Background

The genus Helichrysum (Compositae) consists of more than 600 species with a major center of distribution in South Africa [1]. Several Helichrysum species have been used in folk medicine of different countries for the protection of post-harvest food [2]. Moreover, Helichrysum species are used extensively for stress-related ailments and as dressings for wounds normally encountered in circumcision rites, bruises, cuts and sores [3]. It has been also reported that Helichysum species are used to relief abdominal pain, heart burn, cough, cold, wounds, female sterility, menstrual pain [4] and to treat some diseases such as gastric [5–7], gastroduodenal, gastric ulcers and gastritis [8], stomach damage [9, 10], acute hepatitis, fever, or oedema [11], diuretic, inflammatory, allergic [12, 13]. In addition, some of these species have been reported to possess antimicrobially active compounds [14–16].

Chemical studies on Helichrysum species have been carried out by many investigators and the presence of flavonoids, phloroglucinols, α-pyrones, coumarins and terpenoid compounds has been reported [17–25]. H. foetidum has been assessed to treat influenza, infected wounds, herpes, eye problems, menstrual pains and to induce trance and possess antifungal properties [2, 26]. H. mechowianum is used for the treatment of stomach damage, cephalgy [9, 27] and possesses ulcerogenic activity [28, 29]. In continuation of these studies, we extended our search for biologically active compounds from Helichrysum species [17, 18] to the protease-inhibiting activity of extracts and isolated compounds from Helichrysum foetidum and Helichrysum mechowianum using a fluorescence resonance energy transfer (FRET) protease pepsin inhibition assay as pharmacological model for anti-ulcer compounds [30]. Beside excessive stomach acid and Helicobacter pylori, pepsin is one of the major factors in the pathophysiology of peptic ulcer disease and reflux oesophagitis. In addition, the antibacterial and antifungal effects of both species against Bacillus subtilis and the yeast Cladosporium cucumerinum were evaluated respectively.

The chemical profile of methanol extracts of H. mechowianum and H. foetidum was investigated. To our knowledge, this is the first report about constituents of H. mechowianum. The compounds identified have been reported previously from other Helichrysum species in different compositions.

In order to assess the drug-likeness profiles of the isolated metabolites, low energy computer models were generated and a number of ADMET-related descriptors calculated, with the view of drug metabolism and pharmacokinetics (DMPK) evaluation.

Results and discussion

Biological tests

The methanol leaf extracts of Helichrysum foetidum and Helichrysum mechowianum showed significant activity in the pepsin protease FRET assay while no activity was detected against the serine protease subtilisin (Table 1). The extract of H. foetidum exhibited the higher pepsin protease inhibition (37.4 and 35.6 % inhibition at 50 and 25 μg/ml) (Table 1). Therefore also the previously isolated constituents 1–6 (Fig. 1) from H. Foetidum and 12–13 from H. mechowianum were tested. The best results at a concentration of 50 μg/ml were obtained with apigenin-7-O-β-D-glucoside (5) and 6′-methoxy-2′,4-dihydroxychalcone-4′-O-β-D-glucoside (3) with a moderate inhibition activity of 46.3 and 37.4 % respectively (Table 1) while 3,5,7-trihydroxy-8-methoxyflavone (12), 4,5-dicaffeoyl quinic acid (13) showed weak activity. These results suggest that the inhibition activity on the aspartate protease observed with H. foetidum extract could be mainly attributed to the glycosidic compounds (3) and (5). Contrarily, in the inhibition assay with the serine protease subtilisin, neither the crude extracts, nor the isolated substances of both species show significant activity (Table 1). We can conclude that the substances present in the crude extract of H. foetidum are selective for aspartate proteases. Observed negative results may be due to the auto-fluorescence debris of subtilisin cleavage of these compounds resulting in fluorescent residues or the absence of bioaffinity interactions between the substances present in the crude extract of H. foetidum with the serine protease subtilisin [31]. The observed protease inhibiting activity may have mucosal protective effects and therefore may help to reduce peptic ulceration. From the Black birch fungus (Inonotus obliquus), which is used in folk medicine in Russia for the treatment of gastrointestinal tract disorders, also the flavonoidal fraction was shown to possess antiulcerous activity [31].

