The making of a leaf tip: how cell division angles define shape

Abstract

Leaves come in an extraordinary range of shapes: they can be simple or dissected, can have smooth or toothed margins and can display a variety of lobes, points and indentations. These shape variations have important functional consequences, affecting light capture, water loss, wind resistance and even herbivore interactions (Chitwood & Sinha, 2016). Leaves originate as small primordia on the flanks of the shoot apical meristem, from which they grow and expand until they reach their mature form. The rate, duration and spatial patterning of cell division and expansion are key determinants of leaf size and shape. These processes are finely tuned by complex gene regulatory networks that respond to both internal and external signals (Schneider et al., 2024).

Hirokazu Tsukaya has long studied the mechanisms underpinning leaf development. Starting with the simple ovate leaves of the model plant Arabidopsis thaliana, his group now explores leaf shape from an evo-devo perspective, in a wide range of model and non-model species. Among these is Triadica sebifera, the Chinese tallow tree – a member of the spurge family and an oil-producing species of economic significance. Its leaves are distinctly shaped, featuring a rounded base and a sharply tapering (acuminate) tip, linked by concave joint regions (Figure 1a). This shape can be quantified by its curvature, that is, how much the contour deviates from a straight line (Figure 1b): the base shows moderately positive curvature (convex, rounded), the joint regions display negative curvature (concave), and the tip reaches a maximum curvature with a sharp point. Until now, it has remained unknown how this distinctive apex forms.

Zining Wang, a Ph.D. student in Tsukaya's lab and first author of the highlighted study, combines mathematical modelling with developmental biology to investigate leaf shape formation. Surprised by the lack of research into acuminate tip formation, Wang took on the challenge of dissecting leaf shape formation in T. sebifera – made more complex by it being a woody, non-model species that required significant effort to establish reliable germination and growth protocols. Wang began by describing leaf growth using contour growth mapping, an approach that represents the leaf outline as a series of contour points. Growth is simulated by moving these points outward at defined speeds. When growth was isotropic (equal in all directions), the model produced a circular leaf. Vertical anisotropic growth yielded an elliptical form. However, combining isotropic growth at the base with vertical anisotropic growth at the tip generated a shape resembling T. sebifera leaves, indicating that distinct regional growth patterns are required.

To assess whether this biregional growth is reflected at the cellular level, Wang and colleagues analysed cell shapes and division patterns during leaf development. Interestingly, the sharp apex was already present in very young primordia, despite apical cells lacking elongation, and no elongation was observed after this stage. This indicated that cell elongation was not responsible for tip formation. Instead, the orientation of cell division emerged as a key factor. Using the thymidine analog 5-ethynyl-2′-deoxyuridine (EdU) to label replicating cells (Yin & Tsukaya, 2016), the researchers tracked the angles of subsequent divisions relative to the medial-lateral axis of the leaf. They observed vertically aligned (tip-to-base) division angles in the apex, leading to horizontally oriented cells, whereas division angles at the base changed from vertical to random orientation as the primordium matured.

These findings prompted Wang et al. to test whether a biregional pattern of cell division angles alone could generate the observed leaf shape. They turned to a more detailed vertex model (Nagai & Honda, 2001), which simulates cellular growth by representing cells as polygons, and cell divisions as the splitting of these shapes. Random division angles yielded a circular form, while uniformly vertical divisions created a rod-like shape. However, fixing the division angle to be vertical only in the apex region produced a form closely resembling T. sebifera leaves (Figure 1c). In these simulations, the sharpness of the apex was primarily determined by the degree of bias towards vertical division, while the length of the apex depended on the position of the apex–base boundary.

Spatial control of division angles thus appears sufficient to explain T. sebifera's leaf shape. Yet Wang's previous observations showed that division angles in the base region change over time and that an ‘arrest front’ forms, progressively restricting division to the basal region of the leaf. This suggested a temporal component to leaf shape control. Wang et al. therefore incorporated this idea into their model, adding dynamic changes in division orientation and an advancing arrest front. Remarkably, this version of the model was also able to reproduce the characteristic leaf shape, even without predefined regional differences in the division angle (Figure 1d).

In conclusion, Wang et al. show that both spatial and temporal regulation of division angles can generate leaves with sharply pointed tips. Tsukaya views these not as competing hypotheses, but as complementary aspects of a unified developmental process, where spatial patterns represent the outcome of shifting division angles and division activity over time. The modelling framework established in this study could apply to other species with pointed leaves, and Tsukaya speculates that species with multiple leaf tips, such as Japanese maple, may repeatedly use this same developmental module to generate their intricate shapes.