Quantitative imaging of plants: multi-scale data for better plant anatomy

| 343 species to better follow the mechanisms controlling seed production under changeable environments, which will be a continuing trait of importance for sustainable grain production of our most important crops. Keywords: Fertilization, flower, ovary, pericarp, pollination, wheat. Journal of Experimental Botany, Vol. 69, No. 3 pp. 341-343, 2018 doi:10.1093/jxb/erx430 References Insight Quantitative imaging of plants: multi-scale data for better plant anatomy David Legland*, Marie-Françoise Devaux and Fabienne Guillon UR1268 Biopolymères, Interactions et Assemblages, INRA, France * Correspondence: The ongoing development of imaging systems continuously brings novel possibilities for the exploration of plant anatomy at different scales. However, increasing resolution often results in a smaller field of view, limiting the scope for wider conclusions. Staedler et al. (2018) got round this problem by making use of 3D images acquired at two different scales to estimate the number of pollen grains within flowers. It is a powerful approach, providing much more information than with a single scale. An understanding of the biological functions, development, or evolution of plants requires an accurate description of their anatomy at various scales: the whole organism, its organs, tissues within each organ, cells within a tissue, the cell walls, or the organelles within a cell. Depending on the representative scale of the structures of interest, various image acquisition devices can be employed to investigate their morphology, chemical composition, or spatial organization (Rousseau et al., 2015) (see Box 1). Historically, microscopy has been the usual technique for investigating plant anatomy at the cellular or tissue scale, and the rise of confocal microscopy has allowed us to perceive the 3D structure of tissues or organs with a resolution at the micron level (Truernit et al., 2008). But new technologies – such as the recent development of super-resolution techniques (e.g. PALM or STORM) or the introduction of optical coherence tomography (OCT) (Lee et al., 2006) – continuously bring novel imaging possibilities. For imaging cell walls or organelles within the cells, electron microscopy has often been the method of choice, reaching resolutions at the nanometre scale. The 3D structure can also be assessed, either by combining scanning electron microscopy with serial sectioning of the specimen (Bhawana et al., 2014), or by adapting tomography algorithms to transmission electron microscopy. Magnetic resonance imaging (MRI) and X-ray computed tomography are popular methods for the non-destructive investigation of the 3D architecture of biological specimens, without the need for staining, sectioning or inclusion. The high resolution reached by computed tomography (below the micron) often makes it the best method for the investigation of plant organs (Stuppy et al., 2003; Cloetens et al., 2006; Dhondt et al., 2010; Staedler et al., 2013). Staedler et al. (2018) took advantage of this resolution to quantify the 3D anatomy of orchid inflorescences, and through this showed differences in reproductive © The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Experimental Biology. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Chaudury A, Ming L, Miller C, Craig S, Dennis E, Peackok J. 1997. Fertilization-independent seed development in Arabidopsis thaliana. Plant Biology 94, 4223–4228. Hucl P. 1996. Out-crossing rates for 10 Canadian spring wheat cultivars. Canadian Journal of Plant Science 76, 423–427. Longin CFH, Mühleisen J, Maurer HP, Zhang H, Gowda M, Reif JC. 2012. Hybrid breeding in autogamous cereals. Theoretical and Applied Genetics 125, 1087–1096. Martin TJ. 1990. Outcrossing in 12 hard red winter-wheat cultivars. Crop Science 30, 59–62. Mette MF, Gils M, Longin CFH, Reif JC. 2015. Hybrid breeding in wheat. In: Ogihara Y, Takumi S, Handa H, eds. Advances in wheat genetics: from genome to field. Tokyo: Springer. Okada T, Jayasinghe R, Nansamba M, et al. 2017. Unfertilized ovary pushes wheat flower open for cross-pollination. Journal of Experimental Botany 68, 395–408. Tester M, Langridge P. 2010. Breeding technologies to increase crop production in a changing world. Science 327, 818–822. Whitford R, Fleury D, Reif JC, Garcia M, Okada T, Korzun V, Langridge P. 2013. Hybrid breeding in wheat: technologies to improve hybrid wheat seed production. Journal of Experimental Botany 64, 5411–5428. Wuest SE, Philipp MA, Guthörl D, Schmid B, Grossniklaus U. 2016. Seed production affects maternal growth and senescence in Arabidopsis. Plant Physiology 171, 392–404. 344 | Box 1. Multi-scale imaging of wheat grain investment between inflorescences of rewarding and deceiving orchids. The physical properties of image acquisition devices limit the total quantity of information that can be gained, and so there is a compromise between a high resolution and a large field of view. When the spatial resolution is too low, the smallest structures are difficult to identify. On the other hand, the smaller the field of view, the more difficult it is to cover the totality of the organ of interest with a reasonable acquisition time. The increase in resolution of reconstructed images therefore often corresponds to a reduction in the size of the field of view. This difficulty was encountered in the work of Staedler et al. (2018): the structures of interest (pollen grains) could not be imaged with a resolution that allowed their identification while taking into account the whole reference structure (the pollinium, an aggregate of pollen grains). The strategy adopted to circumvent this difficulty was to acquire images at two different resolutions. Images acquired at finer resolution were used for segmentation and counting the pollen grains; images at a coarser resolution were used for assessing the size and shape of the reference structure. The total number of pollen grains was then estimated by combining their numerical density with the volume of the pollinium. It is an approach which exemplifies how data obtained at different resolutions may be used together to provide much more information than data at a single resolution. The multi-scale investigation of plant tissues is undoubtedly a promising strategy for a better description and understanding of plant anatomy. However, investigation and integration of images, obtained both at different scales and using different imaging modalities (see below), raise new methodological questions (Rousseau et al., 2015). Joint exploration of multi-scale images Investigating plant anatomy at different scales often relies on different imaging modalities. A common approach in microscopy for combining these (...truncated)