. Genome-wide association mapping revealed 107 marker trait associations, and hence genotypic effects on seed ageing. Abiotic and biotic stresses were found to affect seed longevity. To address aspects of abiotic, including oxidative, stress, two major antioxidant groups were analysed. No correlation was found between seed deterioration and the lipid-soluble tocochromanols, nor with oil, starch and protein contents. Conversely, the water-soluble antioxidant glutathione and related thiols were converted to disulphides, indicating a strong shift towards more oxidizing intracellular conditions, in seeds subjected to long-term dry storage at two temperatures or to two artificial ageing treatments. The data suggest that intracellular pH and (bio)chemical processes leading to seed deterioration were influenced by the type of ageing or storage. Moreover, seed response to ageing or storage treatment appears to be significantly influenced by both maternal environment and genetic background.
In addition to the QTL analysis, genome-wide association mapping (GWAS) was conducted for seed storability using 116 two-row and 68 six-row barley lines with widely ranging total germination that had been stored since 1974. Genotyping was conducted using Illumina chips. In total 7,864 markers were analysed, of which 4,343 markers were mapped for the seven barley chromosomes and structured genotypes into four major clusters (six-row Ethiopian landraces, two-row German landraces, six-row cultivars, two-row cultivars) after phylogenetic analysis.
In 2013 and 2014, the spring barley collection was multiplied in the greenhouse and in the field to investigate the link between key agronomic traits, dormancy and storability characteristics. Seeds from both years were used to evaluate days to flowering and maturity, plant height, 1000-seed weight as well as dormancy and vigour characteristics. The different growth conditions affected the time to flowering, maturity, plant height and 1000-seed weight and showed a great variation between the genotypes. Dormancy was tested at maturity at 10 °C and 20 °C; germination was drastically reduced at 20 °C, indicative of thermodormancy in the spring barley collection. Only weakly significant correlations were found between days to maturity and dormancy at maturity, and no significant correlation was shown when dormancy was linked with germination after 40 years of dry storage. However, dormancy was significantly different between the four major phylogenetic clusters. Six-row Ethiopian landraces showed strongly reduced and six-row cultivars strongly developed dormancy.
The 184 spring barley accessions multiplied in the field in 2014 were used to assess seed storability after controlled deterioration. Here, fresh seeds were kept at 60% relative humidity and 45 °C for 15 days and showed great variation between germination of genotypes. Results were used for GWAS analysis and revealed 200 highly significant marker-trait associations for percentage of total germination after controlled deterioration and 169 associations after 40 years of dry storage. Although the correlation coefficient between total germination after 40 years of dry storage and controlled deterioration was low, 25 marker-trait associations for 40 years of dry storage and controlled deterioration were localised at similar positions.
THE IMPACT OF KNOWN FACTORS ON SEED QUALITY IN ARABIDOPSIS
The review paper 14 by Née, Xiang and Soppe (Current Opinion in Plant Biology, 2017; http://www.sciencedirect.com/science/article/pii/S1369526616301339) focuses on the control of dormancy induction by plant hormones and dormancy-specific proteins, the influence of environmental conditions during seed maturation and storage, non-enzymatic oxidative reactions, adjustment of dormancy in the soil seed bank through temporal and spatial sensing and the decision of the seed to germinate during early imbibition.
INFLUENCE OF THE TRANSCRIPTIONAL NETWORK ON SEED QUALITY TRAITS
Transcriptional control is the single most important mechanism of regulation of gene activity in all organisms and is ultimately mediated by transcription factors that recognise cis-regulatory elements within their target genes. Because of their key role in the regulation of coordinated gene activity and consequently complex traits (i.e. seed germination), transcription factors are thought to play an important role in the onset of germination and have been the targets for breeding and domestication. One of the aims of WP4 was to characterize the transcriptional network implicated in seed quality. For this purpose, transcription factors involved in the nitric oxide signal transduction pathway during seed germination and seedling establishment were identified.
Paper 3 by Albertos, Romero-Puertas, Tatematsu, Mateos, Sanchez-Vicente, Nambara and Lorenzo (Nature Communications, 2015; http://www.nature.com/articles/ncomms9669) showed that S-nitrosylation triggers ABI5 degradation to promote seed germination and seedling growth. ABI5 (AT2G36270) was found to be involved in the crosstalk between abscisic acid and nitric oxide during seed germination; abi5 mutants were insensitive to abscisic acid and to nitric oxide scavenging by cPTIO during germination. Nitric oxide-production and ABI5 expression co-localised in young seedlings. Using genetic (nitric oxide-deficient and nitric oxide-over accumulating lines) and pharmacological approaches (nitric oxide donors and nitric oxide scavengers, it was shown that ABI5 accumulation is altered in genetic backgrounds with impaired nitric oxide homeostasis. Cysteine-153 was identified as the central point of ABI5 regulation by nitric oxide. Mutation of ABI5 at cysteine-153 deregulates protein stability and inhibition of seed germination by nitric oxide depletion. These findings suggest an inverse molecular link between nitric oxide and abscisic acid signalling through distinct posttranslational modifications of ABI5 during early seedling development. Overall, abscisic acid regulates ABI5, a central hub of growth repression, while the reactive nitrogen molecule nitric oxide counteracts abscisic acid-mediated effects during seed germination. S-nitrosylation of ABI5 at cysteine-153 facilitates its degradation through CULLIN4-based and KEEP ON GOING E3 ligases, and promotes seed germination.
Furthermore, paper 11 by Sanz, Albertos, Mateos, Sánchez-Vicente, Lechón, Fernández-Marcos and Lorenzo (Journal of Experimental Botany, 2015; https://academic.oup.com/jxb/article/66/10/2857/534941/Nitric-oxide-NO-and-phytohormones-crosstalk-during) reviews our knowledge of the interactions between nitric oxide and phytohormones during seed dormancy and germination, and seedling development.
Paper 15 by Sanz, Fernández-Marcos, Modrego, Lewis, Muday, Pollmann, Dueñas, Santos-Buelga and Lorenzo (Plant Physiology, 2014; http://www.plantphysiol.org/content/166/4/1972.short) further explored the mechanisms by which nitric oxide modulates root development. A pharmacological approach and NO-deficient mutants were used to unravel the role of nitric oxide in establishing auxin distribution patterns necessary for stem cell niche homeostasis.
EVALUATION OF THE POTENTIAL OF ACTIVITY-BASED PROTEIN PROFILING (ABPP) AND HIGH FIELD EPR (HF-EPR) SPECTROSCOPY TO ASSESS SEED QUALITY
In addition to genetic markers, novel technologies like ABPP and HF-EPR have the potential to evaluate the quality of harvested seeds at low cost and have been tested in this context in EcoSeed. ABPP is a novel technology that displays the active proteome, using biotinylated or fluorescent small molecule probes that irreversibly bind to the active site residues of various protein classes such as lipases, proteases, glycosidases and phosphatases. The feasibility of ABPP to assess seed quality was analysed using seed lots with different qualities. These were obtained by CD of Arabidopsis seeds. For the ABPP assay, proteins were extracted from seeds that had been subjected to CD for different time periods, representing several seed quality levels (ranging from highly viable to dead). Four ABPP probes, which targeted different enzymes, including proteases, were evaluated. Two of them exhibited a differential band pattern between seed lots with different qualities. One of the probes showed a strong increase in signal intensity for two different bands during seed ageing, both in the Ler and Col accessions. In contrast, several bands that were detected with the other probe decreased in intensity during the ageing treatment.
Heat shock, which resembles the conditions used for the artificial ageing treatment, can trigger activation of the enzyme targeted by one of the probes. In order to verify that the observed pattern is a result of the ageing treatment and not caused by a heat response, additional labelling was performed using naturally aged seeds. Eight-year-old seeds from Ler and the longevity-defective mutant dog1-1 were used. A germination test showed that aged Ler exhibited 40 % germination after eight days of imbibition, whereas the dog1-1 mutant was dead. The naturally aged seeds showed the same increased activity pattern that we observed previously using artificially aged seeds.
HF-EPR provides an opportunity to directly monitor potentially important radical chemistry in seeds. HF-EPR spectra were obtained from seed lots of control and aged seeds from A. thaliana, Brassica, sunflower and barley. Different signals were dominating the spectra of the four species investigated. In Arabidopsis, a flavonoid radical was identified. In sunflower, the HF-EPR spectra were dominated by a melanin radical. In B. oleracea, surprisingly high manganese contents were detected, both in the pericarp and in the seeds when the pericarp had been removed. In the genotypes of barley seeds used for the measurements no signal was observed that could be attributed to an organic radical, whereas wild Hordeum species contained similar radicals as Arabidopsis. The number of different signals obtained by HF-EPR excludes this method as a general one to assess seed quality. Furthermore, there was no simple correlation between signal size and ageing treatment as described below.
The spectra of control and aged Arabidopsis seeds were examined for flavonoid organic radical content. The hypothesis was that the radical content should reflect exposure to oxidative conditions. There were typical lot-to-lot variations between and within the non-aged and aged control seeds, but the approximate radical contents in these seeds were not significantly different. The low or high temperature regimes in the maternal environment did not significantly change radical contents. Nonetheless, the high content was indicative of a significant change in the radical chemistry that generated the radical and, equally important, the chemistry that occurs during imbibition. The location of the radical was not identified, but measurements of tt-mutants suggest that in Arabidopsis the radical is located in the seed coat. An effort was made to determine the exact chemical nature of the flavonoid radical in Arabidopsis. In addition to the High Field EPR spectra ENDOR spectra were obtained. The two methods allowed us to identify the organic radical as a catechin radical.
B. oleracea seeds contain high concentrations of manganese (II), but no radicals were found. The manganese (II) content in the seed coat was significantly higher than that in the rest of the seed. HF-EPR measurements suggested that the manganese (II) within the seed was mostly associated with phosphates.
Sunflower seeds were dissected and pericarp, seed coat, cotyledons and embryonic axes were measured separately. A radical signal was only found in the pericarp. Here, the large flavonoid radical signal masked the presence of even larger narrow melanin radical resonances. No radicals were found in cotyledons and embryonic axes. Drought stress to the mother plant or ageing of the seeds did not alter the spectra.
WORKPACKAGE 5:
TRANSLATIONAL BIOLOGY ON WILD SPECIES
WP5 translated knowledge of crop and Arabidopsis responses to environmental stresses imposed during seed development in the context of pre-existing adaptive seed biology traits in eight crop wild relatives. In addition to the characterisation of the seed germination and longevity traits, investigations progressed on the expression of homologous genes / marker molecules for abiotic stress factors. Finally, a conceptual model was constructed of downstream risks to agricultural efficiency and wild species conservation on the basis of changes in seed quality traits anticipated under future climates.
WILD SPECIES SEED REGENERATION AND ASSESSMENT
Partner 3 (IPK) and Partner 2 (RBGK) hold ex situ collections of 98 species in the three targeted crop genera: Brassica (35 species), Helianthus (27 species) and Hordeum (36 species). These unique genetic resources, some harvested decades ago and stored under seed bank conditions, have associated information on (georeferenced) provenance, seed number and quality, verified plant name, etc. The criteria for use of seeds from these collections were: 1) climate of the seed provenance (wild material); 2) permission from the country of origin to use the accession for research purposes; 3) quantity of seed available; 4) regeneration potential. In contrast to domesticated plants, wild plants do not necessarily follow a simple regeneration cycle, with specific requirements for germination, plant and flower development, and maturity. Therefore, we selected accessions, across 10 species or subspecies, for field multiplication / regeneration over two seasons. The material was taxonomically and agronomically evaluated (e.g. plant height, time to flowering); as appropriate, bees were introduced to the greenhouses / tunnels to enhance pollination. Accessions from seven countries were cultivated as follows: Brassica rapa subsp. campestris (Turkey) and Brassica tournefortii (Tunisia, Spain); Hordeum bulbosum (Israel, Iran), Hordeum murinum (Greece) and Hordeum pusillum (USA); Helianthus agrestis (USA), Helianthus angustifolius (USA), Helianthus gracilentis (USA), Helianthus glaucophyllus (USA) and Helianthus praecox (USA). Overall, >500,000 seeds were produced.
Sunflowers require a high number of sunlight hours and a dry climate during flowering and some specific cultivation challenges emerged: some species showed perennial plant characteristics, some plants died before flowering, some plants produced sterile seeds and some had high levels of seed dormancy. Consequently, dormancy-breaking treatments were applied to a selection of Helianthus collections from the Kew seed bank. Scarification of seeds of two species of Helianthus resulted in slightly faster water uptake, meaning that the seeds were not physically dormant. Seed morphology (by light microscopy and X-ray) was compared between nine Helianthus species (eight wild and one crop species) of varying seed dormancy status. Pericarp thickness varied up to 6-fold amongst the wild species and crops. An association with growing site meteorology was assessed. Embryo size and distance from the radicle tip to the pericarp edge were quantified for two species and considered in relation to the speed of germination for individual seeds. The physiological nature of the dormancy was assessed on 11 accessions of 10 species varying in germination from 0 % (fully dormant) to 100% (non-dormant), through the application of gibberellic acid. Six of the dormant seed lots had improved germination after the application of gibberellic acid to seeds scarified at the root axis (micropylar) end. These findings impacted on the experimental designs.
QUANTIFICATION OF SEED QUALITY TRAITS IN WILD SPECIES
Germination characteristics: Thermal time and hydrotime give an indication of how responsive the germination of a species is to temperature and water availability, and can be used to predict the germination behaviour in future environments such as due to climate change. Germination was performed at five or six temperatures and four or five water potentials (created using polyethylene glycol), to give a range of germination rates to describe performance and for model development. Often germination was defined as both radicle emergence (2 mm) and normal seedling production. Cumulative germination curves (i.e. % germination over time) were plotted and times taken to reach 20 %, 50 % and 80 % of the maximum germination interpolated. The reciprocals of these germination rates plotted against temperature or water potential enable predictions of the thresholds (base temperature, Tb; optimum temperature To; ceiling temperature, Tc; base water potential) for growth, and the calculation of thermal time, hydrotime and hydrothermal time. To maximise the translation of knowledge, comparative seed characterisation was performed on 12 species in the Helianthus, Brassica and Hordeum genera. For Helianthus, four wild species were studied, plus five Helianthus annuus genotypes from Partner 11 (Limagrain), and two environmental treatments of sunflower genotype B. We found that amongst Helianthus species and crop genotype seeds, base temperatures varied by c. 12 °C and the thermal-, hydro- and hydro-thermal times varied c. 7- to 10-fold. For Brassica, four wild species / sub-species were investigated, plus two vigour (high and low) genotypes of B. oleracea and two environmental treatments of the same genotypes. We observed that seeds of Brassica species and crops had base temperatures that varied by 9 °C, and the heterogeneity in thermal- hydro- and hydro-thermal times was c. 4- to 7-fold. For Hordeum, four wild species were studied, plus one (long-lived) Hordeum vulgare genotype and two environmental treatments of that genotype. We found that seeds of Hordeum species and crops had base temperatures that varied by 4 °C. Differences in thermal-, hydro- and hydro-thermal times amongst species were c. 4-fold. In conclusion, there are substantial differences between species in stress tolerance during germination, which means that each species will have emergence advantages over the others under different combinations of temperature and water potential. These studies also identified the optimal germination conditions for use in the longevity experiments when assessing seed survival.