Table 1 Activity (% inhibition) of Helichrysum crude extracts and major isolated compounds (1–6) in protease inhibition assays using pepsin and subtilisin
Full size table
Fig. 1
figure 1

compounds 1–6 isolated from leaves and flowers of H. foetidum and compounds 12–13 from leaves of H. Mechowianum

Full size image

In addition, the crude extract of both species and all the isolated compounds were subjected to in vitro antimicrobial assay against the reference strains of bacteria Bacillus subtilis and yeast (Cladosporium cucumerinu).

It has been reported that extracts having MIC values below 8000 μg/ml possess some antimicrobial activity [32]. MIC values below 1000 μg/ml are considered noteworthy [33, 34]. Thus, the crude extract having activities of 1000 μg/ml or lower against all the pathogens studied demonstrated potential anti-infective properties. Compounds having antimicrobial activities less than 64 μg/ml are accepted as having notable antimicrobial activity [33] and those compounds exhibiting activity at concentrations below 10 μg/ml are considered “clinically significant” [32, 33]. According to this observation, the crude leaf and flower extracts of H. foetidum showed a significant and concentration dependent growth inhibition of Bacillus subtilis of of 85.4 % at a concentration of 1 mg/ml and of 21.8 % at a concentration of 0.1 mg/ml whereas the crude extract of H. mechowianum at concentrations of 1 mg/ml and 0.1 mg/ml causes a moderate growth inhibition of 36,2 and 29,8 % respectively (Table 2). Likewise, the crude leave and flower extracts of H. Foetidum also exhibit antifungal activity against Cladosporium cucumerinu shown by the development of inhibition zones on the bioautography plate. In contrast, extracts of H. mechowianum were slightly active against this fungus.

Table 2 Antimicrobial activity ((%)of Helichrysum crude extracts and isolated compounds 1–6, 12–13 from H. Foetidum
Full size table

Furthermore, all of the isolated compounds were subjected to in vitro antimicrobial assay. It was interesting to note that compounds (1–6) from H. Foetidum exhibited notable growth inhibition range of 85.0 to 75.0 % against Bacillus subtilis and a range of 70 to 56 % against the yeast Cladosporium cucumerinu at a concentration of 1 mg/ml whereas compounds (12–13) from H. mechowianum showed a moderate growth inhibition range of 40.2 to 30.8 % at 1 mg/ml against Bacillus subtilis (Table 2). Of all the isolated, compound (6) exhibited the highest sensitivity growth inhibition of Bacillus subtilis of 85.0 % at a concentration of 1 mg/ml and was found to be the most active component of the crude flower extract of H. Foetidum (Table 2).

The results of the work indicate that diterpenoid possess antimicrobial against the gram positive bacterium. This antibacterial activity of H. foetidum extract might be associated to the high content of kaurenoic acid (6). This justifies the use of these plants species in folk medicine and corroborated with the previous reports on the antibacterial activities for Helichrysum species [2, 35, 36]. Kaur-16-en-19-oic acid isolated from extract of the Asteraceae (Senecio erechtitoides and Wedelia calendulaceae) was previously shown to possess high inhibitory activity against several bacterial strains [37, 38].

Chemical constituents

The main constituents of both species were characterized by detailed ESI-MS investigations. The combination of LC-MS, MS/MS and FTICR-HRMS allowed the detection of various components simultaneously. The MS experiments show, that H. foetidum and H. mechowianum possess different chemical compositions. The leaf extract of H. foetidum is dominated by the chalcones 2 and 3, the flavonoids 4 and 5 and by diterpenoids [18] whereas main constituents of H. mechowianum are quinic acid derivatives with a less prominent bioactivity profile.

