CHEMICAL CONSTITUENTS OF OCIMUM KILIMANDSCHARICUM GUERKE ACCLIMATIZED IN KAKAMEGA FOREST, KENYA

The genus Ocimum belongs to the Lamiaceae family and is made of almost 200 species of herbs and shrubs which have potential medicinal properties. The species are native to the tropical and temperate climate zones around the globe. Two new compounds with damarane skeleton namely 2α-hydroxy-3-oxodammara-20,24diene (1) and 2α,3β-dihydroxy dammara-20, 24-diene (2) together with apeginin7-O-neohespeiridoside (3), quercetin (4), turkesterone (5), fesitin (6), apeginin (7), chrysin (8), lupeol (9), stigmasterol (10), friedelin (11), αamyrin acetate (12) and n-octacosonoic acid (13) are reported here from the leaves of Ocimum kilimandscharicum. Their structures were established on the basis of physical and spectroscopic analyses and by comparison with the literature data. Crude extracts and isolated compounds were investigated for contact toxicity and anti-feedant activity against Sitophilus zeamais and Prostephanus truncatus.


General experimental procedure
Melting points were determined using Gallenkamp melting point apparatus (Manchester, UK). Optical rotation was measured on a Jasco P-1020 Polarimeter (Jasco Corporation, Tokyo, Japan). UV spectra were analysed using a Shimadzu UV-2401A spectrophotometer (Shimadzu Corporation, Kyoto, Japan). IR data were recorded on a Bruker Tensor 27 FTIR spectrophotometer (Bruker Corporation, Bremen, Germany) as KBr pellets. NMR data were measured in CDCl 3 and DMSO-d 6 on a JOEL NMR instrument operating 600 and 150 MHz, respectively. Tetramethylsilane (TMS) was used as zero reference. The mass spectral data were obtained using a Varian MAT 8200A instrument. Electron-impact (EI) mass spectra (70 eV) were measured on a Hewlett-Packard 5989B mass Spectrometer. A gas chromatography-mass spectrometry (GC-MS) Thermo Finnigan system fitted with a capillary column SPB-5MS (30 m ×0.32 mm i.d., film thickness 0.25 mm) and splitless injection was used; the oven temperature was programmed from 40 to 250 o C at 10 o C/min; injector temperature and ion source temperature were at 250 and 200 o C, respectively; MS full scan was from 50-650 ms; Helium was used as carrier gas. Silica gel 60 (63-200 μm, Merck, Darmstadt, Germany) was used for gravity column chromatography (CC). TLC was performed on precoated DC Alufoil 60 F254 sheets (Merck, Darmstadt, Germany) and detected by spraying with anisaldehyde spray reagent, UV light and iodine vapour. Paper chromatography was done on standard Whatman No 1 chromatography paper. All solvents used were of analytical grade.

Plant material preparation and solvent extraction
The leaves were spread under shade for one week to dry thereafter pulverized using model 4E grinding mill. The powdered leaves (4.0 kg) were extracted with MeOH at room temperature (6 x 6 L for 7 days). The extracts were combined, filtered and evaporated under reduced pressure to give a dark green MeOH extract. The extract was separately partitioned between H 2 O and nhexane and EtOAc to give the soluble fractions, n-hexane (22.0 g green material) and EtOAc (41.5 g green material). The aqueous fraction was freeze dried to give 300.0 g brownish-green extract.

