Seeds for the future

31 October 2016 - By: Caroline Wood

Seeds for the future

Seeds of wild emmer wheat.
Seeds of wild emmer wheat.

By Caroline Wood

As the proverb says “All the flowers of tomorrow are in the seeds of today”. The same goes for the food crops that need to feed our growing population. Until now, food security has tended to focus on the mature plant and its yield, however we are now beginning to understand the critical influence of the events preceding and during germination. Seeds may be tiny, but they are complex worlds of development and change. During the session ‘Seed Biology – From Laboratory to Field’ at the SEB Annual Meeting in Brighton, we heard from the pioneers of this field, and how they hope to unlock the potential of seeds to sow the harvests of the future. 

All packaged up

We normally assume that once plant embryos are packaged up into seeds and dispersed, they are left to fend for themselves. However new research suggests that mother plants still help their offspring long after they are separated, providing assistance through ‘dead organs’. The outer seed coat enclosing the embryo is derived from maternal cells that undergo programmed cell death (PCD), forming a hard physical layer that protects against damage, microorganisms and temperature extremes. “It is commonly believed that during PCD, cells are completely degraded and their constituents remobilized to other parts of the plant” says Buzi Raviv, a PhD student in Professor Gideon Grafi’s lab (Ben Gurion University, Israel) “But we have found that there is life in the dead organs enclosing the embryo”. Buzi discovered that the dead seed coat releases a range of enzymes – including DNAses, RNAses and proteases- during germination. Furthermore, these enzymes were still active; after separating the enzymes using gel electrophoresis, Buzi incubated them with the appropriate substrates and detected digested products. In planta, hydrolases have a range of functions including DNA replication, cell cycle progression and gene expression; hence these enzymes could be influencing germination in countless ways. Buzi found the same result in a range of species – including Arabidopsis, Cicer arietinum (chickpea), Sinapis alba (white mustard) and cereals such as wild emmer wheat and oat. “This suggests that storing hydrolases in dead organs enclosing embryos and releasing them upon hydration is a general theme in plants that may increase the survival rate of germinating seeds” says Buzi.

The released enzymes also included chitinases, which degrade fungal cell walls, suggesting a role in protecting the embryo from pathogens. Buzi tested this by incubating spores of the model pathogen Alternaria alternata with samples of the substances released by the seed coat of white mustard. “When Alternaria was incubated with the aqueous extract, 85% of the spores showed increased branching, a known stress response” Buzi says.  The substances released from seed coats of other species also had activity against bacteria, including Escherichia coli and Staphylococcus aureus.

This exciting work could help develop new methods to protect seeds without coating them in hazardous chemicals, and improve how seeds are stored in seed banks. “There is a lot of work ahead in this project” says Buzi. “In my next experiments, I will address how the mother plant’s growth conditions affect the composition and quantities of the substances stored in the dead organs enclosing the embryo.”

Primed for action

When it comes to seeds, one of the biggest problems for farmers is non-synchronous or poor germination. This results in non-uniform development and can cause the farmer to under- or over-sow, costing time and money. Unlike yield and disease resistance, it can be time-consuming to accurately phenotype crops for germination, as Steven Penfield (John Innes Centre, United Kingdom) explains: “Germination is typically scored by eye by counting at various intervals. There are some protocols for extracting data from images but they are mostly old and cumbersome”. Consequently, germination times are often only approximate, as no one tends to watch the seedlings overnight, at weekends or during their tea beak. To combat this, Steven with his collaborator Ji Zhou at the Earlham Institute, has helped develop an automated phenotyping platform for germination and seedling establishment, powered by Raspberry Pi computers (single-board computers the size of a credit card). “This very simple platform uses time-lapse photography to automatically score seed and seedling traits during germination” says Steven. “We can score germination accurately to 30 minutes and measure the rate the root and shoot appear”. 

Together, these parameters provide a good assessment of “seed vigour”, which ultimately determines the quality of the crop. “Vigorous seeds germinate rapidly and uniformly, and produce healthy seedlings under the widest possible range of conditions” says Steven.

Yet certain techniques do exist that can synchronise germination, such as ‘osmopriming’. “This restricts the amount of water a seed can take up and therefore limits its ability to germinate” says PhD student Jack Mitchell (University of Birmingham, United Kingdom). The seeds are soaked in enough water to allow the early stages of germination but not the full transition, before being dried. Once sown and rehydrated with water, ‘primed’ seeds germinate faster and more uniformly than unprimed seeds. This simple method, which was even used by the ancient Greeks, is effective on a wide range of crops, including lettuce, carrot, pepper and onion. But the underlying mechanisms remain largely unknown. Jack’s research is investigating the gene expression changes during osmopriming in Arabidopsis seeds. Already his work has documented clear changes in genes related to Gibberellic Acid (GA) and Abscisic Acid (ABA). However standard molecular biology techniques require him to physically destroy each seed to quantify gene expression which limits the number of time points he can test. 