A more detailed ESI-MS investigation of the crude extract of Helichrysum mechowianum indicates (Additional file 1) the presence of quinic acid (7, ESI-FTICR-MS: [M - H]−, m/z 191.05578, calc. for C7H11O6 − 191.05556, ferulic acid (8, ESI-FTICR-MS: [M - H]−, m/z 193.05028, calc. for C10H9O4 − 193.05008, chlorogenic acid (9, ESI-FTICR-MS: [M - H]−, m/z 353.08751, calc. for C16H17O9 − , 353.08726, three isomers of dicaffeoyl quinic acid (10, ESI-FTICR-MS: [M - H]−, m/z 515.11949, calc. for C25H23O12 − 515.11895, and three isomers of methyl derivatives of 10 (11, ESI-FTICR-MS: [M - H]−, m/z 529.13539 calc. for C26H25O12 − 529.46950.

Compounds 7–11 were detected before in other Helichrysum species. Mono- and dicaffeoyl quinic acids are the main constituents of the Mediterranean herb H. italicum [38, 39] and are also present in the French H. stoechas var. olonnense [40]. Both are used as digestive. A similar compound composition is known for the Artichoke; Cynara scolymus L., which is used for its choleretic, lipid-lowering, hepatostimulating, and appetite-stimulating actions [41]. Extracts and constituents of artichoke were also shown to possess antibacterial and antifungal activities, however, Extracts and constituents of H. mechowianum showed least efficiency antifungal properties against the yeast Cladosporium cucumerinum. The observed quinic acid derivatives might be responsible for the stomach protecting effects of H. mechowianum.

Chromatographic separation of the partitioned extracts of H. mechowianum resulted in the isolation of a phytosterol mixture from the n-heptane fraction, 3,5,7-trihydroxy-8-methoxyflavone (12, ESI-FTICR-MS: [M + Na]+ m/z 323.05297 calc. for C16H12O6Na 323.05261) from the ethyl acetate fraction and 4,5-dicaffeoyl quinic acid (13, ESI-FTICR-MS: [M - H]−, m/z 515.12168 calc. for C25H23O12 − 515.11950) from the water fraction. The relative composition of the phytosterol fraction was determined by GCMS as campesterol (2 %), stigmasterol (9.3 %), campest-7-en-3-ol (61.4 %), chondrillasterol (18.9 %), β-sitosterol (61.4 %) and an unidentified sterol (1.2 %). The compounds (12) [42] and (13) [43] were identified by comparison of spectral data with literature data. In addition, the position of the caffeoyl residues in compound (13) was determined by 2D NMR measurements. In particular, HMBC correlations from H-4 and H-5 of the quinic acid to C-9′, C-8′ and C-7′ of the caffeoyl residues indicates the substitution at position 4 and 5 (Table 3). Since this compound is reported to possess cytotoxic and apoptose inducing activity [44, 45], the anticancer activity against the prostate cancer cell line PC-3 was tested. However in concentrations of 50 nM or 50 μM no effect was observed with this cell line (Table 4).

Table 3 1H and 13C NMR assignments for compounds (1–5, 12)
Full size table
Table 4 LC/MS data, deprotonated and protonated molecules (m/z) for peaks, including the retention times (Rt), MS/MS experiments of the constituents found in MeOH extract of Helichrysum mechowianum and Helichrysum foetidum
Full size table