Longevity properties and biophysics: Seed storage lifespan is affected by temperature and moisture content, with the latter being a feature of the seeds' water binding properties which vary with relative humidity and with oil content. Therefore, seeds of eight wild species (three Helianthus, three Brassica and two Hordeum) were used to construct water sorption isotherms and to determine seed oil contents (using time-domain NMR). The Brassica species had high oil contents, slightly higher than for wild Helianthus and much greater than the wild Hordeum species. Generally, the crop seeds were similar in oil content to their wild relatives. We noted that the relative humidity required to achieve the same moisture content will vary greatly with oil content. The implication for the seed trade is that when storing seeds at the same moisture content there is a risk that the different seed lots (accessions) will not be at the same water activity and thus not have the same longevity characteristics.
Seed prehydration history also has a potential impact on the seed biophysical properties. One feature of sorption isotherms is that the relationship between seed moisture content and relative humidity achieved when hydrating a sample (adsorption) is different to that achieved during dehydration (desorption). Seeds of the same eight wild species used for oil content determination were equilibrated to different relative humidities at 20 °C in an IgaSorp machine (c. 50 mg samples) and seed mass measured as relative humidity was ramped up from 15 to 75% relative humidity and back down. On average, each adsorption / desorption isotherm run took three weeks to construct. Seed moisture content was determined at the end of each run. Seeds were also equilibrated at 20 °C above LiCl solutions at various saturations. Although there were some differences between the IgaSorp and LiCl data (e.g. the isotherms of the later tend to be a little more concave), the goodness of fit for the isotherms suggests a general robustness of the two methods.
Seed sorption properties of the three species of Helianthus were found to be broadly similar, with 7% seed moisture content achieved during adsorption at c. 50-60% RH, whether in the IgaSorp or over LiCl solutions. The adsorption result for Brassica seeds (three species) was similar to Helianthus seeds. In contrast, the two Hordeum species had much lower oil contents and this was reflected in a different moisture content-relative humidity relationship such that 7% seed moisture content was achieved at 30-40% RH. One clear feature of the isotherm study was the hysteresis between adsorption (a lower seed moisture content for a given RH) and desorption (a higher seed moisture content for a given relative humidity) curves. The differences in the mid-range of relative humidities was about 10% for Helianthus, and 7% for Brassica and Hordeum. This means that drying wet seeds to 7% seed moisture content (desorption) will result in a seed equilibrium relative humidity that is lower than that reached by seeds that were pre-dried and allowed to rehydrate (adsorp) to the same moisture content. The implication for the seed trade is that the "desorbed" seed will have lower chemical / water activity and potentially a better lifespan than "adsorbed" seed at the same MC.
Seeds of eight wild species were aged under a range of relative humidity conditions at 40 °C. The aim of this study was to characterise the longevity of a range of wild species to see if their seeds have similar keeping properties to the main crops in the same genera. Seed survival lines were subjected to probit analysis to allow estimates of lifespan (the distribution of seed deaths in time, sigma; and the half-life, P50). It was originally intended to additionally store seeds of all eight species at 30 and 20 °C under a range of relative humidity conditions, but limitations on the number of seeds available meant this was only possible at one temperature and five relative humidities (15 – 75%). Seed viability after storage was assessed as germination under optimal conditions. As a benchmark, a wider range of conditions was used to age the Helianthus annuus (control and drought regime) seed lots (linking to WP2).
For all seed lots (accessions) investigated (wild species and crops) the benefits of decreasing seed relative humidity on seed longevity were noted. For example, Helianthus angustifolius seed pre-equilibrated to 75% relative humidity and aged at 43 °C (average) lost most viability in 10 days. One feature of some of the ageing curves for the Helianthus crop wild relatives was an increase in germination value during the early stages of treatment, indicative of seed dormancy loss (after-ripening). Across the eight wild species, P50s (half-lives) varied about 8-fold for seeds at 75% relative humidity and 43 °C. Published data sources for the crops in the three investigated genera indicate values of 6.7-9.1 for KE and 4.2-5.3 for CW. The terms KE and CW describe the dependency of seed lifespan on moisture content. By using the universal temperature constants (CH = 0.0329; CQ = 0.000478) it was possible to generate the first estimates of the viability constants for eight wild species; their KE and CW value ranges overlapped with those of the crop values. There are a number of possible errors introduced when estimating / predicting seed lifespan, including assumptions about equilibrium moisture contents. Consequently, we recommend to the seed trade and conservationists that seed oil content and sorption properties are determined for each seed lot subjected to seed longevity assessment.
Drier (and cooler) seeds live longer as the rate of chemical reactions contributing to ageing is slowed down when the intracellular viscosity is high. When seeds are stored in the visco-elastic (i.e. glassy) state, perceptible losses in viability may take many months or years. To tie in with one of the seed ageing conditions applied, we equilibrated seeds to 75% relative humidity (and some to 15% RH) and ran the samples in a Dynamic Mechanical Analyser (DMA). Two parameters were measured during rewarming from around -100 °C to 80 °C: storage modulus (E’) and tan-delta (damping). A decrease in E’ indicates that the material is becoming less rigid, and tan-delta provides greater resolution of molecular relaxation events. One of the challenges of applying DMA to seeds is the precise orientation of the sample, its architecture and size. Another is that the interpretation of such traces is clearly complex. A comparison of thermal scans for six wild species / subspecies confirmed that seeds aged at 40 °C and 75% relative humidity are in the non-glassy state, explaining their rapid loss of viability under these conditions.
VALIDATION OF MARKERS IDENTIFIED IN MODEL PLANTS AND CROP ON WILD SPECIES
The aim was to identify marker molecules and genes that influence seed quality in response to abiotic stress during seed development. These markers / genes were selected based on the results of the analyses in other work packages and used for the validation in a range of wild species.
Firstly, Activity-Based Protein Profiling (ABPP) was used for evaluating seed quality. ABPP relies on the use of small, reporter-tagged probes that specifically and covalently react with the active site of an enzyme. These probes are designed to specifically bind certain enzymes (such as lipases, proteases, glycosidases or phosphatases). We used two probes and found that they could be used as a marker to evaluate seed quality of wild species, but these need to be optimized depending on the species.
Secondly, high magnetic-field electron paramagnetic resonance spectroscopy (HF-EPR) was used to identify and characterize radicals and transition metal ions in seeds of wild species of Helianthus and Hordeum. Whether the variations in transition metal ions and radicals in seeds of diverse species reflects growth conditions or species remains to be elucidated.
Thirdly, the presence of abscisic acid and nitric oxide signal transduction pathways transcription factors were assessed during seed germination and seedling establishment. Eight wild species were tested by immunoblot analysis of protein levels in seed extracts from dry and imbibed seeds, with variable detection levels. In summary, we were able to identify proteins and radicals in both the crop and related wild species, some of which have the potential to be used for seed quality assessment.
TRAIT CHARACTERISTICS AND DOWNSTREAM RISKS TO AGRICULTURAL EFFICIENCY AND WILD SPECIES CONSERVATION
Based on our understanding of how the environment might impact on seed quality traits, we aimed to develop for the seed trade and conservationists a preliminary climate-related model that predicts the risks to food security and to wild species survival, including through natural seed regeneration.
The wild species seed phenotypes were compared with germination and storage responses determined on the main crops of interest in WP2 to develop a model / matrix of risks associated with anticipated changes in climate. The scale of this risk depends on many factors, including how success is measured, because seeds reaching early germination do not always convert to seedlings, and this step can be temperature- and water potential-dependent. Generally, warmer conditions will result in quicker germination for all seed lots assuming that water is not limiting. Under conditions of water restriction, germination of some wild Helianthus species and selected lines of the Brassica crop will be affected the greatest. Although not always the case, the crops generally have either similar or shorter thermal, hydro- or hydro-thermal times for germination compared with the wild species; meaning that the crop seeds may be able to cope better with erratic rainfall. Generally, higher seed mass of the crops may buffer environmental extremes by being able to emerge from deeper in the soil. The maternal environment of the wild species, inferred from historical climate data for these long-term stored seeds, also correlated with numerous seed germination traits. Concerning lifespan, the higher temperature regime in the Brassica crop and the drought regime in Helianthus crop affected seed oil content, with potential implications for the equilibrium water content in seed storage (in agricultural or conservation settings). Potential downstream effects of the maternal environment on seed lifespan were also evident. Some treatments (temperature or drought regime) to the crops and Arabidopsis during seed development changed the initial quality (total germination) of the seed lot with implications for subsequent half-lives (P50s). All aspects of seed processing (field to store) should be optimised, so as to improve retention of seed quality in the agricultural setting (breeding programmes, native seed industry) or to meet conservation objectives. In this context, the benefits to lifespan of drying below c. 45-50% relative humidity (c. 5 – 10 % moisture contents) often associated with commercially available seed lots was very evident.
Paper 5 by Dürr, Dickie, Yang and Pritchard (Agricultural and Forest Meteorology, 2015; http://www.sciencedirect.com/science/article/pii/S0168192314002482) reported on the changes of critical temperature and water potential values for the germination of species worldwide. The paper contributes to a seed trait database, providing a wide spectrum of species' seed germination traits, in the form of a set of parameter values describing germination responses to variations in temperature and water potential. Major differences in germination traits were seen to depend on the climatic conditions where the species grow or originated, with some species able to germinate on ice and others unable to germinate below 18 °C. By contrast, within the different plant groups, similar ranges of threshold values were found, linked to the species geo-climatic origin. Crops germinate faster, however, their range of threshold temperatures and water potential values is wider, and some crops have higher optimum and maximum temperatures as well as lower water potential threshold values. The collected data also form a valuable database, enabling plant establishment to be better taken into account in modelling and simulation studies of vegetation boundaries (wild or cultivated) under changing land-use and climate.
Potential Impact:
THE IMPORTANCE OF PLANT SEEDS TO HUMAN AND ANIMAL NUTRITION
Plants and hence seeds are the basis of life on earth. Most human and animal food is derived directly or indirectly from seeds. Crop production starts with the sowing of high quality seeds and generally concludes with seeds that are stored for human food or animal feed production, or for plant propagation and breeding. Hence, seed quality underpins European agriculture and food security. Equally important to European society are the seeds of wild species that fulfil a pivotal role in the maintenance and evolution of natural ecosystems. The seeds of wild species serve as barometers of environmental change and are a source of adaptive functional traits in breeding. Temperature and water availability as seeds develop on the parent plants, i.e. "the maternal environment", impact on quality traits, such as seed dormancy, longevity, germination and vigour and successful seedling establishment by mechanisms that are not fully understood. EcoSeed aimed at providing a step change in our knowledge of how seed quality traits are modified by the environment in crop and wild species, creating new opportunities for increased sustainability of European agriculture and strengthen the competitiveness of the European knowledge-based bio-economy.
Seed quality inexorably changes with time whether seeds are stored in gene banks - as a "genetic insurance" against losses of species and cultivars - or in the natural environment that represents the "genetic memory" of our landscapes. Currently, considerable economic losses result from sub-optimal seed performance. The resulting seed wastage undermines food security and livelihoods, and losses may worsen as climate change is predicted to intensify abiotic (and concomitant biotic) stress events that negatively impact on the seed quality traits of vigour and longevity. The Inter-governmental Panel on Climate Change predicts that more frequent extreme weather events in Europe will have profound implications for agricultural efficiencies (decreasing seed yields). These will be realised through reduced total harvest and lower individual seed mass; and smaller seeds have lower vigour and produce weaker plants. One estimate of lost yield from major cereals from rising temperatures between 1981 and 2002 is $5 billion per year. It is known that the seed longevity of model, crop and wild species is species- and genotype-specific and responsive to pre- and post-harvest environments. Consequently, subtle differences in the quality of seeds at harvest are known to have profound downstream implications for successful storage for conservation and agricultural use.
Furthermore, sub-optimal conditions of post-harvest storage in both large scale agriculture or by SMEs can accelerate seed ageing resulting in additional losses in vigour and viability that further undermine food security. Post-harvest seed losses can account for c. 5-25% of grains in just the first few months of storage (http://www.phlosses.net/) representing billions of seeds wasted and revenue lost in inventory. Whilst most of these losses are on-farm, the seed trade also uses suboptimal storage conditions, as evidenced by information provided by the laboratories of the International Seed Testing Association (ISTA).
In addition, paper 16 by Foyer et al. pointed out that the current lack of coordinated focus on grain legumes has compromised human health, nutritional security and sustainable food production.
A greater understanding of the factors that control seed quality traits is therefore essential to maintain the competitiveness of EU agriculture, to enhance resilience through advanced breeding programmes and to provide guidelines for optimal seed production, treatment and storage. Classical genetic and physiological studies on seed traits, particularly dormancy, have focused largely on spatial and temporal changes in plant hormones (e.g. "the hormonal balance theory"). The EcoSeed consortium characterised the upstream signalling hub that mediates seed sensitivity to the environment, including the development of vigour, longevity-ageing and dormancy traits. There is now compelling evidence that such upstream signalling is modulated to a large extent by the internal redox (oxidative) environment.
An unfortunate consequence of "globalisation" is the disappearance from mainstream agriculture of the old landraces and crop wild relatives that have, over millennia, adapted to adverse environmental conditions. Recognising that the complete loss of this genetic variation will impact on future crop production, about 100 years ago plant specialists initiated collecting missions to safeguard plant genetic resources (PGR) in ex situ collections (seed banks or field gene banks). Presently 7.4 million PGR accessions are available for food and agriculture worldwide (FAO, 2010). The German ex situ genebank at the IPK in Gatersleben holds the largest and most diverse collection for cultivated crops within the EU with >150,000 accessions from 890 genera and 3,032 species (http://www.ipk-gatersleben.de/). The Royal Botanic Gardens, Kew (UK) holds the largest and most diverse collection of wild plant seeds globally and is an internationally acknowledged Centre of Excellence for wild species conservation, with unparalleled expertise in collection management and seed science. The ex situ seed collections exceed 50,000 across >30,000 species, including species in the target genera of the EcoSeed project: Hordeum, Helianthus, Brassica and Arabidopsis. EcoSeed was thus well positioned to take full advantage of these resources. By bringing together the skill base and unique collections of these two world class institutions to support the scientific discovery programme on seed traits, EcoSeed aimed at supporting 1) longer-term demands of the EU agricultural industry; and 2) conservation objectives of the EU as signatory countries to the Convention on Biological Diversity (CBD).
MAIN DISSEMINATION ACTIVITIES AND EXPLOITATION OF RESULTS
One of the EcoSeed Workpackages (WP7) was dedicated to support dissemination and exploitation of project knowledge and results. This included the publication of project results in scientific journals, presentations at scientific meetings, transfer of information to other relevant research programs, potential industrial users and agricultural and conservational stakeholders, training of early career scientists and knowledge transfer to the general public.
