Sir

The generation of leaf pattern (phyllotaxis) has long been a topic of interest and debate among plant biologists and mathematicians, as observed by Amar Klar in his Concepts essay1. As Klar points out, the various models proposed for the production and maintenance of such biological patterns lack full supporting experimental data. These models, nonetheless, are based on (and reflect) biologically relevant observations and, contrary to Klar's assertion, lead to testable predictions. In contrast, Klar's new model based on asymmetric stem-cell division does not take into account some relevant biological observations that do not support his hypothesis.

During the maintenance of phyllotactic patterning, nascent multicellular organs (leaf primordia) determine the position of incipient primordia. For example, experimental data show that primordia can be initiated in the 'wrong' position by modifying either the hormonal2 or biophysical3,4 context of the tissue, and that these induced primordia provide a feedback loop to determine the position of the next leaf. Most evidence supports short-distance chemical signalling as the mechanism involved in designating the site of primordium initiation, with biophysical alteration in cell-wall extensibility as a key executor of the morphogenic programme initiated. The biochemical nature of the 'morphogen' is unknown, but progress is being made in its identification.

With respect to the initiation of phyllotactic pattern, the first primordia are generated from a multicellular ball, the plant embryo. Mathematical modelling convincingly demonstrates how an asymmetric pattern of a theoretical morphogen can be generated from an initially uniform field and how such asymmetry could lead to phyllotactic patterns5. Again, the biochemical identity of the proposed morphogen is unproven, but candidate molecules have been suggested6. In contrast, histological and clonal analysis of higher plant apical meristems has so far failed to reveal any consistent pattern of stem-cell division or stem-cell lineage in relation to leaf formation7. A phyllotactic model using asymmetric cell division would predict the existence of such a prepattern.

Moreover, disruption of cellular patterning in the meristem does not lead to disruption of phyllotactic pattern8. It may be possible to imagine asymmetric division guiding leaf position in bryophytes (in which a single apical cell does undergo repeated asymmetric division to generate daughter cells that become incorporated into leaf-like organs), but in angiosperms (flowering plants), Klar's phyllotactic hypothesis represents a model too far.