Isolation , characterization and biological activity of organic extractives from Calodendrum capense ( L . f . ) Thunb . ( Rutaceae )

1 Chemistry Department, School of Pure and Applied Chemistry, Kenyatta University, P. O. Box 43844, Nairobi, Kenya. 2 Department of Chemistry and Biochemistry, University of Eldoret, P. O. Box 1125 30100, Eldoret, Kenya. 3 Department of Pharmacy and Complementary/Alternative medicine, School of Medicine, Kenyatta University, P. O. Box 43844, Nairobi, Kenya. 4 Department of Chemistry, Faculty of Natural Science, University of the Western Cape, Modderdam Road, Private bag X1, Belleville, South Africa. 5 Department of Applied Sciences, Maseno University, Private Bag Maseno, Kenya.


INTRODUCTION
Calodendrum capense (Linnaeus Filius) Thunberg belongs to the family Rutaceae.It is a semi-deciduous tree with tough timber used in house building, for tool handles and poles and as fuelwood (Dharani, 2011).The bark is used as an ingredient of skin ointments and is sold at traditional medicine markets in South Africa.Previous analytical studies carried out on seed kernels reported, antioxidant element copper, magnesium, manganese and zinc composition of the oil (Nawiri et al., 2012), performance of domestic cooking wick stove using fatty acid methyl ester of oil (Wagutu et al., 2010).Phytochemical studies revealed the composition of fatty acids in the isolated oil (Munavu, 1983) and isolation from the seeds of limonin, limonin diosphenol and rutaevin *Corresponding author.E-mail: rokwesh@yahoo.com.Tel: +274 0722 293 770.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License (Dreyer, 1967).A degraded limonoid, calodendrolide has been isolated from the root bark (Cassady and Lui, 1972).The C. capense metabolites, limonin and limonin diosphenol have been reported to exhibit biological activity in larvicidal activity in Aedes aegypti (Kiprop et al., 2005) and the wood extract showed mimicry of juvenile hormone activity in Oncopeltus fasiatus (Jacobson et al., 1975).In the chemotaxonomy of rutaceous plants, biologically active metabolites reported include; alkaloids with quinoline skeleton, coumarins, limonoids, essential oils and coumarins.Alkaloids and coumarins have not been reported and no phytochemical work has been done on the leaves and fruit pericarp of C. capense.As part of phytochemical investigation in screening for biological activity of crude extracts in medicinal plants and presence of these secondary metabolites, the isolation and characterization of two furoquinoline alkaloids, two furocoumarins and the common rutaceae limonoids, limonin and limonin diosphenol from organic extracts obtained from leaves, stem bark, fruit pericarp and seed kernels of C. capense were reported.

Instrumentation
Melting points of the isolated compounds were determined using Sanyo GallenKamp (UK) electronic melting point apparatus.Infrared spectra were obtained using NaCl pellets with acetone as solvent from Shimadzu Fourier Transform (FTIR-8400) Spectrophotometer.Nuclear Magnetic resonance NMR spectra were measured using Bruker Avance 400 Proton 1 H NMR(400MHz); Carbon 13 C NMR (100 MHz).Solvents used were deuterated chloroform CDCl3, methanol CD3OD and acetone (CD3)2CO.Chemical shifts were given in ppm values with trimethylsilane (TMS) used as the internal standard.Mass spectra were measured on electron impact mass spectra (EI-MS) using a Finnigan Gas chromatography Mass spectrometer GC-MS analyses.The samples were analysed on an Agilent GC-MS apparatus equipped with DB-5SIL MS (30 m × 0.25 mm i.d., 0.25 μm film thickness) fused-silica capillary column; helium (at 2 ml/min) was used as a carrier gas.For vacuum liquid chromatography (VLC), silica gel 60 (0.063-0.2 mm, Merck Chemicals Ltd, South Africa) was used.For column chromatography, silica gel 60 (70-230 mm, Merck Chemicals Ltd, South Africa) was used.For analytical thin layer chromatography TLC, aluminium plates coated with fluorescence indicator F254 (Alugram Sil G/UV 254, Macherey-Nagel, Germany) were used.Sephadex LH-20 was used as filter gel.

