|Home | About | Journals | Submit | Contact Us | Français|
The flowering plant genus Hypericum (Hypericaceae) contains the well-known medicinally valuable species Hypericum perforatum (common St. John’s wort). Species of Hypericum contain many bioactive constituents, including proanthocyanins, flavonoids, biflavonoids, xanthones, phenylpropanes and naphthodianthrones that are characterized by their relative hydrophilicity, as well as acylphloroglucinols and essential oil components that are more hydrophobic in nature. A concise review of the scientific literature pertaining to constituents of Hypericum essential oils and volatile fractions is presented.
The genus Hypericum contains, by the most recent count, 469 species that have been classified into 36 taxonomic sections [1,2]. Hypericum perforatum L. (Common St. John’s wort), the most thoroughly studied taxon of the genus, is well known for its historical and contemporary use as a medicinal plant, particularly for the treatment of mild to moderate depression . Monographs for the crude drug (Hyperici herba), extracts of which are prepared from the aerial flowering portions of the plant, have been included in the European Pharmacopoeia (fully integrated as of Ph Eur 4) . An infused oil of the flowers (Oleum hyperici), which is prepared by macerating fresh flowers in olive or sunflower oil and exposing the mixture to sunlight for two to three weeks, has a history of traditional use in Europe for treatment of burns and ulcers. The topical application is included in the monograph of the German Commission E and the Swiss Pharmacopoeia [5,6].
Oleum hyperici has a red color when either fresh (i.e. water-containing) flowers are extracted or heat is applied during the maceration process, although the naphthodianthrones (specifically hypericin and pseudohypericin) are not extracted into the oil. It has been proposed that a related emodin-derivative(s), specifically a degradation product of hypericin upon exposure to sunlight, is responsible for this coloration, but these have not yet been isolated [7-10]. The lipophilic phloroglucinol derivative, hyperforin, considered to be one of the primary bioactive (e.g. antibacterial, antidepressant) constituents of H. perforatum, along with a related compound, adhyperforin, are extracted into the oil [3,9,11-12].
Maisenbacher and Kovar  reported that these compounds degraded quickly in the presence of air, heat and light, and Isacchi et al.  demonstrated that when stored in the dark at −20°C, the hyperforin content had decreased by more than 60% and adhyperforin was no longer detectable after one year.
Primary production areas for H. perforatum in Europe include Germany, Italy and Romania, and the majority of the harvest goes toward the production of the crude drug . A small fraction of the plant material is sold for the production of Oleum hyperici, which is also used as a carrier oil in aromatherapy, and for the distillation of the essential oil. The applications of the infused oil in aromatherapy and massage therapy mirror to some extent that of Hyperici herba, being used to treat anxiety, depression, bruises, mild wounds, rheumatic pain and sunburn [5,6,14]. Studies have shown that the infused oil demonstrates anti-inflammatory activity and speeds wound healing through the stimulation of epithelial tissue production when applied ex vivo, and exhibits gastroprotective effects when taken orally [10,15]. The steam-distilled essential oil, which has only recently become available on a broader scale on the market, is largely produced by small companies in Serbia, Croatia, Poland, Bulgaria, France, Canada and the United States . Although evidence for the absorption of certain components of essential oils through the skin or lungs exists , clinical data for the efficacy of Hypericum essential oil applied topically in a pure form or diffused in a carrier oil are still lacking.
Hypericum species are generally classed as essential oil-poor plants (generally oil yield <1%, w/w) [17,18]. The content of essential oils in healthy H. perforatum plants is highest during the full-bloom stage versus pre-bloom or fruiting stage (0.35% versus 0.12 and 0.18%, respectively) [18-20]. Research has shown that essential oil extraction efficiency with liquid or supercritical CO2 (yield 1.00%, w/w) is higher than with steam distillation (0.06%, w/w). Hyperforin and derivatives can be extracted using liquid or supercritical CO2, but not with steam distillation [21,22].
Exudate-containing glands and canals are characteristic features of members of Hypericaceae, as well as of the related families Clusiaceae and Calophyllaceae (Guttiferae sensu lato), and are found on most organs of the plants [23,24]. In Hypericum L. (Hypericaceae), these glands appear as lines or dots, and are of two types: translucent (pale yellow to amber in color) and dark (red to nearly black), found variously on the stems, leaves, sepals, petals and the anther connective . Anatomical and histochemical studies of H. perforatum have shown that the translucent glands are actually sub-epidermal cavities, lined with two layers of flattened, thin-walled secretory cells . Dark glands, meanwhile, are specialized clusters of cells containing wax and the intensely red naphthodianthrone pigments hypericin and pseudohypericin [26-28]. The number and size of the dark glands in H. perforatum have been shown to correlate positively with naphthodianthrone content .