In silico pharmacokinetics assessment

Many bioactive compounds do not make it to clinical trials because of adverse pharmacokinetic properties. It therefore becomes imperative to access the pharmacokinetic profiles of potential drugs early enough in order to access their potential for further development. A summary of twenty two of the computed molecular descriptors used to assess the drug-likeness profiles of the isolated metabolites have been summarized in Table 2. These include the #stars or ‘drug-likeness’ parameter, the molecular weight (MW), the solvent accessible surface area (SASA), along with its hydrophobic component (FOSA) and hydrophilic component (FISA), the molecular volume, the number of hydrogen bond acceptors (HBA) and donors (HBD), the n-octanol/water partition coefficient (log P), the solubility parameter (log S), the predicted IC50 values for the blockage of the human-ether-a-go-go potassium ion (HERG K+) channels (logHERG), predicted permeability of Caco-2 cells, the blood–brain barrier partition coefficient (log BB), permeability of Madin-Darby canine kidney (MDCK) monolayers, skin permeability (log Kp), the number of predicted primary metabolites (#metab), the binding affinity to human serum albumin (log KHSA), the percentage human oral absorption (PHOA), the number of violations of Lipinski’s ‘Rule of Five’ (Ro5) and Jorgensen’s ‘Rule of Three’ (Ro3) and the polar surface area (PSA). The range of values of each parameter for 95 % of known drugs have been given beneath Table 2. Five of these compounds (1, 2, 3, 5 and 12) showed #star = 0, which indicates that all the computed parameters fell within the recommended range for 95 % of known drugs. Meanwhile, compounds 4 and 6 showed only #star = 1. An overall ADME-compliance score, drug-likeness parameter (indicated by #stars), was used to assess the pharmacokinetic profiles of the isolated compounds. The #stars parameter indicates the number of property descriptors computed by QikProp [46], which fall outside the optimum range of values for 95 % of known drugs. The methods implemented were developed by Jorgensen et al. [47–49] (Table 5).

Table 5 Computed molecular descriptors for the assessment of the DMPK profiles of the major isolated metabolites and the recommended range for 95 % of known drugs
Full size table

Materials and methods

General methods

Silica gel (Merck, 63–200 μm) and Sephadex LH-20 (Supelco) were used for column chromatography. Fractions were monitored by TLC using precoated silica gel plates 60 F254 (Merck). Spots were visualized by heating silica gel plates sprayed with vanillin-H2SO4 in MeOH. The 1H and 13C NMR spectra were recorded on a Varian Mercury 300 spectrometer at 300.22 and 75.50 MHz, respectively. 1H and 2D NMR spectra were recorded on a Varian VNMRS 600 system operating at a proton NMR frequency of 599.83 MHz equipped with a 5 mm inverse detection cryoprobe using standard CHEMPACK 4.1 pulse sequences (COSY, ROESY, 1DNOESY, HSQCAD, HMBCAD) implemented in Varian VNMRJ 2.2C spectrometer software. Chemical shifts were referenced to internal TMS (δ = 0 ppm, 1H) and CDCl3 (δ = 77.0 ppm, 13C) or CD3OD (δ = 49.0 ppm, 13C), respectively. The high resolution ESI mass spectra were obtained from a Bruker Apex III Fourier transform ion cyclotron resonance (FTICR) mass spectrometer (Bruker Daltonics, Billerica, USA) equipped with an Infinity™ cell, a 7.0 Tesla superconducting magnet (Bruker, Karlsruhe, Germany), an RF-only hexapole ion guide and an external APOLLO electrospray ion source (Agilent, off axis spray, voltages: endplate, -3.700 V; capillary, −4.200 V; capillary exit, 100 V; skimmer 1, 15.0 V; skimmer 2, 10.0 V). Nitrogen was used as drying gas at 150 °C. The sample solutions were introduced continuously via a syringe pump with a flow rate of 120 μl/h. All data were acquired with 512 k data points and zero filled to 2048 k by averaging 32 scans. The XMASS Software (Bruker, Version 6.1.2) was used for evaluating the data. The positive ion ESI mass spectra and the collision-induced dissociation (CID) mass spectra were obtained from a TSQ Quantum Ultra AM system equipped with a hot ESI source (HESI, electrospray voltage 3.0 kV, sheath gas: nitrogen; vaporizer temperature: 50 °C; capillary temperature: 250 °C; The MS system is coupled with a Surveyor Plus micro-HPLC (Thermo Electron), equipped with a RP18 column (5 μm, 150 × 1 mm, Hypersil GOLD, Thermo Scientific). For the HPLC a gradient system was used starting from H2O:CH3CN 90:10 (each of them containing 0.2 % HOAc) to 5:95 within 15 min and then hold on 5 % for further 30 min; flow rate 70 μl/min. Sterols were determined by GC-MS (Voyager/Trace GC 2000, Thermo Quest CE Instruments): 70 eV EI, source temp. 200 °C; column ZB-5 (Phenomenex, 30 m × 0.25 mm, 0.25 μm film thickness); inj. temp. 250 °C, interface temp. 300 °C; carrier gas He, flow rate 1.0 ml/min, constant pressure mode; splitless injection, column temp. program: 60 °C for 1 min, then raised to 300 °C at a rate of 10 °C/min to 290 °C for 15 min.