Plant material
The plant samples of C. capense were collected from City Park in Nairobi.Plant authentication was done by a plant taxonomist Mathias Mbale and a voucher RO/001 deposited in plant herbarium, National Museum of Kenya.
A mass of 3.1 kg of dried ground seed kernels used was defatted using hexane and 72 g of a white solid DCM extract was obtained.The white solid was separated by column chromatography using DCM: EtOAc (9:1) and 114 fractions were collected which showed a positive purple test with anisaldehyde reagent.Fractions 109-114 were combined and separated by preparative-TLC DCM:EtOAc; 9:1 to afford compound (5) 41 mg.The fraction103-108 was subjected to Preparative-TLC Hexane: EtOAc 3:2 gave compound (6) 22.8 mg which was UV active observed at 275 nm.

Antimicrobial in vitro assay
Antifungal test on crude extracts from C. capense was done against Candida albicans American Type Culture Collection ATCC 90028, Trichophyton mentagrophytes, Penicillium citrinium and Aspergillus niger using agar diffusion method (Elgayyar et al., 2001).Sterile Petri dishes were filled with sterilized medium of yeast, malt extracts and Potato Dextrose Agar (PDA) to a depth of 4 mm and seeded with the spores of fungi.Two hundred mg extract was dissolved in 1 ml of methanol and 10 µl of solution was loaded onto a 6 mm sterile filter paper, dried (2 mg/disc) and placed on the Petri dishes containing the medium which was sterilized using an autoclave set at 121 o C for 15 min, allowed to cool and seeded with fungi.Fluconazole was used as standard, negative control plates had discs with sterile distilled water and methanol.The Petri dishes containing Aspergillus niger were incubated at room temperature (about 25°C) and the zones of inhibition were read after 72 h (Chabbra and Uiso, 1992).
Antibacterial test on crude extracts from C. capense was done on Escherichia coli American Type Culture Collection ATCC 25923, Bacillus subtilis, and Staphylococcus aureus ATCC 25922, by Plate diffusion method (Elgayyar et al., 2001).Chloramphenicol was used as standard antibiotic.For each drug, 100 mg was dissolved in 1 ml of dimethylsulphoxide (DMSO).Twenty eight grams per liter of nutrient Agar in distilled water was sterilized by autoclaving at 121°C for 15 min.Fifteen milliliter agar was poured into the Petri dishes in a lamina flow machine under sterile conditions and 0.1 ml of the bacterial solution was added to it.Filter papers containing the drug test were put on the Petri dish and incubated at 37°C for 24 h.After which diameter of the zone of inhibition from the initial 6 mm was measured (Chhabra and Uiso, 1991).
The active extracts from the antimicrobial screening were tested for minimum inhibitory concentrations (MIC).The MICs were determined using a two-fold serial dilution method in a peptone water solution for bacteria and potato dextrose agar (PDA) broth for yeast and fungi for active extracts to give a final extract concentration of between 1.95 and 8000 µg/ml.The negative control of the disc diffusion testing was done by the use of methanol that showed no inhibition, while the negative control was done by the use of standard antibiotic discs.The average zone of inhibition was calculated for three replicates.A clearing zone of 9 mm for bacteria and 10 mm or more for fungi was used as a criterion for designating significant antibacterial and antifungal activity (Faizi et al., 2003).

