The chemical composition of kernel meal obtained after oil extraction was determined. As shown in Table 1, the meal is high in protein (61.8% w/w) but low in neutral detergent fiber (NDF) (9.7% w/w) and acid detergent fiber (ADF) (4.8% w/w). These results are comparable to those of J. curcas kernel meal from Mexico, India, Nicaragua and Cape Verde as reported by Makkar and Becker [1]. The phytochemical analyses of J. curcas kernel meal showed that total phenolics was 3.9 ± 0.23 mg tannic acid equivalents/g DM and total flavonoid was 0.4 ± 0.15 mg rutin equivalents/g DM (Table 1). Total phenolic content is comparable to the value of 3.6 mg tannic acid equivalents/g reported by Makkar and Becker [1], while total flavonoid content of kernel meal has not been reported so far.

Phenolic and flavonoid compounds occur ubiquitously in plants and a variety of biological activities have been attributed to them. These advantageous properties include antimicrobial, antiviral, anti-ulcerogenic, cytotoxic, anti-neoplastic, mutagenic, antioxidant, anti-hepatotoxic, anti-hypertensive, hypolipidemic, anti-platelet and anti-inflammatory activities. Many of these biological functions corresponded to their free radical scavenging and antioxidant activities [5].

Total saponin content was 19.0 ± 0.48 mg diosgenin equivalents/g DM. This value was lower when compared to kernel meal of J. curcas from Mexico, Cape Verde and India with the values of 28.5, 26.0 and 27.3 mg diosgenin equivalents/g DM, respectively [6]. Saponins are widely distributed amongst plants and it is believed that saponins naturally act to protect the plant against pathogens. Their main biological activities include haemolytic, molluscicidal, anti-inflammatory, antimicrobial, anti-parasitic, cytotoxic and anti-tumor [7]. The presence of both polar (sugar) and nonpolar (steroid or triterpene) groups provide saponins with strong surface-active properties that are responsible for many of their biological activities such as antioxidant, antimicrobial, anti-inflammatory and cytotoxicity [8].

Phorbol esters are esters of tetra-cyclic diterpenes which are widely distributed in plant species of the families Euporbiaceae and Thymelaceae and biological activities such as anti-HIV, anti-malaria and antimicrobial have been reported by Goel et al. [9]. The phorbol esters content of kernel meal was 3.0 ± 0.16 mg phorbol-12-myristate 13-acetate (PMA) equivalents/g DM. Phorbol esters are known to mimic the action of diacylglycerol to activate the protein kinase C, which regulates different signal transduction pathways and other cellular metabolic activities including proliferation, malignant transformation, differentiation and cell death [9]. The multiplicity of effects of phorbol esters on biological systems is associated with the presence of phorbol esters receptors [10] and the types of phorbol esters. Fujii et al. [10] and Park et al. [11] reported the induction of apoptosis using phorbol esters in prostate cancer cells and gastric cancer cells.

In this study, high performance liquid chromatography (HPLC) was used to determine the phenolic and flavonoid compounds in J. curcas kernel meal extracts. The analyses showed the presence of gallic acid and pyrogallol as phenolics (Figure 1); rutin and myricetin as flavonoids (Figure 2) and daidzein as isoflavonoid compound (Figure 1) with the concentrations shown in Table 2. Compounds that have been reported from J. curcas leaves including apigenin and its glycosides, vitexin and isovitexin, stigmasterol, β-sitosterol and gallic acid [12] while the root and stem contained gallic acid, ellagic acid, quercetin, coumaric acid, benzoic acid and salicylic acid [13].