High hyperforin content (ca. 7 mg/g fresh tissue) has been measured in isolated translucent glands, but detected in only minor amounts (ca. 0.4 mg/g fresh tissue) in dark glands (possibly due to cross-contamination), and not detected at all in non-secretory tissue . The isoprenoid moiety of hyperforin is produced via the same biosynthetic pathway leading to the monoterpenes, which are commonly major constituents of essential oils in Hypericum (e.g. α- and β-pinene, limonene). Evidence suggests that essential oil components and hyperforin (as well as derivatives) are biosynthesized within chloroplasts present in the thin cells surrounding the translucent glands and then secreted into the cavities [28,30,31].
Secretory canals in the floral and vegetative tissue in H. perforatum are of three types. Type A canals generally have a narrow diameter (although this can be variable in the pistil and root tissue) and are delimited by four or more polygonal cells. Type B and C canals are structurally similar to the translucent glands, but have different patterns of ontogeny . Piovan et al.  identified hyperforin in sepal tissue containing Type B canals in H. elodes L., which is interesting in light of the shared ontogeny between this canal type and the translucent glands [25,28,32]. The distribution of translucent and dark glands; Type A, B and C canals; and hypericin and hyperforin in H. perforatum are summarized in Table 1. Anatomical and histochemical examinations of the translucent and dark glands have been performed in more than 20 other Hypericum species, including H. androsaemum, H. erectum, H. seniawinii and H. richeri [33-37].
Methods undertaken to prepare samples prior to detection of either naphthodianthrones or acylphloroglucinols included excision of individual glands and/or pieces of tissue using razor blades , syringe tips , silica microcapillaries, and the technique of laser microdissection . Once samples were excised from the respective tissue, they were either extracted in an organic solvent that was analyzed using HPLC-MS [28,32] or they were directly analyzed using laser desorption/ionization mass spectrometric (LDI-MS) imaging . The solid phase micro extraction (SPME) technique, in which volatile and semi-volatile organic compounds are directly adsorbed on to a polymer coating on a thin fiber and subjected to GC-MS or HPLC-MS analysis, could potentially be used to examine the chemical content of stalked glands that occur in some species of Hypericum (e.g. on the margins of the sepals and/or petals), the glands on the anther connectives, or the glands on the outer surface of the ovaries.
Essential oil and volatile constituents that have been most frequently reported from Hypericum include the aliphatic hydrocarbons n-nonane and n-undecane; the monoterpenes α- and β-pinene; and the sesquiterpenes β-caryophyllene and caryophyllene oxide. The results of essential oil and volatile constituent analyses for representatives from 22 of the 36 taxonomic sections of Hypericum have been published, and a summary of major constituents is provided in Table 2 (except H. perforatum). A number of major components have been identified from these species that have a relatively limited occurrence among higher plants. Because some of these compounds may have a potential as food and/or beverage additives (flavoring agents), in cosmetics, or as aroma chemicals (e.g. see uses of β-caryophyllene in ), an interest in further research on targeted breeding programs for selected Hypericum species exists.
Many studies of the essential oil content of H. perforatum, in which samples from a single population were taken, have been performed. The results of these studies are summarized in Table 3, and indicate the enormous variability inherent in the volatile chemistry of this species. Typical essential oil constituents for this species include the monoterpenes α- and β-pinene, limonene and myrcene; the sesquiterpenes β-caryophyllene and caryophyllene oxide; and hydrocarbons such as n-decane, C16- and C29 alkanes and C24-, C26-, and C28-alkanols . Much variation between subspecies has also been documented, which is not surprising given the broad range of morphological variability encompassed by this species . When the essential oils from flowers alone from two subspecies of H. perforatum collected in central Italy were directly compared, α-pinene and 2,6-dimethyloctane were identified as dominant in ssp. veronense, while β-caryophyllene and 2,6-dimethylheptane dominated in ssp. perforatum . The oil yield from plants identified as H. perforatum ssp. perforatum growing in Serbia was markedly greater than from those identified as ssp. angustifolium, indicating that variability in both content and composition exists at this taxonomic level .
