Hyperforin Accumulates in the Translucent Glands of Hypericum perforatum

Abstract

Background and Aims

Hypericum perforatum contains the therapeutically important compounds hypericin and hyperforin. Hypericin is known to accumulate in the dark glands. This investigation aimed to determine the accumulation site of hyperforin.

Methods

Dark and translucent glands as well as non-secretory tissue in leaves were manually isolated under the microscope. Hyperforin content was quantified by UV HPLC. Secretory structures were surveyed anatomically.

Key Results

The hyperforin content of intact leaves was found to be about 3 mg g⁻¹ fresh tissue, whereas a content of about 7 mg g⁻¹ fresh material was found in isolated translucent glands. Hyperforin was found only to occur in minute amounts in dark glands (approx. 0·4 mg g⁻¹ fresh tissue). In non-secretory tissue no hyperforin was detected.

Conclusions

The accumulation of hyperforin detected in the translucent glands supports the proposed hypothesis that hyperforin is synthesized by the same biosynthetic machinery as monoterpenes in the chloroplasts of cells delimiting the gland.

INTRODUCTION

Hypericum perforatum is a common perennial herb of the northern hemisphere. In current phytomedicine, the aerial parts are established to be effective against mild to moderate depression (Chatterjee et al., 1998; Philipp et al., 1999). Two compounds, hypericin (a naphthodianthrone) and hyperforin (a prenylated phloroglucinol), have received widespread attention as a result of their pharmacological activities (Kubin et al., 2005; Beerhues, 2006). Hypericum perforatum is characterized by having different secretory structures (see Fig. 1A): dark glands, also known as (black) nodules, translucent glands and secretory canals (Ciccarelli et al., 2001a, b).

Fig. 1.

(A) Tip of axillary leaf of H. perforatum; note translucent and dark glands. (B) Cross-section of leaf with translucent gland and (C) paradermal section of leaf with translucent gland; note chloroplast-containing cells delimiting the gland. (D) Compounds of hyperforin synthesis (compounds originating from the synthesis of monoterpenes taking place in chloroplasts are enclosed by stippled lines).

It has generally been accepted that hypericin primarily accumulates in the dark glands in both leaves and flowers (Mathis and Ourisson, 1963; Robson, 1981; Ciccarelli et al., 2001a; Maffi et al., 2005) but only recently was it demonstrated that there exists a positive correlation between the size and number of dark glands and the overall content of hypericin in the plant organ (Zobayed et al., 2006). In Hypericum elodes hypericin has been shown to accumulate in the ‘red glands’ peripherally on sepals; removal of these red glands indicated that hyperforin accumulated elsewhere (Piovan et al., 2003).

The present study was undertaken to investigate the site of accumulation of hyperforin in H. perforatum.

MATERIAL AND METHODS

Sampling

On 20 July, 2006, the aerial parts of Hypericum perforatum L. (accession number E5464-0018) were collected in the Botanical Garden of the University of Copenhagen. The samples were stored in a cool, dark location for no more than 24 h. Healthy axillary leaves from the upper third of the flowering plant were used for HPLC analysis and for anatomical survey of the leaves and isolated secretory structures.

Chemicals

Reference standard hyperforin (>95 % purity) and (2-hydroxypropyl)-β-cyclodextrin were purchased from Sigma-Aldrich. Methanol, acetonitrile, ortho-phosphoric acid, phosphoric acid and water were of HPLC purity.

Extraction procedures

Prior to analysis of intact leaves, 250 mg of fresh leaves of H. perforatum was sliced finely before being added to a centrifuge glass with 2·5 mL of extraction solvent [80 % aqueous methanol with 0·073 m (2-hydroxypropyl)-β-cyclodextrin adjusted to pH 2·5 with ortho-phosphoric acid; procedure adapted from de los Reyes and Koda (2001)], and placed in an ultrasonificator for 10 min. After centrifugation of the vials (10 min at 1000 g) the supernatant was decanted and the extraction procedure repeated with the pellet three times. To the collected extracts, approximately 1 mL of extraction solvent was added to produce a final extract of 10·0 mL solvent, of which 300 µL was filtered into HPLC vials for analysis.

For manual isolation of the secretory structures, dark and translucent glands, in leaves of H. perforatum a number of syringes (diameters of 0·80 and 0·45 mm for black and translucent glands, respectively) were cut perpendicular at the sharp end. Aided by a stereo microscope, it was possible to prick out discs of leaf material and thus transfer single dark or translucent glands (still unruptured and encompassed in the adjoining leaf material) into the cavity of the syringe (Fig. 1A). From there the material was transferred to the extraction solvent. In addition to the secretory structures, non-secretory tissue was isolated for comparison. The amount of material taken from the leaves was calculated based on the weight of the axillary leaves before and after gland removal. A parallel perforation procedure was performed without removing leaf material to take into account the average loss of weight due to evaporation. For dark and translucent glands, a variable approximate number of perforations were performed, 50 and 100 respectively. On average, 0·016 mg material encompassing dark or translucent glands or non-secretory tissue was transferred into Eppendorf tubes containing 0·5 mL of the extraction solvent. Subsequently, the extracts were ultrasonicated (10 min), filtered and transferred to HPLC vials for analysis.

All procedures were performed at least in duplicate.

HPLC analysis

The HPLC system consisted of a Waters pump (600E), an in-line degasser, injector loop, auto sampler and photodiode (PDA) detector (200–800 nm). The column was a Waters Symmetry C18 (3·9 × 150 mm, 5 µm) analytical column fitted with a suitable guard column.