The GC-MS analysis is a rapid and simple method to determine the compounds present in the plant materials. Table 3 shows the major organic compounds detected by GC-MS in the methanolic and hot water extracts. More than half of the organic compounds in the methanolic extract consists of 2-(hydroxymethyl)-2 nitro-1,3-propanediol, β-sitosterol, 2-furancarboxaldehyde, 5-(hydroxymethy) and acetic acid. In the hot water extract, the most common compounds present are 2-furancarboxaldehyde, 5-(hydroxymethy), acetic acid and furfural (2-furancarboxaldehyde) adding up to about 60% of the total extracted materials. The 2-(Hydroxymethyl)-2 nitro-1,3-propanediol has been reported as microbicidal and it is used as a bacteriostat in disinfectants [14]. β-sitosterol, a plant sterol, shows antimicrobial [15], anti-inflammatory and cytotoxic activities [16,17]. The compound 2-furancarboxaldehyde, 5-(hydroxymethy) also shows antibacterial and antifungal activity applicable in pharmaceutical, cosmetics and pesticides [18]. Furfural is used as a flavor in foods, and in other products, such as cosmetics, fragrance, pesticide, herbicide, fungicide, insecticide and germicide [19]. According to Ryssel et al. [20], acetic acid also showed excellent bactericidal effects toward Proteus vulgaris, Pseudomonas aeruginosa and Acinetobacter baumannii. Therefore, the GC-MS results showed the presence of compounds with known activities indicating the J. curcas kernel meal as a potential source of functional metabolites.

Antibacterial activity of methanolic and hot water extracts of J. curcas kernel meal were tested against four Gram negative and four Gram positive pathogens at two concentrations (1.0 and 1.5 mg/disc). The results shown in Table 4 indicate significant difference (P < 0.05) in inhibitory activity of methanolic and hot water extracts. The diameter of inhibition zones of methanolic extract ranged from 0.35 to 1.25 cm while for hot water extract ranged from 0.00 to 0.75 cm at 1.0 mg/disc. Antimicrobial activity of the methanolic extract appeared to be more effective than water extract, since methanol could extract a wide variety of active components compared to hot water. The size of inhibition zones increased with 1.5 mg/disc. At this concentration, the antibacterial effect of methanolic extract was similar for both Gram negative and Gram positive bacteria, except for P. aeruginosa, which showed moderate zone of inhibition. This Gram negative bacteria seemed to be less sensitive to other plant extracts [21].

Escherichia coli seemed to be the most susceptible bacterial species to methanolic and hot water extracts at 1.0 and 1.5 mg/disc extract concentration. The antibacterial activities of both kanamycin and streptomycin were as expected and some of the clearing zones of pathogens were comparable to that of antibiotics, particularly when 1.5 mg/disc extracts were tested.

Various reports have shown the antimicrobial activity of extracts from different parts of J. curcas plant. However, the efficacy of extracts to inhibit microbial growth seemed to vary according to extraction procedures and assay methods. Igbinosa et al. [22] have reported the in vitro antimicrobial activity of crude methanolic and water extracts of the J. curcas stem bark against a wide spectrum of microorganisms with zones of inhibition ranging from 0.8 to 2 and 0 to 0.8 cm for methanol and water extracts at the concentration of 1.0 mg per well (0.6 mm) respectively. In another study, Atindehou et al. [23] reported the antimicrobial activity of the leaves and roots of J. curcas against E. coli, P. aeruginosa, S. aureus and E. faecalis.

Total phenolics generally possess antimicrobial activities which provide chemical barriers for invading microorganisms. On the other hand, flavonoids inhibit bacterial growth by inhibition of DNA gyrase, cytoplasmic membrane function and energy metabolism [24]. Gallic acid and pyrogallol as phenolic compounds, daidzein as isoflavonoid and rutin and myricetin as flavonoids (Table 2) are substantially effective as antibacterial agent [24]. Additionally, the saponins present in J. curcas kernel meal could also contributed to the antimicrobial activity as saponins have been shown to possess antibacterial activity against several Gram positive and Gram negative bacteria [7]. Furthermore, phorbol esters (Table 1) could also act as antimicrobial agent as this property has been demonstrated by Chumkaew et al. [25].

