Hypericin: the active ingredient in Saint John's Wort

by A.Y. Oubre

Hypericin, a photochemical extracted from St. Johns Wort (Hypericum perforatum) and related species, has been shown to have potent, broad spectrum antimicrobial activity. This compound is an aromatic polycyclic anthrone, a class of colored or pigmented chemical substances which have photosensitizing activity. In both in vitro (laboratory) and in vivo (animal) studies, low, non-toxic doses of hypericin significantly inhibited the replication of several viruses, including HIV, influenza A, cytomegalovirus (CMV), Herpes simplex 1 and 2 (HSV-1 and HSV-2), and Epstein-Barr virus (EBV). Hypericin and its chemical relative, pseudohypericin, produce antiviral activity through a different mechanism of action than do AZT and other nucleoside antiviral agents. Hypericin does not appear to directly alter the activity of reverse transcriptase although it does block the formation of HIV synctium. Recent findings have shown that the antiretroviral action of this compound disrupts uncoating of the lipid envelope of both DNA and RNA viruses, thus preventing infected cells from releasing HIV copies. Theoretically, hypericin and AZT, in combination, may have synergistic antiviral effects against HIV. On the other hand, hypericin actually may increase the toxicity of antiretroviral nucleosides such as AZT, ddI, or ddC.

Traditionally, extracts of St. Johns Wort (which contain hypericin) have been used as an antidepressant, possibly by acting as a MAO inhibitor. The psychotropic effects attributed to hypericin in St. Johns Wort extract suggest that the pigment compound can cross the blood brain barrier (possibly treating neuropsychological symptoms such as dementia). Laboratory investigations indicate that hypericin may be beneficial as an HIV therapy. However, its administration should be carefully monitored by a physician. The levels of hypericin found in most commercially available extracts of St. Johns Wort generally are not sufficient to be therapeutically effective against viral infections.

Liver function should be tested periodically in persons taking hypericin. Also, extreme photosensitivity has been observed in a few cases of people taking this high doses (in excess of 10 mg per day) of this compound. Finally, there is a very small possibility that adverse reactions could occur on occasion between hypericin and other foods or drugs which interfere with MAO inhibitors.

Active principles in St. Johns Wort

The quantity and quality of active principles in Hypericin species vary according to geographical locale, climate, time of day, and time of year. St. Johns Wort contains dianthrone derivatives, mainly in the form of hypericin and pseudo-hypericin as well as flavonoids. Small amounts of coumarins, phenolic carboxylic compounds, phloroglucinol derivatives, monoterpenes, sesquiterpenes, n-alkanes, n-alkanols, carotenoids, and beta-sitosterol are present. The roots contain zanthones. Practitioners and consumers should note that St. Johns Wort extracts, whether standardized or not, consist of other active ingredients in addition to hypericin and pseudohypericin.

Pre-clinical studies

Both in-vitro (test tube) and in-vivo (animal) pre-clinical studies suggest that hypericin (and, to a lesser extent, pseudo-hypericin) may have therapeutic benefits for HIV infection and other retroviral diseases. Certain compounds other than hypericin extracted from Hypericum species have antibiotic activity. marked antiretroviral effects, however, have been reported primarily for hypericin which is isolated mainly from Hypericum perforatum. However, synthetic hypericin has been used in recent studies.

In in-vitro and in-vivo studies, both hypericin and pseudohypericin (extracted from Hypericum triquetifolium) had antiviral activity against several retroviruses. In one experiment, mice were simultaneously injected with low doses of the compounds and with Friend leukemia virus (FV). This aggressive retrovirus normally causes rapid splenomegaly (swelling of the spleen) and acute erythroleukemia in mice. However, these symptoms were effectively suppressed by the addition of hypericin. Splenomegaly had not occurred ten days after infection at the close of the study. No infectious virus could be recovered from the spleen. Also, viremia normally associated with FV was absent. Mice treated with hypericin and pseudohypericin survived a much longer time than mice treated with a toxic antiviral (N3dthd). Unlike most antiretroviral drugs, hypericin (given in a single dose of low concentration) was effective without being cytotoxic. Even when it was administered after viral infection had already started, it still inhibited the onset of disease.