Contact toxicity of plant leaf extracts
The experiment tested the hypothesis that topically applied plant extracts and pure isolates solutions exhibit contact toxicity to the adult maize weevil Sitophilus zeamais (Curculionidae) and the adult larger grain borer Prostephanus truncatus (Bostrichidae). The contact toxicity assay on leaf extracts and pure isolates of O. kilimandscharicum were carried out according to a method previously described [25] by topical application using 3rd instar larvae. The extract stock suspensions and pure isolate solutions were prepared immediately prior to the assays by dissolving them in acetone to obtain solutions of concentrations 100.0, 300.0, 500.0 and 1000.0 µg/mL. Pure isolates were dissolved in acetone to obtain solutions of concentrations 50.0, 100.0, 250.0 and 500.0 µg/mL. The experiment was done with two replicates. For each replicate, 10 larvae were transferred to a Whatman No. 1 filter paper disc in a 90.0 mm disposable Petri dish. Larvae were treated topically with a 1.0 μL droplet of the respective solution applied onto the pronotum using a Hamilton's syringe (700 series, Microliter TM Hamilton Company, USA). For the negative control, larvae were exposed to 1.0 μL of acetone, for the positive control larvae were exposed to 0.5 μL of deltamethrine, and for the test runs the larvae were exposed to 0.5 μL of acetone solution of each concentration. After treatment the larvae were confined in Petri dishes containing 5 corn kernels each-within metal rings and maintained in the dark at 26±2 o C and 60±5% relative humidity on a 16:8 (L:D) photocycle for 48 h, after which mortality was assessed. The percentage mortality values were subjected to analysis of variance (ANOVA). The LC 50 values, which is the concentration at which 50% of the larvae died, the confidence upper and lower limits, the regression equations and chi-square (χ2) values were calculated using probit analysis [26].

Feeding deterrence assay
The activity of extracts and pure isolates was studied using leaf disc feeding deterrence choice bioassay method [27]. The experiment tested the hypothesis that larval feeding is deterred by plant extracts and pure isolates incorporated in fresh maize leaf discs (1350.0 mm 2 ). Crude extracts were tested at 2000.0 µg/mL and pure isolates at 100.0, 200.0, 500.0 and 1000.0 µg/mL. Larvae were exposed to leaf discs treated with a plant extract or pure compound versus an equal amount of solvent treated (control) diet in Petri dishes of 50.0 mm diameter. Each extract and concentration treatment combination was tested individually versus the control. The larvae were inserted into Petri dishes individually on a piece of Whatman N0 1 filter paper (1.0 x 1.0 cm) and placed centrally on portions of either treated or control diet pieces. Larvae were incubated for 72 h at 25 ± 2 ºC and on a 16:8 light:dark (L:D) photocycle. In the case of the extracts and pure isolates treatment 100.0 μL of each concentration was added to the diet. For the negative control treatment, 100.0 μL of HPLC grade acetone was added to the diet while for the positive control, 100 μL of azadirachtin was added to the diet. The insects were allowed to feed on treated discs for 24 hours. At the end of the experiment, unconsumed area of leaf disc was measured with the aid of a leaf area meter and per cent antifeedant activity calculated [28] and data subjected to analysis of variance. Each experiment was repeated three times. Insect mortality was also recorded.

Data analyses
For the initial screening bioassay, data were corrected for mortality in the controls using Abbott's formula [29] and then normalized using an arcsine transformation. Transformed data were submitted to a randomized complete block analysis of variance (ANOVA) (p < 0.05) and differences between treatments were compared using Tukey's test (p < 0.05). For the feeding deterrence choice assay, the numbers of larvae feeding on extract-treated or pure compoundtreated versus control portions of the diet were compared.