The method was adopted from that of Brolis et al. (1998). The mobile phase consisted of (A) 0·3 % phosphoric acid, (B) acetonitrile and (C) methanol at times: 0 min (A) 100 %; 10 min (A) 85 % and (B) 15 %; 30 min (A) 70 %, (B) 20 % and (C) 10 %; 40 min (A) 10 %, (B) 75 % and (C) 15 %; 55 min (A) 5 %, (B) 80 % and (C) 15 %. The flow rate was 1 mL min⁻¹, and the temperature in the column compartment was held at 30 °C.

Identification of hyperforin was done by internal spiking and UV spectrum comparisons with reference compounds. A calibration curve for hyperforin was constructed by analysing a series of hyperforin stock solutions: 125, 62·5, 31·3, 15·6 and 7·8 µg mL⁻¹. The curve was linear in this area, and had an r² value of 0·99.

Anatomical survey

Thin slices of H. perforatum leaf containing secretory structures were subjected to fixation with 3 % glutaraldehyde in 0·1 m phosphate buffer, pH 7, dehydrated and embedded by polymerization in glycolmethacrylate monomer mixture. Microtome sections 4 µm thick cut with a glass knife were stained by the Periodic Acid – Schiff reaction followed by Aniline Blue Black, for identification of polysaccharides and proteins, respectively (Feder and O'Brien, 1968).

A Reichert-Jung Polyvar light microscope was used for surveying the anatomy of the secretory structures. Images was captured with an ‘Evolution LC’ CCD camera, as TIFF files, and processed via the computer program Image-Pro Plus.

RESULTS AND DISCUSSION

The results of the HPLC analysis of intact leaves and isolated leaf components are presented in Fig. 2. The hyperforin content of intact leaves was found to be about 3 mg g⁻¹ fresh tissue. The amount of hyperforin g⁻¹ fresh material found in intact leaves was within the range of the values found by Zobayed et al. (2005) for hyperforin content in H. perforatum shoots (grown at 15 °C/20 °C), i.e. 1–1·5 mg g⁻¹ fresh weight of both stem and leaf tissue.

Fig. 2.

Hyperforin content detected in (IL) intact leaves, (DG) dark glands, (TG) translucent glands and (NS) non-secretory tissue.

A high hyperforin content of about 7 mg g⁻¹ fresh material was found in isolated translucent glands (Fig. 2). Hyperforin was found only to occur in minute amounts in dark glands (about 0·4 mg g⁻¹ fresh tissue). In non-secretory tissue no hyperforin was detected, which indicates that the hyperforin content found in the secretory structures derives solely from the gland and/or the immediately adjacent tissue. The concentrations of hyperforin found in translucent glands exceeded that of intact leaves by more than 100 %, indicating that the translucent glands are the site of accumulation.

The dark glands were found to contain small amounts of hyperforin, indicating that dark glands could be a minor site of accumulation, but this result could potentially be due to a single instance of contamination by the numerous translucent glands.

There were strong indications that hyperforin is synthesized in the translucent glands and their delimiting cells: the lipophilic isoprenoid moiety of the hyperforin molecule is synthesized by the same pathway as the monoterpenes, which are constituents of the essential oil present in the translucent glands (Ciccarelli et al., 2001b; Adam et al., 2002). Several steps in the non-mevalonate pathway are in common for the synthesis of hyperforin and monoterpenes (Fig. 1D). In the biosynthesis of hyperforin, three molecules of dimethyllalyl diphosphate and one of geranyl diphosphate are adjoined to the phloroglucinol moiety by electrophilic substitution, which by an additional ring closure produces the structure of hyperforin (Adam et al., 2002). In the synthesis of monoterpenes, the molecules dimethyllalyl diphosphate and geranyl diphosphate are hydrolysed to form a simple monoterpene (Croteau, 1987).

Thus, it is evident that the biosyntheses of essential oil and hyperforin are intimately connected, and it can be proposed that their syntheses take place at the same site.

At the mature stage the translucent glands are cavities spanning from the upper to lower epidermis (Fig. 1B). The cavity is surrounded by an inner layer of collapsed secretory cells, which are delimited towards the mesophyll by an outer layer of large, turgescent, mesophyll-like, chloroplast-containing cells (Fig. 1C). This is in accordance with the observations of Curtis and Lersten (1990) and Ciccarelli et al. (2001b). The translucent glands in H. perforatum were found by Ciccarelli et al. (2001b) to have the same ontogeny – by schizogeny – as the type B canals in the sepals and petals. In this regard, it is noteworthy that Piovan et al. (2004) for H. elodes found hyperforin in sepals in tissues containing type B canals.

Synthesis of essential oil is linked to chloroplasts (Croteau, 1987). Thus, the essential oil in the translucent cavity is likely to be produced in these chloroplast-containing cells delimiting the translucent glands, and it can thus be proposed that these cells also constitute the site of synthesis of the isoprenoid moieties of the hyperforin molecule.

The other significant secretory structure in H. perforatum, the dark gland, has been shown to be the site of accumulation of hypericin (Zobayed et al., 2006). The demonstrated hypericin accumulation in dark glands, combined with the results of the current study, indicates that there exists an unequivocal division and confinement of these two therapeutically important bioactive substances.

The apparent overall division of hyperforin and hypericin into two distinct secretory structures could be due to an antagonism of the underlying pathways of synthesis, or the molecules' differing preferences for the polarity of the medium in which they accumulate. It might also be an adaptation to prevent the breakdown of the unstable hyperforin molecule by the singlet oxygen and super-oxide radicals that hypericin produce upon light activation (Schmitt et al., 2006).

It would be interesting in a future study to investigate the potential correlation between the numbers and size of the translucent glands and the content of hyperforin, for example by assessing the amount of light transmitted through the translucent glands.