Antifungal properties of a new terpernoid saponin and other compounds from the stem bark of Polyscias fulva Hiern (Araliaceae)

Background In our previous studies, it was evident that the dichloromethane-methanol (1:1 v/v) stem barks extract of Polyscias fulva and fractions (ethyl acetate, n-butanol and residue) demonstrated interesting antidermatophytic activities. So, as a continuity of that, this work aimed at identifying active principles with antifungal properties from P. fulva that could be used as markers for possible standardization of this plant as phytomedicine. Methods The ethyl acetate, n-butanol and residual fractions of the dichloromethane-methanol (1:1 v/v) stem bark extract of Polyscias fulva were further fractionated by column chromatography and the structures of isolated compounds elucidated based on their spectroscopic data in comparison with existing literature information. Antifungal activity was assayed by broth microdilution techniques on yeasts and dermatophytes spores. Results The fractionation of the crude dichloromethane-methanol (1:1 v/v) stem bark extract of Polyscias fulva led to the isolation of 10 known compounds (1 to 10) and one new saponin (11: 3-O-[α-L-rhamnopyranosyl (1–2)-α-L-arabinopyranosyl]-28-O-[α-L-4-O-acetyl-rhamnopyranosyl (1–4)-β-D-glucopyranosyl-(1–6)-β-D-glucopyranosyl]-hederagenin). Among these compounds, 3-O-α-L- arabinopyranosyl-hederagenin and 3-O-[α-L-rhamnopyranosyl (1–2)-α-L-arabinopyranosyl]-hederagenin were the most active on the tested fungi with MIC values ranging from 0.78 to 100 μg/ml against both yeasts and dermatophytes. Conclusion The results of this work constitute a step forward in the possible development of an antidermatophytic phytomedicine from Polyscias fulva stem bark, the isolated compounds being possible markers for the standardisation. Electronic supplementary material The online version of this article (doi:10.1186/s12906-015-0541-7) contains supplementary material, which is available to authorized users.


Background
Exploring the healing power of plants is an ancient concept. For many centuries people have been trying to alleviate and treat diseases with different plant extracts and formulations [1]. The interest in plants with antimicrobial properties has been revived because of current problems associated with the use of antibiotics [2]. The fact that microorganisms among which fungi nowadays tend to develop resistance towards drugs, coupled to the undesirable side effects of certain antibiotics is a real problem of concern. Medicinal plants constitute a prolific source of antimicrobial substances. The valorization of medicinal plants through the production of phytomedicine implies the isolation of active compounds that can be used in the standardization process of such drugs.
Polyscias fulva is a medium size and fast growing deciduous tree of the tropical forests of sub-Saharan Africa which is found at an altitude range of 1,180-2,500 m, with annual rainfall of 1,500-2,000 mm [3]. In Cameroon, decoction of its bark is orally administered to cure venereal infections [4] while paste from its stems barks and leaves are used topically against dermatoses. In previous studies, the dichloromethane extract from the bark of Polyscias fulva appeared to possess a weak antiplasmodial against Plasmodium falciparum (IC 50 = 9.8 μg/ml) and antitrypanosomial activities against Trypanosoma rhodesiense (MIC = 100 μg/ml) [5]. Furthermore, its dichloromethanemethanol (1:1 v/v) extract showed interesting in vitro and in vivo antidermatophytic properties [6]. With the aim of producing a standardized phytomedicine from the plant species, the dichloromethane-methanol of the stem bark was fractionated to isolate and characterize the antifungal active principles.

Plant material
The stem bark of Polyscias fulva (Hiern) was collected in April 2008 at Bazou (Nde Division, West Region, Cameroon). Botanical identification was done at the Cameroon National Herbarium in Yaoundé by Mr Tadjouteu Fulbert, where a voucher specimen was kept under the reference number 43546/HNC.

Microorganisms
The antimicrobial activities of different substances were carried out on seven yeasts and eleven dermatophytes. The references strains ATCC were obtained from the American Type Culture Collection (Rockville, MD, USA), IP from "Institut Pasteur" of Paris-France, E from "Ecole Nationale Vétérinaire d'Aford" in France, CBS from the Centraalbureau Voor schimmelcultures (Central office for fungal cultures) in Netherlands, whereas the clinical isolate were obtained from the Laboratory of Bacteriology and Mycology of the "Centre Pasteur" of Yaoundé-Cameroon. The strains have been maintained in the refrigerator at 4°C on agar slant.
UV spectra were measured using a Shimadzu UV-2401 PC spectraphotometer. IR spectra were obtained on Bruker Tensor-27 infrared spectrophotometer with KBr pellets. ESI-MS spectra were recorded on a Bruker HTC/Esquire spectrometer, HRESIMS spectra were recorded on an API Qstar Pulsar instrument. NMR, 1 H-1 H COSY, HMBC, and HSQC experiments were performed on Bruker AM-400, DRX-500, and Avance III 600 instruments with TMS as the internal standard. Chemical shifts (δ) were expressed in ppm with reference to the solvent signals.