SETUP AND MAINTENANCE OF A PROJECT WEBSITE: A project website was set up at the start of the project (http://www.uibk.ac.at/botany/ecoseed/). This website served as a direct portal to deliver information to the public. Detailed information is provided on project objectives, the Partners of the consortium and the Workpackages. The members of the Scientific Advisory Board comprising two distinguished scientific experts in seed science research as well as industrial Partners are listed in order to address the needs of the seed industry at all stages of the project. In addition, press releases and public status reports were regularly presented to the public. Furthermore, peer reviewed publications emanating from EcoSeed (24 so far) are listed on the public website. The website also contains a restricted area, based on SharePoint, for EcoSeed members only, enabling efficient project management and data exchange between the Partners. The website was continually updated throughout the project lifetime and is still active.
EXPLOITATION FACILITATION: Establishment of industrial contacts: A list of relevant industrial, conservation and agricultural end-users was compiled. To aid communication targeted to the different needs of the various stakeholders, contacts were established with national seed production and plant breeding associations in Austria, the UK, France, Spain and Germany as well as the European umbrella organisation, the European Seed Association, and their permission obtained to add on the public EcoSeed website a link to the European Seed Association webpage. In this way, national stakeholders can navigate to their respective associations through the European Seed Association. In the Consortium Agreement, access rights to background made available to all Beneficiaries, and backgrounds excluded from access rights were listed. A document on adequate protection of IP (inventions) was prepared.
KNOWLEDGE TRANSFER: Knowledge transfer to the scientific community and the public: Throughout the project life time, the EcoSeed consortium was, and still is, proactive to transfer knowledge to the scientific community, to agronomic and conservational stakeholders, the public and importantly, to early career scientists. In summary, 9 PhD thesis (4 of which have been defended and 3 are close to completion), 20 MSc theses and 18 BSc thesis were supported by the consortium, 24 papers were published in peer-reviewed journals; 83 oral presentations were delivered at scientific conferences and 57 posters were presented; 11 articles were published in the popular press, 5 press releases and 4 web site contributions were released. In addition, 11 popular lectures were given to the wider public, 4 scientific conferences were organised, 3 interviews were given to the international press, 20 activities were dedicated to knowledge transfer and 9 exchanges visits of early career scientists were supported.
Specifically, four international conferences/workshops were organized by project partners including the Seed Longevity Workshop of the International Society for Seed Science (ISSS) held in Wernigerode (Germany), 5 – 8 July, 2015. This workshop was jointly organized by partner 3 (IPK), 1 (UIBK), 2 (RBGK), 5 (MPIPZ) and 6 (Warwick). The workshop was attended by 212 participants from 44 countries. The consortium is also happy to report that with 140 oral and poster presentations at numerous international scientific conferences EcoSeed knowledge was extensively transferred to the international scientific community. In addition, publications in peer-reviewed journals are indicators for research excellence, and work published so far included papers in Nature Communications, Nature Plants, PNAS, New Phytologist and other scientific journals with high impact factors.
During the period of the project three Public status reports were published on the EcoSeed webpage in 2014, 2015 and 2016, and also on ResearchGate. A project leaflet was displayed by the partners attending conferences, during public days, etc., to showcase and advertise the EcoSeed project. Furthermore, a compilation of press releases is shown on the public EcoSeed website.
Links to international seed science organisations: Strong links between the EcoSeed consortium and the two most important international seed science organisations existed prior to project start, and were further strengthened: the International Seed Science Society (ISSS) - a mainly scientific organisation - and the International Seed Testing Association (ISTA) - dedicated to the transfer of knowledge from science to the seed trade and plant breeders. Links to the ISSS and ISTA websites are available on the public EcoSeed website.
Knowledge transfer to agronomic and conservational stakeholders: EcoSeed knowledge was transferred 1) through the ISTA to agronomic stakeholders; 2) to members of the Scientific Advisory Board from the seed companies, including one SME (Syngenta, ENZA Zaaden and Gautier Semences) and Partner 11 (LIMAGRAIN); and 3) by contacting national and international stakeholders as mentioned earlier.
KNOWLEDGE TRANSFER TO EARLY CAREER SCIENTISTS: Led by 11 Principal Investigators (all senior researchers or University Professors) a total workforce of 127 researchers (head count basis), 79 of which were female, contributed to the EcoSeed project. The consortium encouraged short-term exchange visits of early-career scientists between the project Partners in order to foster their technological knowledge and aid their personal development. In total 9 early career scientists (PhD students and Postdocs) received such training in the Partner labs during short-term visits. Of the 34 experienced researchers (postdoctoral level) contributing to EcoSeed 5 Postdocs were funded by EcoSeed as well as 9 PhD students and 1 work experience student. In addition, training by the EcoSeed consortium was received by 4 further postdocs, 23 MSc students, 18 BSc students and 7 work experience students (e.g. Erasmus students). The consortium also gratefully acknowledges excellent support by 28 technicians and engineers (one of which was fully funded by EcoSeed) and 2 volunteers.
List of Websites:
http://www.uibk.ac.at/botany/ecoseed/home/
Beneficiary 1 (Coordinator): University of Innsbruck (UIBK), Austria. Contact: Prof. Ilse KRANNER, University of Innsbruck, Department of Botany, Sternwartestraße 15, A-6020 Innsbruck, Austria; Tel: +43-0512-507-51035; Fax: +43-0512-507-2715; E-mail: ilse.kranner@uibk.ac.at
Beneficiary 2: Royal Botanic Gardens, Kew (RBGK), UK. Contact: Prof. Hugh W. PRITCHARD, h.pritchard@kew.org
Beneficiary 3: Leibniz-Institut für Pflanzengenetik und Kulturpflanzenforschung, IPK Gatersleben (IPK), Germany. Contact: Dr. Andreas BÖRNER, boerner@ipk-gatersleben.de
Beneficiary 4: Université Pierre et Marie Curie (UPMC), France. Contact: Prof. Christophe BAILLY, christophe.bailly@upmc.fr
Beneficiary 5: Max Planck Institute for Plant Breeding Research (MPIPZ), Germany. Contact: Dr. Wim SOPPE soppe@mpipz.mpg.de
Beneficiary 6: Warwick University (WARWICK), UK. Contact: Prof. William FINCH-SAVAGE, Bill.Finch-Savage@warwick.ac.uk
Beneficiary 7: Institute National de la Recherche Agronomique (INRA), France. Contact: Dr. Loic RAJJOU, loic.rajjou@versailles.inra.fr
Beneficiary 8: University of Leeds (ULEED), UK. Contact: Prof. Christine FOYER, C.Foyer@leeds.ac.uk
Beneficiary 9: Universidad de Salamanca (USAL), Spain. Contact: Prof. Oscar LORENZO-SANCHEZ, oslo@usal.es
Beneficiary 10: Commissariat à l’énergie atomique et aux énergies alternatives (CEA), France. Contact: Anja KRIEGER-LISZKAY, anja.krieger-liszkay@cea.fr
Beneficiary 11: Limagrain Europe: Limagrain (Limagrain), France. Contact: Philippe CAYREL, philippe.cayrel@limagrain.com
PROJECT CONTEXT AND OBJECTIVES
Seed quality is of paramount importance to agriculture, food security and the conservation of wild plant species. Considerable economic losses result from sub-optimal seed performance, undermining food security and livelihoods. Seed quality is strongly influenced by the environmental stresses experienced by the mother plant. Climate change will further exacerbate economic losses and decrease the predictability of seed yield and quality for the farmer. The imminent challenges of climate change and food security require new knowledge of how stress impacts on seed quality, as well as a re-appraisal of optimal seed storage conditions. The EcoSeed project addressed these challenges by bringing together a group of distinguished European experts in seed science to characterise seed quality and resilience to environmental stress factors. State-of-the-art "omics", epigenetics, and post-"omics" approaches were brought together with the ambition to unravel the molecular switchboards that determine seed fate from development, through storage, germination and seedling development. Three crop plants were studied, Hordeum vulgare (barley), Brassica oleracea (B. oleracea) and Helianthus annuus (sunflower) as well as the model plant Arabidopsis thaliana (Arabidopsis), also with the aim to transfer knowledge gained from studying the model and crop plants to wild relatives in their genera, to the benefit of agriculture and conservation alike.
MAIN RESULTS ACHIEVED
Model and crop plants were subjected to environmental stresses such as those predicted in climate change scenarios, i.e. temperature stress and drought stress, and the impact of these stresses on seed quality were assessed. In the first year (2013), the consortium conducted pilot experiments to develop and optimize experimental procedures, and produced most of the biological material needed for molecular analyses. In 2014, key seed quality traits were assessed in the different seed lots and the first molecular analyses were conducted on selected samples. The year 2015 was mainly dedicated to the detailed molecular analysis of seeds selected after further in-depth seed phenotyping. Lab work was completed by the end of 2016. Importantly, it was found that drought and temperature stress applied to the mother plant impacted on seed yield, mean seed size and quality, including dormancy and vigour, in a species- and genotype-specific manner.
FINAL RESULTS AND THEIR POTENTIAL IMPACT AND USE, INCLUDING THE SOCIO-ECONOMIC IMPACT AND THE WIDER SOCIETAL IMPLICATIONS OF THE PROJECT
Novel markers for seed quality and information generated in this project will assist plant breeders, the seed trade and conservationists alike. In this way, the EcoSeed consortium intends to help solve problems related to seed quality and storability, and also play a leading role in enabling associated industries to better capture current and emerging markets. EcoSeed knowledge is transferred to the scientific community through peer-reviewed publications. As yet, 24 papers were published in scientific journals, including in Nature Communications, Nature Plants and PNAS, although the main results are yet to be published. Knowledge is also transferred to the public through the internet, the press and other activities such as YouTube videos, Podcasts and an EcoSeed Flyer. The consortium also recognized the importance of training and knowledge transfer to, and support of early career scientists. Nine PhD students and five Postdocs (on a head count basis) were employed by EcoSeed, and another 52 students (MSc, BSc, work experience and Erasmus exchange students) were trained by the consortium. Scientific exchange between partner labs was strongly encouraged, and 10 short scientific exchange visits of early career scientists were supported by the consortium.
Project Context and Objectives:
PROJECT CONTEXT
Seeds are pivotal to agricultural productivity and ecosystem conservation, yet there remain substantial gaps in our understanding of the critical role that the environment plays during seed development and storage, and its effect on seed quality. The EcoSeed project addressed this important knowledge gap by bringing together 11 EU Partners with well-established track records in seed biology and converging sciences. The consortium adopted an integrated approach with the aim to ultimately help to provide improved food security in times of increased population growth, greater pressure on cultivated land and changing climates. The EcoSeed partners aimed to deliver a step change in our understanding of seeds in relation to the environment, providing a basis for improving seed quality traits and seed handling protocols required for ensuring the availability of future plant genetic resources across Europe. Research focussed on the two major stresses, drought and elevated temperature, that are predicted to have a major impact on European crop production as a result of global climate change. Another ambition of the consortium was to provide young world class researchers with the skills needed to address the significant global challenges of food security for a population expected to reach 9.7 billion in 2050.
State-of-the-art scientific techniques were used to identify the important molecular regulatory pathways and key cellular switches that mediate seed environmental sensing and signalling, and their importance for downstream seed quality and seedling establishment. The intention was to provide a framework for the assessment of environmental effects on seed performance, urgently needed for improving agricultural production and for optimising conservation strategies. "Omics" techniques, epigenetic and post-"omics" approaches such as specific analysis of DNA repair and post-translational modifications (PTMs) of proteins, were used to define the signalling pathways that underpin the seed phenome. Special emphasis was put on redox regulation and signalling involving reactive oxygen species (ROS), and the interactions of the redox signalling hub that influence quality traits. The cellular redox hub processes information from metabolism and the environment to regulate plant growth and defence through integration with signalling pathways. Studies into redox regulation are necessary to provide a better understanding of the cross-talk between these signalling pathways that control dormancy, longevity, germination and vigour of model, crop and wild seeds. EcoSeed did not include research on the nutritional properties of seeds, which are the objectives of plant breeding programmes and have no direct relevance for conservation activities. The main focus was on understanding how seed stress incurred during development on the mother plant and during dry post-harvest storage impact on seed quality, germination and seedling establishment.
THE SPECIFIC OBJECTIVES OF ECOSEED WERE
1. to determine how environmental stresses on the mother plant during seed development affect key seed quality traits (WP1);
2. to determine how environmental stresses during seed development influence seed physiology during dry storage and define the molecular switches that trigger downstream biochemical pathways (WP2);
3. to determine the effects of the above-mentioned pre-harvest stresses and storage conditions on subsequent seed germination, and identify the signalling pathways involved (WP3);
4. to identify genes and technologies that contribute to the prediction of seed quality, storage and germination (WP4);
5. to translate the knowledge gained from the study of model and crop plants to wild species and provide a model for seed quality trait dependency on environmental factors as required by the seed trade and conservationists (WP5).
Two additional workpackages were dedicated to project management (WP6) and to the dissemination and exploitation of project results, and training of early career scientists (WP7).
SPECIES SELECTION
The consortium took advantage of the fully sequenced model plant Arabidopsis thaliana (Arabidopsis), representative of an endospermic seed type within the dicotyledonous plants. Several partners provided various Arabidopsis lines, including several mutants suitable for studying the molecular pathways implicated in plant and seed resilience to perturbation.
Responses to the environment gained in Arabidopsis were compared with those of its close relative B. oleracea, representative of vegetable crops.
Hordeum vulgare (barley) is a starch-accumulating, endospermic seed type within the monocotyledonous plants. Barley is one of the four most important cereal crops worldwide, with about one third of the global production of barley (62 million tons) produced in the EU. Due to its importance to agriculture and the brewing industry, barley germination has been extensively investigated and it is therefore a suitable crop model. Partner 3 provided selected varieties chosen from over 23,245 barley accessions.
Helianthus annuus (sunflower) is representative of a non-endospermic, dicotyledonous seed type in which the oily reserves are stored in the cotyledons. Sunflower is the second most important oil crop in the EU economy after Brassica napus (canola). Over seven million tons of sunflower seeds are harvested globally every year (http://faostat.fao.org). Partner 2 provided selected varieties chosen from over 700 wild sunflower accessions. Commercial sunflower seeds were provided by Partner 11.
Trait improvement that enables crops to withstand harsh climatic conditions with altered growing seasons, and the growth of crops on more marginal land is fundamentally dependent on gene mining from wild species. Translational research in WP5 aimed at gaining an in-depth understanding of the underlying physiology of seeds from wild species in the three crop genera, and validating seed quality markers developed for crops in WP4 for use on wild species. Using the unique germplasm collections available through Partners 2 and 3, it was tested if the mechanisms of stress resistance identified in WPs 1-4 are evolutionarily conserved and can be found in crop wild relatives that have adapted to stressful environments. New markers of seed quality could help with the regular monitoring of the quality of seeds in storage and the results could be valuable to provide a better basis for the use of genetic material of wild plants.
TO THE BENEFIT OF LAY READERS, A LIST OF THE MOST IMPORTANT DEFINITIONS IS PROVIDED BELOW:
After-ripening: a naturally occurring process, and also a commonly used method, to alleviate dormancy and to promote germination by dry storage of freshly harvested, mature seeds.
Ageing: the process of seed deterioration during seed storage, whereby vigour and then total germination gradually decrease over time, finally resulting in the death of all seeds. All seeds age and eventually die, whether stored in agricultural seed banks or in gene banks. However, gene bank storage at low temperature (e.g. 20 °C or lower) and low seed moisture content (e.g. 5% moisture content) increases seed longevity, maintaining viability in the range from decades (in short-lived species) to millennia (in long-lived species).