Plant material

The plant materials were collected and identified by Elias Ndive, a botanist from Limbé Botanic Garden, on March 2009 near the town of Buea on the eastern slopes of Mount Cameroon in the South West Province of Cameroon. Voucher specimens (H. foetidum (L.) Moench: SCE2463, H. mechowianum Klatt: SCE2467) are deposited in the Herbarium of Limbé Botanic Garden.

Extraction and isolation

Leaves and flowers of H. foetidum and leaves of H. mechowianum were extracted exhaustively with 90 % methanol for a period of 72 h. The solvent was removed by evaporation under reduced pressure. From the crude flower extract of H. foetidum, by purification with successive column and preparative TLC chromatography on silica gel using a chloroform/methanol gradient systems,the compounds 7,4′-dihydroxy-5-methoxy-flavanone (1) and kaur-16-en-18-oic acid (6) were obtained while the compounds 6′-methoxy-2′,4,4′-trihydroxychalcone (helichrysetin) (2), 6′-methoxy-2′,4-dihydroxychalcone-4′-O-β-D-glucoside (3), apigenin (4), apigenin-7-O-β-D-glucoside (5) were isolated from the leaves and flowers of H. Foetidum.

The aqueous residue of the crude extract of H. mechowianum leaves was partitioned successively with n-heptane and ethyl acetate. The n-heptane and the ethyl acetate extracts were further purified by silica gel column chromatography using n-hexane/ethyl acetate gradient systems resulting in the isolation of a phytosterol fraction and of 3,5,7-trihydroxy-8-methoxyflavone (12), respectively. The water fraction was further separated using Diaion HP20 eluted with water, methanol, ethyl acetate and acetone followed by chromatography of the methanol fraction on Sephadex LH20 to give 4,5-dicaffeoyl quinic acid (13). The crude extracts of H. foetidum and H. mechowianum were analyzed by LC-ESI-MS, MS/MS and FTICR-HRMS.

7,4′-dihydroxy-5-methoxy-flavanone (1): 1H NMR (DMSO-d6): δ 9.60 (1H, brs, OH), 7.28 (2H, d, 8.6, H2′/6′), 6.77 (2H, d, 8.6, H3′/5′), 6.04 (1H, d, 2.2, H6), 5.94 (1H, d, 2.2, H8), 5.32 (1H, dd, 12.6/2.9, H2), 3.72 (3H, s, OMe), 2.98 (1H, dd, 16.3/12.6, H3b), ca. 2.5 (1H, m, superimposed by DMSO, H3a). 13C NMR (DMSO-d6) δ 79.0(C2), 42.4(C3), 196.3(C4), 163.0(C5),104.6(C6),167.1(C7),95.4(C8),162.8(C9),101.6(C10), 129.6(C1′), 127.6(C2′/C6′), 115.0(C3′/C5′), 157.6(C4′), 55.4 (OMe).

