Background

The worldwide use of natural products including medicinal plants has become more and more important in primary health care especially in developing countries. Many pharmacognostical and pharmacological investigations are carried out to identify new drugs or to find new lead structures for the development of novel therapeutic agents for the treatment of human diseases such as cancer and infectious diseases [1]. In developing countries and particularly in Yemen, a large segment of the population still rely on folk medicine to treat serious diseases including infections, cancers and different types of inflammations.

The Soqotra Archipelago in Yemen has long been a land of mystery. Over the centuries travelers returned from the Indian Ocean isles with bizarre tales of trees yielding dragon's blood and cucumbers, forests of frankincense and towering pinnacles shrouded in mist [2]. Soqotra is considered the "jewel" of biodiversity in the Arabian Sea. The long geological isolation of the Soqotra archipelago and its fierce heat and many droughts have combined to create a unique and spectacular endemic flora. Surveys have revealed that more than a third of the 800 or so plant species of Soqotra are found nowhere else [2]. Botanists rank the flora of Soqotra among the ten most endangered island flora in the world.

Currently, there is insufficient scientific research on the plants from Soqotra. In previous studies we described investigations of some endemic and non-endemic plants from the island Soqotra for their antimicrobial [3], antiviral [4], enzyme-inhibitory [5] and anticancer activity [6]. Since a literature search indicated the absence of further information regarding biological and phytochemical investigations of plants from Soqotra, this study was carried out as a part of our continued exploration of Yemeni medicinal plants for interesting biological activities. Thus, the main aim of the present project was to carry out a phytochemical and biological investigation on selected plants from the island of Soqotra, especially on those that are endemic and those that find use in the traditional medicine. In this study, 26 plants belonging to 17 families were collected for evaluation of their cytotoxic, antimicrobial and antioxidant activities and the main chemical contents.

Methods

Plant materials

The plants (Table 1) were collected from different locations of the island of Soqotra in the Winter of 2006 and identified at the Pharmacognosy Department, Faculty of Pharmacy, Sana'a University. Part of the identification of the investigated plants was done by Dr. Anthony G. Miller from the Royal Botanic Garden at Edinburgh, UK. Voucher specimens were deposited at the Pharmacognosy Department, Faculty of Pharmacy, Sana'a University.

Table 1 List of plants screened in the study.
Full size table

Extraction of plant materials

The air-dried and powdered plant materials (10 g of each) were extracted with 400 ml methanol (CH3OH) by Soxhlet extraction for 8 hours. The residue was dried over night and then extracted with 250 ml water (H2O) by using a shaking water-bath at 70°C for 2 hours. The obtained methanolic and water extracts were filtered and evaporated by using a rotary evaporator and freeze dryer. The dried extracts were stored at -20°C until used.

Determination of antimicrobial activity

Test organisms

The following microorganisms were used as test organisms in the screening: 3 Gram-positive strains namely, Staphylococcus aureus (ATCC 6538), Bacillus subtilis (ATCC 6059), Micrococuss flavus (SBUG 16), 2 Gram-negative strains namely, Escherichia coli (ATCC 11229), Pseudomonas aeruginosa (ATCC 27853) and one fungal strain Candida maltosa (SBUG 700). In addition, three multiresistant Staphylococcus strains namely, Staphylococcus epidermidis 847, Staphylococcus haemolyticus 535, and Staphylococcus aureus North German Epidemic Strain (supply from the Institute of Hygiene of Mecklenburg-Vorpommern, Greifswald, Germany) were also employed as test organisms. Stock cultures were maintained at 4°C on slopes of nutrient agar. The cultures were diluted to achieve optical densities corresponding to 2.0 × 106 colony forming units (CFU/ml) for bacterial and 2.0 × 105 spore/ml for fungal strains.

Antimicrobial assay

The disc-diffusion assay [7] was used to determine the antimicrobial potential of the investigated extracts. Nutrient agar (OXOID LTD, Basingstoke, Hampshire, England) was prepared by dissolving of 27 g/l in water. The sterile nutrient agar was inoculated with microbial cells (200 μl of microbial cell suspension in 20 ml agar medium) and poured into sterile petri dishes. Sterile filter paper discs of 6 mm diameter (Schleicher and Schuell, ref. No. 10321260, lot. DG0274-1) were impregnated with 20 μl of the extract solution (equivalent to 4 mg of the dried extract). The paper discs were dried and placed on the surface of the inoculated agar plates. Plates were kept for 2 hours in refrigerator to enable prediffusion of the extracts into the agar. Then the plates were incubated overnight (18 hours) at 37°C. In contrast, M. flavus was incubated at room temperature for 48 h and C. maltosa was incubated at 28°C for 48 h. Ampicillin, gentamicin and amphotericin B were used as positive control. Negative controls were performed with paper discs loaded with 20 μl of organic solvents (methanol and 5% ethanol) and dried. At the end of the incubation period the antimicobiral activity was evaluated by measuring the inhibition zones (diameter of inhibition zone plus diameter of the disc). An inhibition zone of 15 mm or more was considered as high antibacterial activity.

