The maternal control of progeny seed physiology

In plant ecology it has long been recognised that mother plants have 'bet-hedging' strategies, a mechanism to spread the germination behaviour of progeny seed in space and time to maximise the chances of dispersal and successful progeny reproduction. Such strategies can come at a cost to the life chances of some individual progeny seeds, and thus it has been hypothesised that conflict exists between the optimal dispersal and germination strategy for the mother and that of any single progeny seed. How and whether this conflict manifests itself at the molecular level during seed dormancy control is an important open question in plant reproductive biology. It is well known that the strictly maternal tissues of the seed, such as the seed coat, can play an important role in dormancy and dispersal. More recently, it has become clear that the endosperm plays a critical role in dormancy control, and the endosperm has a 2:1 gene dosage bias in favour of the mother. In addition, specific epigenetic processes in the female and male germline can silence alleles from one parent, and it is known that these processes play an important role in seed development. One recent study implicates some of these processes in dormancy control, but little is known of whether the mother can exert substantial control of progeny germination behaviour by influencing endosperm physiology.

Polycomb Repressive Complex 2 (PRC2) function is an important factor in the maternal control of seed development, via deposition of the epigenetic mark H3K27me3. Loss of PRC2 in seeds leads to maternal effect seed abortion resulting in part from the deregulation of MADS box family transcription factors in early endosperm development. PRC2 requires specific factors to target the complex to specific loci, and during flowering time control by temperature PHD-family transcription factors are known to be important for this task. Notably, we have found that two of this small PHD-finger sub-family, VEL2 and VEL3, which lie adjacent in the Arabidopsis genome, are implicated by GWAS in the control of seed dormancy by temperature. In addition, we show that vel3 mutants exhibit a maternally-inherited seed germination phenotype and share some seed development phenotypes with PRC2 complex mutants.

This proposal describes a workplan which aims to understand the molecular mechanism by which VEL-family PHD finger factors enable seed dormancy control by the mother plant. Through a series of linked objectives, we aim to understand at the molecular level why VEL3 acts maternally, and whether its function is shared by VEL2. We will investigate whether VEL2/3 associate with the PRC2 complex, identify their target genes and test whether these are regulated by H3K27me3 deposition. By building on transcriptome data comparing wild type and vel3 mutant endosperms, we will test the hypothesis that VEL proteins control dormancy through the master regulatory transcription factor of the seed maturation and dormancy programme leafy cotyledon 1 (LEC1). In this way we aim to describe a molecular pathway by which the mother plant retains control of progeny seed physiology, and test whether natural variation in this process occurs within Arabidopsis accessions.

Seed physiology is known to be under a strong parental influence in many species, with the mother controlling progeny seed behaviour via the properties of the two tissues surrounding the embryo, the endosperm and seed coat. Although the role of the maternal seed coat tissue in dormancy control is well established, there is increasing evidence for maternal processes in the endosperm that control progeny seed dormancy. Here we show evidence for a previously undescribed maternal mechanism for controlling progeny seed dormancy, involving VERNALISATION INDEPENDENT 3 (VIN3)-LIKE 3 (VEL3) and possibly also VIN3-LIKE 2 (VEL2), two seed-specific PHD-finger proteins. Preliminary data shows that maternal loss of vel3 leads to low dormancy and frequent seed abortion events that resemble the phenotypes of Polycomb Repressive Complex 2 mutants in seeds. This is interesting because two proteins closely related to VEL3, VIN3 and VEL1, are important for PRC2 targeting during flowering time control. Thus, we hypothesise that VEL2 and VEL3 may be important for PRC2 targeting during seed development. Here we describe an experimental programme to test whether the mother plant controls seed dormancy through VEL2/3 protein activity in the endosperm. We show that maternal loss of VEL3 leads to strong down-regulation of LEAFY COTYLEDON 1 (LEC1) expression in the endosperm/seed coat fraction, a master regulator of seed maturation and seed dormancy. Our aim is to identify the mechanism by which VEL3 and PRC2 affect seed dormancy and LEC1 expression, thus exerting maternal control over the progeny seed physiology. Finally, we will explore the significance of natural variation at the VEL2/3 locus which we can link using GWAS to dormancy variation among natural Arabidopsis accessions.

Although this research project is fundamental in its objectives and scope, its findings are highly relevant to the seed industry. While bet-hedging strategies are important for the fitness of wild plants, for crops they generate unwanted variation in seed properties within and between genetically identical seed lots. Much of the field of seed technology is concerned with mitigating these effects. For instance, during routine seed processing seeds may be sorted based on colour, size and density with the aim of recovering physiologically similar seed fractions for further processing. These seed fractions may have different vigour properties and therefore different market values. In addition, in species where seed vigour is an important trait and use of hybrids is common, understanding parental control of seed properties is important in selecting lines with good combinability for the selection of high vigour F1s. Thus understanding the genetics of parental control over seed properties, including size and vigour, is important for high quality seed production.

We aim to ensure impact of the work in the following ways. Firstly, we will have meetings with representatives of different seed companies on a rolling basis, exchanging knowledge on areas of our research that we believe may be of interest, and learning for ourselves about the commercial priorities of the companies they represent. We will present our work to industry at events including Morley Open Day, JIC Plant Breeders Day. We will present the work at one meeting of the International Seed Testing Association, and the Plantum seed industry workshop, both of which pull together representatives of the seed technology industry. Finally, we will use the project as an opportunity to introduce children to the area of seed biology by hosting a Nuffield Foundation student in lab in the summer of the second year, where they will undertake original experiments aligned with the project outcomes under the direction of the requested PDRA.