Anti-Influenza A Virus Effect of Hypericum perforatum L. Extract ^*

Hypericum perforatum L. extract was provided by Lanzhou Institute of Animal and Veterinary Phar-maceutics Sciences, Chinese Academy of Agricultural Sciences and the ribavirin (H19993462) was purchased from Tianjin Pharmaceutical Group Xinzheng Co., Ltd, Henan, China.

The human influenza A virus (IAV) strain (H₁N₁) (A/Gansu/1771/2006, Gansu Center for Disease Control and Prevention, Lanzhou, China) was used in this study. The virus used in cell culture experiments was passaged through MDCK cells at least once to prepare pools. The pools were then titrated in MDCK cells before use and were then propagated in the allantoic cavities of 9-day-old chicken eggs for 72 h at 35 ℃. Then the virus was passaged five times through mice, after which the recovered virus from the lungs was used to prepare a pool in MDCK cells, which were provided by the Gansu Center for Disease Control and Prevention. MDCK cells were grown in DMEM (Dulbecco's Modified Eagle Medium) (from Sigma Company) supplemented with 0.55% w/v NaHCO₃ (Jianghai Biological Engineering Co., Ltd, Beijing, China), 100 units/mL penicillin, 100 mg/mL strep-tomycin and 10% fetal calf serum (FCS) (Sijiqing Company, Hangzhou, China). FCS was reduced to 2% for the viral infection. The cell suspension containing 100 000 cells per mL was obtained after enzymatic dissociation with trypsin (1 g/L, 1:250, Lvshengyuan Biotechnology). 0.1 mL of this cell suspension was added to each well of a 96-well cell culture plate (Corning, USA, NY. 14831). The cells were incubated (at 37 ℃ in a 5%CO₂ incubator) for 3 d and after forming a continuous layer, they were inoculated with the virus.

Female specific pathogen-free BALB/c mice weig-hing 18-21 g were purchased from The Cancer Hospital of Gansu Province, China.

SaO₂ was determined using the Ohmeda Biox 3800 pulse oximeter (Medi Tech Group Co., Ltd, Qingdao, China). The ear probe attachment was used with the probe placed on a thigh of the mice. Readings were made after a 30 s stabilization time on each animal (32). All SaO₂ determinations were made on days 3 through 11 after virus exposure.

Each mouse lung was homogenized and varying dilutions assayed in triplicate for infectious virus in MDCK cells grown in 96-well flat-bottomed micro-plates (34). Each lung homogenate was centrifuged at 2500 rpm for 30 min and the supernatants were used in these assays.

Three methods were used to assay antiviral activity in vitro: inhibition of virus-induced cytopathic effect (CPE) determined by visual (microscopic) exami-nation of the cells, increase of neutral red (NR) dye uptake into cells and virus yield reduction (35). Eight concentrations of the test agents, each concentration varying in 2X steps, were evaluated in MDCK cells. Standard virus controls, toxicity controls and normal medium controls were included in all assays. CPE inhibition data were expressed as 50% effective (viral CPE-inhibitory) concentration (EC₅₀), 50% cytotoxic (cell-inhibitory) concentration (CC₅₀) and selectivity index (SI), determined as the CC₅₀/EC₅₀. Virus yield reduction data were expressed as the concentration inhibiting virus yield by one log10 (EC₉₀); the SI for virus yield results was calculated as CC₅₀/EC₉₀.

HPE and ribavirin were each evaluated for the dose considered lethally toxic to mice, which was done by treating ten mice per dose with HPE and ribavirin. The ribavirin doses studied were 320, 160, 80, 40, and 20 mg/kg/day and HPE doses were 2000, 1000, 500, 250 and 125 mg/kg/day. The mice were treated by oral gavage (po) twice daily for 5 days and their weights were determined prior to the first treatment and again 18 h after the final treatment. Animals were checked daily for 14 d to check whether any had died.

