Final Report Summary - HEMOTIONAL (HEteroblasty MOdelling: the TImetable of ONtogeny in Arabidopsis Leaves)
Developmental biology aims at understanding how small primordia with a few cells turn into differentiated organs with a consistent shape. Many genes have been identified that play an essential role in developmental processes. However, the mechanistic link between gene activity and shape generation remain unclear. One reason for this gap is that the study of growth patterns requires quantitative descriptions.
Tackling this question on plant leaves is important because they are key organs responsible for primary carbon fixation, thus offering important prospects for addressing the global challenge of food security or forecasting the mitigating effect of vegetation on climate changes. Furthermore, plant leaves provide a convenient system to decrypt mechanisms of growth because the first produced leaves, called juvenile leaves, have distinct morphological features than adult leaves. This phenomenon is known as heteroblasty. The major molecular networks responsible for heteroblasty have been characterized in the model plant Arabidopsis thaliana (Arabidopsis). In this species, juvenile leaves (e.g. leaf 1) have a small, round leaf blade without serrations at the margin, a long petiole, and a glabrous abaxial side, while adult leaves (e.g. leaf 6) have a large, elliptic and serrate blade, a short petiole and trichomes on the abaxial side. In addition, adult leaves have a higher cell number but a lower cell size.
Our long-term objective was to understand the dynamic establishment of these heteroblastic variations. Our group has established a quantitative, parsimonious model of Arabidopsis leaf 1 in which leaf shape arises from early events in tissue deformation, involving a single polarity organizer that rules the growth direction, and a network of morphogens that modulate growth rate locally. This modelling framework was validated against growth maps obtained using devices specially developed in our group for long-term live imaging of leaf 1. We aimed to develop a quantitative model of Arabidopsis leaf 6 using this modelling framework established for leaf 1. The published version of the model considered only the early stages of leaf development, meaning that the growth arrest had to be implemented. Thus, as a first step towards analysing heteroblasty, we focused on modelling development arrest on leaf 1, using previously available tracking data obtained at later stages of leaf development.
We hypothesised that growth slows down at the same rate over the whole leaf through a global signal that would increase with time and inhibit growth. To test this hypothesis, we started by extracting quantitative information from growth patterns obtained on unexplored, later developmental stages. Then, we used this information to cross-validate the quantitative model we were developing. By changing subtly the spatiotemporal distribution of the morphogens and testing new relationships and parameters, we produced a model that accounts for the growth arrest and remains valid on the early stages. Thus, we managed to simplify the model while extending its validity. Overall, the model and the data are consistent with the idea that growth arrests uniformly over the leaf through a single inhibitor that superimposes gradually with a distribution of morphogens that is fixed initially.
Developing this model led us to investigate how cell division could coordinate with tissue expansion. Noting that the regions which have the highest growth rate also show the most intense cell division at any developmental stage, we hypothesized that growth and division share common determinisms. We performed a thorough, simultaneous analysis of cell division and growth by direct tracking over several days. To analyse cell division data, we have developed an extension of a software previously developed in the group to quantify growth. We obtained precise developmental maps of cell division parameters such as duration of cell cycle, cell area at division, or cell growth rate prior division. Then, we incorporated cell division explicitly in our model and implemented different rules that we tested using the parameter values estimated from the biological data. Our simulations do not support a causal relationship between growth and cell division, but rather suggest that they are controlled by the same set of morphogens.
The completion of this model provides a robust, quantitative framework to study heteroblastic changes in the shape and size of the leaf and its cells. Because direct tracking is not available for adult leaves such as leaf 6, a sector analysis has been performed as an indirect method to determine how growth patterns differ from those observed in leaf 1. Modelling these heteroblastic variations still needs to be done. Additionally, we set up a genetic platform to analyse in vivo how genes involved in heteroblasty are able to modulate growth. We contributed to the construction of a novel gene activation system suitable for time-lapse microscopy. This was done by combining Gateway and Cre/lox recombination technologies in a single plasmid, that confers the expression of either a blue fluorophore marking cell outlines by default, or any gene of interest fused to another fluorophore available in some cells upon heat shock-induced recombination. The construct is currently being tested.
We also explored the effect of changing environmental conditions on heteroblasty. It is known that Arabidopsis plants grown in agar plates are much smaller than in soil. In both conditions, plants showed similar heteroblastic variations when leaves were observed at 1 mm width, and a significant increase in final size between juvenile and adult leaves. However, the final shape as well as the absence of abaxial trichomes in adult leaves grown in agar plates indicated a loss of heteroblasty in this environment. These results led us to hypothesise that the heteroblastic changes in leaf shape are autonomous until a certain leaf developmental stage, while the changes in leaf size are conditional on the plant’s capacity to exploit its environmental resources.
To further test this hypothesis, we transferred Arabidopsis plants from agar to soil at several developmental stages of leaf 6 and analysed the final size and shape. Strikingly, final leaf area could be fully restored even when plants were transferred at very late stages of leaf 6 development, while leaf shape was virtually fixed after a few days following leaf emergence. Furthermore, we aimed at testing whether the loss of heteroblasty in agar was due to an upregulation of the genes known to maintain juvenility. Fresh material have been harvested and frozen for future comparative analyses of gene expression.
As a conclusion, our work suggest i) that leaf growth arrests through a global inhibitor that accumulates over time, ii) that cell division and expansion are control by the same set of morphogens, and iii) that heteroblastic variations in leaf shape are fixed early in leaf development while changes in leaf size show much more environmental plasticity.