Broth micro-dilution assay for minimum inhibitory concentrations (MIC)

The broth micro-dilution method described by [8] with modifications was used to determine the MIC of extracts against the three standard Gram-positive strains. With sterile round-bottom 96-well plates, duplicate two-fold serial dilutions of extract (100 μl/well) were prepared in the appropriate broth containing 5% (v/v) DMSO to produce a concentration range of 2000 to 15.6 μg of extract/ml. Two-fold dilutions of ampicillin were used as a positive control. A bacterial cell suspension (prepared in the appropriate broth) of 100 μl, corresponding to 1 × 106 CFU/ml, was added in all wells except those in columns 10, 11 and 12, which served as saline, extract and media sterility controls, respectively. Controls for bacterial growth without plant extract were also included on each plate. The final concentration of bacteria in the assay was 5 × 105 CFU/ml. The final concentration of extracts ranged between 1000 to 7.8 μg/ml. Plates were then incubated at 37°C for 18 h overnight. After incubation, the MIC of each extract was determined as the lowest concentration at which no growth was observed in the duplicate wells. Twenty microliters of a p-iodonitro-tetrazolium violet solution (0.04%, w/v) (Sigma, USA) was then added to each well. The plates were incubated for a further 30 min, and estimated visually for any change in color from yellow to pink indicating reduction of the dye due to bacterial growth. The highest dilution (lowest concentration) that remained yellow corresponded to the MIC. Experiments were performed in duplicate.

Determination of anticancer activity-Cytotoxicity assay on human cancer cell lines

For the estimation of the in vitro cytotoxic potency of the investigated extracts, an established microtiter plate assay [9] was used with three human cancer cell lines: one lung cancer (A-427), one urinary bladder cancer (5637) and one breast cancer (MCF-7) line. Cell lines were obtained from the DMSZ, Braunschweig, Germany, and culture in RPMI 1640 medium with 10% FCS. Cytotoxicity determinations are based on cellular staining with crystal violet and were performed as previously described in detail. Briefly, a volume of 100 μl of a cell suspension was seeded into 96-well microliter plates at a density of 1000 cell/well. Twenty-four hours later, cells were treated with the plant extracts at five dilutions and exposed continuously to the extracts for the next 96 h. Etoposide was used as a positive control. At the end of the exposure time, the medium was removed and the cells were fixed with a glutaraldehyde solution. The cells were then stained with crystal violet and the optical density (OD) was measured at λ = 570 nm with a plate reader. The percent growth values were calculated by the following equation:

Growth (%) = ODT - ODc, 0/ODc - ODc, 0 × 100

Where ODT is the mean absorbance of the treated cells, ODc is the mean absorbance of the controls, ODc, 0 is the mean absorbance at the time the extract was added. The IC50 values were estimated by a linear least-squares regression of the growth values versus the logarithm of the extract concentration; only concentrations that yielded growth values between 10% and 90% were used in the calculation.

Determination of antioxidant activity (Scavenging Activity of DPPH Radical)

The DPPH free radical scavenging assay was carried out for the evaluation of the antioxidant activity. This assay measures the free radical scavenging capacity of the investigated extracts. DPPH is a molecule containing a stable free radical. In the presence of an antioxidant, which can donate an electron to DPPH, the purple colour typical for free DPPH radical decays, and the absorbance change at λ = 517 nm is measured. This test provides information on the ability of a compound to donate a hydrogen atom, on the number of electrons a given molecule can donate, and on the mechanism of antioxidant action. The method was carried out as previously described by [10]. The methanolic and aqueous extracts were redissolved in methanol and 5% ethanol, respectively, and various concentrations (10, 50, 100, 500 and 1000 μg/ml) of each extract were used. Similar concentrations of ascorbic acid were used as positive control. The assay mixture contained in a total volume of 1 ml, 500 μl of the extract, 125 μl prepared DPPH (1 mM in methanol) and 375 μl solvent (methanol or 5% ethanol). After 30 min incubation at 25°C, the decrease in absorbance was measured at λ = 517 nm. The radical scavenging activity was calculated from the equation:

% of radical scavenging activity = Abscontrol - Abssample/Abscontrol × 100

Phytochemical screening of the methanolic extracts

The screening of chemical constituents was carried out with the methanol extracts by using chemical methods and thin-layer chromatography (TLC) using different mixtures of organic solvents as mobile phases. Several chemical reagents e.g. Dragendorf's reagent for alkaloids, borntraeger reagent for anthraquinons and etc. were used in the detection according to previously published methodology [11].

Results

In the course of our screening for the antimicrobial, anticancer, and antioxidant and activities, a number of plants from different locations of the island Soqotra used in Yemeni traditional medicine were evaluated. A total of 52 extracts representing 26 plant species belonging to 17 families were submitted to biological screening. The botanical names, plant part used and the traditional uses of the plants in the collected areas are presented in Table 1.

Antimicrobial activity

Table 2 shows the results of the antimicrobial activity of the investigated extracts in agar diffusion method. An inhibition zone > 15 mm was considered as a high antimicrobial activity. It was observed that the antimicrobial activity of the studied plant extracts was exhibited mainly against the Gram-positive bacteria. Consequently, the MIC-values were determined only against Gram-positive bacteria. The MIC values are reported in Table 3. In a general, among the investigated extracts the methanolic extracts exhibited the greatest antimicrobial effect.

Table 2 Antimicrobial activity of the investigated plants in agar diffusion test.
Full size table

Only rarely do plants shown antimicrobial activity against Gram-negative bacteria and Candida maltosa except Acacia pennivenia, Boswellia species and Euclea divinorum. It was apparent that the multiresistant Staphylococcus strains demonstrate more sensitivity to the investigated extracts than the other antibiotic susceptible Gram-positive bacteria (Table 2). Moreover, it was found that with the exception of Boswellia species, Euclea divinorum and Hypoestes pubescens no extract showed antifungal activity against Candida maltosa (Table 2). The most pronounced activity with inhibition zones greater than 15 mm was found with the methanolic extracts of 10 plants (Table 2). The most effective plant was Boswellia socotrana, of which the methanolic and aqueous extracts demonstrated the greatest antimicrobial effect against all tested microorganisms (Table 2). The lowest MIC values were obtained against Staphylococcus aureus and Micrococuss flavus by the methanolic extracts of Boswellia species (125 μg/ml) (Table 3).

Table 3 Free radical scavenging activity and MIC of the investigated plants.
Full size table

In vitro anticancer activity

Table 4 presents the IC50 values for the in vitro cytotoxic activity of the investigated 26 methanolic extracts. From our experience, the most aqueous extracts don't show any notable in vitro anticancer effect at the highest concentration tested (50 μg/ml). Generally, the aqueous extracts may exhibit only a weak cytotoxic effect starting with a concentration of 250 μg/ml. Thus, these extracts were excluded in our screen. The results in Table 4 demonstrate that out of the 26 investigated plants (only methanolic extracts) three plant extracts had noteworthy cytotoxic effects in all three cell lines tested, namely: Ballochia atro-virgata (IC50 between 1.5 and 2.8 μg/ml), Eureiandra balfourii (IC50 between 0.8 and 3.4 μg/ml) and Hypoestes pubescens (IC50 between 4.8 and 8.2 μg/ml) (Table 4). Moreover, the extracts of Acanthospermum hispidum, Boswellia dioscorides, Boswellia socotrana, Commiphora ornifolia and Euphorbia socotrana exhibited a pronounced cytotoxic effect against all tested cell lines ((IC50 between 9.3 and 38.5 μg/ml). The other plant extracts demonstrated no significant cytotoxic effect at the highest tested concentration of 50 μg/ml.

Table 4 IC50 values (μg/ml) for cell growth inhibition of the investigated methanolic extracts on three human cancer cell lines and phytochemical screening.
Full size table

Radical scavenging activity

The methanol extracts of seven plants, namely: Acacia pennivenia, Acridocarpus socotranus, Boswellia dioscorides, Boswellia socotrana, Commiphora ornifolia, Euphorbia socotrana, and Lannea transulta showed a high effective free radical scavenging in the DPPH assay (Table 3). These extracts exhibited a noticeable antioxidant effect at low concentrations (Table 3). So the methanolic extracts of Acacia pennivenia, Boswellia dioscorides, Boswellia socotrana and Commiphora ornifolia exhibited a great antioxidant effect at 50 μg/ml (78, 89, 88% and 85%, respectively), comparing with the effect of ascorbic acid at this concentration (Table 3). The hot aqueous extracts of all investigated plants showed only weak antioxidant effect (Table 3).