Controlled Deterioration: a method to accelerate seed deterioration, in this report also termed "ageing" for simplicity, by subjecting seeds with defined, relatively low moisture contents (e.g. 10 % seed moisture content) to high temperatures (e.g. 40 °C), resulting in relatively rapid decline in total germination (in the range of weeks to months).
Cultivar: A crop genotype produced by modern plant breeding techniques for selecting plants with desirable characteristics.
Dormancy: the inability of an imbibed seed to germinate under favourable conditions before certain environmental cues (e.g. low temperatures for an extended time) have been received. Seed dormancy controls the timing of seedling establishment by preventing germination of a viable, imbibed seed during (temporary) favourable conditions in an unfavourable season.
Germination: the process by which a quiescent seed develops into a seedling, starting with water uptake and ending with the protrusion of the radicle (the embryonic root) through the surrounding covers in conjunction with cell elongation in the region immediately behind the radicle.
ISSS: International Society for Seed Science.
ISTA: International Seed Testing Association.
Landrace: a local variety of a crop genotype that has adapted to tolerate biotic and abiotic stress factors typical for the natural environment in which it grows, resulting in a high yield stability and an intermediate yield level under a low-input agricultural system.
Non-viable: dead.
Priming: a technique used by the seed trade to enhance seed vigour, involving a transient increase of seed moisture content followed by re-drying and seed storage.
PGR: plant genetic resources.
ROS: reactive oxygen species.
Seed: the product of the double fertilization in angiosperm plants, which results in the formation of a small embryonic plant enclosed by the seed coat. For simplicity, in this report the term "seed" is also used for the caryopses of the Poaceae family (botanically, these are fruits in which the fruit coat, or pericarp, is fused with the thin seed coat) and for de-coated seeds (particularly for sunflower achenes without pericarp).
Seed quality: here defined as "the sum of all seed traits that are acquired from the time of seed development on the mother plant to seed germination". Seed quality determines storability and subsequent performance during germination and also has significant impacts on seedling, and hence crop, establishment.
Viability (adjective viable): capability of a seed to live and germinate (alive).
Vigour: the sum total of those properties of a seed that determine the potential level of activity and performance of the seed during germination and seedling emergence, often best observed under stress conditions.
WP: Workpackage.
Project Results:
WORKPACKAGE 1:
ENVIRONMENTAL AND GENETIC FACTORS INFLUENCING SEED QUALITY
The purpose of WP1 was
1) to characterise the impact of environmental stresses applied during grain filling and maturation on seed physiology and quality;
2) to use high-throughput "-omics" technologies (i.e. transcriptomics, proteomics, metabolomics) and study nuclear size and chromatin compaction, to explore the effects of theses stresses on seed metabolism;
3) to produce sufficient seeds for use in other WPs.
Regarding point 2) we want to point out that vast amounts of –omics data were produced for seeds grown under different environmental conditions (WP1), then subjected to storage and ageing conditions (WP2), followed by imbibition (WP3), but these cannot be described in this final public report that is limited to 25 pages, and only brief summaries are given below.
The plant species studied in WP1 were Arabidopsis thaliana (Arabidopsis), Brassica oleracea (B. oleracea), Hordeum vulgare (barley) and Helianthus annuus (sunflower). In the initial experiments the effects of the maternal environment ("drought stress" and "temperature stress") on plant and seed performance were tested. The stress conditions for further experimentation were chosen according to the following criteria: 1) changes in water availability and temperature should be close to those predicted in climate change scenarios, and 2) the stresses applied to the mother plants should not be too detrimental to the plant or to seed production, so that the requirements to produce sufficient seed for other project tasks could be met. The stress conditions were evaluated for each species. To test the stress factor "temperature", a "control temperature" for good plant growth and seed quality was chosen and compared to a regime with "lowered temperature" (3-5 °C lower than the control temperature) and one with "elevated temperature" (3-5 °C higher than control temperature). "Drought" was simulated by maintaining a defined level of plant water potential in growth cabinets (Arabidopsis, B. oleracea) and by a controlled water deficit in glasshouses (barley) or by withholding irrigation in field experiments (sunflower).
Both stresses applied individually strongly influenced seed yield, so that it was not always possible to produce enough seeds when drought and temperature stress were combined. Therefore, the consortium focused on the effects of the individual stresses and produced seeds of selected genotypes for each species in two or three environmental regimes that provided a good environmental contrast, and also allowed a sufficient yield of seeds for analyses in WPs 1, 2 and 3. Following harvest, the seeds produced from all four species were dried and stored under defined conditions to minimise the impacts of non-experimental variables and changes to physiological behaviour during storage. A detailed analysis of yield characteristics was conducted and seed quality was assessed using a range of species-specific measures. These included speed of germination at appropriate temperature, dormancy characteristics, and germination following controlled deterioration. In general, stress factors in the maternal environment had a negative effect on seed yield, mean seed size and quality.
ARABIDOPSIS
Seeds of the model plant Arabidopsis (Col-0 accession) and a range of mutant genotypes were produced in growth cabinets under three selected temperature regimes, a "control temperature" regime, a "lowered temperature" regime and an "elevated temperature" regime. Measurements of stomatal conductance and photosynthesis were taken to evaluate the stress effects under the lowered and elevated temperature regimes. Plants grown under lower temperatures produced more seeds. However, seeds produced in the lowest temperature regime exhibited a high degree of thermodormancy and seed quality was poorer. Seed performance was similar in those produced in the two other temperature regimes, but the highest temperature regime significantly reduced the numbers of seeds produced.
Furthermore, seeds from a large panel of Arabidopsis mutants were produced and screened for quality traits and seed longevity. These included a range of hormone and dormancy mutants and mutants affected in reactive oxygen species (ROS) metabolism with an assumed role in dormancy regulation. Dormancy assays were performed on freshly harvested dry seeds of all genotypes. From this, 19 mutant genotypes were selected based on their dormancy index and their sensitivity to the growth temperature of the mother plant. These were used to investigate maternal environmental effects with a focus on the influence of plant hormones on primary dormancy. All seeds were more dormant when produced in colder conditions than in warmer conditions. Seed production temperatures impacted on both seed yield and weight. In general, the higher the growth temperature the greater the reduction in seed yield and the higher the seed weight. Nonetheless, seeds from a subset of the mutants showed the opposite behaviour indicating that sensitivity to the maternal environment during seed maturation was modified in these genotypes. Furthermore, growth temperatures affected seed sensitivity to controlled deterioration (CD).
Transcriptomic analyses of the seeds produced in WP1 revealed that the maternal environment affected transcript regulation. The identified differentially expressed transcripts were further analysed by a Gene Ontology analysis. Transcripts differentially expressed in seeds matured under low temperatures corresponded to a lower numbers of Gene Ontology categories than those found in seeds produced under control or elevated temperatures. Differential proteomic analyses using a gel-free shotgun proteomic approach showed that more than 200 proteins were differentially accumulated in seeds produced under the different temperature regimes in at least one pairwise comparison. A Gene Ontology analysis revealed that proteins that were up-accumulated under elevated or lowered temperatures corresponded to relatively distinct biological and molecular functions. Changes in primary metabolites were studied using a gas chromatography-mass spectrometry (GC-MS)-based metabolomics approach, and flavonoid and phytohormone composition were studied using ultra-high performance liquid chromatography-mass spectrometry (UHPLC-MS). Over 100 primary metabolites showed differential accumulation in seeds produced in the different temperature regimes in at least one pairwise comparison. The effects of temperature and drought in the maternal environment on nuclear size and chromatin compaction was investigated to determine whether these characteristics reflected differences in seed quality.
B. OLERACEA
In paper 1, using a trait to gene analysis Morris, Barker, Walley, Lynn and Finch-Savage (New Phytologist, 2016; http://onlinelibrary.wiley.com/doi/10.1111/nph.14102/abstract) reported that allelic variation in three genes determines the key vigour trait of rapid germination in B. oleracea. A mechanism was proposed in which both seed ABA content and ABA sensitivity determines speed of germination.
Paper 2 by Finch-Savage and Bassel (Journal of Experimental Botany, 2016; https://academic.oup.com/jxb/article/67/3/567/2893341/) considers seed vigour at an ecophysiological, molecular, and biomechanical level. This review discusses how some seed characteristics that serve as adaptive responses to the natural environment are not suitable for agriculture. Past domestication has provided incremental improvements, but further actively directed change is required to produce seeds with the characteristics required now and in the future. Furthermore, ways in which basic plant science could be applied to enhance seed performance in crop production are discussed.
Two B. oleracea genotypes,"A12DHd" (Chinese Kale) and "AGSL101", a substitution line originating from a cross between A12DHd and "GDDH33" (var. italica; broccoli calabrese), were used to assess the impact of drought and temperature stress on seed quality. Seeds from both genotypes are non-dormant under benign production conditions, but A12DHd has inherently lower vigour than AGSL101. Seeds were produced in two temperature regimes, a "control" regime and an "elevated temperature" regime in the absence of water stress. Plants grown under control temperature were also subjected to a "drought" regime following seed set.
The genotype effects were generally consistent, with AGSL101 seeds providing higher quality seeds overall and being less affected by adverse seed production environments. Thus, the difference in quality between the genotypes was greater when seeds were produced under stress. Elevated temperature and drought significantly reduced stomatal conductance and the rate of photosynthesis during seed development. A likely consequence of this was that seeds produced from both genotypes in the elevated temperature regime were generally smaller and had poorer quality. However, the effect can differ between genotypes, e.g. there was a positive impact on germination speed of the elevated temperature regime on seeds of the lower vigour genotype. Furthermore, despite a similarly reduced rate of photosynthesis, seed size was hardly affected by water stress compared to that of seeds produced in the elevated temperature regime. This suggests that the adverse effects do not simply result from reduced resources production in the leaves, but requires a more complex explanation. Transcriptome, proteome and metabolome data were also generated for B. oleracea seeds and statistical analyses were started, revealing more significant differences between seeds produced under the control and elevated temperature regimes than between seeds produced under control and drought conditions.
BARLEY
More than 230 spring barley accessions from the collection of the Federal ex situ genebank for agricultural and horticultural plants at the IPK in Gatersleben, Germany, were screened following storage since 1974. From these, two genotypes with contrasting longevity phenotypes, the long-lived genotype "HOR 4710" and the short-lived genotype "HOR 6489", were chosen for seed production and analyses. Seeds were produced in a "control temperature" regime, and compared to an "elevated temperature" regime and a "drought stress" regime (the latter at control temperature), applied 7 days after flowering.
Barley is a rather stress tolerant crop grown in wide ranges of latitude and altitude, and the maternal environment had no strong effects on seed yield and germination performance, but in both genotypes the "thousand seed weight" was strongly affected by drought, and to a lesser extent by elevated temperature. However, elevating the temperature in the maternal environment reduced thermodormancy. Furthermore, no strong effects of the maternal environment were found on seed performance after CD. Isolated embryos of the produced barley seeds were submitted to transcriptomic, proteomic and metabolomic analyses. Compared to seeds produced under control conditions, those produced under drought showed pronounced molecular and biochemical rearrangements in both genotypes, whereas the elevated temperature regime affected genotype HOR4710 more than HOR2110.
SUNFLOWER
Sunflower seeds from five genotypes differing in oil content, dormancy status and earliness were produced in the field under commercial conditions in Marchena, Andalusia, Spain. At this location, there is almost no rain between June and September. "Drought" was applied during grain filling according to the following protocol: plants were irrigated five times before grain filling, and then irrigation was stopped. For comparison, seeds were produced under "control" conditions by applying two further irrigations during grain filling. The five genotypes tested did not show consistent responses to drought stress. The most pronounced effect was found in the seeds of genotype B (a semi-early linolenic hybrid poor in oil), whereby drought stress during seed maturation led to increased germination at 15 °C. Interestingly, this variety also had greater resistance to ageing, and this genotype was chosen for further investigations.
WORKPACKAGE 2:
BIOCHEMICAL AND BIOPHYSICAL CHANGES IN STORED SEEDS THAT AFFECT SEED QUALITY
In WP2 the seeds of Arabidopsis, B. oleracea, barley and sunflower produced in WP1 were used to evaluate the effects of the maternal environment on seed after-ripening and storability, and to study the associated molecular mechanisms. Seeds produced in WP1 grown in the control, the lowered, the elevated temperature and the drought regime were aged using controlled deterioration (CD) for different time intervals. The seeds were also tested for dormancy. In addition, seeds produced in WP1 were stored under agricultural conditions (by Advisory Board member Syngenta) and seed bank conditions (by Partner 3) so that they are available for future studies into the effects of the maternal environment on long-term storage.
THE EFFECTS OF MATERNAL AND STORAGE ENVIRONMENTS ON SEED RESPONSE TO CD
Seeds of Arabidopsis and B. oleracea, but not barley, produced under elevated temperatures were more susceptible to CD (i.e. seed viability and vigour were more negatively affected) compared to seeds grown in the control regime. Arabidopsis seeds grown in the lowered temperature regime had higher levels of thermodormancy, which was released upon CD. Seeds of sunflower and B. oleracea produced under drought stress responded to CD similarly as the seeds grown in the control regime. In sunflower, seed dormancy was not affected by drought stress. To assess the impact of oxygen upon seed storage, seeds were aged under ambient oxygen, termed “normoxia” (around 21% of oxygen), under elevated oxygen (>70% of oxygen) and under low oxygen, termed “hypoxia” (less than 3% of oxygen). For hypoxic conditions seeds were kept in airtight boxes flushed daily with gaseous nitrogen to maintain an oxygen level below 3%.
OXIDATIVE PROCESSES IN SEED AFTER-RIPENING AND LONGEVITY
The combined effects of the maternal and storage conditions on ROS production and antioxidant response were assessed in selected samples of interest prior to and during CD. In the seeds of all four plant species, seed ageing was generally associated with an accumulation of ROS, such as hydrogen peroxide and superoxide, and a breakdown of antioxidant compounds. Different organic radicals were found in the species investigated (reported in WP4), but no significant changes in either organic free radical chemistry or abundance were found in response to the maternal environment upon seed maturation or CD at 75% relative humidity and 40 °C. Changes in the maternal environment affected antioxidant and pro-oxidant status in the seeds of all species. However, the relationship between ROS metabolism, cellular redox state and seed germinability appeared to be, in part, species-dependent. The most consistent results across the species studied were found for the ROS hydrogen peroxide and the water-soluble antioxidant glutathione, which were significantly altered in response to the maternal environment and ageing. For example, the drought regime resulted in a higher concentration of glutathione in seeds of both barley and sunflower, and both species did not appear to display any signs of oxidative stress as a consequence of the maternal seed production environment. The composition of the lipid-soluble tocochromanols (the “Vitamin E family”) differed greatly between species. Barley seeds produced in the elevated temperature regime contained higher concentrations of tocochromanols. In Arabidopsis and B. oleracea seeds, the maternal environment affected the concentrations and redox state of glutathione. In B. oleracea, the adverse effects of the elevated temperature regime on seed vigour were associated with an imbalance between ROS and antioxidants, which can cause oxidative stress. In summary, the maternal environment during seed maturation affected redox status, ROS and antioxidants levels of the seeds.