RESULTS AND DISCUSSIONS
The methanol extract from the dry pulverized leaves of O. kilimandscharicum was partitioned between H 2 O, n-hexane and EtOAc. The EtOAc extract was subjected to silica gel column chromatography and gave several fractions using n-hexane-EtOAc mixtures with increasing polarity of the more polar solvent. The eluants from the n-hexane-EtOAc (3:2) showed three spots of R f values 0.43, on TLC (solvent system: n-hexane-EtOAc, 3:1). This fraction on repeated column chromatography using the same solvent system followed by n-hexane-EtOAc, (3:2) yielded compound 1 (R f = 0.43) as one of the components.  [30]. The 1 H NMR data of compound 1 (Table 1), which were compared with those of the known compounds dammara-20,24-dien-3β-ol [31], dammaradienone [32] and 2-oxo-3β,19α-dihydroxyolean-12-en-28-oic [33] showed the presence of a vinylic proton on a tri-substituted double bond at δ H 5.12 (d, J = 6.4 Hz), two proton doublets resonating at δ H 4.70 (J = 1.4 Hz) and 4.60 (J = 1.4 Hz) assignable to C-21 methylene protons and two vinylic methyl protons at δ H 1.68 (Me-26 and Me-27). The vinylic proton at δ H 5.12 was shown to allylically couple with the vinylic methyls by homonuclear decoupling experiments, thus authenticating the presence of a terminal -CH 2 -CH= C(CH 3 ) 2 group [31]. Furthermore, the spectrum exhibited five quaternary methyls at δ H 0.79, 0.81, 0.94, 1.14 and 1.40 which were ascribable to C-19, C-18, C-30, C-29 and C-28 methyl protons, respectively. An up field double of doublets in the aliphatic region which resonated at δ H 2.84 was assigned to C-2 proton and its orientation was suggested to be α-on the basis of the coupling constant, J =12.0, 3.8 Hz [34,35]. Thus, this inferred that the hydroxyl functionality in this position was in β-orientation. The 13 C NMR spectrum (Table 1) displayed a total of 30 carbon resonances, their multiplicities assigned using DEPT-135 establishing the presence of 7 methyls, 10 methylenes (including one olefinic carbon), six methines (including an oxymethine at δ C 69.7 and one olefinic methine at δ C 124.4) and seven quaternary carbons including a keto group at δ C 216.7. In fact, the 13 C NMR of compound 1 was similar to those of dammara-20,24dien-3β-ol [31] with notable differences being the presence of an oxo group on ring A. In compound 1, the C-3 hydroxyl group was replaced by a keto functionality and the positions of the hydroxyl and the keto groups on ring A were ascertained from HMQC and HMBC experiments showing one-bond H, C-and three bond H, C-correlations, respectively ( Figure 2). In this way it was proved that the hydroxyl moiety was at C-2 and the keto group was at C-3, respectively, a fact cemented by NOESY correlations as outlined in Figure 2. Thus, on the basis of spectroscopic data as well as comparison with literature data, compound 1 was elucidated as 2α-hydroxy-3-oxodammar-20, 24-diene.