In order to get the complete story, he is developing a real-time, non-invasive imaging system using a luciferase reporter system. This introduces a transgene linking the promoter of the gene of interest to the luciferase enzyme from fireflies. As the expression of the gene of interest increases, the luciferase enzyme will also be transcribed at a higher rate. The rate of gene expression can then be quantified by applying the substrate luciferin, which is converted by luciferase into a green light signal. “A luciferase system would be a fantastic tool that could be used to non-invasively monitor gene expression in real time within individual seeds” says Jack. This would enable seed companies to select seeds that will germinate together and give uniform harvests. 

Meanwhile, Julia Buitink (IRHS-INRA Angers, France) has been investigating the gene networks that allow seeds to remain viable in a dry state (known as ‘seed longevity’). She collected over 100 gene-expression timecourses from Medicago truncatula seeds under different growth conditions and assessed seed longevity at each time point. From this she identified key players whose expression level correlated with how long seeds could last in storage. This ‘longevity module’ was found to be enriched with genes that play a role in defence against biotic stress, including the transcription factors WRKY3 and NFXL1. Accordingly, Arabidopsis wrky3 and nfxl1 mutants show increased sensitivity to pathogens and their seeds die faster in storage. For both mutants, the seed coats are more permeable, suggesting that having defective barriers against pathogens critically influences seed longevity. “These data suggest that seed longevity evolved by ‘hijacking’ defence-activated pathways in leaves to provide physical barriers that protect against pathogens whilst seeds are in the dry state and thus unable to elicit any metabolic defences” says Julia. 

Cutting out the middle man – autonomous reproduction

Perhaps one of the greatest frustrations for seed scientists is that genetic material is reshuffled during fertilisation, causing key traits to be lost. A ‘holy grail’ would be to develop crops that could autonomously produce seeds without fertilization, allowing heritable traits to be fixed across time. This process, known as apomixis, has already been demonstrated in the model plant Arabidopsis. However, transferring this trait into crops is made difficult by the ‘double fertilization’ event that normally occurs during plant reproduction. Whilst one sperm cell fuses with the egg cell to form the embryo, another sperm cell fuses with the central cell, which contains two maternal nuclei. The now-triploid central cell gives rise to the endosperm; a food store which surrounds the developing embryo. But the endosperm has dosage sensitivity, meaning that normal development requires the balance of two maternal and one paternal genomes otherwise the seed is aborted. “Transferring apomixis into crops will require us to overcome parental dosage sensitivity of the endosperm, which is very strong in monocot crops like maize, rice, oat, and barley” says Claudia Köhler (Swedish University of Agricultural Sciences). 

However, knock-out Arabidopsis mutants in components of the Fertilization Independent Seed Polycomb Repressive Complex 2 (FIS-PRC2) develop seeds autonomously, without fertilization. “FIS-PRC2 normally represses autonomous replication of the central cell” says Claudia. Her work has shown that this is due to FIS-PRC2 preventing gene expression in the maternal nuclei by modifying chromatin structure. After fertilization, the paternal-specific expression of these genes drives endosperm development. Consequently, these genes show “genomic imprinting”, because they are only expressed from one parental allele. 

It appears that the plant hormone auxin is critical for the endosperm to develop correctly. “Among others, genes involved in auxin biosynthesis are under repressive control of FIS-PRC2 and activation of those genes in FIS-PRC2 mutants induces fertilization-independent auxin formation, which triggers central cell replication” says Claudia. Auxin biosynthesis genes show genomic imprinting across many plant species, suggesting that repressing maternal auxin genes is a conserved mechanism to prevent autonomous endosperm development.  Furthermore, autonomous endosperm formation also occurs in rice knock-out mutants for Fertilisation Independent Endosperm (FIE) – a subunit of the FIS-PRC2 complex. 

These results suggest that auxin is both necessary and sufficient to initiate endosperm development, and hence seed formation. Fine-tuning auxin expression in the central cell therefore could overcome the paternal genome requirement and dosage sensitivity, opening the way towards autonomously fertilising crops.