Paper 9 by Morscher, Kranner, Arc, Bailly and Roach (Annals of Botany, 2015) reported on glutathione redox state, tocochromanols, fatty acids, antioxidant enzymes and protein carbonylation associated with after-ripening and ageing in dormant and after-ripened (non-dormant) sunflower embryos subjected to CD under normoxia (21 % oxygen) or elevated oxygen levels (75 % oxygen). After-ripening was accompanied by a shift in the thiol-based cellular redox environment towards more oxidizing conditions. Controlled deterioration under elevated oxygen led to a faster loss of seed dormancy and significant decreases in glutathione reductase and glutathione peroxidase activities, but viability was lost at the same rate as under normoxia. Irrespective of oxygen concentration, the overall thiol-based cellular redox state increased significantly over 21 d of CD to strongly oxidizing conditions and then plateaued, while viability continued to decrease. Viability loss was accompanied by a rapid decrease in glucose-6-phosphate-dehydrogenase, an enzyme that provides NADPH for reductive processes such as required by glutathione reductase. Protein carbonylation, a marker of protein oxidation, increased strongly in deteriorating seeds. The lipid-soluble tocochromanols, dominated by α-tocopherol, and fatty acid profiles remained stable. In conclusion, after-ripening, dormancy-breaking during ageing, and viability loss appeared to be associated with oxidative changes of the cytosolic environment and proteins rather than the lipid environment. High oxygen concentrations accelerated dormancy alleviation but, surprisingly, did not accelerate the rate of viability loss.
Furthermore, paper 17 by Arc, Galland, Godin, Cueff and Rajjou (Frontiers in Plant Science, 2013; http://journal.frontiersin.org/article/10.3389/fpls.2013.00346/full) gives an overview of the role of nitric oxide in the control of seed dormancy and germination in Arabidopsis with a focus on the implications of this radical in the control of seed dormancy and germination, proposing that nitric oxide-dependent post-translational modifications of proteins are key to nitric oxide signalling during early seed germination (WP3).
OXIDATIVE MODIFICATIONS OF MACROMOLECULES DURING DRY STORAGE
Paper 18 by Meimoun, Mordret, Langlade, Balzergue, Arribat, Bailly, El-Maarouf-Bouteau (PLoS One, 2014; http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0086442) showed that the alleviation of seed dormancy upon dry after-ripening is not related to regulated changes in gene expression. However, oxidative modifications to nucleic acids, proteins and lipids may occur in dry seeds, and are thought to affect seed fate upon seed imbibition (studied in WP3). In WP2, molecular targets of oxidative modification in dry seeds produced in the different maternal environments were identified.
Proteins are rather sensitive to oxidation of certain amino-acid residues, which can be oxidized reversibly or irreversibly. Methionine can be oxidized to methionine sulfoxide which, in turn can be reduced to methionine via the action of methionine sulfoxide reductase. Oxidation of methionine was detected in all protein samples extracted from dry seeds, but was not affected by the production or storage environment. Carbonylation, an irreversible oxidation of proteins, was found at altered levels when conditions in the maternal environment were altered, and upon CD in Arabidopsis and B. oleracea.
The accumulation of DNA single strand breaks (SSB) and double strand breaks (DSB) was quantified by electrophoretic methods and image analysis. For example, in B. oleracea seeds decreasing viability upon CD was associated with significant DNA damage. Moreover, using lipidomics it was demonstrated for Arabidopsis that the temperature during seed maturation had a marked effect on membrane and reserve lipid composition. In conclusion, seed ageing resulted in a more oxidizing cellular environment, although the detection of oxidatively-modified macromolecules depended on the extent of seed ageing and plant species.
MOLECULAR MOBILITY AND WATER SORPTION PROPERTIES DURING DRY STORAGE
It was also determined whether conditions of seed production and storage modulate cellular mobility in the dry state, in order to establish whether a relationship exists between oxidative modifications of biomolecules and molecular mobility. Seed water status (sorption isotherms) were studied using IGAsorp and thermodynamic analyses of water sorption isotherms. Biophysical properties of dry seeds were assessed by DMA (dynamic mechanical analysis) and DSC (Differential Scanning Calorimetry), and oil content was measured using TD-NMR.
Differences in water relations between aged and non-aged seeds of the four species could be explained mainly by variation in seed oil content. Although some isotherm shifts as a result of the maternal environment or ageing could not be fully explained, the consistent fits for the isotherms suggests a general robustness of the IGAsorp and LiCl methods. No obvious impacts of vigour or maternal environment on the glass forming properties of sunflower, B. oleracea or barley seeds were found by DMA. Using DSC, it was shown that the maternal environment had no effects on the glass properties of sunflower and barley seeds. In B. oleracea seeds, the elevated temperature regime slightly modified glass properties but this was not associated with an altered storability upon CD. In sunflower seeds, there was a clear effect of ageing on glass formation, as glass transition temperature appeared to be lower in aged seeds, suggesting that the glass weakened upon CD. No glass transitions were detected in Arabidopsis seeds. In summary, the maternal environment affected some characteristics of molecular mobility and water properties, mainly due to modified seed oil content, and significant effects were found upon CD, which induced changes in water sorption isotherms (B. oleracea, Arabidopsis) and glass properties (sunflower).
In conclusion, novel and original data were produced related to the downstream effects of the maternal environment upon seed ageing in the dry state. The major outcomes of WP2 were that the temperature regime in the maternal environment can modify dormancy alleviation and longevity, whereas drought stress had less clear effects. Furthermore, the maternal environment affected ROS homeostasis and related oxidative processes. In all species, seed ageing was associated with oxidative damage to proteins and DNA.
WORK PACKAGE 3:
MOLECULAR CHANGES DURING IMBIBITION AND GERMINATION
WP3 was dedicated to characterizing how the maternal environment (WP1) followed by post-harvest dry storage (WP2) affects seed germination. High-throughput technologies were used to capture the intricate molecular processes that control seed quality, including dormancy, vigour and longevity. Dry seeds (from WP1 and WP2) were compared with imbibed seeds.
FUNCTIONAL GENOMICS AND METABOLOMICS ASSESSED BY HIGH-THROUGHPUT TECHNOLOGIES
ARABIDOPSIS
Changes in gene expression products (mRNAs and proteins) and low-molecular-weight metabolites of seeds produced under the three temperature regimes were analysed in dry seeds and imbibed seeds. Transcriptomic analyses revealed gene sets displaying different tendencies of transcript regulation in response to the temperature regimes upon seed maturation. The differentially expressed transcripts identified were further analysed by Gene Ontology analysis. Although several Gene Ontology categories were shared between treatments, the elevated temperature regime seemed to enhance stress responses. To characterize the impact of the maternal environment on the seed proteome during germination, a shotgun proteomics approach was conducted allowing us to quantify approximately 1000 individual proteins. For each temperature regime, statistical analyses revealed proteins that were significantly up- or down-accumulated upon imbibition. Changes in the maternal environment also resulted in the differential accumulation of metabolites in dry and imbibed seeds, in accordance with transcriptome and proteome results. Integrative statistics were used to study the complex interactions within transcriptome, proteome and metabolome datasets. The selected variables (transcripts, proteins and metabolites) revealed connexions (correlations) within and between datasets based on the accumulation levels for each variable, highlighting predictive biomarkers of seed quality in Arabidopsis.
Paper 10 by Galland, Huguet, Arc, Cueff, Job, Rajjou (Molecular & Cellular Proteomics, 2014; http://www.mcponline.org/content/13/1/252.short) provides a detailed description of the dynamics of protein synthesis during the time course of germination, demonstrating that mRNA translation is both sequential and selective during this process. The data confirms that during early imbibition, the Arabidopsis translatome keeps reflecting an embryonic maturation program until a certain developmental checkpoint.
Paper 13, an opinion article by Galland and Rajjou (Frontiers in Plant Science, 2015; http://journal.frontiersin.org/article/10.3389/fpls.2015.00284/full) discusses the regulation of mRNA translation, which controls seed germination and is critical for seedling vigour.
Paper 23 by Basbouss-Serhal, Pateyron, Cochet, Leymarie and Bailly (Plant Physiology, 2017; http://dx.doi.org/10.1104/pp.16.01933) shows that the regulation of germination by dormancy in Arabidopsis seeds depends on the specific decay of positive and negative regulators of this complex process. It was demonstrated that 5’ to 3’ mRNA degradation, mRNA secondary structure and specific motifs are implicated in the selectivity of the decay. This work also allowed to reveal several new regulators of seed dormancy and germination.
B. OLERACEA
As described for Arabidopsis, high-throughput omics analyses were also conducted for the seeds of the two B. oleracea genotypes, A12DHd and AGSL101, produced in the control, the elevated temperature and the drought regimes. Again, dry seeds (from WP1) were compared to imbibed seeds. Correlation networks were established between proteins and metabolites revealing major correlations between proteome and metabolome levels as a fingerprint of seed quality in B. oleracea.
BARLEY
As for Arabidopsis and B. oleracea, high-throughput omics analyses were conducted for barley seeds produced in the control, the elevated temperature and the drought regimes. Again, dry seeds (WP1) were compared to imbibed seeds. Metabolomics experiments were conducted on the dry seeds of both barley genotypes initially selected (HOR4710 and HOR2110) and produced under different conditions. The metabolome of HOR2110 seeds was less affected by stress than the metabolome of HOR4710 seeds. Due to the low sensibility of HOR2110 to both drought and heat stress, it was decided to conduct detailed transcriptomic, proteomic and metabolomic analyses only for the HOR4710 genotype. Quantitative and qualitative RNAseq and mass spectrometry data were generated and the statistical analyses are in progress.
SUNFLOWER
Sunflower seeds were produced in the field under control and drought conditions during grain filling, and analysed by high-throughput methods as described for the other three species.
Furthermore, paper 6 by Layat, Leymarie, El-Maarouf-Bouteau, Caius, Langlade and Bailly (New Phytologist, 2014; http://onlinelibrary.wiley.com/doi/10.1111/nph.13002/full) highlighted the importance of post-transcriptional regulation of germination. The data suggests that seed germination requires timely regulated and selective recruitment of mRNAs into polysomes. The polysome-associated mRNAs, that is, the translatome, were fractionated and characterized with microarrays in dormant and nondormant sunflower embryos during imbibition at 10°C. The proteins corresponding to the polysomal mRNAs in nondormant embryos were involved mainly in transport, regulation of transcription and cell wall modifications.
ROLE OF QUINONE REDUCTASES IN THE CONTROL OF ROS LEVELS DURING GERMINATION
Paper 21 by Biniek, Heyno, Kruk, Sparla, Trost and Krieger-Liszkay (Planta, 2016; http://link.springer.com/article/10.1007%2Fs00425-016-2643-y) showed that a NAD(P)H quinone oxidoreductase, NQR, and the cytochrome b AIR12 are involved in controlling superoxide generation at the plasma membrane. Two proteins which are important to control ROS production at the level of the apoplast were identified. NAD(P)H:quinone oxidoreductase (NQR) and cytochrome b (called AIR12 for Auxin Induced in Root culture) are two proteins attached to the plant plasma membrane which may be important for generating and controlling ROS levels in the apoplast. AIR12 is a single gene in Arabidopsis that codes for a mono-heme cytochrome b. The NQR is a two electron-transferring flavoenzyme that contributes to the generation of superoxide in isolated plasma membranes. Arabidopsis double knockout plants of both NQR and AIR12 generated more superoxide and germinated faster than the single mutants affected in AIR12 or in NQR. An electron transfer catalyzed by NQR from cytosolic NAD(P)H to a quinone located in the plasma membrane was demonstrated. The quinol (reduced state of the quinone) is then oxidized by AIR12. In addition, the nature of the quinones in plant plasma membranes was identified and quantified by HPLC. Biochemical analysis showed that purified plasma membranes contained phylloquinone and menaquinone-4 as redox carriers. This is the first report on the occurrence of menaquinone-4 in eukaryotic photosynthetic organisms. It can be concluded from the mutant studies and the in vitro experiments that both NQR and AIR12 play an important role in controlling ROS production at the plasma membrane. Absence of both proteins increased the superoxide level, which may be responsible for the observed stimulation of germination.
ROLE OF ASCORBATE AND NUCLEAR GLUTATHIONE SALICIL ACID-TRANSFERASES (GSTS) AND GLUTAREDOXINS (GRXS) IN CONTROLLING ROS-LEVELS DURING GERMINATION
Redox metabolism and signalling in the emerging radicles of germinating Arabidopsis seeds were studied.
Paper 24 by De Simone, Dong, Vivancos and Foyer (Molecular Physiology and Ecophysiology of Sulfur, 2015; http://link.springer.com/chapter/10.1007%2F978-3-319-20137-5_4) reported on the distribution of the antioxidant glutathione between the nucleus and the cytosol. The glutathione redox potential was similar in the nucleus and cytosol of developing radicles of Arabidopsis seeds after germination. However, in the arrested embryonic root meristem of the root meristemless 1 (rml1) mutant that have less than 5 % GSH of the wild-type, glutathione was predominantly localised in the nuclei. This was also the case in wild-type roots treated with the auxin transport inhibitor, N-1-napthylphthalamic acid, which have decreased root glutathione levels.
Furthermore, paper 20 by Diaz-Vivancos, de Simone, Kiddle and Foyer (Free Radical Biology and Medicine, 2015; http://www.sciencedirect.com/science/article/pii/S0891584915006085) considered the respective roles of ROS and glutathione in the regulation of plant growth, with a particular focus on the regulation of the plant cell cycle. Glutathione, in addition to its roles as a crucial redox buffer, appears to be a flexible regulator of genetic and epigenetic functions. It is proposed that regulated ROS production linked to glutathione-mediated signalling events are the hallmark of viable cells within a changing and challenging environment.
REPAIR MECHANISMS
Paper 12 by Sano, Rajjou, North, Debeaujon, Marion-Poll and Seo (Plant and Cell Physiology, 2016; https://academic.oup.com/pcp/article/57/4/660/2460931/Staying-Alive-Molecular-Aspects-of-Seed-Longevity) reviews molecular aspects of seed longevity. Reduction of seed longevity is often associated with oxidation of cellular macromolecules such as nucleic acids, proteins and lipids. Seeds possess two main strategies to combat these stressful conditions: protection and repair. The protective mechanism includes the formation of glassy cytoplasm to reduce cellular metabolic activities and the production of antioxidants that prevent accumulation of oxidized macromolecules during seed storage (see WP2). The repair system removes damage accumulated in DNA, RNA and proteins upon seed imbibition through enzymes such as DNA glycosylase and methionine sulfoxide reductase.
Paper 7 by Waterworth, Bray and West (Journal of Experimental Botany, 2015; https://academic.oup.com/jxb/article/66/12/3549/524229/The-importance-of-safeguarding-genome-integrity-in) on the importance of safeguarding genome integrity in germination and seed longevity reviewed recent advances that have established molecular links between genome integrity and seed quality. These studies identified that maintenance of genome integrity is particularly important to the seed stage of the plant lifecycle, revealing new insight into the physiological roles of plant DNA repair and recombination mechanisms.
