The use of image analysis and automation for measuring mitotic index in apical conifer meristems

• Published in: Journal of experimental botany Vol. 49; no. 327; pp. 1749 - 1756
• Main Authors: Sundblad, Lars-Göran, Geladi, Paul, Dunberg, Arne, Sundberg, Björn
• Format: Journal Article
• Language: English
• Published: Oxford OXFORD UNIVERSITY PRESS 01.10.1998; Oxford University Press
• Subjects: Agronomy. Soil science and plant productions; Analytical estimating; Biological and medical sciences; Cell cycle; Cell nucleus; Cell physiology; Conifers; Economic plant physiology; Fundamental and applied biological sciences. Psychology; General aspects, methods; Image analysis; Interphase; METHODS PAPER; Mitotic index; Morphology. Anatomy. Histology. Cytology; Pixels; Plant physiology and development; Principal components analysis; Squashes; Image analysis; Automation; Mitotic index; Softwood forest tree; Gymnospermae; Coniferales; Spermatophyta; Measurement method; Pinus sylvestris; Apex; Meristem
• ISSN: 0022-0957; 1460-2431; 1460-2431
• DOI: 10.1093/jexbot/49.327.1749

The methodology for determination of mitotic index (MI) from apical meristems of conifers was improved to permit the efficient processing of large sample numbers. Improvements were made at three different stages of the method. Firstly, hydrolysis, staining, cytoplasmic bleaching, washing of samples, and temperature regimes were automated, which reduced the need for labour and improved the standardization of chemical treatment. Secondly, the use of vertical and controlled pressure for squashing improved the quality of the preparations and decreased the fraction of discarded preparations. Thirdly, an interactive image analysis system for estimation of MI from preparations was constructed. This system increased the efficiency of analysis of preparations, but did not eliminate subjective manual classification of nuclei into cell cycle stages. The possibility of using fully automated image analysis for estimation of MI was investigated using a standard image processing sequence and by multivariate analysis of image analysis parameters. For this, principal component analysis (PCA) was used to detect cell cycle stage related clustering of nuclei in score plots. PCA was also used to construct a model based on interphase nuclei that enabled correct classification of 25 nuclei from five cell cycle stages as either dividing or non-dividing.