The CURLY LEAF gene encodes a histone methyltransferase enzyme. Mutants are early flowering and have curled leaves. These defects arise because the gene AGAMOUS, normally expressed only in flowers, is activated precociously during vegetative development.


Double mutants for CURLY LEAF and SWINGER show a more extreme phenotype. Plants are only viable if grown in tissue culture and form callus with somatic embryos. Cytogenetic analysis indicates that these plants have gross defects in histone methylation patterns.


Role of a novel Polycomb protein complex in plant epigenetic regulation

The Polycomb group of genes control many important aspects of plant development, for example time of flowering, flower formation, and stem cell proliferation.  Four of the Polycomb proteins associate in a  complex – Polycomb Repressive Complex 2 (PRC2) – which is an enzyme that catalyses an epigenetic modification, namely methylation of lysine 27 of histone H3 (H3K27me3).  H3K27me3 methylation is strongly associated with transcriptional silencing of developmental genes in higher plants and animals, but recent studies suggest that its original role may have been in silencing transposons.  We identified the ALP genes through a genetic screen for factors that oppose Polycomb silencing.  Surprisingly, our proteomic analysis revealed that the ALP proteins are part of a variant PRC2 complex, ALP-PRC2.  In addition, we found that ALP proteins do not resemble known chromatin components but rather have evolved from transposase proteins that catalyse proliferation of transposons.  Our hypothesis is that the ALP proteins change the enzymatic activity or specificity of PRC2, and that this may be used by plants as a way to regulate their transposons.  In collaboration with Frank Wellmer’s group (Trinity College, Dublin) and Philipp Voigt (Wellcome Trust for Cell Biology, Edinburgh) and Juri Rappsilber (Wellcome Trust Institute for Cell Biology, University of Edinburgh) we are using genomic approaches to test how ALP genes affect the epigenome, and biochemistry to purify the ALP-PRC2 complex and characterize its enzymatic activity.  The study is published in PLoS Genetics (see Publications).



Fig. 1. Genetic interaction between ALP1 and Polycomb genes.  The Curly Leaf (CLF) gene encodes one component of the plant Polycomb machinery.  Plants with a mutation in clf have reduced leaf development.  However if plants have mutations in both ALP1 and CLF, normal growth is restored.  Genetic interaction is a result of the interaction between ALP1 and Polycomb proteins.


Evolution of gene regulatory pathways controlling cuticle formation and water transport in plants

All land plants have evolved from aquatic ancestors (green algae).  The move to a terrestrial environment required many adaptations to a more arid environment, including the formation of a waterproof covering, the cuticle, to prevent water loss and also the ability to transport water from soil to aerial parts via specialised conducting cells.  Working with the flowering plant Arabidopsis thaliana we discovered a gene ZHOUPI (ZOU) which controls the formation of the cuticle during embryo development in seed.  It does this by regulating a signalling pathway between endosperm (a nutritive tissue that surrounds the embryo)  and embryo that monitors the integrity of the cuticle.  ZOU also regulates genes which soften the endosperm, which allows the growing embryo to expand and crush the endosperm, triggering programmed cell death and endosperm breakdown.  We discovered that the ZOU gene is ancient and present in all land plants, even those without seed or endosperm.  We are using genetic analysis in the liverwort Marchantia polymorpha to work out what ZOU does in this early diverging plant lineage.  We found that ZOU also regulates cuticle integrity in Marchantia, even though the bryophyte cuticle is thought to be biochemically distinct from that of angiosperms.  ZOU is also needed for formation of pegged rhizoids, an external water conducting tissue in some liverworts that is formed of bundles of hollow, dead cells with reinforced cell walls.  We are identifying the genes regulated by ZOU that control pegged rhizoid differentiation and have found genes involved in programmed cell death as well as cell wall modification.  Our work is part of a long term collaboration with Gwyneth Ingram (ENS, Lyon) and recently Moritz Nowack (VIB, Ghent) and Takayuki Kohchi (Univ of Kyoto, Japan).


Fig. 2. Marchantia thallus (left) can either produce female gametophores, the archegoniophores (middle) or male gametophores, the antheridiophores (right).