In in-vitro studies, mouse cell lines were infected with radiation leukemia virus (Rad LV) and then incubated with hypericin. The activity of reverse transcriptase in these cells was suppressed through indirect mechanisms. In contrast to nucleoside analogues, polycyclic diones such as hypericin interfere directly with the viral replication cycle during stages in which virions are assembled or intact virions are shedded from immature cores. Alternatively, these aromatic compounds may directly inactivate mature retrovirus that contains normal, assembled cores. Other findings indicate that hypericin is able to inactivate virions and block viral release from infected cells by interacting with the cell membrane.

Unpublished data show that hypericin disrupts the formation of synctia in HIV disease as well as in de novo infection of cells. In in-vitro studies, hypericin showed selective activity against HIV and modest inhibition of reverse transcriptase. In vitro research also revealed that hypericin lowered viral activity in whole human blood taken from HIV infected persons. “Wild” strains of HIV taken directly from infected patients are sometimes more resistant to antiviral agents than are viral strains bred in the laboratory.

Other investigations indicate that the antiviral effects of hypericin on murine cytomegalovirus (MCMV), Sindbus virus (SV), and HIV are enhanced by exposure to fluorescent light. Hypericin and to some degree, pseudohypericin, were effective against FV and HSV-1 when the viruses were first incubated with the compounds for one hour at 37 degrees C before mice were infected. Pre-incubation for one hour at 4 degrees C, however, produced no antiviral effects. The authors of this study (who are scientists at Lilly Research Laboratories) reported that hypericin and pseudohypericin were effective in vitro against enveloped viruses such as HSV and influenza when the cultures were pre-incubated with these agents at 37 degrees C. They also correctly showed that hypericin and its analogue inhibit DNA and RNA viruses, but not viruses which lack a lipid envelope. The Lilly researchers, who call AZT a preferred therapy for HIV, however, claim that single dose administration of hypericin is not efficacious. Human clinical trials are needed to evaluate the appropriate dose ranges at which hypericin is therapeutic but nontoxic, and to assess the differences, if any, between natural and synthetic hypericins. In animal studies, natural sources of hypericin (in combination with its analogue, pseudohypericin), showed greater antiviral activity that did synthetic hypericin. Preliminary findings thus far strongly indicate that the wide spectrum antiviral properties of hypericin, its experimental effectiveness at low concentrations, and its unconventional mechanisms of antiviral action make it a promising candidate for a new class of HIV therapies.

Mechanisms of action

Hypericin and pseudohypericin had no effect on purified reverse transcriptase alone. They did not alter levels of intracellular viral mRNA. Instead, hypericin lowers the number of mature viral particles without suppressing intracellular levels of viral mRNA. The concentrations of viral antigens on the cell surface were also unaffected by hypericin. These findings, as a whole, imply that the compounds interfere with viral assembly, budding, shedding or stability at the level of the cell membrane When hypericin was added to viral-infected cell cultures, red fluorescence appeared at localized areas on the lipid surface membrane.

Unlike nucleoside analogues, polycyclic diones such as hypericin have no effects on transcription, translation, or transport of viral proteins to the cell membrane. They are not directly active against reverse transcriptase even though reverse transcriptase activity was reduced in infected cells that had first been incubated with hypericin. Cells treated with hypericin form immature or abnormally assembled cores. This indicates that hypericin may block the processing of gag-encoded precursor polypeptides. Hypericin, whether in the intracellular medium or bounded to the membrane, is thought to lower the activity of reverse transcriptase by interfering with protein synthesis. (It is noteworthy that the antiviral effects of harmine, a photoactive alkaloid, involve disrupted kinase activity in enveloped RNA viruses.)