The results shown in Figure 3 indicate the 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging activity of J. curcas kernel meal extracts at different concentrations. The inhibition of DPPH radical scavenging activity increased in a dose-dependent manner. Table 5 shows the IC₅₀ (required concentration to scavenge 50% of DPPH radicals) concentration of extracts. Free radical scavenging activity of methanolic extract showed a value of 1.6 mg/mL but hot water extract did not show an IC₅₀ value, even at the highest concentration used in this study. Methanolic extract of J. curcas kernel meal appeared to be more active as compared to the hot water extract in scavenging the free radicals. Reference antioxidants vitamin C, BHT and β-carotene showed IC₅₀ values of 0.3, 0.3 and 1.5 mg/mL respectively. Both extracts showed lower activities than vitamin C and BHT due to the significantly (p < 0.05) higher IC₅₀ values required to inhibit the free radicals (Table 5). The methanolic extract activity was similar to β-carotene as no significant difference was observed between their IC₅₀ values (Table 5). It has been reported that the strong free radical scavenging activities in latex, root and stem of J. curcas was attributed to the presence of phenolic compounds [2,13].

Ferric reducing antioxidant power (FRAP) of extracts and reference antioxidants (vitamin C, BHT and β-carotene) is presented in Figure 4. Reference antioxidants showed higher reducing power followed by methanolic and hot water extract. The IC₅₀ value (the concentration to reduce 50% of Fe³⁺ to Fe²⁺ as an index for antioxidant potential) in FRAP for vitamin C, BHT and β-carotene were 0.3, 0.3 and 2.6 mg/mL, respectively (Table 5), while for the methanolic and hot water extracts were 3.0 and 8.0 mg/mL, respectively. As observed in the DPPH assay, FRAP assay also showed similar IC₅₀ values between methanolic extract and β-carotene (Table 5).

The results indicate that the extracts were more active in scavenging free radicals compared to ferric reducing power activity. In FRAP assay, the general ability of the extracts to donate electrons is tested whereas, in the DPPH assay, reduction by hydrogen atoms are also involved. Hence, the free radical scavenging activity assay was a better indicator of antioxidant activity of kernel meal than FRAP based on the reaction mechanisms proposed. In complex systems, such as food, various different mechanisms may contribute to oxidative processes. Therefore, it is important to characterize the extracts by a variety of antioxidant assays [26].

The presence of phenolics (gallic acid, pyrogallol), flavonoids (rutin, myricetin) and isoflavonoid (daidzein) in the kernel meal (Table 2) may contribute to the antioxidant activity of extracts. These compounds have been identified as some of the major compounds involved in antioxidant activity [27]. However, saponins in the kernel meal (Table 1) might be also involved, because Sparg et al. [7] reported the antioxidant activity of these compounds.

Phytochemical analyses, antimicrobial and antioxidant activities showed that the methanolic extract was more active compared to hot water extract. Therefore, cytotoxicity assay was performed using the methanolic extract. It is known that different cell lines might exhibit different sensitivities while treated with the same plant extract, therefore screening plant crude extract against various cell lines is a first step in determining the most susceptible cancer cells. Apart from that, breast and cervical cancers are very common manifestations, hence the MCF-7 and HeLa cell lines were used to investigate the effects of the plant extract. Chang liver cells (human hepatocytes) were also used to determine the hepatoxicity of extract used in this study. Cytotoxicity assay results demonstrated that the methanolic extract of J. curcas kernel meal inhibited the growth and proliferation of MCF-7, HeLa and Chang liver cells in a dose-dependent manner (Figure 5). Methanolic extract at the concentration of 6.25 μg/mL and above decreased the cell viabilities significantly (P < 0.0001) in all cell lines used in this study as compared to the untreated control.

In this cytotoxicity activity study, Tamoxifen was used as a positive control. As shown in Figure 6, Tamoxifen also inhibited the growth and proliferation of MCF-7, HeLa and Chang liver cells in a dose dependent manner. At 6.25 μg/mL and above, cell viabilities decreased significantly (P < 0.0001) in all cell lines when compared to the untreated control.

The CC₅₀ values of methanolic extract and Tamoxifen are presented in Table 6. The MCF-7 cells were found to be more susceptible to the methanolic extract as compared to the HeLa and Chang cells. Most anticancer drugs have been discovered through random screening of plant materials. The National Cancer Institute requires a CC₅₀ value of less than 30 μg/mL for the active crude extracts against cancer cells [28]. Since the methanolic extract showed CC₅₀ values less than National Cancer Institute requirement toward MCF-7 cells, methanolic extract could be a promising source of bioactive compounds that could be purified for the development of anticancer drugs in order to treat breast cancer.

