Investigation of the biological activities of Siphonochilus aethiopicus and the effect of seasonal senescence

Siphonochilus aethiopicus (Schweinf.) B.L. Burtt (Zingiberaceae), commonly known as wild ginger, is one of the most important and threatened medicinal plants in South Africa. A study of the pharmacological properties of S. aethiopicus and the effect of seasonal senescence on antibacterial and anti-inflammatory properties was undertaken. Water, ethanol and ethyl acetate extracts were prepared from the leaves, rhizomes and roots of S. aethiopicus plants. The extracts were tested in a variety of pharmacological assays. Results for the general screening showed antibacterial and anti-inflammatory activity. Some cytotoxicity was observed with the aqueous extracts of the rhizome. However, no significant activity against the herpes simplex virus types 1 and 2, the influenza A virus, and in the anthelmintic, antischistosomal and biochemical induction assays were observed. In the microdilution antibacterial assay, no inhibitory activity against the test bacteria was detected with the aqueous extracts. The ethanol and ethyl acetate extracts tested showed greater antibacterial activity at minimal inhibitory concentrations ranging from 0.78 to 3.13mg ml against the Gram-positive bacteria (Bacillus subtilis, Staphylococcus aureus) than the Gram-negative bacteria (Escherichia coli, Klebsiella pneumoniae). Little difference was observed between the ethanol and ethyl acetate extracts, or between the different plant parts. Ethanol extracts were prepared from the different plant parts before and after seasonal senescence, and tested for antibacterial and anti-inflammatory activity. There appeared to be a loss of antibacterial activity in the leaves with senescence, concomitant with an increase of activity in the alpha-roots. In the cyclooxygenase-1 assay, the aqueous extracts showed no significant prostaglandin synthesis inhibition. For the ethanol and ethyl acetate extracts, the leaves showed the highest levels of activity at a concentration of 250μg ml per test solution, in both the cyclooxygenase-1 and -2 assays. Some differences in the levels of anti-inflammatory activity in the roots following senescence were also observed. There was a slight loss of activity as a result of drying the rhizome material prior to extraction. This suggests that fresh rhizome material may be more effective for medicinal use, although it should be noted that the aqueous rhizome extracts displayed moderately high levels of cytotoxicity, and may require further investigation.


Introduction
. None of the terpenoids of the oil of true ginger are present in the essential oil of S. aethiopicus (Van Wyk et al. 1997).
Extracts of the leaves, rhizomes and roots were screened in the microdilution antibacterial assay, cyclooxygenase-1 and -2 anti-inflammatory assays, vervet monkey kidney cell cytotoxicity assay, herpes simplex virus type 1 and 2 and influenza A antiviral assays, anthelmintic assay, antischistosomal assay and the biochemical induction assay.

Extracts for general screening
S. aethiopicus plants were obtained from stock plants at the University of Natal Botanical Garden, Pietermaritzburg. A voucher specimen was deposited in the University of Natal Herbarium (Light 17 NU). Mature plants were harvested in summer, divided into leaves, rhizomes and roots and dried at 50°C for 2 days. The dried, ground material (10g) was extracted in 100ml solvent (water, ethanol or ethyl acetate), in an ultrasound bath (Branson 5210) for 60min. The extracts were vacuum filtered through Whatman No. 1 filter paper, and dried under vacuum, using a rotary evaporator.

Extracts for comparison of seasonal activity
Plants of S. aethiopicus were harvested prior to senescence (green leaves), and after the plants had fully senesced. For each set of extracts, plants were divided into leaves, rhizomes and roots, differentiating between αand β-rhizomes and roots. The mature rhizomes, and associated roots, from the previous year's growth were classified as α-rhizomes and roots, and the younger rhizomes, and roots, which develop from the α-rhizome at the start of the growing season were classified as β-rhizomes and roots. Part of the freshly harvested material was extracted in 100ml ethanol using a Wareing blender, and then placed in an ultrasound bath (Branson 5210) for 60min. The remaining material was dried at 50°C for 2 days prior to extraction in the same manner. The extracts were vacuum filtered through Whatman No. 1 filter paper, and dried under vacuum, using a rotary evaporator.

