ERA-CAPS: From genes to shape: Towards development of a computable flower

California Institute Of Technology, Pasadena CA

Investigators

Abstract

Plants have typical shapes, and one part of each plant, the flower, has a shape and structure characteristic of the species. How growing plants create flower shapes is unknown. Understanding how flowers acquire shape is important flowers are major sources of human food. Earlier work has shown that the shape that flowers take depends on the expression of regulatory genes in the growing flowers, and on the mechanical properties of the walls of the floral cells. This project will create computer models of flowers that show all of the cells and cell walls at several stages of floral growth and add to those models the physical properties of the walls (measured by several biophysical methods) and the patterns of gene expression that create the cell walls and regulate floral shape. By comparing these patterns, the team will develop hypotheses about how the combination of changing gene expression and changing mechanical properties lead to changes in cell shape and size, and therefore to the development of flower shape. They will test the hypotheses and the models by changing wall mechanical properties experimentally and seeing if the computer model predicts the shape changes seen in the experiments. The outcomes will provide tools with wide applicability to crop plants, enabling researchers and breeders access to genes and processes that contribute to establishing floral shape. Educational efforts focus on training students and post-docs in cutting edge interdisciplinary approaches that merge biology and mathematical modeling. This project will develop a tool called the "Computable Flower" that permits the user to (i) integrate data on geometry, gene expression and biomechanics and (ii) explore, interpret and generate hypotheses based on data supported by mechanistic modeling approaches. The tool therefore provides an integrated description in the form of a 3D dynamic template of the growing flower bud. The Computable Flower will be populated with existing or novel quantitative datasets coming from experimental and computational techniques concerning the spatial distribution of regulatory molecules such as transcription factors and hormones, and the spatial expression patterns of genes involved in cell wall synthesis and remodeling which operate downstream and upstream from these regulatory networks. Information on the spatial organization and properties of structural elements of the developing flower, including cell wall stiffness, microtubule orientation and cellulose microfibril organization will be added to the template. Following this, temporal information will be included, showing correlations of gene expression, mechanical properties of the tissue, and changes in geometry. This will lead to computational models and hypotheses regarding biochemical, physical and geometrical properties with simulation outcomes quantitatively compared with experimental data. Predictions coming from the modeling will guide experiments using domain-specific perturbation of genes that influence microtubule and wall status. These transgenic lines will then be subjected to detailed quantitative growth studies to test the validity of the model or to refine it. The overall result will be a path to calculate floral phenotype from genotype that includes mechanical properties of tissues and their role in morphogenesis. The broader impact will include potential applications in agriculture as well as training of students and postdocs in the combination of experimental and computational developmental biology. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.