Dynamic interactions 

The endosperm has also attracted interest for other reasons; as a highly specialised storage tissue, it could be used to produce valuable recombinant pharmaceutical proteins.  However, the endosperm is a dynamic tissue that changes considerably during seed maturation. The inner layers of starchy endosperm cells undergo programmed cell death whilst the living aleurone cells in the outer layers secrete enzymes to breakdown stored nutrients that fuel germination. These processes require an active and flexible secretory system to coordinate the synthesis and delivery of the storage proteins needed for each stage of development. As such, the Endoplasmic Reticulum (ER) and Golgi body (the main secretory organelles) must be capable of rapid structural reorganisation. In developing maize grain for instance, prolamin-containing protein bodies proliferate from the ER, whilst protein storage vacuoles (PSVs) develop as post-Golgi organelles and merge later during maturation.  

 To date, we have very little understanding of how these changes coordinate the delivery of storage compounds. “Conventional studies of protein storage organelles in cereal seeds involve acquiring static images using immunofluorescence and electron microscopy, revealing little about the dynamic restructuring events” says Eva Stöger (University of Natural Resources and Life Sciences (BOKU), Austria). Her research has been addressing this by developing fluorescent markers to facilitate live-cell imaging in barley. These have already demonstrated that PSVs behave differently depending on the region of endosperm. In the aleurone layer, the PSVs remain spherical whilst those in the subaleurone layer go through several rounds of fusion and rupture to form large composite aggregates.

In txt 2
Barley protein storage vacuole

Understanding these interactions could help develop more efficient molecular farming. “Accumulating proteins in the endosperm has many advantages; for instance proteins with adverse effects on the production host can be expressed because their absence from vegetative tissues means they do not impact plant growth” says Eva. In addition, the ER of cereal storage cells contains many molecular chaperones and isomerase enzymes that facilitate protein folding and maintain stability. The triploid genome of the endosperm also increases the effective number of transgene copies, and thus the transcription of the protein product. So far seed-produced recombinant proteins include lysozyme, human serum albumin and human growth factor. But potentially seeds could be used to produce vaccines and feed additives. According to Eva; “These benefit from encapsulation within storage organelles such as protein bodies, as these provide some protection against proteolysis upon administration and can have an adjuvant effect” (resulting in a more pronounced immune response against the vaccine). She has already demonstrated this principle by producing the model vaccine influenza haemoagglutinin H5 in protein bodies, which were ectopically induced in tobacco leaves.  “In mice, the protein body-formulation achieved a comparable immune response to soluble antigen with a strong adjuvant and still worked at a 100-fold dilution” she says.  


 Spatial control – it’s all about location

Given that 21% of global calories depend on wheat1, even small increases in the quality of this crop could bring great benefits for food security. According to Aakriti Wanchoo-Kohli (Rothamsted Research, United Kingdom), understanding how the balance of plant hormones control grain development could be a key to boost yields. “Whilst the phytohormone  Gibberellin (GA) plays a critical role in wheat grain development  , GA synthesised by the embryo also moves to the aleurone layer to induce the production of  α-amylase and other hydrolytic enzymes that break down starch reserves for growth” she says. However, excessive α-amylase activity causes starch degradation, resulting in poor quality flour and a crop only fit for animal feed. Aakriti’s research has been investigating whether manipulating GA levels in specific areas can increase grain size without compromising quality.

“We generated a range of transgenic lines using tissue-specific promoters to modify GA biosynthesis or signalling in the seed-coat, endosperm, embryo or aleurone of developing wheat grains” she says. The results so far have also demonstrated that GA manipulation is far from straightforward, especially as the hormone can shuttle between the embryo and the endosperm. Increasing GA levels in the endosperm produced larger grains, but also more α-amylase, leading to a poorer quality. When GA in the endosperm was reduced, the grains became smaller but α-amylase levels were unchanged. On the other hand, reducing GA in the embryo and scutellum of developing grains significantly reduced α-amylase levels, showing that it is GA produced here that induces hydrolytic enzymes.


1.   FAOSTAT (2015),

This article captures only a flavour of the work SEB plant researchers are performing to reveal the secrets of seeds. To view all the abstracts from the SEB Brighton session Seed Biology – From Laboratory to Field, visit

Category: Brighton 2016
Caroline Wood-Author Profile

Caroline Wood

Caroline Wood was the SEB’s 2014 science communication intern. Since then, Caroline has been a regular contributor the SEB, reporting on events and writing insightful features for our members.
Caroline has an undergraduate degree from Durham University in Cell Biology and is currently a PHD student at Sheffield University studying parasitic Striga weeds that infect food crops. You can read her blog here.