Paper 4 by Waterworth, Footitt, Bray, Finch-Savage and West (PNAS, 2016; http://www.pnas.org/content/113/34/9647.abstract) demonstrated physiological functions for DNA damage checkpoint kinases (ATM and ATR) in linking genome integrity to germination, thereby influencing seed quality, crucial for plant survival in the natural environment and sustainable crop production.
WORKPACKAGE 4:
BIOTECHNOLOGY AND INNOVATIVE TOOLS
The goals of WP4 were to identify genes and technologies that enable a prediction of seed quality in relation to stress factors experienced in the maternal environment. Some of these genes were identified based on the -omics studies from WP1 and WP3. Additional genes were obtained from studies of natural variation, using natural modifiers, Quantitative Trait Loci (QTL) analysis and Genome Wide Association Study (GWAS) in Arabidopsis, B. oleracea and barley. In addition, seed quality genes were identified by studying the transcriptional network in seeds. Several known genes involved in seed quality were analysed for their tissue-specific roles. Finally, novel techniques like Activity-Based Protein Profiling (ABPP) and High-Field Electron Paramagnetic Resonance (HF-EPR) were investigated for their potential to provide easily scorable markers for seed quality.
IDENTIFICATION OF NOVEL GENES DETERMINING SEED QUALITY
Genes that showed a response to changed temperatures during seed development in Arabidopsis and that have a predicted function indicating potential involvement in seed quality were selected from WP1 and WP3 after combining the datasets produced from transcriptomic and proteomic analyses. These genes participate in different biological processes, such as defence mechanisms, redox regulation or protein and lipid metabolism. Insertion mutants for all selected candidates were obtained from the Arabidopsis Stock Centre. An initial analysis of the seed quality phenotypes of these insertion mutants revealed that most of them showed similar dormancy levels and resistance to CD as wild-type seeds. This suggests that investigated genes could have redundant roles regarding seed quality. However, for a few mutants reduced germination was observed indicating an important role in seed quality. Further analyses will indicate the potential of these genes as markers to predict and as tools to improve seed quality.
The transcriptional network implicated in seed quality was characterized by implementing a high-throughput in-plant expression screening of transcription factors. Key transcription factors involved in abiotic stress tolerance during seed development were identified. Ten independent lines were generated for each transcription factor-open reading frame (ORF) and the transgenic lines were phenotypically characterised in relation to a large number of seed quality-related traits. This screening allowed to identify lines affected in the responses to the plant hormone abscisic acid, salt stress or osmotic stress during seed germination. The results obtained after the analysis of approximately 634 transcription factors, corresponding to 1636 homozygous independent transgenic lines indicated that 13 lines showed insensitive phenotypes to low abscisic acid concentration and 5 lines showed insensitive phenotypes to abscisic acid and abiotic stresses (salt and osmotic stress) compared with the wild-type. Those five transcription factors that showed higher insensitivity to abscisic acid, salt and/or osmotic stress during seed germination and seedling development belonged to different families. The genes selected from the screening are likely to be involved in the mechanism by which different abiotic stresses (i.e. salt and osmotic stress) during seed development influence germination and seedling establishment. Their roles and functions will be further evaluated by the phenotypic analysis of simple and double mutants in these genes.
Furthermore, the impact of the maternal environment on subsequent seed quality was studied in B. oleracea by comparing two genotypes with inherent differences in seed vigour, A12DHd (low vigour) and AGSL101 (high vigour). This study concluded a long-term investigation on the genetic basis of the differences in vigour in these lines by developing a genetic model for the potential regulation of germination. It was demonstrated that allelic variation at three loci influence this key trait. Functional analysis in both B. oleracea and Arabidopsis identified and demonstrated activity of genes at these loci. Two candidate genes were identified at the principle Speed of Germination QTL (SOG1) in B. oleracea. One gene BoLCVIG2 is a homologue of the alternative-splicing (AS) regulator (AtPTB1). The other gene BoLCVIG1 is unknown, but different alleles had different splice forms that were coincident with altered abscisic acid sensitivity. A further QTL, Reduced ABscisic Acid 1 (RABA1), was identified that influenced abscisic acid content and provided evidence that this results from the activity of a homologue of the abscisic acid catabolic gene AtCYP707A2 at this locus (for further details see Paper 1, Morris et al. New Phytologist, 2016). Lines containing beneficial alleles of these three genes had greater seed vigour. This work provides a way forward to understand and manipulate the genetic pathways regulating this key agricultural trait.
Paper 22 by Sechet, Roux, Plessis, Effroy, Frey, Perreau, Biniek, Krieger-Liszkay, Macherel, North, Mireau and Marion-Poll (Molecular Plant, 2015; http://www.sciencedirect.com/science/article/pii/S1674205214000331) showed that the hot ABA-deficiency suppressor2 (has2) mutation increases drought tolerance and the ABA sensitivity of seed germination, in agreement with the above findings.
IDENTIFICATION OF GENES THAT DETERMINE SEED QUALITY IN GENETIC MATERIAL WITH ALTERED LONGEVITY
Previous work identified three natural modifiers from the Arabidopsis Shahdara accession, which were able to enhance seed longevity in a longevity-defective mutant (lec1-3) in the Ler background. Fine-mapping of the strongest modifier led to the identification of a candidate gene explaining the observed enhanced seed longevity phenotype. This candidate gene is involved in the regulation of flowering time, which complicates proper synchronisation of seed maturation periods and harvesting. In order to study seed longevity without the influence of altered flowering time, transgenic plants (which are mutants for the candidate genes) were constructed expressing the coding sequence of a candidate gene under the control of a seed-specific promoter. These transgenic plants showed altered seed longevity without altered flowering time. Further characterisation of these transgenic lines showed that the observed phenotypes perfectly matched transcript levels of the transgene. This work identified an interesting connection between the regulation of flowering time and seed quality which is relevant for further studies and can have practical implications.
Seed longevity loci/genes were also identified and characterised in barley by combining a quantitative genetics approach with -omics technologies. This work is based on a previous Quantitative Trait Loci (QTL) study in recombinant inbred lines (RILs), derived from the short lived Ethiopian spring barley landrace L94 and the long lived Argentinian spring barley landrace Cebada Capa (CC). In total, four putative QTLs for seed longevity were identified and near isogenic lines (NILs) were generated. Using these NILs, the four putative QTLs could be confirmed and further mapping of Cebada Capa introgressions was achieved in the so called L94 NILs using RNA seq. For candidate and downstream target gene identification, a combined transcriptome and proteome analysis was conducted. The analysis of mature, non aged seeds of the two parental lines and the L94 NILs by RNA seq and total seed proteomic profiling identified two putative candidate genes and one possible downstream target gene. To validate these, putative candidate and downstream target genes, T DNA knock out lines of homologous Arabidopsis genes, recently shown to have a seed longevity phenotype, were complemented with the respective Cebada Capa and L94 alleles. Both a candidate and downstream target gene were able to rescue their respective Arabidopsis mutant seed longevity phenotypes when expressed under the control of the 35S or the Ubi promoter. Since one of the identified candidates may affect the redox status in normal and deteriorating seeds, the corresponding Arabidopsis mutant was combined with the redox reporter roGFP2 to visualize the redox potential in seeds using confocal microscopy. First transformant generation (T1) lines, showing strong fluorescence in the cytosol, could be generated in this study and will allow the measurement of the NADP+/NADPH redox potential in future studies. Metabolite profiling, performed on the same genotypes utilized for the transcriptomic and proteomic analyses, identified one monosaccharide sugar and one flavonoid showing a similar trend in abundance in Cebada Capa and the L94 NILs. Sugars have been proposed in seeds to act as signals that regulate and influence seed development. Furthermore, several studies showed that sugars help maintain the structural integrity of membranes and proteins under dry conditions due to the formation of a glassy state to limit deteriorative reactions.
Furthermore, paper 19 by Nagel, Kodde, Pistrick, Mascher, Börner and Groot (Plant Science, 2016; http://journal.frontiersin.org/article/10.3389/fpls.2016.00388/full) analysed the genetic background of seed longevity in barley, revealing a tolerance QTL on Chromosome 5H.
Paper 8 by Nagel, Kranner, Neumann, Rolletschek, Seal, Colville, Fernández-Marín and Börner (Plant Cell and Environment, 2015; http://onlinelibrary.wiley.com/doi/10.1111/pce.12474/full) showed that seed ageing and longevity are intricately affected by genetic background and developmental and environmental conditions in barley. Seeds from 175 barley genotypes from four continents grown in field plots with different nutrient supply were subjected to ageing conditions. Genome-wide association mapping revealed 107 marker trait associations, and hence genotypic effects on seed ageing. Abiotic and biotic stresses were found to affect seed longevity. To address aspects of abiotic, including oxidative, stress, two major antioxidant groups were analysed. No correlation was found between seed deterioration and the lipid-soluble tocochromanols, nor with oil, starch and protein contents. Conversely, the water-soluble antioxidant glutathione and related thiols were converted to disulphides, indicating a strong shift towards more oxidizing intracellular conditions, in seeds subjected to long-term dry storage at two temperatures or to two artificial ageing treatments. The data suggest that intracellular pH and (bio)chemical processes leading to seed deterioration were influenced by the type of ageing or storage. Moreover, seed response to ageing or storage treatment appears to be significantly influenced by both maternal environment and genetic background.
In addition to the QTL analysis, genome-wide association mapping (GWAS) was conducted for seed storability using 116 two-row and 68 six-row barley lines with widely ranging total germination that had been stored since 1974. Genotyping was conducted using Illumina chips. In total 7,864 markers were analysed, of which 4,343 markers were mapped for the seven barley chromosomes and structured genotypes into four major clusters (six-row Ethiopian landraces, two-row German landraces, six-row cultivars, two-row cultivars) after phylogenetic analysis.
In 2013 and 2014, the spring barley collection was multiplied in the greenhouse and in the field to investigate the link between key agronomic traits, dormancy and storability characteristics. Seeds from both years were used to evaluate days to flowering and maturity, plant height, 1000-seed weight as well as dormancy and vigour characteristics. The different growth conditions affected the time to flowering, maturity, plant height and 1000-seed weight and showed a great variation between the genotypes. Dormancy was tested at maturity at 10 °C and 20 °C; germination was drastically reduced at 20 °C, indicative of thermodormancy in the spring barley collection. Only weakly significant correlations were found between days to maturity and dormancy at maturity, and no significant correlation was shown when dormancy was linked with germination after 40 years of dry storage. However, dormancy was significantly different between the four major phylogenetic clusters. Six-row Ethiopian landraces showed strongly reduced and six-row cultivars strongly developed dormancy.
The 184 spring barley accessions multiplied in the field in 2014 were used to assess seed storability after controlled deterioration. Here, fresh seeds were kept at 60% relative humidity and 45 °C for 15 days and showed great variation between germination of genotypes. Results were used for GWAS analysis and revealed 200 highly significant marker-trait associations for percentage of total germination after controlled deterioration and 169 associations after 40 years of dry storage. Although the correlation coefficient between total germination after 40 years of dry storage and controlled deterioration was low, 25 marker-trait associations for 40 years of dry storage and controlled deterioration were localised at similar positions.
THE IMPACT OF KNOWN FACTORS ON SEED QUALITY IN ARABIDOPSIS
The review paper 14 by Née, Xiang and Soppe (Current Opinion in Plant Biology, 2017; http://www.sciencedirect.com/science/article/pii/S1369526616301339) focuses on the control of dormancy induction by plant hormones and dormancy-specific proteins, the influence of environmental conditions during seed maturation and storage, non-enzymatic oxidative reactions, adjustment of dormancy in the soil seed bank through temporal and spatial sensing and the decision of the seed to germinate during early imbibition.
INFLUENCE OF THE TRANSCRIPTIONAL NETWORK ON SEED QUALITY TRAITS
Transcriptional control is the single most important mechanism of regulation of gene activity in all organisms and is ultimately mediated by transcription factors that recognise cis-regulatory elements within their target genes. Because of their key role in the regulation of coordinated gene activity and consequently complex traits (i.e. seed germination), transcription factors are thought to play an important role in the onset of germination and have been the targets for breeding and domestication. One of the aims of WP4 was to characterize the transcriptional network implicated in seed quality. For this purpose, transcription factors involved in the nitric oxide signal transduction pathway during seed germination and seedling establishment were identified.
Paper 3 by Albertos, Romero-Puertas, Tatematsu, Mateos, Sanchez-Vicente, Nambara and Lorenzo (Nature Communications, 2015; http://www.nature.com/articles/ncomms9669) showed that S-nitrosylation triggers ABI5 degradation to promote seed germination and seedling growth. ABI5 (AT2G36270) was found to be involved in the crosstalk between abscisic acid and nitric oxide during seed germination; abi5 mutants were insensitive to abscisic acid and to nitric oxide scavenging by cPTIO during germination. Nitric oxide-production and ABI5 expression co-localised in young seedlings. Using genetic (nitric oxide-deficient and nitric oxide-over accumulating lines) and pharmacological approaches (nitric oxide donors and nitric oxide scavengers, it was shown that ABI5 accumulation is altered in genetic backgrounds with impaired nitric oxide homeostasis. Cysteine-153 was identified as the central point of ABI5 regulation by nitric oxide. Mutation of ABI5 at cysteine-153 deregulates protein stability and inhibition of seed germination by nitric oxide depletion. These findings suggest an inverse molecular link between nitric oxide and abscisic acid signalling through distinct posttranslational modifications of ABI5 during early seedling development. Overall, abscisic acid regulates ABI5, a central hub of growth repression, while the reactive nitrogen molecule nitric oxide counteracts abscisic acid-mediated effects during seed germination. S-nitrosylation of ABI5 at cysteine-153 facilitates its degradation through CULLIN4-based and KEEP ON GOING E3 ligases, and promotes seed germination.
Furthermore, paper 11 by Sanz, Albertos, Mateos, Sánchez-Vicente, Lechón, Fernández-Marcos and Lorenzo (Journal of Experimental Botany, 2015; https://academic.oup.com/jxb/article/66/10/2857/534941/Nitric-oxide-NO-and-phytohormones-crosstalk-during) reviews our knowledge of the interactions between nitric oxide and phytohormones during seed dormancy and germination, and seedling development.
Paper 15 by Sanz, Fernández-Marcos, Modrego, Lewis, Muday, Pollmann, Dueñas, Santos-Buelga and Lorenzo (Plant Physiology, 2014; http://www.plantphysiol.org/content/166/4/1972.short) further explored the mechanisms by which nitric oxide modulates root development. A pharmacological approach and NO-deficient mutants were used to unravel the role of nitric oxide in establishing auxin distribution patterns necessary for stem cell niche homeostasis.