Fractionation and isolation of active compounds from the plant extract
The dichloromethane-methanol (1:1 v/v) extract (263 g) from the stem bark of Polyscias fulva was pre-dissolved in 250 ml of methanol and water (1:9) and shaken vigorously in 500 ml of n-hexane. The n-hexane phase was collected and the process repeated twice. The methanol was then evaporated from the polar phase. The residue obtained after methanol evaporation was partitioned in the ethyl acetate and finally in the n-butanol as above. The n-hexane, ethyl acetate and n-butanol phases were concentrated under reduced pressure in a rotatory evaporator to obtain 24.25 g of the n-hexane (09.22%), 23.77 g of the ethyl acetate (09.03), 16 g of the n-butanol (06.08%) and 195.98 g of the residual (74.52%) fractions after solvent evaporation. The ethyl acetate, n-butanol and residual fractions were fractionated by column chromatography.

Acid hydrolysis and GC analysis of compound 11
Two milligrams of compound 11 were refluxed with 2 M HCl (1, 4 dioxane/H2O 1:1, 2 ml) on water bath for 2 h at 95°C. After cooling, the reaction mixture was extracted with CHCl 3 (3 × 5 ml). The aqueous layer was evaporated to dryness with MeOH until neutral. The dried residue was dissolved in 1 ml anhydrous pyridine and treated with L-cysteine methyl ester hydrochloride (1.5 mg) stirred at 60°C for 1 h. Trimethylsilylimidazole (1.0 ml) was added to the reaction mixture, and kept at 60°C for 30 min.
Column temperature: 100-230 with the rate of 10°C/ min, and the carrier gas was N 2 (2 ml/min); injector temperature: 250; split ratio: 1/20. The standard monosaccharides were subjected to the same reaction and GC-MS analysis. Under these conditions, the derivatives of Dglucose, L-rhamnose and L-arabinose were detected at 10.942, 7.645, and 8.138 min respectively.

Preparation and standardisation of inocula Preparation of yeasts inocula
Inoculum of each yeast was prepared from a 48 hours Sabouraud Dextrose Agar culture. Isolated colonies from this culture were diluted in 0.9% NaCl to match the 0.5 Mc Farland standard turbidity, corresponding to about 1.5 × 10 8 colony forming unit (CFU) per mL. This microbial suspension was diluted to match the 0.09 optical density at 600 nm corresponding to 2.5 × 10 5 spores.mL −1 using a Jenway 6105UV/Vis spectrophotometer (50 Hz/ 60 Hz) [7].

Preparation of dermatophyte inocula
The inoculum of each dermatophyte was prepared from a 15 days old culture on Sabouraud Dextrose Agar (Conda, Madrid, Spain). The culture surfaces were gently scraped and introduced in test tubes containing 10 mL of sterile saline, homogenized for 5 minutes and filtered. The absorbance of the spore suspensions (filtrates) were read at 530 nm and adjusted with sterile distilled water between 0.15 and 0.17 (Jenway 6105UV/Vis spectrophotometer, 50Hz/60Hz) to match 0.6 × 10 6 -1.4 × 10 6 CFU.mL −1 [8].

In vitro antimicrobial assay
The broth microdilution method [9] was used to determine the minimum inhibitory concentration (MIC) and minimum fungicidal concentration (MFC) of the tested substances using 96 well microplates (Nunclon, Roskilde, Denmark). 96-well plates were prepared by dispensing into each well 100 μL of Sabouraud Dextrose broth for both yeasts and dermatophytes. A volume of 100 μL of each test sample was added into the first wells of the micro-titre plate. Serial two-fold dilutions of these test samples were made. A volume of 100 μL of the above standardized inocula was then added into each well to match approximately 2.5×10 3 CFU.mL −1 for yeast and 10 4 CFU.mL −1 for dermatophytes in a total volume of 200 μL. This gave final concentration range of 0.1 to 0.00078 mg.mL −1 for compounds as well as for reference drug (positive control). For every experiment, sterility control (5% v/v aqueous DMSO and broth) and negative control (broth plus inoculum) were included. The content of each well was mixed thoroughly and the micro well plates were covered with the sterile lips and incubated at 37°C for 48 h for yeasts and at 28°C for 5 days for dermatophytes on a plate shaker (Flow Laboratory, Germany) at 300 rpm. After incubation, fungal growth in each well was monitored by observing and comparing the turbidity of the test wells to that of the positive and negative controls. MIC was the lowest concentration of the test substances that prevented visible growth of the microorganisms.
The MFC values were determined by subculturing 50 μL aliquots of the preparations, which did not show any visible growth of the micro-organisms during MIC determinations, into 150 μL of test substance-free SDB. These preparations were further incubated as indicated above. Microbial growth in each well was determined as