In vitro antiproliferative assay
This was carried out following a modified rapid calorimetric assay (Mosmann, 1983) using two cell lines; Vero-199 cells and Vero-E6 cells sourced from American Type Culture Collection (ATCC), by Kenya Medical Research Institute KEMRI.These cells were maintained in Eagle's Minimum Essential Medium (MEM) supplemented with 10% foetal bovine serum (FBS) and 2 mM Lglutamine.On 96 well micro titer plates 5 × 10 3 cells/well suspension were seeded and incubated at 37°C/5% CO2 (Kamuhabwa et al., 2000).Samples were added to the cultured cells over a concentration range of 1000 to 0.14 µg/ml dimethylsulphoxide (DMSO).The plates were incubated for 48 h at 37°C and 5% CO2, after which 10 µL of MTT (Thiazoil Blue Tetrazolium Bromide) dye was added and incubated for another 4 h.Media was removed from all wells and 100 µl of DMSO was added.The plates were then read (colour absorbance) on an enzyme-linked immunosorbent assay ELISA scanning multiwell spectrophotometer (Multiskan Ex labs systems) at 562 and 690 nm as reference.Data was transferred onto a graphic programme (EXCEL) and expressed as percentage of the untreated controls.The 50% inhibition concentration (IC50) value was evaluated by linear regression analysis.Podophylotoxin (PPT) drug with an Initial concentration of 100 µg/ml DMSO was used as the control standard for the experiment.

Antimicrobial assay in vitro
The organic crude extracts of leaves, stem bark and fruit pericarp exhibited weak antimicrobial activity against S.  Mandeel and Taha, 2005).
Ethylacetate crude extracts of C. capense leaves showed moderate MIC of 1250 µg/ml and fruit pericarp hexane extract showed weak MIC of 2500 µg/ml; the extracts show potential antifungal use.Grewia asitica leaves used on pustular eruptions showed MIC of 1500 µg/ml when methanolic extracts of the leaves were used (Sangita et al., 2009).
The hexane stem bark and leaves extract and ethylacetate stem bark extract showed inhibition zone triplicates in Appendix 2 moderate activity with mean inhibition zone of 14 mm against S. aureus with chloroamphenicol as positive control drug showing 22 mm (Table 2).This shows potential of these extracts in antibacterial application.The results are in close agreement with those carried out on Acalypha wilkesiana leaves used in the treatment of gastrointestinal disorders; ethanol extract showed of 11.5 mm with positive control drug ciprofloxacin showing an inhibition zone 27.3 mm (Akinyemi et al., 2006).The hexane leaves crude extract and ethylacetate stem bark showed moderate activity on S. aureus with MIC of 1250 µg/ml (Table 2).Acalypha fruticosa used in treatment of skin infection and diarrhea showed MIC of 512 µg/ml against S. aureus for methanolic leaves extract (Sama Fonkeng et al., 2015).
Sequential leaves and fruit pericarp extracts of C. capense in Table 3 and 4 showed moderate activity against B. subtilis with mean inhibition zones between 9.0 and 13.0 mm with standard drug showing 24 mm, the triplicate inhibition zones are shown in Appendix 3 and 4 In comparison to a similar study carried on Cinnamomum tamala methanol and ethylacetate extracts showed close agreement between 11.7 and 12.5 mm in inhibition zone against B. subtilis and 34.2 mm using tetracycline as positive control drug (Goyal et al., 2009).Hexane leaves and ethylacetate fruit pericarp extract showed moderate   (Goyal et al., 2009).

