Evolution of Shoot Development


Land plants evolved from green algae over 450 million years ago.  Phylogenetic analyses based on both morphological and molecular data suggest that within the green algae, charyophytes in the order Charales are the sister group to the land plants. Within the land plants, the least controversial phylogenetic distinction is between the bryophytes (liverworts, mosses and hornworts) and the tracheophytes (vascular plants). The tracheophytes are monophyletic, with seed plants (gymnosperms and angiosperms), monilophytes (ferns and horsetails) and lycophytes each forming monophyletic groups within that clade.  In contrast, the more basal bryophytes comprise a paraphyletic grade with the liverworts resolved as the most basal group (and thus sister group to all other land plants), and the hornworts as most likely sister group to the vascular plants.

As land plants evolved from green algae, developmental mechanisms were either generated de novo, or were recruited from existing toolkits and adapted to facilitate changes in form. Some of these changes occurred once, others on multiple occasions, and others were gained and then subsequently lost in a subset of lineages. In the context of land plant evolution, a few key innovations facilitated the variations in morphology that are seen in extant plant groups. These include the development of multi-cellular embryos, branched shoots, vasculature, leaves, roots, seeds and flowers. Through both forward and reverse genetic analyses in model species, a reasonable understanding of how these processes are regulated in flowering plants has emerged. However, most of the processes evolved before the flowering plants and are thus also features of non-flowering plant species. Our work aims to investigate how key processes are regulated in non-seed plants in order to provide insight into how developmental mechanisms involved.

Our work is currently focussed in 4 key areas:

  1.  The evolution of 3D growth

Multicellular eukaryotes exhibit 3-dimensional (3D) body plans that result from the elaboration of two or three growth axes. Although studies in a range of organisms have investigated the mechanisms underpinning the establishment of individual axes, we have little understanding of how 3D growth per se is initiated. In flowering plants the onset of 3D growth occurs within the first few divisions of the fertilized zygote. As such, it is virtually impossible to genetically dissect the underlying mechanisms because mutants would be embryo lethal and a compromised switch to 3D growth would be difficult to distinguish from many other causes of lethality. In early divergent plant lineages such as the mosses, however, the production of 3D shoots is often preceded by an extended 2D growth phase. We are exploiting this diphasic growth profile in the moss Physcomitrella patens to define gene regulatory networks that trigger the transition from 2D to 3D growth.  Specifically we have devised a novel strategy to isolate P. patens mutants that fail to make the transition to 3D growth (and hence cannot reproduce), and to identify the causative mutations.


Side branches from a P. patens 2D filament. The one on the left will develop into a 3D bud, the one on the right will develop into a 2D secondary filament.

  1. Leaf evolution & development

Leaves evolved independently and sequentially in the four major land plant lineages (bryophytes, lycophytes, monilophytes and seed plants). Using cell lineage analyses, we have demonstrated that leaves are initiated in distinct cellular contexts in different land plant groups. In the leafy shoots of the bryophyte moss Physcomitrella patens (Harrison et al. 2009), and in a monilophyte fern (Sanders et al. 2011), leaves are initiated from a single cell. In contrast, leaves of the lycophyte Selaginella kraussiana are initiated from two cells (Harrison et al., 2007). Leaf initiation in lycophytes and seed plants therefore requires intercellular communication whereas in mosses and ferns, initiation processes can be cell-autonomous. These observations have overturned the idea of a progressive increase in complexity during evolution and instead point towards ancestral mechanisms of leaf initiation in bryophytes and monilophytes, and derived mechanisms in lycophytes and seed plants.

We demonstrated that a developmental module comprised of interactions between ARP and KNOX proteins operates in both S. kraussiana and flowering plants to regulate the switch from iterative shoot to determinate leaf growth (Harrison et al., 2005). As such, the module was recruited in parallel on two independent occasions during land plant evolution.


KNOX gene expression in S. kraussiana apices

  1. Evolution of land plant embryos

In charophycean algae the majority of the life cycle is represented by the haploid gametophyte and only the unicellular zygote is diploid, undergoing meiosis rather than mitosis after formation.  Embryo development represents a major growth transition in that meiotic division of the zygote is delayed and cells divide by mitosis giving rise to a multicellular diploid sporophyte. This feature is a synapomorphy (derived character state) of land plants yet nothing is known about the underlying mechanisms that facilitated the transition. To provide insight in this area, we are currently using genome-wide expression profiling to compare developmental trajectories in the unicellular zygotes of charophycean algae and very young (4-8 cell) embryos of bryophytes.

Four cell embryo of the liverwort Marchantia polymorpha (left) and mature sporophyte of the moss Physcomitrella patens (right)

Four cell embryo of the liverwort Marchantia polymorpha (left) and mature sporophyte of the moss Physcomitrella patens (right)

  1. Shoot morphogenesis from single apical initials.

The specification of plant body plans occurs throughout the lifeccyle. As land plants evolved, so did bodyplan complexity, with particularly conspicuous elaboration in the shoot. In all vascular plants, shoot development is iterative with apical initials (stem cells) generating multiple above ground organs. Our current understanding of shoot apical initial function is derived almost entirely from studies of multicellular meristems in the most recently diverged flowering plant lineage. In these meristems, intercellular communication plays a central role both in maintenance of a group of initials cells, and in specification of organ founder cells. In earlier diverging vascular plants, however, shoot growth frequently occurs from just a single apical initial that divides asymmetrically to produce derivatives that form organs and to replenish itself. As such, mechanisms of patterning and organogenesis are conceptually very different.

We are investigating how single apical initials regulate shoot architecture, using the fern Ceratopteris richardii as an experimental organism. Until recently it was technically impossible to test gene function in vascular plant species that develop from single apical initials but we have developed a transformation system for C. richardii (Plackett et al. 2014). Our experimental strategy therefore utilizes mutagenesis, candidate gene and transcriptome approaches for gene discovery and transgenic methods for analyses of gene function.


Gametophyte (top left) and sporophytes of Ceratopteris richardii