Anthocyanin inhibits propidium iodide DNA fluorescence in Euphorbia pulcherrima: implications for genome size variation and flow cytometry

Abstract

Background: Measuring genome size by flow cytometry assumes direct proportionality between nuclear DNA staining and DNA amount. By 1997 it was recognized that secondary metabolites may affect DNA staining, thereby causing inaccuracy. Here experiments are reported with poinsettia (Euphorbia pulcherrima) with green leaves and red bracts rich in phenolics.

Methods: DNA content was estimated as fluorescence of propidium iodide (PI)-stained nuclei of poinsettia and/or pea (Pisum sativum) using flow cytometry. Tissue was chopped, or two tissues co-chopped, in Galbraith buffer alone or with six concentrations of cyanidin-3-rutinoside (a cyanidin-3-rhamnoglucoside contributing to red coloration in poinsettia).

Key results: There were large differences in PI staining (35-70 %) between 2C nuclei from green leaf and red bract tissue in poinsettia. These largely disappeared when pea leaflets were co-chopped with poinsettia tissue as an internal standard. However, smaller (2.8-6.9 %) differences remained, and red bracts gave significantly lower 1C genome size estimates (1.69-1.76 pg) than green leaves (1.81 pg). Chopping pea or poinsettia tissue in buffer with 0-200 microm cyanidin-3-rutinoside showed that the effects of natural inhibitors in red bracts of poinsettia on PI staining were largely reproduced in a dose-dependent way by this anthocyanin.

Conclusions: Given their near-ubiquitous distribution, many suspected roles and known affects on DNA staining, anthocyanins are a potent, potential cause of significant error variation in genome size estimations for many plant tissues and taxa. This has important implications of wide practical and theoretical significance. When choosing genome size calibration standards it seems prudent to select materials producing little or no anthocyanin. Reviewing the literature identifies clear examples in which claims of intraspecific variation in genome size are probably artefacts caused by natural variation in anthocyanin levels or correlated with environmental factors known to induce variation in pigmentation.

Fig. 1.

Relative DNA staining in nuclei from green leaf (A) or red bract tissue (B) of Euphorbia pulcherrima chopped alone, or in ‘co-chopped’ mixtures processed in the order green leaf first and red bract second (C), red bract first and green leaf second (D), or when samples C and D were combined and run immediately after (E), or run 2 h later (F). The mean relative absorbance for nuclei in each peak is given.

Fig. 2.

Relative DNA staining in 2C nuclei of Pisum sativum co-chopped with green leaf (A) or red bract tissue (B) of Euphorbia pulcherrima and of Pisum sativum suspended in Galbraith buffer with no added anthocyanin (C), or with 200 µm of added cyanidin-3-rutinoside (D), or (E) after stained nuclei of C and D were mixed. The mean relative absorbance for nuclei in each peak is given.

Fig. 3.

Relative DNA staining in 2C nuclei of Pisum sativum, green leaf or red bract of Euphorbia pulcherrima suspended in Galbraith buffer with no added anthocyanin or with five concentrations of added cyanidin-3-rutinoside.

Fig. 4.

Model suggesting how differences in key geographical and topological factors, and in many associated abiotic environmental characters commonly reported to enhance anthocyanin levels in plants, may produce environmentally correlated pseudo-intraspecific genome size variation. Note, this may apply to many cytosolic compounds beside anthocyanin which interfere with DNA staining for nuclear DNA content estimation, for example furanocoumarins (Walker et al., 2006).