Viral particles are not formed when gag-related polyproteins fail to be cleaved or synthesized. Gag-related polyproteins, therefore, may play a decisive role in the virucidal actions of hypericin. Reverse transcriptase within the core of the assembled virus probably takes the form of an inactive enzyme or proenzyme. Mechanisms involving viral-encoded proteases or kinases might be required to activate reverse transcriptase. These mechanisms could transform the enzyme from a nonfunctional to a functional state. Both hypericin and pseudohypericin are thought to influence protease activity. In turn, altered protease activity could disrupt the cleavage or synthesis of gag-related polyproteins. As a result, immature viral cores would be formed. Alternatively, by selectively binding to viral polyproteins, hypericin could interfere with the gag and gag-pol polyproteins needed for viral assembly. Thus, hypericin could block the process whereby RNA packages encapsulated viral particles.

Some investigators, however, propose that hypericin lyses infectious virion by interacting directly with the viral envelope instead of disrupting gag-encoded precursor polyproteins or modifying other proteins. In any case, the antiviral properties of hypericin appear to involve its interactions with the cell membrane or cell surface recognition sites. Molecular modifications at or near the surface provide a model for rationally designing a new class of anti-HIV agents. Such therapies may be able to block HIV-encoded protease located in the gag-pol region. Importantly, drugs of this type would not be toxic like AZT and other agents whose pharmacological actions are based on direct inhibition of reverse transcriptase.

The aromatic, ringed structure encircled by six phenolic hydroxy groups seems critical to the antiviral activity of the hypericin molecule. Quinone groups, which often have antiviral properties, also exert photodynamic effects. Hypericin is thought to generate singlet oxygen. However, free radical quenchers can interfere with singlet oxygen reactions involving hypericin thereby reducing its antiviral properties.

Hypericin has a unique molecular structure in which one-half of the molecule is hydrophilic (water loving) while the other half is hydrophobic (water repelling). The top, bottom and side (non-polar) of the hypericin molecule which contains the methyl groups are hydrophobic. It is thought that the molecule might bond to the outer surface of the cell membrane. Presumably, the hydrophobic side would be immersed in fat Singlet oxygen, though less reactive than triplet oxygen, binds with two-electron targets, including, for example, the double bonds found in polyunsaturated fatty acids. The hydrophilic sides, in contrast, could hydrogen-bond to the aqueous media.

Discrepancies between in vivo findings from different studies on the antiretroviral effects of hypericin and pseudo-hypericin may be due partly to variations in light. However, differences in hypericin isolation methods and in the strains of mice used also could account for variable findings in several investigations. The antiviral effects of hypericin are largely but not completely, attributed to its photodynamic properties. In the presence of light, hypericin completely inhibited infection of cell cultures of equine infectious anemia virus (EIAV). On exposure to fluorescent light, hypericin inactivated MCMV, Sindbis virus (SV), and HIV-1. Both membrane virions and virus-infected cells were more strongly inactivated by visible light. (Polyacetylene phenylheptatriyine (PHY), a substance purified from the plant, Bidens pilosa, also shows antiviral activity against membrane-bound viruses such as MCMV. The antiviral effects, which involved interactions between PHY and membrane, occurred in the presence of long wave ultraviolet light). Significant advances in photobiological research have been made in recent years. New findings demonstrate clearly that photodynamic action accounts for the antiviral properties of several natural product derivatives, including hypericin.

Light is required for the photosensitization of hypericin. The compound absorbs light quanta and generates it in the form of singlet oxygen. In so doing, hypericin triggers the photo-oxidation of cellular components, including, for example, the photohemolysis of red blood cells. The underlying mechanism of photodynamic reactions is not fully understood. It is thought to involve interactions between oxygen and light as well as sensitizing pigment which binds to the cell membrane.

