A visualization framework for cell division activity and orientation in pre-anthesis ovaries of <i>Prunus</i> species

Abstract Fruit size and shape, which influence horticultural quality, are determined by the number and the size of the cells in the local region. In fruit trees, however, the difficulty of applying molecular genetic approaches has hindered a detailed understanding of the localization and orientation of cell division in developing fruit tissues. In this study, we established a novel framework to visualize cell division in pre-anthesis ovaries of three drupe crops, peach ( Prunus persica ), Japanese apricot ( P. mume ) and the interspecific hybrid Japanese apricot ( P. salicina x P. mume ), providing clear insight into the spatial distribution and orientation of dividing cells. We systematically optimized a 5-ethynyl-2′-deoxyuridine (EdU) labeling protocol for thick ovary tissues by adjusting infiltration conditions and fixation methods. In addition, electron microscopy combined with wide-view tiling visualization was applied to directly identify dividing cells, including those undergoing chromosome segregation and cell plate formation. By combining with machine learning-based detection, we efficiently and objectively identified dividing cells. Using these complementary approaches, we found that cell division activity was broadly distributed throughout pre-anthesis ovaries in all three crops, without pronounced spatial restriction. In contrast, analysis of division orientation revealed region-specific patterns: cells in the outermost exocarp divided predominantly anticlinally, whereas cells in the mesocarp divided largely periclinally, consistent with subsequent ovary (fruit) enlargement. The integrated framework presented here provides a foundation for understanding the spatial and three-dimensional regulation of fruit development and for future studies in fruit morphogenesis and horticulture.