Antiproliferative assay in vitro
Cytotoxicity screening of plant extracts is a preliminary aspect of safety evaluation for crude extract and isolated compounds ensuring that bioactivity is not due to general toxic effect of the extracts or compounds.The 96 well plate replicate results for podophylotoxin and fruit pericarp in Appendix 5 and 6 were analyzed and IC 50 values determined from which the mean IC 50 values were determined and the variation of cell viability against represented graphically in Figures 1 and 3 respectively.The regression curves were represented in Figures 2 and  4. The hexane fruit pericarp extract and podophylotoxin both showed weak cytotoxicity IC 50 of 81.49 ± 0.689 µg/ml and 65.13 ± 0.0.599µg/ml respectively (Tables 5  and 6).Significant cytotoxicity is considered when the IC 50 value is ≤20 µg/ml; however their MIC values against B. subtilis differ significantly 2500 g/ml for hexane pericarp and 50 µg/ml for podophylotoxin.The selectivity index (SI) is calculated as a ratio of IC 50 value of Vero cells to MICs (Vicente et al., 2009) and is a measure of tolerability of cells in vitro to extracts or compounds.The selectivity indices greater than 1.0 indicate safety of the  drug on the host as compared to the pathogen, for SI less than 1.0, high amount of extractible will be required to be applied in eradication of the pathogen.When selectivity index is ≥10 the compound is considered suitable for further investigations (Oliveira et al., 2014).The SI values presented in Table 7 is slightly low for the hexane fruit pericarp extract 0.033 µg/ml compared to the standard drug podophylotoxin, a medical cream applied topically to treat genital warts which is 1.243 µg/ml against B. subtilis, this means that lower amount of standard drug can be used in eradication of the pathogen since larger amounts could be toxic.In a previous investigation using plant extracts, the selectivity indices ranged between 0.02 and 0.68 µg/ml against S. aureus with one plant extract showing a value of 2.87 µg/ml, while doxorubicin used as positive control showed cytotoxicity of 8.3 ± 1.76 µg/ml (Elisha et al., 2017).The antimicrobial activities of the extracts on P. citrinium, S. aureus and B. subtilis, was therefore not due to toxic effect of the extracts.
Compound (2) was isolated as a white crystalline solid.The 1 H NMR spectrum Appendix 12a and b showed six protons characteristic of a linear furoquinoline alkaloid; two proton doublets with signals δ 7.89 (1H, J = 2.4, H-2) and δ 7.35 (1H, J = 2.4, H-3) which showed correlation according to COSY and NOESY spectrum.A prominent proton signal at δ 4.26 (3H, -OCH 3 ) integrating to three proton singlets characteristic of a methoxy group was observed.The aromatic protons were observed at δ 8.24 (1H, d, J = 9.7, H-5) diagnostic of the type of substitution in the benzenoid ring, the ortho-proton δ 6.28 (1H, d, J = 9.9, H-6) showing correlation as observed in the COSY spectrum and NOESY spectrum.The proton at paraposition resonated at δ 7.34 (1H, d, J = 0.84, H-8).The 13 C NMR spectrum in Appendix 13 and 14 exhibited 12 signals resolved by DEPT spectrum as five methine, one methyl and six quarternary carbons.The chemical shifts were characteristic of a furoquinoline alkaloid.The positions of the 12 carbon atoms were confirmed from HSQC spectrum and HMBC spectrum.There was HMBC correlation of H-2 to C-3a (δ 113.5) and C-9a (δ 159.3).There was HMBC spectra revealed correlation of H-2 to C-3a (δ 113.5) and C-9a (δ 159.3).The proton H-3 shows HMBC correlation with C-2 (δ 146.2),C-3a and C-9a hence confirming C-3a and C-9a to be the bridge carbon atoms between the furan ring and the heterocyclic ring.There was HMBC correlation between the methoxy hydrogen atoms and with a highly desheilded oxygenated C-4 (δ 150.8).The aromatic proton H-5 showed HMBC correlation with C-4, C-8 (δ 153.8), and the highly deshielded oxygenated centre C-7 (δ 160.8).There was observed correlation between H-6 to C-4a and C-7 from the HMBC spectrum.The proton H-8 exhibited correlation with carbon atoms C-4a, C6, C-7 and C8a from the HMBC spectrum.+ .The proposed fragmentation pattern of compound 2 was rationalized by mass spectrometric studies carried out on substituted furoquinoline alkaloids, where those with methoxy substituent at position C-4 and C-8 lose the methyl group leaving behind a carbonyl at these positions, followed by the subsequent loss of carbon monoxide molecules (Glugston and Maclean, 1965;O'Donnell et al., 2006).
1 H-NMR and MS values were in close agreement with literature values for confusameline (Kang and Woo, 2010).
The connectivity of proton to carbon atoms were derived unambiguously from the heteronuclear single quantum correlation (HSQC) spectrum.The proton H-5 exhibited cross peaks with methyl C-25a (δ C 80.6) and quaternary C-10 (δ 46.8) from the heteronuclear multiple bond correlation (HMBC) spectrum.The proton H-19b showed HMBC correlation to C-10 (δ 46.8) indicating the β-orientation of the proton.H-6a showed long range HMBC correlation with C-5 (δ 60.2), while H-6b showed correlation with C-10 (δ 46.8).There were cross peaks observed between H-9 and H-11 in COSY spectra.In ring C, a cross peak correlation occurs between H-11 and H-12 in COSY spectra.There was long range connectivity from HMBC between H-18 and C-12.In ring D, the 14, 15 epoxide moiety was determined by long range HMBC correlation between H-17 (δ H 5.43) and C-14 (δ C 67.31), H-15 and C-14.