In photodynamic reactions mediated by hypericin, singlet oxygen serves as the main oxidant. Singlet oxygen has a strong affinity for pi electron-systems found in compounds such as polycyclic diones. The pi electrons, responsible for the photoactive properties of hypericin, absorb visible and ultraviolet light and then reemit it within the range of green and red light. The two hydroxy groups and the two methyl groups flanking each side of hypericin’s eight ringed structure do not lie within the same plane. Instead, they repel each other, placing strain on the benzenoid structure. This causes the hypericin molecule to twist and become unstable. Hypothetically, the steric strain could increase the energy state of the pi electrons. This would allow them to form temporary bonds with singlet oxygen which, at a later point, could be released to disrupt mechanisms of viral replication. Pi electrons therefore, seem to play a major role in the antiviral activity of hypericin.

The “impressive light-mediated antiviral activities” of hypericin have been shown in several studies. Sindbus virus (SV), for example, was 99% inhibited in the presence of light. In the dark, however, the antiviral effects of hypericin were reduced by more than two orders of magnitude. On exposure to light (650-700nm.), hypericin undergoes type II photosensitization in which singlet oxygen and other reactive molecular species are produced. Though not as destructive as free radicals (which are generated in Type I photosensitization, singlet oxygen could damage viral membranes, thereby interfering with proteins and nucleic acids. Nonetheless, hypericin also has some degree of virucidal activity in the dark, though much less so than in light. It is thought that the antiviral effects produced in the absence of light take place through a different mode of action than light-mediated virucidal activity. Protein kinase C, for example, may represent an alternative target for hypericin’s antiviral action in the absence of light.

One of the therapeutic advantages of hypericin is that it works by multiple steps. Hypericin is probably able to interrupt various phases in the replication of enveloped retroviruses. These stages include polyprotein cleavage and protein alterations as well as assembly, budding and shedding of viral components. The compound’s disadvantages as an anti-HIV agent involve its potential toxicity when patients are exposed to sunlight. Also, it has been reported that in humans, hypericin is more effective in suppressing HSV-1 and HSV-2 than HIV. Future research is required to elucidate the mechanisms through which hypericin generates and reacts with singlet oxygen in triggering antiviral effects.


Clinical trials were conducted the early 1990’s by Dr. Bihari in New York City recommending the following regimen. For the first two weeks, patients take 10 mg once a day for two weeks. During the second two weeks, the dosage alternates between 10 mg per day on the first day, 20 mg on the second day, 10 mg on the third day, and so forth. The ultimate dosage is given during the fifth and sixth weeks when patients are given 20 mg per day.

It has been proposed that beta-carotene, a known free radical quencher, be administered in combination with hypericin. Theoretically, this might reduce any toxic side-effects associated with hypericin’s generation of free radicals. Some investigators, however, have warned that quenchers also may lessen the anti-viral properties of hypericin. The role of beta-carotene and other free-radical scavengers in hypericin therapy deserves further clarification.


Hypericin, a pigment molecule with photodynamic activity, has dramatic antiviral activity, especially in the presence of light. Like several other plant derived substances, hypericin only inhibits viruses with membranes. It has antiviral effects against a wide range of retroviruses, including HSV-1, HSV-2, (murine) CMV, and HIV-1. In contrast to nucleoside agents such as AZT, hypericin and its chemical relative, pseudohypericin, do not directly affect the activity of reverse transcriptase. Rather, these agents seem to disrupt various stages of viral replication, including assembly, budding, shedding and possibly protein synthesis, all of which depend on the integrity of the viral membrane.

Hypericin and related compounds, with their alternative targets for virucidal activity, comprise a new class of potential anti-HIV drugs. It is not yet known how effective hypericin will be in human AIDS. Preliminary findings, however, strongly suggest that this compound is one of the most promising, new anti-HIV prototype molecules currently under investigation.

Leave a Reply

Your email address will not be published. Required fields are marked *