Microdilution antibacterial assay
Antibacterial activity was assessed using the microdilution bioassay, as described by Eloff (1998). The residues from the plant extracts were redissolved to a concentration of 50mg ml -1 in water (for aqueous extracts) or ethanol (for ethyl acetate or ethanol extracts). Each extract was bioassayed against bacterial strains obtained from the bacterial collection of the Microbiology Department, University of Natal, Pietermaritzburg and maintained on Mueller-Hinton nutrient agar (Biolab) at 4°C. The bacteria used were Bacillus subtilis, Staphylococcus aureus (Gram-positive) and Escherichia coli, Klebsiella pneumoniae (Gram-negative).
For each test bacterium, 100μl of redissolved extract was serially diluted 2-fold with 100μl sterile distilled water in a sterile 96-well microtitre plate. A similar 2-fold serial dilution of neomycin (100μg ml -1 ) was used as a positive control for each bacterium. Extract and bacteria-free negative controls were also included. Suspension cultures of the bacteria were inoculated in Mueller-Hinton (MH) broth (Oxoid) from stock cultures and incubated overnight at 37°C in a waterbath on an orbital shaker. Prior to use in the bioassay, the saturated suspension cultures were diluted 1:100 with sterile MH broth. To each of the wells containing the test and control solutions, 100μl of the bacterial cultures were added. The plates were covered and incubated overnight at 37°C. To indicate bacterial growth, 40μl of 0.2mg ml -1 p-iodonitrotetrazolium chloride (Sigma), were added to each well, and the plates were incubated for a further 30min. The wells which displayed no change in colour represented antibacterial activity. The minimum inhibitory concentration (MIC) was taken as the lowest concentration of plant extract to elicit a bacteristatic or bactericidal effect against the tested bacterium.

Cyclooxygenase-1 assay
The cyclooxygenase-1 (COX-1) assay was performed as described by White and Glassmann (1974), with slight modifications (Jäger et al. 1996). The COX-1 enzyme (10μl microsome suspension from sheep seminal vesicles, 0.3μg protein) was activated with 50μl of co-factor solution (0.3mg ml -1 adrenaline and 0.3mg ml -1 reduced glutathione in 0.1M Tris buffer, pH 8.2) on ice for 15min. The enzyme solution (60μl) was added to the sample solution (2.5μl ethanolic or aqueous plant extract and 17.5μl water) and incubated at room temperature for 5min. The plant extracts were tested at a final concentration of 250μg ml -1 per test solution. Indomethacin was tested as a positive control. The reaction was started by adding 20μl [ 14 C]-arachidonic acid (30μM, 16Ci mol -1 ). Samples were incubated for 8min at 37°C and the reaction terminated by adding 10μl 2M HCl. Prostaglandins and unmetabolised arachidonic acid were separated by column chromatography over silica gel, after addition of 4μl unlabeled prostaglandins (PGE 2 :PGF 2 1:1) as a carrier solution. Arachidonic acid was eluted first with n-hexane:1,4-dioxan:glacial acetic acid (70:30:0.2). The prostaglandin products were then eluted with ethyl acetate:methanol (85:15) and collected. After mixing with scintillation solution, the samples were counted in a Beckman LS 6000LL scintillation counter. Inhibition refers to reduction of PGE 2 formation in comparison to an untreated sample (2.5μl ethanol in 17.5μl water).

Cyclooxygenase-2 assay
The cyclooxygenase-2 (COX-2) assay was performed as described by Noreen et al. (1998), with slight modifications (Zschocke and Van Staden 2000). The COX-2 assay follows a very similar protocol as the COX-1 assay. Purified COX-2 enzyme from sheep placental cotyledons was purchased from Cayman chemical. The enzyme (10μl containing 3 units) was activated with 50μl co-factor solution (0.6mg ml -1 adrenaline, 0.3mg ml -1 reduced glutathione and 1μM hematin in 0.1m Tris buffer, pH 8.0) on ice for 5min. The enzyme solution (60μl) was added to the sample solution (2.5μl ethanolic or aqueous plant extract and 17.5μl water) and incubated at room temperature for 5min. The plant extracts were tested at a final concentration of 250μg ml -1 per test solution. Positive control measurements were carried out with indomethacin and nimesulide at concentrations of 200μM. The reaction was started by adding 20μl [ 14 C]arachidonic acid (30μM, 16Ci mol -1 ). Samples were incubated for 10min at 37°C and the reaction terminated by adding 10μl 2M HCl. Prostaglandins and unmetabolised arachidonic acid were separated and COX-2 inhibition was determined as described for the COX-1 assay.

