Flow Cytometry for Estimating Plant Genome Size: Revisiting Assumptions, Sources of Variation, Reference Standards, and Best Practices

Abstract

Abstract Background Flow cytometry has been widely used to estimate relative and absolute genome sizes (DNA contents) of plants for over 50 years. However, the accuracy of these estimates can vary widely due to many factors, including errors in the genome size estimates of reference standards and various experimental methods. The objectives of this study were to reassess genome sizes of commonly used reference standards and to quantify sources of variation and error in estimating plant genome sizes that arise from buffers, confounding plant tissues, tissue types, and plant reference standards using both DAPI (4′,6-diamidino-2-phenylindole) and PI (propidium iodide) fluorochromes. Results Five separate studies were completed to elucidate these objectives. Revised estimates of genome sizes of commonly used plant reference standards were determined using human male leukocytes and updated estimates of the genome size of human male leukocytes (6.15 pg, 12.14% lower than earlier studies) with both DAPI and PI fluorochromes. Comparison of six different extraction buffers (Galbraith’s, LB01, MB01, MgSO4, Otto’s, and Sysmex) resulted in variation in genome size estimates by as much as 18.1% for a given taxon depending on the buffer/fluorochrome combination. The addition of different confounding plant tissues (representing 10 diverse taxa and associated secondary metabolites) resulted in variation in genome size estimates by as much as 10.3%, depending on the tissue/fluorochrome combination. Different plant tissue types (leaf color/exposure and roots) resulted in a variation in genome size estimates of 10.7%, independent of the fluorochrome. The selection of different internal reference standards introduced additional variation in genome size estimates of 5.9% depending on the standard/fluorochrome combination. The choice of fluorochrome (DAPI vs. PI) had one of the largest impacts on variation in genome size and differed by as much as 32.9% for Glycine max ‘Polanka’ when using human male leucocytes as an internal standard. A portion of this variation (~10.0%) can be attributed to the base pair bias of DAPI and variation in AT:CG ratios between the sample and standard. However, as much as 22.9% of the variation in genome size estimates may result from how effectively these fluorochromes stain and report the genome. The combined variation/error from all these factors (excluding variation from base pair bias for different fluorochromes and assuming variation from confounding tissues and tissue types to both result from secondary metabolites) the additive experimental error totaled 57.6%. Additional details of how selected factors impact accuracy, precision, and the interaction of these factors are presented. Conclusions Overall, flow cytometry can be precise, repeatable, and extremely valuable for determining the relative genome size and ploidy of closely related plants when using consistent methods, regardless of fluorochrome. However, accurate determination of absolute genome size by flow cytometry remains elusive and estimates of genome size using flow cytometry should be considered gross approximations that may vary by ± 29% or more as a function of experimental methods and plant environment. Additional recommendations on best practices are provided.