Chloroplast volume is underestimated from two-dimensional cross sections

Harwood RH and Barbour MB

University of Sydney, The Center For Carbon Water and Food, 380 Werombi Rd, Brownlow Hill NSW 2570.

Cell ultrastructure is predominantly studied by transmission electron microscopy (TEM), providing the user detailed two-dimensional (2D) information. Recent advances in microscopy have streamlined the acquisition of three-dimensional (3D) images. Scanning electron microscopy with an automated microtome (SBFSEM) produces serial micrographs that can be stacked and segmented to produce a three-dimensional volume data set. We are using SBFSEM on leaf cells to explore organelle size, shape and position, along with cell density and packing. The 3D anatomical data produced is being used to explore the relationship between leaf form and function. Models of key leaf processes, such as photosynthesis, sit at the heart of crop productivity and climate change models but include significant assumptions regarding the structure of leaves that ignore 3D complexity. SBFSEM allows us to challenge the idea of “textbook” leaf cell. To date we have focused on wheat and found that chloroplast volume estimates from 2D cross sections underestimated volume by 54% in mesophyll cells and 44% in bundle sheath cells. Chloroplast in meosphyll cells were 35% bigger then those in bundle sheath cells when measured in 3D, when estimated from 2D cross sections the difference was only 20%. Size, shape and relative distance of cells and organelles are critical in addressing leaf form and function, our research shows over simplification from 2D quantification limits our understanding of leaf function. Using volume microscopy, we can demonstrate uniform geometry assumptions can no longer be made in plant physiology.