Phytochemical screening

The results of the phytochemical screening of the investigated methanolic extracts indicated the presence of different types of active constituents like flavonoids, terpenoids, tannins, volatile oils, etc... (Table 4).

Discussion

In continuation of our search for substances of plant origin with pharmacological effects, we have screened 26 plants collected from the island Soqotra, Yemen for their antimicrobial, cytotoxic and antioxidant activities and for their chemical content. It is important to mention that this work represents the first report on the antimicrobial, cytotoxic and antioxidant activities of extracts from 20 endemic plants (Table 1). The existing knowledge about the other six investigated plants is in many cases very limited.

A correlation was found between the antibacterial activity observed by agar diffusion assay and MIC determination. It is interesting to note that these plant extracts showed more activity on multi-drug resistant Staphylococcus strains than the antibiotic susceptible Gram-positive bacteria.

It was demonstrated previously that alcoholic extract of different species of Acacia like A. arabica, A. nilotica and A. auriculiformis have antibacterial activity [12–15], and that the isolated saponines from A. auriculiformis were responsible for the noticed effect [15]. Previous work indicated that both bark and heartwood extracts of A. confusa clearly have strong antioxidant effects [16]. Phenolic compounds were isolated from the bark of A. confuse, which were mainly responsible for the antioxidant effect of this plant [17]. Furthermore, triterpenoid saponins were isolated from A. victoriae, which induced decreased tumor cell proliferation and induced apoptosis [18]. In addition, a significant reduction in the values of tumor burden and tumor incidence was observed in mice treated by oral gavage with the A. nilotica gum and leaf extracts [19]. The antimicrobial and antioxidant effects we have found with extracts from A. pennivenia are in accordance with these data. The phytochemical screening revealed the presence of saponins and phenolic compounds like tannins in the methanolic extract of A. pennivenia, which could be responsible for these activities. On the other hand, our results of cytotoxic activity were not in agreement with the anticancer effect noted with other Acacia species; i.e., at the highest concentration used in our screening (50 μg/ml), the extracts of A. pennivenia exhibit no cytotoxic activity.

The antimicrobial effect of leaves and flowers extracts of Acanthospermum hispidum has been previously investigated [20]. Moreover, the isolation of sesquiterpene lactones was reported [21, 22]. Our screening results confirmed the antimicrobial effect of the alcoholic extract. The hot aqueous extract exhibited a moderate antibacterial effect in contrast with data presented earlier [20]. In addition, the methanolic extract brought about a pronounced growth inhibitory effect against all tested cancer cell lines we tested it against. These effects are most likely due to the presence of sequiterpene lactones identified in our phytochemical screening and in literature data.

Whereas the investigation of Acridocarpus vivy showed cytotoxic effect against some cancer cell lines [23], the methanolic extract of Acridocarpus socotranus demonstrated no cytotoxic activity in our screen. However the plant extract exhibited a strong antioxidant and moderate antibacterial effects. This finding may be correlated with the presence of flavonoids and terpenoids.

In contrast to Ali and co-workers [24], who found a promising antimicrobial effect for Aloe perryi, our extracts of Aloe perryi only showed activity against multiresistant bacteria. Although the phytochemical screening illustrated the presence of anthraquinons, which are mostly responsible for cytotoxic and antioxidant activity [25], our results do not indicat any cytotoxic activity at the highest concentration tested (50 μg/ml) and exhibited a moderate antioxidant effect only at high concentrations (500 and 1000 μg/ml)

It is important to consider that we found no published data on the genus Ballochia and Eureiandra. So this is the first report on pharmacological and chemical investigation of plants of these genera. The remarkable cytotoxic effect of both plants with IC50 values between 0.8 and 3.4 μg/ml is significant. According to the criteria of the American National Cancer Institute, 30 μg/ml is the upper IC50 limit considered promising for purification of a crude extract [26]. Therefore, the highest concentration tested (50 μg/ml) in our screening was slightly above this limit.