EVALUATION OF THE POTENTIAL OF ACTIVITY-BASED PROTEIN PROFILING (ABPP) AND HIGH FIELD EPR (HF-EPR) SPECTROSCOPY TO ASSESS SEED QUALITY
In addition to genetic markers, novel technologies like ABPP and HF-EPR have the potential to evaluate the quality of harvested seeds at low cost and have been tested in this context in EcoSeed. ABPP is a novel technology that displays the active proteome, using biotinylated or fluorescent small molecule probes that irreversibly bind to the active site residues of various protein classes such as lipases, proteases, glycosidases and phosphatases. The feasibility of ABPP to assess seed quality was analysed using seed lots with different qualities. These were obtained by CD of Arabidopsis seeds. For the ABPP assay, proteins were extracted from seeds that had been subjected to CD for different time periods, representing several seed quality levels (ranging from highly viable to dead). Four ABPP probes, which targeted different enzymes, including proteases, were evaluated. Two of them exhibited a differential band pattern between seed lots with different qualities. One of the probes showed a strong increase in signal intensity for two different bands during seed ageing, both in the Ler and Col accessions. In contrast, several bands that were detected with the other probe decreased in intensity during the ageing treatment.
Heat shock, which resembles the conditions used for the artificial ageing treatment, can trigger activation of the enzyme targeted by one of the probes. In order to verify that the observed pattern is a result of the ageing treatment and not caused by a heat response, additional labelling was performed using naturally aged seeds. Eight-year-old seeds from Ler and the longevity-defective mutant dog1-1 were used. A germination test showed that aged Ler exhibited 40 % germination after eight days of imbibition, whereas the dog1-1 mutant was dead. The naturally aged seeds showed the same increased activity pattern that we observed previously using artificially aged seeds.
HF-EPR provides an opportunity to directly monitor potentially important radical chemistry in seeds. HF-EPR spectra were obtained from seed lots of control and aged seeds from A. thaliana, Brassica, sunflower and barley. Different signals were dominating the spectra of the four species investigated. In Arabidopsis, a flavonoid radical was identified. In sunflower, the HF-EPR spectra were dominated by a melanin radical. In B. oleracea, surprisingly high manganese contents were detected, both in the pericarp and in the seeds when the pericarp had been removed. In the genotypes of barley seeds used for the measurements no signal was observed that could be attributed to an organic radical, whereas wild Hordeum species contained similar radicals as Arabidopsis. The number of different signals obtained by HF-EPR excludes this method as a general one to assess seed quality. Furthermore, there was no simple correlation between signal size and ageing treatment as described below.
The spectra of control and aged Arabidopsis seeds were examined for flavonoid organic radical content. The hypothesis was that the radical content should reflect exposure to oxidative conditions. There were typical lot-to-lot variations between and within the non-aged and aged control seeds, but the approximate radical contents in these seeds were not significantly different. The low or high temperature regimes in the maternal environment did not significantly change radical contents. Nonetheless, the high content was indicative of a significant change in the radical chemistry that generated the radical and, equally important, the chemistry that occurs during imbibition. The location of the radical was not identified, but measurements of tt-mutants suggest that in Arabidopsis the radical is located in the seed coat. An effort was made to determine the exact chemical nature of the flavonoid radical in Arabidopsis. In addition to the High Field EPR spectra ENDOR spectra were obtained. The two methods allowed us to identify the organic radical as a catechin radical.
B. oleracea seeds contain high concentrations of manganese (II), but no radicals were found. The manganese (II) content in the seed coat was significantly higher than that in the rest of the seed. HF-EPR measurements suggested that the manganese (II) within the seed was mostly associated with phosphates.
Sunflower seeds were dissected and pericarp, seed coat, cotyledons and embryonic axes were measured separately. A radical signal was only found in the pericarp. Here, the large flavonoid radical signal masked the presence of even larger narrow melanin radical resonances. No radicals were found in cotyledons and embryonic axes. Drought stress to the mother plant or ageing of the seeds did not alter the spectra.
WORKPACKAGE 5:
TRANSLATIONAL BIOLOGY ON WILD SPECIES
WP5 translated knowledge of crop and Arabidopsis responses to environmental stresses imposed during seed development in the context of pre-existing adaptive seed biology traits in eight crop wild relatives. In addition to the characterisation of the seed germination and longevity traits, investigations progressed on the expression of homologous genes / marker molecules for abiotic stress factors. Finally, a conceptual model was constructed of downstream risks to agricultural efficiency and wild species conservation on the basis of changes in seed quality traits anticipated under future climates.
WILD SPECIES SEED REGENERATION AND ASSESSMENT
Partner 3 (IPK) and Partner 2 (RBGK) hold ex situ collections of 98 species in the three targeted crop genera: Brassica (35 species), Helianthus (27 species) and Hordeum (36 species). These unique genetic resources, some harvested decades ago and stored under seed bank conditions, have associated information on (georeferenced) provenance, seed number and quality, verified plant name, etc. The criteria for use of seeds from these collections were: 1) climate of the seed provenance (wild material); 2) permission from the country of origin to use the accession for research purposes; 3) quantity of seed available; 4) regeneration potential. In contrast to domesticated plants, wild plants do not necessarily follow a simple regeneration cycle, with specific requirements for germination, plant and flower development, and maturity. Therefore, we selected accessions, across 10 species or subspecies, for field multiplication / regeneration over two seasons. The material was taxonomically and agronomically evaluated (e.g. plant height, time to flowering); as appropriate, bees were introduced to the greenhouses / tunnels to enhance pollination. Accessions from seven countries were cultivated as follows: Brassica rapa subsp. campestris (Turkey) and Brassica tournefortii (Tunisia, Spain); Hordeum bulbosum (Israel, Iran), Hordeum murinum (Greece) and Hordeum pusillum (USA); Helianthus agrestis (USA), Helianthus angustifolius (USA), Helianthus gracilentis (USA), Helianthus glaucophyllus (USA) and Helianthus praecox (USA). Overall, >500,000 seeds were produced.
Sunflowers require a high number of sunlight hours and a dry climate during flowering and some specific cultivation challenges emerged: some species showed perennial plant characteristics, some plants died before flowering, some plants produced sterile seeds and some had high levels of seed dormancy. Consequently, dormancy-breaking treatments were applied to a selection of Helianthus collections from the Kew seed bank. Scarification of seeds of two species of Helianthus resulted in slightly faster water uptake, meaning that the seeds were not physically dormant. Seed morphology (by light microscopy and X-ray) was compared between nine Helianthus species (eight wild and one crop species) of varying seed dormancy status. Pericarp thickness varied up to 6-fold amongst the wild species and crops. An association with growing site meteorology was assessed. Embryo size and distance from the radicle tip to the pericarp edge were quantified for two species and considered in relation to the speed of germination for individual seeds. The physiological nature of the dormancy was assessed on 11 accessions of 10 species varying in germination from 0 % (fully dormant) to 100% (non-dormant), through the application of gibberellic acid. Six of the dormant seed lots had improved germination after the application of gibberellic acid to seeds scarified at the root axis (micropylar) end. These findings impacted on the experimental designs.
QUANTIFICATION OF SEED QUALITY TRAITS IN WILD SPECIES
Germination characteristics: Thermal time and hydrotime give an indication of how responsive the germination of a species is to temperature and water availability, and can be used to predict the germination behaviour in future environments such as due to climate change. Germination was performed at five or six temperatures and four or five water potentials (created using polyethylene glycol), to give a range of germination rates to describe performance and for model development. Often germination was defined as both radicle emergence (2 mm) and normal seedling production. Cumulative germination curves (i.e. % germination over time) were plotted and times taken to reach 20 %, 50 % and 80 % of the maximum germination interpolated. The reciprocals of these germination rates plotted against temperature or water potential enable predictions of the thresholds (base temperature, Tb; optimum temperature To; ceiling temperature, Tc; base water potential) for growth, and the calculation of thermal time, hydrotime and hydrothermal time. To maximise the translation of knowledge, comparative seed characterisation was performed on 12 species in the Helianthus, Brassica and Hordeum genera. For Helianthus, four wild species were studied, plus five Helianthus annuus genotypes from Partner 11 (Limagrain), and two environmental treatments of sunflower genotype B. We found that amongst Helianthus species and crop genotype seeds, base temperatures varied by c. 12 °C and the thermal-, hydro- and hydro-thermal times varied c. 7- to 10-fold. For Brassica, four wild species / sub-species were investigated, plus two vigour (high and low) genotypes of B. oleracea and two environmental treatments of the same genotypes. We observed that seeds of Brassica species and crops had base temperatures that varied by 9 °C, and the heterogeneity in thermal- hydro- and hydro-thermal times was c. 4- to 7-fold. For Hordeum, four wild species were studied, plus one (long-lived) Hordeum vulgare genotype and two environmental treatments of that genotype. We found that seeds of Hordeum species and crops had base temperatures that varied by 4 °C. Differences in thermal-, hydro- and hydro-thermal times amongst species were c. 4-fold. In conclusion, there are substantial differences between species in stress tolerance during germination, which means that each species will have emergence advantages over the others under different combinations of temperature and water potential. These studies also identified the optimal germination conditions for use in the longevity experiments when assessing seed survival.
Longevity properties and biophysics: Seed storage lifespan is affected by temperature and moisture content, with the latter being a feature of the seeds' water binding properties which vary with relative humidity and with oil content. Therefore, seeds of eight wild species (three Helianthus, three Brassica and two Hordeum) were used to construct water sorption isotherms and to determine seed oil contents (using time-domain NMR). The Brassica species had high oil contents, slightly higher than for wild Helianthus and much greater than the wild Hordeum species. Generally, the crop seeds were similar in oil content to their wild relatives. We noted that the relative humidity required to achieve the same moisture content will vary greatly with oil content. The implication for the seed trade is that when storing seeds at the same moisture content there is a risk that the different seed lots (accessions) will not be at the same water activity and thus not have the same longevity characteristics.
Seed prehydration history also has a potential impact on the seed biophysical properties. One feature of sorption isotherms is that the relationship between seed moisture content and relative humidity achieved when hydrating a sample (adsorption) is different to that achieved during dehydration (desorption). Seeds of the same eight wild species used for oil content determination were equilibrated to different relative humidities at 20 °C in an IgaSorp machine (c. 50 mg samples) and seed mass measured as relative humidity was ramped up from 15 to 75% relative humidity and back down. On average, each adsorption / desorption isotherm run took three weeks to construct. Seed moisture content was determined at the end of each run. Seeds were also equilibrated at 20 °C above LiCl solutions at various saturations. Although there were some differences between the IgaSorp and LiCl data (e.g. the isotherms of the later tend to be a little more concave), the goodness of fit for the isotherms suggests a general robustness of the two methods.
Seed sorption properties of the three species of Helianthus were found to be broadly similar, with 7% seed moisture content achieved during adsorption at c. 50-60% RH, whether in the IgaSorp or over LiCl solutions. The adsorption result for Brassica seeds (three species) was similar to Helianthus seeds. In contrast, the two Hordeum species had much lower oil contents and this was reflected in a different moisture content-relative humidity relationship such that 7% seed moisture content was achieved at 30-40% RH. One clear feature of the isotherm study was the hysteresis between adsorption (a lower seed moisture content for a given RH) and desorption (a higher seed moisture content for a given relative humidity) curves. The differences in the mid-range of relative humidities was about 10% for Helianthus, and 7% for Brassica and Hordeum. This means that drying wet seeds to 7% seed moisture content (desorption) will result in a seed equilibrium relative humidity that is lower than that reached by seeds that were pre-dried and allowed to rehydrate (adsorp) to the same moisture content. The implication for the seed trade is that the "desorbed" seed will have lower chemical / water activity and potentially a better lifespan than "adsorbed" seed at the same MC.
Seeds of eight wild species were aged under a range of relative humidity conditions at 40 °C. The aim of this study was to characterise the longevity of a range of wild species to see if their seeds have similar keeping properties to the main crops in the same genera. Seed survival lines were subjected to probit analysis to allow estimates of lifespan (the distribution of seed deaths in time, sigma; and the half-life, P50). It was originally intended to additionally store seeds of all eight species at 30 and 20 °C under a range of relative humidity conditions, but limitations on the number of seeds available meant this was only possible at one temperature and five relative humidities (15 – 75%). Seed viability after storage was assessed as germination under optimal conditions. As a benchmark, a wider range of conditions was used to age the Helianthus annuus (control and drought regime) seed lots (linking to WP2).
For all seed lots (accessions) investigated (wild species and crops) the benefits of decreasing seed relative humidity on seed longevity were noted. For example, Helianthus angustifolius seed pre-equilibrated to 75% relative humidity and aged at 43 °C (average) lost most viability in 10 days. One feature of some of the ageing curves for the Helianthus crop wild relatives was an increase in germination value during the early stages of treatment, indicative of seed dormancy loss (after-ripening). Across the eight wild species, P50s (half-lives) varied about 8-fold for seeds at 75% relative humidity and 43 °C. Published data sources for the crops in the three investigated genera indicate values of 6.7-9.1 for KE and 4.2-5.3 for CW. The terms KE and CW describe the dependency of seed lifespan on moisture content. By using the universal temperature constants (CH = 0.0329; CQ = 0.000478) it was possible to generate the first estimates of the viability constants for eight wild species; their KE and CW value ranges overlapped with those of the crop values. There are a number of possible errors introduced when estimating / predicting seed lifespan, including assumptions about equilibrium moisture contents. Consequently, we recommend to the seed trade and conservationists that seed oil content and sorption properties are determined for each seed lot subjected to seed longevity assessment.
Drier (and cooler) seeds live longer as the rate of chemical reactions contributing to ageing is slowed down when the intracellular viscosity is high. When seeds are stored in the visco-elastic (i.e. glassy) state, perceptible losses in viability may take many months or years. To tie in with one of the seed ageing conditions applied, we equilibrated seeds to 75% relative humidity (and some to 15% RH) and ran the samples in a Dynamic Mechanical Analyser (DMA). Two parameters were measured during rewarming from around -100 °C to 80 °C: storage modulus (E’) and tan-delta (damping). A decrease in E’ indicates that the material is becoming less rigid, and tan-delta provides greater resolution of molecular relaxation events. One of the challenges of applying DMA to seeds is the precise orientation of the sample, its architecture and size. Another is that the interpretation of such traces is clearly complex. A comparison of thermal scans for six wild species / subspecies confirmed that seeds aged at 40 °C and 75% relative humidity are in the non-glassy state, explaining their rapid loss of viability under these conditions.
VALIDATION OF MARKERS IDENTIFIED IN MODEL PLANTS AND CROP ON WILD SPECIES
The aim was to identify marker molecules and genes that influence seed quality in response to abiotic stress during seed development. These markers / genes were selected based on the results of the analyses in other work packages and used for the validation in a range of wild species.
Firstly, Activity-Based Protein Profiling (ABPP) was used for evaluating seed quality. ABPP relies on the use of small, reporter-tagged probes that specifically and covalently react with the active site of an enzyme. These probes are designed to specifically bind certain enzymes (such as lipases, proteases, glycosidases or phosphatases). We used two probes and found that they could be used as a marker to evaluate seed quality of wild species, but these need to be optimized depending on the species.
Secondly, high magnetic-field electron paramagnetic resonance spectroscopy (HF-EPR) was used to identify and characterize radicals and transition metal ions in seeds of wild species of Helianthus and Hordeum. Whether the variations in transition metal ions and radicals in seeds of diverse species reflects growth conditions or species remains to be elucidated.
Thirdly, the presence of abscisic acid and nitric oxide signal transduction pathways transcription factors were assessed during seed germination and seedling establishment. Eight wild species were tested by immunoblot analysis of protein levels in seed extracts from dry and imbibed seeds, with variable detection levels. In summary, we were able to identify proteins and radicals in both the crop and related wild species, some of which have the potential to be used for seed quality assessment.