Structures of isolated compounds
Six compounds (1-6) were isolated from the ethyl acetate fraction ( Figure 1); five others (7-11) from the n-butanol fraction, compounds 10 and 11 were also isolated from the residual fraction. The isolated compounds belong to various chemical groups but mainly triterpenoid saponins. The presence of an acetyl group was revealed by the presence of two signals in the 13 C NMR spectrum at δ 170.9 and 21.3 which were respectively assigned to a carbonyl group and a methyl, this assumption was further confirmed by a signal in the 1 H NMR at δ 2.01 (3H, s). The relatively deshielded shift of C-4""' of the Lrhamnopyranosyl of the C-28 unit compare to that of C-4" of the L-rhamnopyranosyl of the C-3 is due to the attachment of the acetyl unit at this position. This information was further supported by the long-range correlation observed between H-4"" and this carbonyl group. The 1 H and 13 C NMR data of compound 11 (Table 1)  Antifungal properties of compounds isolated from P. fulva stem bark The isolated constituents of P. fulva were screened for their antifungal properties against 8 yeasts (7 Candida species and Cyptococcus neoformans) and 11 dermatophyes (4 Microsporum, 6 Trichophyton and 1 Epidermophyton) ( Table 2). Compounds 7 and 8 were relatively active against all the tested yeasts while 2, 3 and 9 were totally inactive. The antifungal properties of the other compounds were selectively observed on some yeasts. All the tested substances showed a wide range of antidermatophytic activities, inhibiting the growth of almost all the tested dermatophytes (  from P. fulva extract possess MICs less than the reference drug Griseofulvin on selected microorganisms.

Discussion
The isolated compounds globally demonstrated more or less interesting antifungal activities. They were phenolics (1 and 3), steroids (2), triterpene (4) and terpenoid saponins (5-11) secondary metabolites. Most of the antimicrobial substances isolated from Cameroonian medicinal plants belong to these chemical groups [19]. Up to 7 of the 11 isolated compounds from P. fulva were terpenoid saponins; such substances with hederagenin or oleanolic acid as aglycone have been found to possess antifungal activities against yeasts and filamentous fungi [20]. Saponins possess the ability to bind with sterols in fungal membrane and cause pore formation and loss of membrane integrity [21] as antifungal of polyene group [22]. Structure-activity relationship of these types of compounds has been demonstrated and their antifungal properties depend on the number and type of sugar residues, but the increase in sugar length does not enhance the activity [21]. The anti-yeasts activities of saponins 7 and 8 (the only active compounds on yeasts) compared to compounds with the same basic skeleton (4-6 and 9-11) could be ascribed to the presence of a hydroxyl group at position 23 coupled to the presence of a free carboxyl group at position 28. In contrast, all these     A number of studies have been carried out on the antifungal activity of phenolic compounds from natural sources [22]. It was the case of pinoresinol (3) that has previously been described as fungicidal agent from Sambucus williamsii [23] and Methyl atrarate (1) that showed a good antifungal activity (MIC 6 μg/ml) on Candida albicans [24]. Furthermore antifungal properties of phenolic compounds may be due to iron deprivation or hydrogen binding with vital proteins such as microbial enzymes [25]. According to Hwang et al. [23], compound 3 may depolarize or form pores in the fungal bilayer membrane. These two compounds, different in size but with the same number of hydroxyl groups were inactive on yeast and possess almost the same antidermatophytic activities. It is postulated that the site(s) and number of hydroxyl groups of phenolic compounds are closely correlated to their antimicrobial activities [22].
The broad range antidermatophytic activities of the isolated compounds from P. fulvia explains the relatively good in vitro and in vivo antidermatophytic activity of the oil-moistened dichloromethane-methanol (1:1 v/v) crude extract from this plant [6]. They can then serve as markers for the standardization of antidermatophytic phytomedicine from P. fulva.

Conclusion
The tested compounds showed a broad range of antidermatophytic activities while only compounds 7 and 8 inhibited the growth of yeasts. Considering these results and those from our previous studies on the crude extract, these substances may be useful in the standardization of antimicrobial and particularly antidermatophytic phytomedicine from P. fulva.