Preparation of extracts for antiviral testing
A 1 000μg ml -1 (w/v) sterile stock suspension of the extract was prepared by diluting an aqueous suspension of plant extract in serum free Eagle's minimum essential medium (MEM) (National Institute for Virology) followed by filter sterilisation through a 0.45μm membrane (Ministart ® filter unit, Sartorius). Dilutions of the aqueous plant extracts, in serumfree MEM, were tested for cytotoxicity at concentrations ranging from 3.9μg ml -1 up to 1 000μg ml -1 , and from 3.9μg ml -1 up to 500μg ml -1 for the antiviral assays.

Cell culture
Standard cell culture techniques (Grist et al. 1979), were used for all procedures utilising cell cultures. Monolayers of secondary vervet monkey kidney (VK) cells (National Institute for Virology) were prepared by seeding 96-well microtitre plates with 200μl of 10 5 cells/ml cell suspension. MEM supplemented with 8% heat inactivated foetal calf serum (FCS) (Delta Bioproducts) and containing 100U ml -1 penicillin and 100μg ml -1 streptomycin was used for the propagation of the cells. Cell cultures were incubated in a humidified CO 2 atmosphere at 37°C. Maintenance medium was essentially the same as the propagation medium except that it contained only 2% FCS.

Virus stock
The 50 percent tissue culture infectious dose (TCID 50 ) of each virus was calculated according to the Kärber formula as outlined in Grist et al. (1979).
Stock suspensions of herpes simplex virus type 1 (HSV-1) and herpes simplex virus type 2 (HSV-2), with titres of 1 x 10 7 and 1 x 10 6 TCID 50 /ml respectively, were prepared from clinical isolates of HSV-1 and HSV-2 (Department of Medical Virology, University of Pretoria). The viruses were diluted in serum-free MEM and used at a final concentration of 100 TCID 50 /microplate well.
Freshly harvested allantoic fluid containing influenza A virus (Inf A) (strain Panama) was kindly supplied by the National Institute for Virology. A stock suspension, with a titre of 3.16 x 10 5 TCID 50 /ml, was prepared by diluting the allantoic fluid in sterile phosphate buffered saline (PBS) (Sigma) containing penicillin (50μg ml -1 ), streptomycin (50μg ml -1 ) and neomycin (100μg ml -1 ) (PSN Antibiotic Mixture [100X], GibcoBRL). This stock suspension is reportedly stable for one week at 4°C (Barrett and Inglis 1985). For experimental purposes fresh dilutions of the virus, in serum free MEM, were prepared immediately before use. The virus was used at a final concentration of 1 000TCID 50 /microplate well.
As Inf A does not exhibit a cytopathic effect (CPE) in cell culture, viral infection was monitored by direct immunofluorescence (IF) for the detection of viral antigen. Titrations were done in 16-well Lab-Tek ® glass tissue culture chamber slides (Nalge Nunc) instead of 96-well microtitre plates. Forty-eight hours after infection the growth medium was recovered and the slides were washed in PBS and fixed in 100% acetone at -20°C. The direct IF was carried out using standard techniques with a mouse anti-influenza A fluorescein isothiocyanate (FITC)-labeled monoclonal antibody (Chemicon) as the detector. Stained slides were examined for fluorescence using a halogen lamp at 10x magnification. Wells where fluorescent foci were detected were considered to be positive for Inf A infection.

Cytotoxicity assay
Aqueous extracts of the leaves and rhizomes of S. aethiopicus were tested for cytotoxicity by exposing monolayers of VK cells to dilutions of the filter sterilised plant extracts. Serial 2-fold dilutions of the extracts, in serum-free MEM, from a concentration of 3.9μg ml -1 to 500μg ml -1 , were used for testing on 24-hour-old monolayers of VK cells. The cells were monitored visually, by light microscopy, over a period of seven days and on the seventh day tested for cytotoxicity using a tetrazolium salt reduction (MTT) assay (Van Rensburg et al. 1994), based on the method of Hussain et al. (1993). Monolayers of cells exposed to serum-free MEM alone were used as a control.

