Auxin at the Evo-Devo Intersection

See allHide authors and affiliations

Science  26 Jun 2009:
Vol. 324, Issue 5935, pp. 1652-1653
DOI: 10.1126/science.1176526

“Alles ist Blatt.” With this simple but profound assertion (“All is leaf ”), Johann Wolfgang von Goethe launched the modern age of comparative biology (13). By articulating the concept that plants can be broken down into modular and iterative variants of an archetypal structure (the leaf, in the form of bud scales, spines, petals, stamens and so forth), Goethe propelled the analysis of plant and animal structure (4) for the next two centuries and beyond. The idea that variant forms of a basic organ “type” are homologous (morphologically equivalent) within an organism and between organisms has emerged as a central conceptual (and testable) framework in the burgeoning field of comparative molecular analyses of development. On page 1684 of this issue, a study by Pagnussat et al. (5) brings together a remarkable set of experiments that bear on the developmental biology and modular construction of the microscopic egg-producing structure (female gametophyte or embryo sac) buried deep within a flowering plant's reproductive tissues. The findings have great importance for understanding and further examining the evolutionary developmental history of flowering plants.

Pagnussat et al. demonstrate in the flowering plant Arabidopsis thaliana that the phytohormone auxin is a key determinant of cell fates within the angiosperm female gametophyte. The embryo sac contains seven cells and eight nuclei: an egg cell, two synergids (one of which will receive the pollen tube bearing two sperm cells), a binucleate central cell, and three sterile antipodal cells at the opposite pole from the egg (see the figure). The egg cell and central cell serve as female gametes and, upon receipt of the two sperm from a pollen tube during the process of double fertilization, will yield a diploid zygote and a triploid endosperm, the embryo-nourishing tissue within the seed. Pagnussat et al. show that the distribution of auxin within the developing embryo sac is polarized, and they propose that gradient-based variation in the concentration of auxin determines the identity that cells will assume during the transition from a single-celled syncytium to a seven-celled, eight-nucleate mature structure. This finding is seminal, as auxin's centrality to patterning and differentiation in the angiosperm female gametophyte—or, for that matter, any land plant gametophyte—had not been anticipated.

Following the recent discovery that the earliest flowering plants likely produced a female gametophyte with only four cells and four nuclei, it was proposed that the angiosperm female gametophyte is a modular and iterative structure involving quartets of nuclei (69) (see the figure). Development of a basic angiosperm female gametophyte module was hypothesized to involve three ontogenetic stages: positioning of a single nucleus within a developmentally autonomous cytoplasmic domain of the female gametophyte; two nuclear division events to yield four nuclei within that domain; and partitioning of three uninucleate cells such that the fourth nucleus is confined to the central cell of the female gametophyte. Specifically, it was postulated that the first flowering plant female gametophytes (as well as those of some of the most ancient extant angiosperm clades) were constructed of a single fertile module (the egg cell module) and that duplication of this basic module gave rise to the seven-celled, eight-nucleate female gametophyte of most extant angiosperm species (9). This modular duplication event thus would have led to the creation of the antipodal cell module in the common ancestor of monocots, magnoliids, and eudicots (these three clades constitute more than 99% of extant angiosperms) about 115 million years ago.

Module model.

The female gametophyte of most angiosperms, including Arabidopsis, contains seven cells and eight nuclei. (Top) The ancestral angiosperm female gametophyte is hypothesized to form a quartet of nuclei and a single egg cell module within a gradient of auxin. (Bottom) Insertion of a nuclear migration event at the two-nucleate stage of development, in the presence of a gradient of auxin, is predicted to have led to the creation of two autonomous cytoplasmic domains that formed a two-module, seven-celled, eight-nucleate female gametophyte. This would have transformed the fertile egg apparatus into a set of three sterile antipodal cells. The second polar nucleus would behave as the original polar nucleus because both come into direct contact and share the same proximate cytoplasmic environment. In this scenario, an egg cell module and antipodal cell module develop and the evolution of triploid endosperm is the key functional outcome of this transition.


As with all serial homologs, homeosis—the transformation of one form of organ type (e.g., petal leaf) into another form (sepal leaf or stamen leaf)—remains a key test of morphological equivalence. The observation of Pagnussat et al. that the identities of cells of the antipodal cell module can be transformed into cell identities characteristic of the egg cell module when proximate auxin concentrations are raised can be argued to constitute evidence of homeosis and hence modularity (serial homology). Moreover, the findings of Pagnussat et al. generate a clear set of testable predictions. If the first angiosperms produced a female gametophyte with a single egg cell module, and an auxin gradient was present, creation of an additional module at the opposite (low concentration of auxin) pole of the female gametophyte might well have led immediately to the formation of a sterile egg apparatus (egg plus two synergids) that we now recognize as the antipodal cells (5). Contribution of the fourth nucleus of this new ectopically expressed module to the central cell would have placed this nucleus in the same cytoplasmic environment as the original single polar nucleus and hence would have resulted in its participation in the second fertilization event to create triploid endosperm. As such, the origin of triploid endosperms in flowering plants from ancestral diploid endosperms may have a mechanistic explanation in which auxin plays a central role.

What is needed next are comparative studies of auxin concentrations and gradients in the female gametophytes of diverse eudicots and monocots as well as the most ancient lineages of flowering plants that bear four-celled, four-nucleate, single-module female gametophytes (Austrobaileyales and Nymphaeales). For now, we can begin to envision a working model of one of the key evolutionary transitions in flowering plant history: the shift from single-module female gametophytes that produce diploid endosperms to two-module female gametophytes that yield triploid endosperms.

All too often, organismic biology (specifically morphology and embryology) is thought to be a descriptive science. In fact, morphology, as Goethe showed, is a predictive science that generates testable hypotheses. The work by Pagnussat et al. confirms the view that morphological concepts can be supported or undercut by molecular data. The study reminds us of the truly co-dependent (in the best sense) relationship between organismic and molecular biology that remains, although often unacknowledged, at the heart of the field of evolutionary developmental biology.


View Abstract

Stay Connected to Science

Navigate This Article