Planta Med 2010; 76 - P519
DOI: 10.1055/s-0030-1264817

Content of antidepressant hyperforin correlates with translucent glands in Hypericum perforatum L.

J Soelberg 1, B Johansen 1, A Jäger 2
  • 1University of Copenhagen, Biological Institute, Osterfarimagsgade 2D, 1123 Copenhagen, Denmark
  • 2University of Copenhagen, Medicinal Chemistry, Universitetsparken 2, 2100 Copenhagen, Denmark

Hypericum perforatum L. synthesizes two therapeutically important compounds, hypericin and hyperforin, which have recently been shown to accumulate in secretory structures: the dark glands and the translucent glands respectively [1]. Photographic images of H. perforatum leaves were computer-analyzed to produce measures of total gland area and total leaf area. A single leaf was placed between two microscope slides and positioned in a microscope fitted with light externally and from below. Since the translucent glands span the depth of the leaf, they clearly show when lit from behind, and could be captured in high-definition images (CCD camera „Evolution LC“, exposure at 700ms). The images were captured as TIFF files, and processed in computer program Image-Pro Plus 5.1 (conversion to grey tones; masking of anything outside the leaf outline; and finally count and size-measure of any bright objects (the translucent glands) of a defined roundness (0–2 arbitrary units) and within a defined size (0.001–1 mm2). The leaves were then subjected to extraction and HPLC analysis for hyperforin content. The content of hyperforin in a H. perforatum leaf correlated with the total area of translucent glands (correlation 0.71), whereas there was no correlation with the leaf area (correlation 0.09). This means that it is not the biomass of a plant, but the number and size of translucent glands that influence the yield of hyperforin in a crop. This applied method indicates that hyperforin content in H. perforatum leaves could be estimated without use of analytical chemistry, simply by observation and calculation of translucent gland area.

References: 1. Soelberg J. et al. (2007) Ann Bot 99: 1097–1100.