Antiviral assays
Two separate assays were performed for each virus: a)To investigate the effect of the aqueous extract of S. aethiopicus leaves on the replication of HSV-1 and HSV-2, 24-hour-old monolayers of VK cells in 96-well microtitre plates were starved in serum-free MEM for 1h at 37°C in a humidified CO 2 atmosphere. After starvation the serumfree MEM was withdrawn and 100μl (100TCID 50 ) of virus was added to the wells and allowed to adsorb to the cell cultures for 1h at 37°C in a humidified CO 2 atmosphere. After adsorption for 1h the unbound virus was withdrawn and the cells rinsed once with serum-free MEM after which 200μl of the appropriate dilution of plant extract suspension in serum-free MEM was added to each of 6 wells and the cell cultures incubated at 37°C in a humidified CO 2 atmosphere. As a positive control, cells infected with virus were maintained in serum-free MEM, and cells mockinfected with 100μl serum-free MEM and maintained in serum-free MEM served as negative controls. After infection the cell monolayers were examined daily for 7 days, by light microscopy, for the appearance of a typical herpes simplex-like cytopathic effect (CPE) characterised by large refractile cells (Wiedbrauk and Johnston 1993). As the herpes simplex CPE spreads rapidly through the cell monolayer complete destruction of the monolayer was readily discernible by days 4 to 5, post-infection. An obser-vation period of up to 7 days ensured that no further CPE manifested itself and that unaffected monolayers remained intact. The absence of CPE at a specific concentration of the plant extract was considered to be indicative of antiviral activity.
To investigate the effect of the aqueous extract of the rhizomes of S. aethiopicus on the replication of Inf A, the procedure was effectively the same as described above except that the VK cells were grown in 16-well Lab-Tek ® glass tissue culture chamber slides and 1 000TCID 50 Inf A was used per well. Only plant extract concentrations showing no severe cytotoxicity were tested for antiviral activity. Viral infection was monitored by direct IF, in 4 wells per concentration of plant extract, as described previously. Fluorescent foci were counted in three separate fields on each of the 4 wells and the reduction in fluorescent foci in relation to the control wells was calculated. b)To investigate the effect of the aqueous extract of S. aethiopicus leaves on viral adsorption and subsequent replication of HSV-1 and HSV-2 in cell culture, 96-well microtitre plates were prepared and starved as described previously. Equal volumes (100μl) of the appropriate dilution of the plant extract and virus suspension (100TCID 50 ) were added simultaneously to each of 6 wells of the 96well microtitre plate and the cell cultures incubated at 37°C in a humidified CO 2 atmosphere. As a positive control, cells were infected with 100μl (100TCID 50 ) virus in serum-free MEM, and as a negative control, serum-free MEM was added to the cells. Cells were examined daily for 7 days, by light microscopy, for the appearance of a CPE. The absence of CPE at a specific concentration of the plant extract was considered to be indicative of antiviral activity.
To investigate the effect of the aqueous extract of the rhizomes of S. aethiopicus on the adsorption and replication of Inf A, the procedure was effectively the same as described above except that the VK cells were grown in 16-well Lab-Tek ® glass tissue culture chamber slides and 1 000TCID 50 Inf A was used per well. Viral infection was monitored by direct IF, in 4 wells per concentration of plant extract, as described previously. Fluorescent foci were counted in three separate fields on each of the 4 wells and the reduction in number of fluorescent foci in relation to the control wells was calculated.

Anthelmintic bioassay
A simple anthelmintic bioassay, using Caenorhabditis elegans free-living nematodes as test organisms, was carried out as described by McGaw et al. (2000). The water, ethanol and ethyl acetate extracts of the leaves, rhizome and roots of S. aethiopicus were tested at concentrations of 0.5 and 1mg ml -1 . Water and ethanol extracts were redissolved in their extracting solvents, and the ethyl acetate extracts were redissolved in ethanol for use in the assay. A standard concentration of 5μg ml -1 levamisole was used as a control. C. elegans var. Bristol (N2) nematodes were cultured on nematode growth agar seeded with E. coli according to the method of Brenner (1974). For this assay, 500-1000 nematodes (7-10 day-old cultures) in M9 buffer (Brenner 1974) were incubated with the plant extracts for 2h at 25°C in the dark. Nematodes with no plant extracts or levamisole added were included as a control. The percentage of living nematodes, and their movement, was assessed using a dissecting microscope.