The CC₅₀ concentrations of Tamoxifen (Table 6) for MCF-7, HeLa and Chang liver cell lines were 17.3, 8.89, 33.6 μg/mL, respectively. The activity of Tamoxifen toward MCF-7, HeLa and Chang were significantly (p < 0.0001) higher than methanolic extract. However, normal Chang liver cells were also susceptible to both methanolic extract and Tamoxifen. It is interesting to note that the CC₅₀ concentration of Tamoxifen was about half of that methanolic extract. At present it is not known which particular compound(s) present in the kernel meal methanolic extract exert the cytotoxicity effect in both cancer and, to some extent, Chang liver cells.

The presence of bioactive compounds in other J. curcas plant parts has been reported recently. Oskoueian et al. [2] demonstrated the cytotoxic effect of J. curcas root and latex methanolic extract with CC₅₀ values of 18.3 ± 0.98 and 70.1 ± 2.06 μg/mL against HT-29 cloned cancer cells, respectively; and 33.3 ± 0.75 and 60.1 ± 0.89 μg/mL toward Chang liver cells, respectively. Besides root and latex, leaves methanolic extract was also observed to be effective against B16F10 melanoma cells with CC₅₀ value of 24.8 μg/mL [29]. Moreover, Fujii et al. [10] reported that PMA induced apoptosis in prostate cancer cells through overexpression of PKCδ. Park et al. [11] also concluded that the PMA induce apoptosis in gastric cancer cells through activation of PKC and the activation of serine protease(s) and caspase-3/CPP32. All these results indicate the potential of J. curcas plant as a source of bioactive compounds, but, the exact nature of the metabolite involved has to be elucidated, although it has been reported that phenolics, flavonoids, saponins and phorbol esters have been shown to be cytotoxic on different cell lines [9].

The Jatropha curcas L. plant was collected from the farm of Faculty of Agriculture, Universiti Putra Malaysia with the GPS location of 3°0′26.91″N latitude and 101°42′13.24″E longitude for identification by Mr. Shamsul Khamis. A voucher specimen (SK1764/2010) was deposited in the Phytomedicinal Herbarium, Institute of Bioscience, Universiti Putra Malaysia, Serdang, Selangor, Malaysia. Upon confirmation of the plant, the ripe J. curcas seeds were collected from the farm, air dried, dehulled and kept at 4 °C for further analysis. The kernels isolated were ground, using a mechanical grinder and oil was extracted by Soxhlet apparatus, using petroleum ether (boiling point of 40–60 °C) for 16 h [30]. Methanolic extract of J. curcas kernel meal was prepared according to Crozier et al. [27]. Hot water extract was prepared as described by Gulcin et al. [31]. The extracts were filtered through Whatman No.1 paper and evaporated by using vacuum rotary evaporator (Buchii, Switzerland). The crude extracts were re-dissolved in methanol for antioxidant, antimicrobial and HPLC analyses, or in dimethyl sulfoxide for cytotoxicity assay. The extractions were repeated three times.

Chemical constituent analyses were conducted according to AOAC [32]. The fiber fractions which consisted of neutral detergent fiber (NDF) and acid detergent fiber (ADF) were determined according to Van Soest et al. [33]. The analyses were repeated three times.

The phenolic and flavonoid compounds of J. curcas kernel meal were quantitatively measured by HPLC based on the method described by Crozier et al. [27] with some modification. Phenolic standards were gallic acid, syringic acid, vanillic acid, salicylic acid, caffeic acid and pyrogallol. Flavonoid standards used in this research were quercetin, rutin, myricetin, kaempferol, naringin and apigenin and isoflavonoid standards were genistein and daidzein. An aliquot of sample extracts was loaded on the HPLC equipped with an analytical column Intersil ODS-3 (5 μm 4.6 × 150 mm, Gl Science Inc.). Solvents comprising deionized water (solvent A) and acetonitrile (solvent B) were used. The pH of water was adjusted to 2.5 with trifluoroacetic acid. The phenolic and isoflavonoid compounds were detected at 280 nm while flavonoid compounds at 350 nm. The column was equilibrated by 85% solvent A and 15% solvent B. Then the ratio of solvent B was increased to 85% in 50 min followed by reducing solvent B to 15% in 55 min. This ratio was maintained for 60 min for each sample analysis with flow rate at 0.6 mL/min. The analysis was repeated three times.