The chemical screening showed the presence of terpenoids in both plants (mainly diterpenoids and triterpenoids), which could be correlated with this effect. E.balfourii belongs to the family cucurbitaceae, plants of which are famous for their cytotoxic tetracyclic triterpenoids known as cucurbitacins [27–29]. This type of compounds could be responsible for the observed activity in E.balfourii.

In our previous studies, the antimicrobial effects of some Boswellia species namely B. elongata and B. ameero were investigated [3]. It was found that both plants exhibited a strong antimicrobial effect only against Gram positive bacteria. Both plants demonstrated no cell growth inhibition on five human cancer cell lines [6]. Furthermore, it was reported that the extract of B. serrata showed an effect against some bacterial strains [30]. The antimicrobial activity and the chemical content of the volatile oil of some Boswellia species like B. carteri, B. papyrifera, B. serrata and B. rivae were determined [31]. It was shown that the oils consisted of several monoterpenes, sesquiterpenes and diterpenes are responsible for the effect observed. The B. socotrana and B. dioscorides we investigated produced extracts with the most effective activity against all tested microorganisms. The essential oils and the terpenoids determined in our phytochemical screening are mostly responsible for this effect. Besides this noticeable antimicrobial effect, the methanolic extracts manifested a considerable cytotoxic effect against two cell lines (5637 and A-427) with IC50 values between 18 and 29 μg/ml. For this effect, boswellic acids, which may represent a considerable part of the chemical content, could contribute to this observed effect. Furthermore, both plants demonstrated even at low concentration (50 μg/ml) a remarkable radical scavenging effect (88 and 89%). This effect may be attributed to the present flavonoids.

Among the most interesting plants, Commiphora ornifolia showed similar antimicrobial activities with Boswellia plants. The same extract manifested a high antioxidant effect (85%) at 50 μg/ml. The extracts and isolated compounds from Commiphora opobalsamum exhibited similar antimicrobial and antioxidant activities [32]. Thus, the estimated antimicrobial and antioxidant effects of our investigated C. ornifolia are in accordance with these data. Whereas no cytotoxic activity in C. myrrha was found [33], we observed an antiproliferative effect of the methanolic extract against all three cell lines with IC50 values between 30 and 38 μg/ml. The determined effects of C. ornifolia are mostly attributed to the essential oil, flavonoids and tritepenes found in the methanolic extract.

Another interesting plant was Euphorbia socotrana, which demonstrated considerable antimicrobial, cytotoxic and antioxidant activities. Our data are in agreement with literature data of other Euphorbia species as E.thymifolia, E.hirta and E. cheiradenia [34–36]. Studies have demonstrated that some triterpenoids isolated from Euphorbia species are responsible for the antimicrobial effect and that some of these compounds were able to induce moderate apoptosis in at least one cancer cell line [37, 38]. The presence of such terpenoids in our E. socotrana may explain the biological effects seen in our screens.

Some authors reported on investigation with some Hypoestes species, where the isolation of two cell growth inhibitory phenanthroindolizidine alkaloids termed hypoestestatin 1 and hypoestestatin 2 from the East African shrub H.verticillaris was described [39]. In addition, several antifungal diterpenoids from H.serpens were isolated [40]. The methanolic extract of H. pubescens tested in our screens showed a remarkable antifungal effect against Candida maltosa and an extraordinary cytotoxic effect against all tested cancer cell lines. Apparently these observed activities are correlated with the presence of the terpenoids and alkaloids found in the phytochemical screening.

The notably antimicrobial effect of Rhus thyrsiflora and Teucrium sokotranum was consistent with literature data of other Rhus and Teucrium species [41, 42]. On the other hand, other authors noted a strong antioxidant effect of some Teucrium species and R.coriaria, which is not supported by results obtained in our screening [43, 44]. Moreover, a cytotoxic activity for the sap of the lacquer tree R. succedanea was described previously [45], whereas the investigated R. thyrsiflora showed no growth inhibitory effect on the tested cancer cell lines.

Conclusion

In conclusion, the results in the present study are agreed to some extent with the traditional uses of the plants investigated. Our results further support the idea that medicinal plants can be promising sources of potential anticancer and antimicrobial agents and antioxidants. The present results will form the basis for selection of plant species for further investigation in the potential discovery of new natural bioactive compounds. Studies aimed at the isolation and structure elucidation of anticancer, antibacterial and antioxidant active constituents from some plants e.g. Acacia pennivenia, Ballochia atro-virgata, Boswellia dioscorides, Boswellia socotrana, Commiphora ornifolia, Euphorbia socotrana, Eureiandra balfourii and Hypoestes pubescens are in progress.