6′-methoxy-2′,4,4′-trihydroxychalcone (2): 1H NMR (MeOD + CDCl3): δ 7.73 (1H, d, 16.0, Hα), 7.68 (1H, d, 16.0, Hβ), 7.49 (2H, d, 8.6, H2/6), 6.84 (2H, d, 8.6, H3/5), 5.99 (1H, d, 2.2, H3′),5.96 (1H, d, 2.2, H5′), 3.93 (OMe) 13C NMR (MeOD + CDCl3): δ 193.6 (C = O), 168.2 (C4′), 165.9 (C2′), 164.2 (C6′), 160.5 (C4), 143.5 (CH-β), 131.0 (C2/6), 128.0 (C1), 125.2 (CH-α), 116.6 (C3/5), 106.3 (C1′), 96.9 (C5′), 92.2 (C3′), 56.2 (OMe).

6′-methoxy-2′,4-dihydroxychalcone-4′-O-β-D-glucoside (3): 1H NMR (MeOD): δ 7.72 (1H, s, Hα), 7.71 (1H, s, Hβ), 7.51 (2H, d, 8.7, H2/6), 6.83 (2H, d, 8.7, H3/5), 6.31 (1H, d, 2.2, H3′), 6.24 (1H, d, 2.2, H5′), 3.95 (3H, s, OMe), glucose moiety: δ 4.99 (1H, d, 7.5, H1″), 4.24(H2″), 3.61(H3″), 3.60(H4″), 3.53(H5″), 3.80–3.90 (H6″), 13C NMR (MeOD) δ 193.6 (C = O), 168.2 (C4′), 165.9 (C2′), 164.2 (C6′), 160.5 (C4), 143.5 (CH-β), 131.0 (C2/6), 128.0 (C1), 125.2 (CH-α), 116.6 (C 3/5), 106.3 (C1′), 96.9 (C5′), 92.2 (C3′), 56.2 (OMe) glucose moiety: δ 73.7(C1″), 71.0(C2″), 77.2(C3″), 78.9(C4″), 79.8(C5″), 60.6(C6″),

apigenin (4): 1H NMR (DMSO-d6): δ 12.96 (1H, s, OH), 7.93 (2H, d, 8.8, H2′/6′), 6.93 (2H, d, 8.8, H3′/5′), 6.78 (1H, s, H3), 6.48 (1H, d, 2.0, H8), 6.19 (1H, d, 2.0, H6). 13C NMR (DMSO-d6): δ 180.4 (C4), 164.94 (C5), 164.4 (C2), 162.6 (C4′), 160.7 (C9), 160.1 (C7), 129.3 (C2′/ C6′), 123.1 (C1′), 117.0 (C3′/ C5′), 109.3 (C10), 106.5 (C3), 104.8 (C6), 99.3 (C8).

apigenin-7-O-β-D-glucoside (5): 1HNMR(CD3OD, 500 MHz): aglycon moiety:δ 7.83 (2H,d, 8.8,H-2′/ H-6′), 6.92 (2H,d, 8.8, H-3′/ H-5′), 6.83 (1H,d, 2.1 Hz, H-6), 6.71 (1H,d, 2.1 Hz, H-8), 6.58 (1H,s, H-3), glucose moiety: δ 4.90 (1H,d, 7.6, H-1″’), 3.50-3.25 (4H,m,H-2″, 3″, 4″, 5″), 3.87 (1H,dd, 11.9/ 2.2, H-6b″), 3.73 (1H,dd,11.9/ 5.4, H-6a″ 13C NMR (DMSO-d6): δ 182.0 (C4), 164.3 (C5), 163.0 (C2), 161.5 (C4′), 161.1 (C9), 157.0 (C7), 128.7 (C2′/C-6′), 121.0 (C1′), 116.0 (C3′/C5′), 105.4 (C10), 103.1 (C3), 99.9 (C6), 99.6 (C8), glucose moiety: δ 105.1 (C1″), 78.6 (C5″), 77.5 (C3″), 74.8 (C2″), 71.8(C4″), 62.6 (C6″)