The general procedure of initial evaluation for the relative efficacies of HPE and ribavirin was to infect groups of mice intranasally with the IAV. This was done by anesthetizing them with aether and instilling 50 μL of 25 LD₅₀concentration of virus on the nares. The human influenza A virus (IAV) strain was studied in vivo. The mice were treated with varying dosages (as described in Table 3) of either drug po twice daily for 5 d beginning 4 h prior to virus ex-posure. Parameters for determining the effects of treatment included prevention of death through 14 d, slowing of SaO₂decline, inhibition of lung consolidation [scores ranging from 0 (normal) to 4 (maximal plum coloration) and lung weight], and lessening of lung virus titers. The lung parameters were assayed on d 3, 6 and 9 of the infection, with 3-4 mice being killed at each time point. Ten mice were used per treated group to assay for death and SaO₂. Virus control was used in each experiment with the same numbers of animals used for each disease parameter. Toxicity control was also run in parallel for each dose; three mice were used per dose, with their weights taken before the start of the treatment and again 18 h after the last treatment and the animals were observed for overt signs of toxicity and death for 14 d. A normal control was run in parallel; these animals were weighed along with the toxicity controls and SaO₂ levels were ascertained on the same days as the animals were infected. The normal control was sacrificed along with the infected mice to provide background lung data.

Table 3.  Inhibition of influenza A virus infections in mice treated po with HPE or ribavirin

To determine the relative efficacy of ribavirin and HPE against the IAV infection when initiation of therapy was delayed, groups of 10 mice infected with the IAV were treated po twice daily for 5 d with either agent beginning 4, 12 or 20 h post-virus exposure. Effects on mortality and slowing of SaO₂ decline were determined, with the numbers of animals per group described in Table 4. The dose of ribavirin was equivalently non-toxic (1/3 LD₅₀ dose), which was 70 mg/kg/day, and the dose of HPE was equiva-lently nontoxic, 500 mg/kg/day. The animals were held for 14 d. When it was determined that all periods of treat-ment were highly effective, the experiment was repeated with treatments delayed until 28, 36 or 48 h post-virus exposure.

Table 4.  Effect of delay of p.o. HPE or ribavirin treatment on an influenza A virus infection in mice

Increases in number of survivors were evaluated using χ² analyses. Differences in the mean days to death, mean SaO₂ values, lung scores and mean lung virus titers were analyzed by t-test.

Both HPE and ribavirin were inhibitory to the in-fluenza A virus evaluated, the EC₅₀ values for ribavirin being 5.0 μg/mL, whereas HPE was approximately five-fold less potent, its EC₅₀ being 24 μg/mL (Table 1). Multiple cytotoxicity assays using neutral red uptake indicated the mean CC₅₀ values for ribavirin and HPE was 520 μg/mL and 1.5 mg/mL respectively.

Table 1.  In vitro inhibition of influenza A virus by HPE and ribavirin

Oral gavage treatment with ribavirin and HPE using therapies twice daily for 5 d indicated that the ap-proximate LD₅₀ dose of ribavirin was 200 mg/kg/day, whereas HPE was better tolerated since below a dosage of 2 000 mg/kg/day, almost all treated mice survived. It should also be noted that some weight gain was seen at dosages below 2000 mg/kg/day. No attempt was made to determine the cause of death in the mice in this range finding study (Table 2).

Table 2.  Comparison of toxicity of po administered HPE and ribavirin in mice

Multiple dosages of ribavirin and HPE were utilized and both were evaluated against the IAV in young adult mice. The results of each treatment against the IAV infections are summarized in Table 3. Due to the large amount of data available, only the last day of SaO₂ determination, day 11, is shown in the tables; it was on this day that the values usually reached their lowest levels in virus control animals and thus differences resulting from therapy were the most pronounced. The day 6 lung consolidation data are shown in the tables, since it was at this time that lung scores and weights were maximal; day 3 lung virus titers are indicated because these titer differences are usually best seen early in the infection. According to these parameters each agent was significantly inhibitory to the virus.

Against the influenza A virus infection, HPE was effective in inhibiting death and SaO₂ decline at doses down to 31.25 mg/kg/day. Ribavirin's inhibition of this virus infection was seen down to 40 mg/kg/day.

Each agent has approximately the same therapeutic index. But HPE represented the minimum effective dose and had the least toxicity in the mice, as described in Table 3.

