Competitive trans-generational control of seed dormancy via stable inheritance of gametophytic epigenomes

During reproduction evolutionary theory has shown that parents can have conflicting optimal strategies for maximising fitness of their offspring. One example where this situation arises is when males give rise to a fraction of the female brood: in this case the optimal maternal strategy is to distribute resources among progeny and minimise competition between kin, while fathers gain an advantage by allowing progeny to sequester resources and outcompete kin.

In plants, even in predominantly selfing species, there are clear examples of parental conflict such as in the regulation of seed size. Increasing the maternal genome dose in Arabidopsis seeds leads to smaller seeds, whereas increasing the paternal genome dose leads to larger seeds. These conflicts do not play out in the embryo but in a second seed tissue called the endosperm. Flowering plants are distinguished from other plant lineages by the process of double fertilisation: two sperm cells in each pollen grain fertilise the egg cell to create the embryo, but also the central cell of the ovule to create the endosperm. The endosperm plays a key role in the control of seed size, and the conflicting influences of the mother and father on seed size are played out through distinctive epigenomes which are established in the gametes and inherited into the endosperm. These epigenomes lead to so-called 'imprinted' gene expression where the allele inherited from only one parent is expressed during endosperm development. For instance paternally expressed imprinted genes arise if the paternal allele is inherited in an epigenetic 'on' state, whereas the maternal alleles are inherited in an epigenetic 'off' state.

The endosperm is retained in mature seeds of most species and plays a key role in seed dormancy and germination, after which it undergoes programmed cell death. Germination terminates seed dispersal and so plays a role in distributing progeny in space and time. One further prediction of mathematical theories of parental conflict during plant reproduction is that mothers should favour increased dormancy to disperse progeny and reduce competition, whereas fathers should favour fast germination to outcompete progeny and secure favourable niches closer to the mother plant. It remains an open question whether parental conflicts really occur in the control of seed dormancy but recently it has been shown that imprinted genes can play a role in seed dormancy control in the endosperm.

Recently we have shown that the maternal epigenome is retained stably in the endosperm even after seed dispersal and is essential for the development of seed dormancy and the control of embryo behaviour, and identified specific chromatin remodelling complexes that impose seed dormancy from the maternal genome. Here we also show that increasing the paternal genome dose in the endosperm leads to an identical low dormancy phenotype. This raises the novel hypothesis that parental epigenomes compete in the endosperm to control progeny seed dormancy and germination behaviour. This proposal describes a detailed characterisation of the roles of parental epigenomes in the endosperm that regulate seed dormancy and aims to identify the target genes and processes by which the father's epigenetic state promotes low dormancy, via the study of imprinted genes at the critical developmental stages. Because seed dormancy is promoted by changes in environmental temperature during seed set, this requires understanding how temperature and whether temperature acts through the same chromatin remodelling processes to affect seed dormancy and germination.

Arabidopsis and many crops exhibit coat imposed physiological dormancy with a prominent role for the endosperm in imposing and maintaining dormancy during seed maturation and imbibition. Specialised epigenetic landscapes are established in the male and female gametophytes that lead to imprinting of gene expression the endosperm and enable parents to influence seed development and resource sequestration, often in conflicting ways. Whether parental conflicts extend to control of seed dormancy is an open question. Here we show that the maternal heterochromatic epigenetic states known to be established in the central cell persist into seed maturity and are necessary for low temperatures to induce primary seed dormancy. Failure of maternal heterochromatin formation in the central cell in mutants lacking VEL3-histone deacetylase complexes leads to loss of PRC2-mediated H3K27me3 deposition and abnormally high expression of paternally-expressed imprinted genes (PEGs) in the endosperm that we hypothesise arises from maternally-derived genome. Further supporting a role for increased PEG expression in preventing seed dormancy induction we show that seeds with an extra dose of the paternally-derived genome also fail to enter dormancy at low temperatures.

The first aim of this proposal is to test whether maternal VEL3 imposes dormancy via repression of PEGs and uncover the mechanism by which PEGs act to prevent dormancy induction by low temperatures. We will profile in detail the epigenetic landscape of the mature endosperm in both parentally-derived genomes separately and aim to identify specific PEGs that control dormancy and the underlying mechanism. The second aim is to understand whether the association of VEL-HDAC complexes with phytochrome B in the endosperm is important for the transduction of temperature information by uncovering the relationships between seed maturation temperature, PHYB and the endosperm epigenome.