TRAIT CHARACTERISTICS AND DOWNSTREAM RISKS TO AGRICULTURAL EFFICIENCY AND WILD SPECIES CONSERVATION
Based on our understanding of how the environment might impact on seed quality traits, we aimed to develop for the seed trade and conservationists a preliminary climate-related model that predicts the risks to food security and to wild species survival, including through natural seed regeneration.
The wild species seed phenotypes were compared with germination and storage responses determined on the main crops of interest in WP2 to develop a model / matrix of risks associated with anticipated changes in climate. The scale of this risk depends on many factors, including how success is measured, because seeds reaching early germination do not always convert to seedlings, and this step can be temperature- and water potential-dependent. Generally, warmer conditions will result in quicker germination for all seed lots assuming that water is not limiting. Under conditions of water restriction, germination of some wild Helianthus species and selected lines of the Brassica crop will be affected the greatest. Although not always the case, the crops generally have either similar or shorter thermal, hydro- or hydro-thermal times for germination compared with the wild species; meaning that the crop seeds may be able to cope better with erratic rainfall. Generally, higher seed mass of the crops may buffer environmental extremes by being able to emerge from deeper in the soil. The maternal environment of the wild species, inferred from historical climate data for these long-term stored seeds, also correlated with numerous seed germination traits. Concerning lifespan, the higher temperature regime in the Brassica crop and the drought regime in Helianthus crop affected seed oil content, with potential implications for the equilibrium water content in seed storage (in agricultural or conservation settings). Potential downstream effects of the maternal environment on seed lifespan were also evident. Some treatments (temperature or drought regime) to the crops and Arabidopsis during seed development changed the initial quality (total germination) of the seed lot with implications for subsequent half-lives (P50s). All aspects of seed processing (field to store) should be optimised, so as to improve retention of seed quality in the agricultural setting (breeding programmes, native seed industry) or to meet conservation objectives. In this context, the benefits to lifespan of drying below c. 45-50% relative humidity (c. 5 – 10 % moisture contents) often associated with commercially available seed lots was very evident.
Paper 5 by Dürr, Dickie, Yang and Pritchard (Agricultural and Forest Meteorology, 2015; http://www.sciencedirect.com/science/article/pii/S0168192314002482) reported on the changes of critical temperature and water potential values for the germination of species worldwide. The paper contributes to a seed trait database, providing a wide spectrum of species' seed germination traits, in the form of a set of parameter values describing germination responses to variations in temperature and water potential. Major differences in germination traits were seen to depend on the climatic conditions where the species grow or originated, with some species able to germinate on ice and others unable to germinate below 18 °C. By contrast, within the different plant groups, similar ranges of threshold values were found, linked to the species geo-climatic origin. Crops germinate faster, however, their range of threshold temperatures and water potential values is wider, and some crops have higher optimum and maximum temperatures as well as lower water potential threshold values. The collected data also form a valuable database, enabling plant establishment to be better taken into account in modelling and simulation studies of vegetation boundaries (wild or cultivated) under changing land-use and climate.
Potential Impact:
THE IMPORTANCE OF PLANT SEEDS TO HUMAN AND ANIMAL NUTRITION
Plants and hence seeds are the basis of life on earth. Most human and animal food is derived directly or indirectly from seeds. Crop production starts with the sowing of high quality seeds and generally concludes with seeds that are stored for human food or animal feed production, or for plant propagation and breeding. Hence, seed quality underpins European agriculture and food security. Equally important to European society are the seeds of wild species that fulfil a pivotal role in the maintenance and evolution of natural ecosystems. The seeds of wild species serve as barometers of environmental change and are a source of adaptive functional traits in breeding. Temperature and water availability as seeds develop on the parent plants, i.e. "the maternal environment", impact on quality traits, such as seed dormancy, longevity, germination and vigour and successful seedling establishment by mechanisms that are not fully understood. EcoSeed aimed at providing a step change in our knowledge of how seed quality traits are modified by the environment in crop and wild species, creating new opportunities for increased sustainability of European agriculture and strengthen the competitiveness of the European knowledge-based bio-economy.
Seed quality inexorably changes with time whether seeds are stored in gene banks - as a "genetic insurance" against losses of species and cultivars - or in the natural environment that represents the "genetic memory" of our landscapes. Currently, considerable economic losses result from sub-optimal seed performance. The resulting seed wastage undermines food security and livelihoods, and losses may worsen as climate change is predicted to intensify abiotic (and concomitant biotic) stress events that negatively impact on the seed quality traits of vigour and longevity. The Inter-governmental Panel on Climate Change predicts that more frequent extreme weather events in Europe will have profound implications for agricultural efficiencies (decreasing seed yields). These will be realised through reduced total harvest and lower individual seed mass; and smaller seeds have lower vigour and produce weaker plants. One estimate of lost yield from major cereals from rising temperatures between 1981 and 2002 is $5 billion per year. It is known that the seed longevity of model, crop and wild species is species- and genotype-specific and responsive to pre- and post-harvest environments. Consequently, subtle differences in the quality of seeds at harvest are known to have profound downstream implications for successful storage for conservation and agricultural use.
Furthermore, sub-optimal conditions of post-harvest storage in both large scale agriculture or by SMEs can accelerate seed ageing resulting in additional losses in vigour and viability that further undermine food security. Post-harvest seed losses can account for c. 5-25% of grains in just the first few months of storage (http://www.phlosses.net/) representing billions of seeds wasted and revenue lost in inventory. Whilst most of these losses are on-farm, the seed trade also uses suboptimal storage conditions, as evidenced by information provided by the laboratories of the International Seed Testing Association (ISTA).
In addition, paper 16 by Foyer et al. pointed out that the current lack of coordinated focus on grain legumes has compromised human health, nutritional security and sustainable food production.
A greater understanding of the factors that control seed quality traits is therefore essential to maintain the competitiveness of EU agriculture, to enhance resilience through advanced breeding programmes and to provide guidelines for optimal seed production, treatment and storage. Classical genetic and physiological studies on seed traits, particularly dormancy, have focused largely on spatial and temporal changes in plant hormones (e.g. "the hormonal balance theory"). The EcoSeed consortium characterised the upstream signalling hub that mediates seed sensitivity to the environment, including the development of vigour, longevity-ageing and dormancy traits. There is now compelling evidence that such upstream signalling is modulated to a large extent by the internal redox (oxidative) environment.
An unfortunate consequence of "globalisation" is the disappearance from mainstream agriculture of the old landraces and crop wild relatives that have, over millennia, adapted to adverse environmental conditions. Recognising that the complete loss of this genetic variation will impact on future crop production, about 100 years ago plant specialists initiated collecting missions to safeguard plant genetic resources (PGR) in ex situ collections (seed banks or field gene banks). Presently 7.4 million PGR accessions are available for food and agriculture worldwide (FAO, 2010). The German ex situ genebank at the IPK in Gatersleben holds the largest and most diverse collection for cultivated crops within the EU with >150,000 accessions from 890 genera and 3,032 species (http://www.ipk-gatersleben.de/). The Royal Botanic Gardens, Kew (UK) holds the largest and most diverse collection of wild plant seeds globally and is an internationally acknowledged Centre of Excellence for wild species conservation, with unparalleled expertise in collection management and seed science. The ex situ seed collections exceed 50,000 across >30,000 species, including species in the target genera of the EcoSeed project: Hordeum, Helianthus, Brassica and Arabidopsis. EcoSeed was thus well positioned to take full advantage of these resources. By bringing together the skill base and unique collections of these two world class institutions to support the scientific discovery programme on seed traits, EcoSeed aimed at supporting 1) longer-term demands of the EU agricultural industry; and 2) conservation objectives of the EU as signatory countries to the Convention on Biological Diversity (CBD).
MAIN DISSEMINATION ACTIVITIES AND EXPLOITATION OF RESULTS
One of the EcoSeed Workpackages (WP7) was dedicated to support dissemination and exploitation of project knowledge and results. This included the publication of project results in scientific journals, presentations at scientific meetings, transfer of information to other relevant research programs, potential industrial users and agricultural and conservational stakeholders, training of early career scientists and knowledge transfer to the general public.
SETUP AND MAINTENANCE OF A PROJECT WEBSITE: A project website was set up at the start of the project (http://www.uibk.ac.at/botany/ecoseed/). This website served as a direct portal to deliver information to the public. Detailed information is provided on project objectives, the Partners of the consortium and the Workpackages. The members of the Scientific Advisory Board comprising two distinguished scientific experts in seed science research as well as industrial Partners are listed in order to address the needs of the seed industry at all stages of the project. In addition, press releases and public status reports were regularly presented to the public. Furthermore, peer reviewed publications emanating from EcoSeed (24 so far) are listed on the public website. The website also contains a restricted area, based on SharePoint, for EcoSeed members only, enabling efficient project management and data exchange between the Partners. The website was continually updated throughout the project lifetime and is still active.
EXPLOITATION FACILITATION: Establishment of industrial contacts: A list of relevant industrial, conservation and agricultural end-users was compiled. To aid communication targeted to the different needs of the various stakeholders, contacts were established with national seed production and plant breeding associations in Austria, the UK, France, Spain and Germany as well as the European umbrella organisation, the European Seed Association, and their permission obtained to add on the public EcoSeed website a link to the European Seed Association webpage. In this way, national stakeholders can navigate to their respective associations through the European Seed Association. In the Consortium Agreement, access rights to background made available to all Beneficiaries, and backgrounds excluded from access rights were listed. A document on adequate protection of IP (inventions) was prepared.
KNOWLEDGE TRANSFER: Knowledge transfer to the scientific community and the public: Throughout the project life time, the EcoSeed consortium was, and still is, proactive to transfer knowledge to the scientific community, to agronomic and conservational stakeholders, the public and importantly, to early career scientists. In summary, 9 PhD thesis (4 of which have been defended and 3 are close to completion), 20 MSc theses and 18 BSc thesis were supported by the consortium, 24 papers were published in peer-reviewed journals; 83 oral presentations were delivered at scientific conferences and 57 posters were presented; 11 articles were published in the popular press, 5 press releases and 4 web site contributions were released. In addition, 11 popular lectures were given to the wider public, 4 scientific conferences were organised, 3 interviews were given to the international press, 20 activities were dedicated to knowledge transfer and 9 exchanges visits of early career scientists were supported.
Specifically, four international conferences/workshops were organized by project partners including the Seed Longevity Workshop of the International Society for Seed Science (ISSS) held in Wernigerode (Germany), 5 – 8 July, 2015. This workshop was jointly organized by partner 3 (IPK), 1 (UIBK), 2 (RBGK), 5 (MPIPZ) and 6 (Warwick). The workshop was attended by 212 participants from 44 countries. The consortium is also happy to report that with 140 oral and poster presentations at numerous international scientific conferences EcoSeed knowledge was extensively transferred to the international scientific community. In addition, publications in peer-reviewed journals are indicators for research excellence, and work published so far included papers in Nature Communications, Nature Plants, PNAS, New Phytologist and other scientific journals with high impact factors.
During the period of the project three Public status reports were published on the EcoSeed webpage in 2014, 2015 and 2016, and also on ResearchGate. A project leaflet was displayed by the partners attending conferences, during public days, etc., to showcase and advertise the EcoSeed project. Furthermore, a compilation of press releases is shown on the public EcoSeed website.
Links to international seed science organisations: Strong links between the EcoSeed consortium and the two most important international seed science organisations existed prior to project start, and were further strengthened: the International Seed Science Society (ISSS) - a mainly scientific organisation - and the International Seed Testing Association (ISTA) - dedicated to the transfer of knowledge from science to the seed trade and plant breeders. Links to the ISSS and ISTA websites are available on the public EcoSeed website.
Knowledge transfer to agronomic and conservational stakeholders: EcoSeed knowledge was transferred 1) through the ISTA to agronomic stakeholders; 2) to members of the Scientific Advisory Board from the seed companies, including one SME (Syngenta, ENZA Zaaden and Gautier Semences) and Partner 11 (LIMAGRAIN); and 3) by contacting national and international stakeholders as mentioned earlier.
KNOWLEDGE TRANSFER TO EARLY CAREER SCIENTISTS: Led by 11 Principal Investigators (all senior researchers or University Professors) a total workforce of 127 researchers (head count basis), 79 of which were female, contributed to the EcoSeed project. The consortium encouraged short-term exchange visits of early-career scientists between the project Partners in order to foster their technological knowledge and aid their personal development. In total 9 early career scientists (PhD students and Postdocs) received such training in the Partner labs during short-term visits. Of the 34 experienced researchers (postdoctoral level) contributing to EcoSeed 5 Postdocs were funded by EcoSeed as well as 9 PhD students and 1 work experience student. In addition, training by the EcoSeed consortium was received by 4 further postdocs, 23 MSc students, 18 BSc students and 7 work experience students (e.g. Erasmus students). The consortium also gratefully acknowledges excellent support by 28 technicians and engineers (one of which was fully funded by EcoSeed) and 2 volunteers.
List of Websites:
http://www.uibk.ac.at/botany/ecoseed/home/
Beneficiary 1 (Coordinator): University of Innsbruck (UIBK), Austria. Contact: Prof. Ilse KRANNER, University of Innsbruck, Department of Botany, Sternwartestraße 15, A-6020 Innsbruck, Austria; Tel: +43-0512-507-51035; Fax: +43-0512-507-2715; E-mail: ilse.kranner@uibk.ac.at
Beneficiary 2: Royal Botanic Gardens, Kew (RBGK), UK. Contact: Prof. Hugh W. PRITCHARD, h.pritchard@kew.org
Beneficiary 3: Leibniz-Institut für Pflanzengenetik und Kulturpflanzenforschung, IPK Gatersleben (IPK), Germany. Contact: Dr. Andreas BÖRNER, boerner@ipk-gatersleben.de
Beneficiary 4: Université Pierre et Marie Curie (UPMC), France. Contact: Prof. Christophe BAILLY, christophe.bailly@upmc.fr
Beneficiary 5: Max Planck Institute for Plant Breeding Research (MPIPZ), Germany. Contact: Dr. Wim SOPPE soppe@mpipz.mpg.de
Beneficiary 6: Warwick University (WARWICK), UK. Contact: Prof. William FINCH-SAVAGE, Bill.Finch-Savage@warwick.ac.uk
Beneficiary 7: Institute National de la Recherche Agronomique (INRA), France. Contact: Dr. Loic RAJJOU, loic.rajjou@versailles.inra.fr
Beneficiary 8: University of Leeds (ULEED), UK. Contact: Prof. Christine FOYER, C.Foyer@leeds.ac.uk
Beneficiary 9: Universidad de Salamanca (USAL), Spain. Contact: Prof. Oscar LORENZO-SANCHEZ, oslo@usal.es
Beneficiary 10: Commissariat à l’énergie atomique et aux énergies alternatives (CEA), France. Contact: Anja KRIEGER-LISZKAY, anja.krieger-liszkay@cea.fr
Beneficiary 11: Limagrain Europe: Limagrain (Limagrain), France. Contact: Philippe CAYREL, philippe.cayrel@limagrain.com