The results of delaying the initiation of therapy with ribavirin and HPE on the influenza A virus infection are presented in Table 4. In this experiment, 500 mg/kg/day of HPE and 70 mg/kg/day of ribavirin were used, including a nontoxic dose of HPE and an approximately one-third of the LD₅₀ dose of ribavirin. Both agents remained highly protective of the mice despite the delay in starting the treatment, and a highly significant difference in reduction of mortality was seen as late as the 20 h post-virus exposure initiation time, which was the longest delay evaluated in the first experiment. One infected animal treated with ribavirin and HPE respectively died at this latest therapy initiation time. All mice survived in the other infected and treated groups. In the second experiment, with the therapy initiation being further delayed, both agents were highly protective of the mice even when the treatment was begun 48 h after virus exposure, which was the latest time evaluated.

As a herbal remedy known since Greek and Roman times, Hypericum perforatum L.(HP) (Fam: Clu-ciaceae = Guttiferae) has been used against ulcers, diabetes mellitus, the common cold, gastrointestinal disorders, jaundice, hepatic and biliary disorders in Turkish medicine (3). Antidepressant actions of HP has been demonstrated in experimental (22, 23, 24) and clinical (37) studies. Antitumoral effects are also cited (19). Hepatoprotective activity of its extracts on rodent species has been reported too (7, 25).

HPE was considered highly effective against in-fluenza A virus in this study, as was shown both in vitro and in experimentally infected mice. And its cytotoxicity was less than that of ribavirin (CC₅₀ values of 1.5 mg/mL for HPE versus 520 μg/mL for ribavirin)

Efficacy in infected mice was demonstrated by all disease parameters studied in these experiments, including prevention of death, slowing of the mean day to death, lessening of SaO₂ decline, inhibition of lung consolidation as seen by lessened lung scores and lower lung weights and reduction in titers of infectious virus recovered from the lungs. In some experiments, the lung virus titers were not markedly affected by therapy with either agent, although a strong disease inhibitory effect was observed. It is probable that in such cases, examination of the lungs at an earlier time, e.g. day 1, would have demonstrated a greater effect. This was shown in the case with neuraminidase inhibitors (33). This early time period was not selected for the present study because signifi-cant lung consolidation is not manifested at this time.

The relative toxicities of the two agents in mice did not correlate well with what was seen in vitro; in the animals, HPE was approximately no more toxic than ribavirin (below the dosage of 2000 mg/kg/day, HPE was almost non-toxic, while an LD₅₀ of 200 mg/kg/ day was observed for ribavirin).

Both agents appeared to work equally well when the initiation of therapy was delayed as late as 48 h after virus exposure. This was the latest time con-sidered. In a previously study, ribavirin was not efficacious against another IAV infection when treat-ment began 24 h after virus exposure; however, in that study the viral challenge was highly lethal to the placebo-treated animals, causing 100% mortality and a mean day to death of 4.9±1.1 days. It was apparent that when the viral challenge resulted in reduced mortality, such as in the present experiment where 90% of the placebo treated mice had a greater mean time to death (6.2±0.6 days), treatment could be delayed much longer and still be efficacious.

The high efficacy and tolerability of the herb's extract is unquestionable and, above all, HPE does not produce major side effects, making it a safe herbal therapeutic agent. But hypericin, one of the active ingredients within HPE, was a sort of photosensitizer, which might cause allergic reaction. The daily dose of the herb, which is about 500 mg extract, corresponds to a total dose of 5 mg of hypericins, the main photo-sensitizers in the herb. These doses of hypericins, given orally, do not provoke skin phototoxicity to patients.

The use of HPE as a novel agent for viral disease offers the advantages of easy production, low cost, extensive availability, adequate water solubility and minimal side effects.

The present study suggests that HPE may have potentials for treatment of human influenza A virus infections. The agent's in vitro efficacy against the strain of the IAV would suggest that it may be of use against the current outbreaks of influenza occurring in human populations in Asia (18). The lack of sufficient supplies of active drugs for treatment of such in-fections, in the event of a pandemic occurrence, would justify further studies on HPE to investigate into this possibility.