Phorbol esters were determined based on Makkar et al. [8]. An aliquot of sample extracts was loaded on a high-performance liquid chromatography (HPLC Agilent-1200) series instrument equipped with a UV-Vis photodiode array (DAD) detector, binary pump, vacuum degasser, auto sampler and analytical column Agilent Reverse-phase C₁₈ LiChrospher 100 (250 × 4 mm I.D and 5 μm pore size). The solvents were deionized water (solvent A) and acetonitrile (solvent B). The absorbance was read at 280 nm and results were expressed as equivalents to phorbol-12-myristate 13-acetate standard. The analysis was done in triplicate.

The chemical composition of the extracts was characterized by GC-MS based on the method described by Hossain and Rahman [34]. The GC-MS used was a Schimadzu QP2010PLUS system. Six μL were analysed on a BPX-5 SGE ultra-low-bleed 5% phenyl polydimethylsiloxane capillary column (30 m × 0.25 mm i.d. × 0.25 μm film thickness). Splitless injection was performed with a purge time of 1.0 min. The carrier gas was helium at a flow rate of 1 mL min⁻¹. The column temperature was maintained at 50 °C for 3 min, then programmed at 5°C min⁻¹ to 80 °C and then at 10 °C min⁻¹ to 340 °C. The inlet and detector temperature was 250 °C and 340 °C, respectively, and the solvent delay was 4 min. The identification of the peaks was based on computer matching of the mass spectra with the National Institute of Standards and Technology (NIST 08 and NIST 08s) library and by direct comparison with published data. The analyses were carried out in triplicate.

The antibacterial assay of J. curcas kernel meal extracts was carried out by the disc diffusion method as described by Boussaada et al. [21] against Gram positive (Micrococcus luteus, Bacillus subtilis B145, Bacillus cereus B43 and Staphylococcus aureus S1431) and Gram negative (Enterobacter aerogenes, Klebsiella pneumonia K36, Escherichia coli E256 and Pseudomonas aeruginosa PI96) pathogens. All the bacteria were purchased from the Institute of Malaysian Research (IMR) and maintained in the Department of Microbiology, Faculty of Biotechnology and Biomolecular Sciences, Universiti Putra Malaysia. The positive control without extracts (solvent) and reference controls using two standard antibiotics (kanamycin and streptomycin) were included in the assay. The extracts inhibitions were corrected based on positive control and compared with those of reference control. The experiments were conducted in triplicate.

Two cancer cell line, including human breast cancer cell line (MCF-7 ATCC: HTB-22) and human cervical adenocarcinoma (HeLa ATCC: CCL-2) and one normal cell line, human hepatocytes (Chang ATCC: CCL-13) were purchased from the American Type Culture Collection (ATCC). Cell lines were grown as monolayers in a T-75 cm² culture flask. The Dulbecco’s Modified Eagle Medium (DMEM) was supplemented with 2.0 g/L sodium bicarbonate and 10% fetal bovine serum. The cultured cell lines were maintained in a humidified atmosphere of 5% CO₂ at 37 °C and were harvested when they reached 80% confluency for the cytotoxic test. Cytotoxicity of J. curcas kernel meal extract was measured by MTT cell proliferation assay [36]. Cytotoxic concentration was expressed as CC₅₀ (concentration of the sample that decreased the number of viable cells to 50% of the control cell) which was calculated by Graph Pad Prism 5 software. All determinations were performed in triplicate.

Inhibition zones of antibacterial assay subjected to the general linear models procedure of SAS [37] in a completely randomized design and the means were compared with Duncan’s Multiple Range test. The difference was considered significant when the P value was <0.05. GraphPad Prism 5 software (GraphPad Software Inc., San Diego, CA) was used for all the statistical analyses in cytotoxic assay.