Antischistosomal bioassay
The antischistosomal assay was performed according to Sparg et al. (2000). Infected Bulinus africanus snails were placed into test tubes under a 60W electric light to promote the shedding of cercariae. The cercariae were collected and transformed into schistosomula worms by subjection to a shearing stress, using a syringe with an 0.8ml needle. The bioassay was run in 96-well microtitre plates. The aqueous extracts of the leaves, rhizomes and roots (100μl) were serially diluted 2-fold, giving an initial concentration of 25mg ml -1 per test solution. A similar 2-fold serial dilution of praziquantel (Sigma) was used as a positive control. A culture medium blank was included as a negative control. Three schistosomula, in 100μl culture medium, were added to each well. After incubation at 25°C for 1h, the survival of the schistosomula were assessed.

Biochemical induction assay (BIA)
The BIA was performed according to the method described by White et al. (1986). Water, ethanol and ethyl acetate extracts of the leaves, rhizomes and roots were resuspended to 100mg ml -1 and applied (10μl) to filter paper discs which were placed onto agar assay plates. Plates were incubated for 5 hours at 37°C, after which the chromogenic substrate was added and zones of induction recorded. The strong mutagen 4-nitroquinoline 1-oxide was used at concentrations of 1, 5, 50 and 100μg per spot as positive controls. Appropriate solvent controls were included as negative controls.