Numerous observations of intra- and interpopulation variability in H. perforatum oil content and composition for plants have been made. Because the plants for these studies were collected from the field rather than grown under controlled conditions, environmental factors are proposed to have contributed to this observed variability. For example, variation in oil content (0.04-1.93%) and dominant components was reported for plants from six different localities in Serbia. Terpenoid constituents, particularly sesquiterpenes, were described as primary constituents in the oil of the plants collected from mountainous regions, while long-chain waxes and fats were reported as dominating in plants from lowland areas . Previous studies with other plant species have shown similar results, with concentrations of monoterpenoids and sesquiterpenoids increasing with increasing altitudes, hypothetically due to the role of these compounds in helping the plant deal with abiotic stress factors (e.g. UV radiation) . An examination of H. perforatum plants growing in 10 defined habitat types in Lithuania allowed the identification of three distinct chemotypes, dominated respectively by β-caryophyllene, caryophyllene oxide and germacrene D [116,126]. A similar study performed in southeastern Poland identified significant differences among both the content and composition of essential oils from plants growing in 16 habitats, although two major constituents (2-methyloctane and α-terpineol) were produced by representatives of most populations . It is not yet clear whether these reports indicate general trends for plants growing at particular altitudes or in particular habitats. Transplantation studies, in which plants from one altitude (or habitat) are moved to another altitude (or habitat) and later examined for their essential oil content, have not yet been conducted with Hypericum, but would be of value to test such hypotheses. Carefully controlled experiments examining the influence of genotypic background on the essential oil composition and yield in H. perforatum have not yet been made, but would be of significant interest.
In addition to the interspecific, inter- and intrapopulation variability observed, each organ selected from a particular plant for extraction may display a unique chemical profile. An examination of 11 accessions of H. perforatum leaves and flowers growing in a single population in Lithuania indicated that β-caryophyllene and caryophyllene oxide dominated in leaves, while spathulenol, tetradecanol and viridiflorol were dominant constituents of the flowers . A similar study conducted on plants growing in northeastern Iran revealed α- and β-pinene and α- and β-selinene as the primary volatile constituents of the leaves and flowers, while germacrene D was predominant in the oil extracted from the stems and roots . In an examination of H. androsaemum, a comparison of the volatile components present in translucent glands distributed just within the margin of the leaves (marginal glands) and glands distributed on the lamina (laminar glands) was made. The marginal glands contained β-caryophyllene and germacrene B as their dominant volatile components, while the laminar glands contained mainly β-pinene and limonene .
The antimicrobial and antioxidant properties of plant essential oils have been well-documented [127-130]. Lipophilic compounds, including terpenoid derivatives, have been shown to disrupt cellular membranes in bacteria and fungi, thus inhibiting cellular respiration and ionic transport [131,132]. Table 4 summarizes bioactivity data reported to date for essential oils and volatile fractions from Hypericum species. The antimicrobial activities of α- and β-pinene, as well as β-caryophyllene, have been well-documented and, as these compounds represent dominant components in the essential oils of many Hypericum species (see Tables Tables11 and and2),2), such effects are not unexpected. Further investigations with essential oils, volatile fractions and infused oils from Hypericum species would be of interest due to the ex vivo anti-inflammatory activity and in vivo gastroprotective effects that have been demonstrated with H. perforatum infused oils [10,15].
A recent study additionally revealed that healthy H. perforatum plants produced higher amounts of essential oil (0.75%) than plants infected with a ribosomal group 16SrVII phytoplasma known as Ash Yellows (0.11%), and that a higher sesquiterpene to monoterpene and aliphatics ratio was observed in infected plants . Increased accumulation of hyperforin has been observed after elicitation of H. perforatum seedlings with the fungal pathogen Colletotrichum gloeosporioides . How bacterial (e.g. Agrobacterium tumefaciens) or fungal infections alter the essential oil (or other lipophilic constituent) profile of H. perforatum, or other Hypericum species, is a promising area of future investigation.