kaur-16-en-18-oic acid (6): 1H NMR (CDCl3): δ 4.79 (1H, s, H17), 4.73 (1H, s, H17), 2.63 (1H,brs), 2.15 (1H, brd,13.9), 2.05 (2H, d, 2.0), 1.24 (3H, s, Me), 0.95 (3H, s, Me) 13C NMR (CDCl3): δ 184.1 (C = O), 155.9 (C), 103.0 (CH2), 57.0 (CH), 55.1 (CH) 48.9 (CH2), 44.2 (C), 43.8 (CH), 43.7 (C), 41.3 (CH2), 40.7 (CH2), 39.7 (CH2), 39.6 (C), 37.8 (CH2), 33.1 (CH2), 28.9 (CH3), 21.8 (CH2), 19.1 (CH2), 18.4 (CH2), 15.6 (CH3).

3,5,7-trihydroxy-8-methoxyflavone (12): 1H NMR (CDCl3): δ 11.49 (1H, s, OH), 8.24 (2H, d, 7.1), 7.60-7.49 (3H, m), 6.70 (1H, s, OH), 6.46 (1H, s, OH), 6.33 (1H, s, H6), 4.05 (3H, s, OMe). 13C NMR (CDCl3): δ 175.6 (C = O), 156.6 (C), 155.4 (C), 148.0 (C), 144.9 (C), 136.5 (C), 130.7 (C), 130.4 (CH), 128.8 (2 CH), 127.5 (2 CH), 126.9 (C), 98.2 (CH), 61.9 (OMe).

4,5-dicaffeoyl quinic acid (13): 1H NMR (MeOD): δ 7.585/7.507 (1H, d, 16.0, H7′/7″), 7.014/6.994 (1H, d, 2.0, H2′/2″), 6.905/6.890 (1H, dd, 8.2/2.0, H6′/6″), 6.739/6.733 (1H, d, 8.2, 5′/5″), 6.273/6.190 (1H, d, 16.0, H8′/8″), 5.661 (1H, m, H5) 5.108 (1H, dd, 9.8/3, H4), 4.337 (1H, d, 3, H3), 2.284 (1H, brdd, 14.1 2.1, H2a), ca. 2.22 (2H, H6), 2.046 (1H, brd, 12.5, H2a).13C NMR (MeOD): δ 178.9 (C7), 168.6/168.4 (C9′/9″), 149.6 (C4′/4″), 147.6/147.4 (C7′/7″), 146.75/146.73 (C3′/3″), 127.7/127.6 (C1′/1″), 123.1 (C6′/6″), 116.4 (C5′/5″), 115.1 (C2′/2″), 114.8 (C8′/8″), 76.9 (C1), 76.6 (C4), 70.2 (C3), 69.3 (C5), 40.2 (C6), 38.7 (C2).

Bioassays

Protease inhibition assay

Procedure

Initially the extracts or purified constituents were dissolved in DMSO and the dilutions of samples tested were made in the respective buffer for each enzyme, i.e., 0.1 M sodium phosphate (pH 7.5) for subtilisin and 0.1 M sodium acetate (pH 4.4) for pepsin. Samples (0.01 - 50 μg/ml) were pre-incubated with subtilisin (37 nM) or pepsin (1.7 nM) for 30 min and then, transferred to a black opaque microplate. The substrate EDANS-DABCYL (2 μM), prepared in the specific buffer for each protease, were automatically injected. The final volume was 100 μl. Experiments were performed separately for each protease, which was prepared at the day of experiment. Reads were made for a period of 5 min, with 1 min intervals, and temperature controlled at 37 °C. The mean, standard deviation and relative standard deviation (RSD) of triplicates and the percentage of inhibition were calculated using the final fluorescence intensity measured.