Results and Discussion
Results for the general screening of extracts from the different plant parts for antibacterial activity are shown in Table 1. No inhibitory activity against the test bacteria was detected with the aqueous extracts. The ethanol and ethyl acetate extracts tested showed greater antibacterial activity against the Gram-positive bacteria than the Gram-negative bacteria. No distinct differences were observed between the activity obtained with the ethanol and ethyl acetate extracts in the general screening of different plant parts, or between the activity of the different plant parts.
Antibacterial activity has been shown for a number of other species of the Zingiberaceae. In a study by Ahmad and coworkers (1998) (Arambewela et al. 1999). In screening 13 species of Alpinia, Costus and Zingiber, most of the dichloromethane and methanol extracts showed some antibacterial activity against B. subtilis, methicillin-resistant S. aureus (MRSA) and P. aeruginosa. The strongest inhibitory activity of a dichloromethane extract was shown by Alpinia mutica with the minimum inhibitory dose of 125μg per disc against both B. subtilis and MRSA (Habsah et al. 2000).
Results for the study of seasonal effects on antibacterial activity are shown in Table 2. Little difference was observed between the extracts prepared from fresh and dry material for activity against S. aureus, as well as for the other test bacteria (results not shown). Extracts prepared from the leaves gave a MIC value of 0.2 (fresh) and 0.1(dry)mg ml -1 before senescence and 3.13mg ml -1 after senescence, indicating a loss in activity. Extracts prepared from α-roots gave MIC values of 3.13 and 1.56mg ml -1 before senescence and 0.1 and 0.2mg ml -1 after senescence, indicating an increase in activity.
The results for the COX-1 inhibition assay are shown in Figure 1A. Aqueous extracts showed no significant activity. Ethanol and ethyl acetate extracts of the leaves showed high levels of anti-inflammatory activity. These results confirm the findings of Zschocke et al. (2000b) that the leaves of S. aethiopicus showed the highest levels of anti-inflammatory activity. The rhizome and root extracts showed much lower levels of activity, with the ethyl acetate extracts having slightly higher activity than the ethanol extracts. Furthermore, the level of activity in the rhizomes was lower than in the roots.
Inhibition of the various extracts in the COX-2 assay are shown in Figure 1B. As was seen in the COX-1 assay, the ethanol and ethyl acetate extracts of the leaves showed the highest levels of activity. Similarly, the rhizome and root extracts showed lower levels of activity, with the ethyl acetate extracts having slightly higher activity than the ethanolic extracts. The aqueous extracts of the roots showed no noteworthy activity, although those of the rhizomes and leaves were slightly higher.
A variety of compounds which demonstrate anti-inflamma-tory activity have been previously isolated from other species of the Zingiberaceae. Claeson and co-workers (1993) isolated three non-phenolic diarylheptanoids from hexane extracts of the rhizomes of Curcuma xanthorrhiza. These compounds showed significant anti-inflammatory activity in the assay of carrageenin-induced hind paw oedema in rats. Bioassay-guided fractionation of extracts of the rhizomes of Zingiber cassumunar led to the isolation of three cassumunins with anti-inflammatory activity (Masuda and Jitoe 1994). In further studies on Z. cassumunar by Pongprayoon and co-workers (1996), 5 compounds with topical anti-inflammatory activity were isolated from hexane extracts of the rhizome. These compounds gave ID 50 values ranging from 2 to 62μg/ear in the model of 12-O-tetradecanoylphorbol-13-acetate-induced ear oedema in rats. Two pimarane diterpenes have been isolated from Kaempferia pulchra, and were found to have ID 50 values estimated at 330 and 50μg/ear in the above-mentioned rat ear oedema bioassay .
Results for the effect of senescence on COX-1 inhibition are shown in Figure 2. As observed with the general screening, the leaves showed the highest levels of activity, with the exception of senesced leaf material that was dried prior to extraction. No differences were observed between activity of the αand β-rhizomes, before and after senescence. However, there did appear to be a slight loss of activity as a  result of drying the rhizome material before extraction. This does suggest that fresh rhizome material may be more effective for use as a traditional medicine. For the α-roots, there was a slight decrease in activity following senescence. This was seen in both the fresh and dried material. The βroots showed a trend opposite to this, in that there was an increase in activity following senescence.
In the cytotoxicity assay the integrity of the VK cell monolayers, treated with concentrations of the S. aethiopicus aqueous leaf extract, from 3.9μg ml -1 up to 1 000μg ml -1 , was maintained. Only minimal morphological changes, indicative of cytotoxic effects, were observed during the 7 day observation period at extract concentrations greater than 250μg ml -1 . This low level of toxicity was confirmed in the MTT assay (Table 3). Furthermore, the aqueous leaf extracts which were tested against HSV-1 and HSV-2, at concentrations ranging from 3.90μg ml -1 to 500μg ml -1 , exhibited no antiviral activity when the virus was inoculated onto cell cultures simultaneously with the plant extract (results not shown). The same extracts also showed no activity against the replication of either HSV-1 or HSV-2 after viral adsorption had taken place (results not shown).
Cytotoxicity testing of the aqueous extracts of the rhizome revealed high levels of cytotoxicity (Table 3), as evidenced by a percentage survival of the VK cells of less than 50% at concentrations of 125μg ml -1 and higher. In performing the assay, toxicity of the extract was evident by day 3 after inoculation. This extract was also tested in the influenza assay, but did not indicate any noteworthy activity against Inf A (results not shown).
No activity was observed in the anthelmintic, antischistosomal and BIA assays (results not shown). For the BIA test, the S. aethiopicus leaf extracts prepared with ethanol and ethyl acetate produced a zone of bacterial growth inhibition around the site of sample application in the BIA assay. However, no red ring of β-galactosidase induction was observed, so it is likely the extracts did not cause any DNA damage, but rather simply exhibited antibacterial activity.    Indomethacin gave 71% inhibition in the COX-1 assay. In the COX-2 assay, indomethacin and nimesulide standards inhibited the synthesis of prostaglandins by 54% and 33% respectively Although no activity in the antiviral, anthelmintic, antischistosomal or biochemical induction assays was detected, anti-inflammatory activity was confirmed, and antibacterial activity observed against both Gram-positive and Gram-negative bacteria. From the experiments on seasonal senescence, it is clear that the time of harvest, and state of the material for extraction may only have a minimal influence on the degree of antibacterial and anti-inflammatory activity. Generally, little differences were observed between the levels of activity of the rhizomes before and after senescence, which is the plant part most used by traditional healers. This would suggest that the rhizomes could be harvested some time just before senescence or after the leaves have fully died back. Results from this study also support the use of the roots, which are sometimes used in traditional medicine. The results from the cytotoxicity assay indicate that the rhizomes are potentially harmful. This is an aspect which would require further investigation in terms of the medicinal use of this plant. These are important factors in the use of S. aethiopicus by traditional healers, and should be taken into consideration with regards to cultivation, commercialisation and product quality.