The identification of particular chemotypes of H. perforatum, displaying a dominant production of β-caryophyllene, caryophyllene oxide, germacrene D, α- and β-pinene [61,116,126] has led to the development of a hypothesis that particular populations (or chemotypes) of this species are rich in either β-caryophyllene/caryophyllene oxide (e.g sesquiterpene hydrocarbons) or α-/β-pinene (e.g. monoterpene hydrocarbons), but not produce both groups of compounds in higher amounts simultaneously [see 52,55,58,68,71,91,106,108,114,115]. This hypothesis, however, has not been supported by the results of essential oil analyses from other Hypericum species (see Tables Tables11 and and2),2), or by data from particular samples of H. perforatum from France and Italy (see Table 3). Another hypothesis states that there are Hypericum populations (e.g. chemotypes) that produce higher amounts of non-terpenoid constituents to the exclusion of terpenoids [68,120]. An interesting study that compared volatiles from the essential oils, air and/or heat-dried, and lyophilized tissue of 13 Hypericum species from Portugal using instrumental olfactroscopy indicated that a decrease in n-methyloctane and α-pinene frequently correlated with an increase in β-caryophyllene and germacrene D, and vice-versa , but again such observations are not consistently supported (see Table 1).
Hypotheses of this nature, therefore, while potentially applicable when examining a limited number of populations, collected or originating from a defined geographic region, and examined within a specific time window (e.g. state of phenology), are not supported when examining evidence upon a broader scale (e.g. geographic range, taxonomic rank, seasonality). This point is emphasized because several recent publications have proposed using volatile constituents as chemotaxonomic markers, examining data with multivariate statistical techniques such as principal coordinate analysis (PCO), principal component analysis (PCA) and canonical discriminate analysis (CDA) [58,61]. Phylogenetic reconstruction using CDA resulted in the division of the analyzed Hypericum species into taxonomic sections in accordance with those delineated by morphological characters [1,58], while analyses using PCO and/or PCA either did  or did not result in such separations [58,61]. It is important to note that, despite the excellent scientific quality of these studies, the plants were collected across a relatively restricted geographic range (respectively, Portugal , southeastern Serbia  and the Hellenic arc ). The authors noted that, although samples were taken during different phenological stages, from different populations within the restricted geographic range and various extraction procedures were used, that the results of their analyses allowed the grouping of samples into accepted taxonomic clusters [58,64]. At the same time, the volatile oil compositions reported for particular species in these studies differed significantly from those reported for the same species from other geographic regions (e.g. H. perfoliatum, compare  and ). Petrakis et al.  discussed the implications of such findings and described in detail specifically why such methods can be potentially used to separate species samples (at least collected within a restricted geographic range), but should not be used to infer phylogeny.
Most studies with Hypericum have been conducted almost exclusively on wild-collected material and without repetition, limiting their usefulness. They do, however, provide an initial platform from which more detailed studies may be launched. In addition to the wealth of information available for H. perforatum, a considerable amount of data are available for essential oil content and composition in H. androsaemum, H. hyssopifolium, H. maculatum, H. perfoliatum, H. scabrum and H. triquetrifolium (see Table 1), providing a basis for comparison. H. androsaemum and H. triquetrifolium are cultivated to a limited extent, the former for the production of cut-flower stems and the latter, for the isolation of hypericin [140-141]. These species along with H. perforatum would, therefore, be ideal targets for further controlled studies on the influence of genotype on essential oil yield and constitution. The implications of the multi-dimensional (i.e. at the subspecies, population, plant, tissue, and even cellular level) phytochemical variability, such as that which has been reported in the Hypericum literature, have long since been appreciated by researchers in the fields of metabolomics and systems biology. Correspondingly, much energy has been invested in the development of appropriate experimental designs to ask specific questions about tracking variation in plant biochemistry in these fields [142,143]. Genetic engineering studies, aimed toward modifying aspects of the biosynthesis of essential oil components [144,145], are highly relevant to the field of essential oil research. The immediate future of research on the essential oil and volatiles chemistry of Hypericum, as for many other plant genera, lies in the development of targeted breeding programs for species with the potential to biosynthesize medicinally and economically valuable constituents. A slightly more distant future prospective includes the design of studies to genetically modify such biosynthetic pathways, to tailor both content and constituent output.
Brian Lawrence (Winston-Salem, NC, USA), Filippo Maggi (Univ. of Camerino, Italy) and Chlodwig Franz (Univ. of Veterinary Medicine, Vienna, Austria) are gratefully acknowledged for providing selected reference literature. Xin Liu (Karl-Franzens-Univ. Graz, Austria) is thanked for providing selected Chinese references and aid in translation. K. H. C. Başer (Anadolu Univ., Eskisehir, Turkey) is sincerely thanked for several years of fruitful collaboration on essential oil research. Funding for the author was provided through a Hertha-Firnberg Stipend (T345) from the Fonds zur Förderung der Wissenschaftlichen Forschung (FWF) in Austria.