and the furanyl [C4H3OCO+H]
+ 96 could be rationalized.The spectral data, both 1 H and 13 C NMR of compound (5) and were in close agreement with literature data of limonin (Breksa et al., 2006;Teranishi et al., 1999;Hasegawa et al., 1986).The compound 5 structure shown in Figure 9 was proposed to be limonin and this is the first report of isolation from stem bark of C. capense.

Conclusions
The reported bioactivities of the crude extracts and compounds isolated shows the profound reported use of rutaceous plants in traditional and alternative medicine; the reported use of the stem bark in traditional medicine is justified by presence of alkaloids, coumarins and copius amounts of limonoids.This is the first reported isolation of alkaloids; capensinin (1), confusameline (2), coumarins; psolaren (3), bergapten (4) from leaves, fruit pericarp and stem bark and limonin (5) from stem bark of C. capense.

RECOMMENDATION
Further studies should be carried out on the fruit pericarp of C. capense.

Figure 1 .
Figure 1.Cytotoxicity profile of standard podophylotoxin (PPT) used as control for the Vero cell experiments.IC50 was 65.13±0.48.

Figure 2 .
Figure 2. Linear regression analysis for cell viability with increase in concentration of podophylotoxin.

Figure 4 .
Figure 4. Linear regression analysis of cell viability with increase in concentration of hexane fruit pericarp extract.

1H-
NMR spectrum exhibited a total of eight proton signals, five of which were characteristic

Table 2 .
Antibacterial activities of crude extracts against Staphylococcus aureus.

Table 3 .
Antibacterial activities of crude extracts against B. subtilis.

Table 4 .
Antibacterial activities of crude extracts against B. subtilis.
Standard chloroamphenicol showed inhibition zone of 24 mm against B. subtilis.

Table 5 .
The IC50 for each replicate and mean for standard podophylotoxin (PPT).

mL) Percentage Cell Viability of Test Replicates 1 st 2 nd 3 rd 4 th
activity with MIC of 1250 µg/ml against B. subtilis while the other sequential extracts of leaves and fruit pericarp had weak activity with MIC of 2500 µg/ml.In comparison, C. tamala methanol extracts showed MIC value of 4096 µg/ml

Table 5 .
The IC50 for each replicate and mean for standard podophylotoxin (PPT).

Table 6 .
The IC50 for each replicate and mean for pericarp Hexane extract.

Table 7 .
The selectivity index (SI) from IC50 of Vero E-199 cell and MIC for extracts of C. capense.

Plant extract / compound Solvent Test specie MIC (µg/ml) Cytotoxicity [IC50(μg/ml)] SI
,from the side chain gave rise to the fragment ion m/z 200[C 11 H 6 NO 3 ] + in agreement with reported fragmentation of linear and angular furoquinoline alkaloids, where a large substituent group attached to the furoquinoline nucleus such as prenyloxy side-chain [C 5 H 8 ] Donnell et al., 2006).The peak at m/z 172 [C 10 H 6 NO 2 ] + + is lost leaving a carbonyl group at C-7 position as reported in 7-O-dimethyallyl-γ-fagarine (O'