Compound 2
The compound was obtained as white crystals from CH 2 Cl 2 -MeOH mixture. Its IR spectrum revealed the presence of hydroxyl (3340 cm -1 ) and carbon-carbon double bond (1589 cm -1 ) functionalities. The compound exhibited an HRESI-MS molecular ion peak at m/z 465.3430 [M+Na] + (calcd for C 30 H 50 O 2 +Na, 465.3424). The 13 C NMR spectrum (Table 1)  substantiated by the molecular ion peak which is 2 atomic mass units more than that of compound 1. Furthermore, from the 1 H NMR spectrum, the methine bearing hydroxyl group exhibited a doublet at δ H 3.23 (d, J = 9.0 Hz) assignable to H-3α on the basis of the coupling constant characteristic of H-3α and H-2β axial-equatorial interaction [36], a fact further evidenced by HMBC correlations between the non-protonated carbon C-10 (δ C 39.6) and H-2 (δ H 2.81, m). The methine proton H-2 in turn also showed cross-peak with the C-4 (δ C 38.8). Again comparing both the 1 H and 13 C NMR data of compound 2 with those of dammara-20,24dien-3β-ol [30] revealed a shift of a C-1 methylene peak δ C 43.0 with corresponding 1 H NMR peaks at δ H 2.20 (m) and 1.01 (m). This suggests that the second hydroxyl group was possibly at C-2, a fact augmented by COSY spectrum which revealed the 1 H-1 H proximity between H-2 and H-3. Additional evidence from comparative studies between the two structures revealed the presence of an exomethylene with 1 H NMR peaks appearing at δ H 4.70 (d, J = 1.5 Hz) and 4.58 (d, J = 1.5 Hz). Its position on the side chain was ascertained by HMBC cross-peaks between H-17 (δ H 1.75, m) and C-21 (δ C 108.0). Thus, on the basis of spectroscopic evidence as well as comparison with various literature data [31], compound 2 was established to be 2α,3βdihydroxydammar-20,24-diene. Contact toxicity assay using n-hexane, EtOAc and aqueous MeOH extracts of O. kilimandscharicum against insect pests S. zeamais and P. trunctus showed that the activities were dependent on concentration and time of exposure. Significant mortalities (p < 0.05) were observed for both the post harvest pests (Figure 3). The aqueous MeOH extract exhibited the highest activity (p ≤ 0.05) with an Lc 50 of 22.55±1.57 μg/mL followed by EtOAc extract with an LC 50 of 78.05±1.26 μg/mL against S. zeamais. The same extracts when tested on P. truncatus gave LC 50 values of 31.95±1.35 and 56.29±1.23 μg/mL, respectively. The results showed that the extracts exhibit different levels of mortality against S. zeamais and P. truncatus at different concentrations which could be attributed to the chemical constituents in the extracts. Chemical compounds present in extracts are known to often act synergistically against physiology of many insects [37]. A previous study on the blend effect of essential oil constituents of O. kilimandscharicum on the two post harvest pests, S. zeamais and P. truncatus showed that the toxicity could be attributed to combined effect of the different compounds [37]. From the probit analysis, the LC 50 values of 22.55±1.57 μg/mL and 31.95±1.35 μg/mL from aqueous MeOH extract for S. zeamais and P. truncatus, respectively compared favorably with an LC 50 of 38.05 µL/40 g grain [37]. In the antifeedant activity assay of the crude extracts of O. kilimandscharicum against S. zeamais and P. truncatus, all the three extracts were active against the two post harvest insects with the activities increasing with increase in concentrations ( Figure 4). The aqueous MeOH extract exhibited promising antifeedant activity against. S. zeamais and P. truncatus with an AFI 50 (concentration that causes 50% feeding deterrence) values of 26.39±1.43 μg/mL and 31.85±1.23 μg/mL. The results obtained in this study compare favorably with previous investigations [38] which showed that methanol and EtOH extracts O. kilimandscharicum had high activity against Sitophilus granarius also known as the grain weevil or granary weevil. These findings also suggest that the cumulative effect of various chemical constituents present in the extracts is responsible for the activity.
When the pure isolates of O. kilimandscharicum were tested against S. zeamais and P. truncatus for their contact toxicities, some of them showed promising activities against the two post harvest insects, though most of the activities were not significantly different (p > 0.05). Turkesterone (5) had the highest activity against both S. zeamais and P. truncatus at LC 50 21.64±1.25 μg/mL and 16.06±1.34 μg/mL, respectively. Other compounds that exhibited relatively moderate activities against the pests included apeginin7-O-neohesperidoside (3) with LC 50 value of 23.41±1.42 μg/mL and 26.20 μg/mL, respectively. The two new compounds 1 and 2 were also fairly active ( Figure 5). The relatively high contact toxicity assay result of turkesterone (5) is in agreement with the results reported previously for this compound in reference with [39], which showed promising efficacy against Callosobruchus maculatus.   Pure isolates from O. kilimandscharicum exhibited interesting antifeedant activities against S. zeamais and P. truncatus. The activities were concentration dependent, increasing with increase in concentration. Activities of the compounds against the two insects were not significantly different (p > 0.05). Turkesterone (5) showed the highest activities with AFI 50 17.50±1.73 µg/mL and AFI 50 21.33±1.74 µg/mL against S. zeamais and P. truncatus respectively ( Figure 6) and are comparable to previous results [40].