Reagents

Pepsin from porcine gastric mucosa, Recombinant Type VIII Subtilisin Carlsberg, Arg-Glu-(EDANS)-Ser-Gln-Asn-Tyr-Pro-Ile-Val-Gln-Lys-(DALBCYL)-Arg fluorogenic substrate (EDANS-DABCYL), DMSO spectrophotometric grade were from Sigma-Aldrich, Sao Paulo, Brazil. Fluorescence bioassay data were collected with a multi detection microplate reader SynergyTM HT (Bio-Tek® Instruments Inc., Winooski, Vermont, USA), with 360 nm excitation and 460 nm emission filters, and analyzed using KC4 software (Bio-Tek®Instruments) and a Microsoft Windows XP.

Antibacterial assay

The antibacterial activity against a fluorescent Bacillus subtilis [50] was determined with a fluorescence based antibacterial growth inhibition assay. The fluorescence was measured on a microtiter plate reader GENios Pro (Fa. Tecan, excitation 510 nm; emission 535 nm). The Bacillus subtilis strain 168 (PAbrB-IYFP) was maintained on TY (tryptone-yeast extract) medium supplemented with 1 % Bacto-tryptone, 0.5 % Bacto-yeast extract, 1 % NaCl and Chloramphenicol (5 μg/ml). Erythromycin was used as positive control for growth inhibition.

Antifungal assay

The antifungal activity against the phytopathogenic fungus Cladosporium cucumerinum was tested by bioautography on silica gel plates [51] in concentrations of 50, 100, 200 and 400 μg/cm. Amphotericine B was used as positive control for growth inhibition.

Cytotoxicity assay

The cytotoxicity was determined by XTT method, using the Cell Proliferation Kit II (Roche). The human prostate cancer cell line PC-3 was maintained in RPMI 1640 medium supplemented with 10 % fetal bovine serum, 1 % L-alanyl-L-glutamin (200 mM), 1 % penicillin/streptomycin and 1,6 % hepes (1 M). For the measurement of cytotoxicity the same medium was used without antibiotics. For PC-3 500 cells/well were seeded overnight into 96-well plates and exposed to serial dilution of each compound for three days.

Molecular modeling

All molecular modelling was carried out on a Linux workstation running on a 3.5 GHz Intel Core2 Duo processor (Santa Clara, USA). Low energy 3D structures of the thirteen isolated compounds were generated using the MOE software package [52] and the Merck molecular forecefiled [53] and saved in mol2 format. These were initially treated with LigPrep [54], distributed by Schrodinger, Inc (Camberley, UK). This implementation was carried out with the graphical user interface (GUI) of the Maestro software package (New York, USA) [55], using the OPLS force field [56–58]. Protonation states at biologically relevant pH were correctly assigned (group I metals in simple salts were disconnected, strong acids were deprotonated and strong bases protonated, while topological duplicates and explicit hydrogens were added). A set of the ADMET-related properties (a total of 46 molecular descriptors) were calculated using the QikProp program (New York, USA) [46] running in normal mode. QikProp generates physically relevant descriptors and uses them to perform ADMET predictions. An overall ADME-compliance score, drug-likeness parameter (indicated by #stars), was used to assess the pharmacokinetic profiles of the compounds. The #stars parameter indicates the number of property descriptors computed by QikProp, which fall outside the optimum range of values for 95 % of known drugs. The methods implemented were developed by Jorgensen et al. [47–49].

Conclusion

Eight known compounds have been identified from the extracts of two species from the genus Helichrysum (Compositae) harvested from the South West of Cameroon (Central Africa). The results showed that the flavonoid glycosides (3, 5) from H. foetidum exhibited protease inhibition, while the compound (13) from H. mechowianum contribute to the stomach protecting effects. In addition, the antibacterial and antifungal activities of compound (6) was demonstrated by the fact that it was found to possess a potent inhibitor effect against the tested microorganisms. The differential bioactivities and determined constituents support the traditional use of the species. Molecular modelling studies showed that five of the isolated compounds showed physicochemical properties that completely within the recommended range for more that 95 % of known drugs, while two compounds have only one violation.