Quality assessment of Moringa concanensis seed oil extracted through solvent and aqueous-enzymatic techniques

En este estudio se compara la composición y la calidad del aceite de semilla de M. concanensis extraído mediante enzimas, utilizando tres enzimas comerciales (Natuzyme, Kemzyme, y Feedzyme) con las de un control extraído sin enzimas y con las del aceite extraído con disolvente. El contenido en aceites de las semilla extraídas con enzimas osciló entre 23,54 a 27,46% y fue significativamente más elevado (P 0.05) que el del control sin enzimas (15.41%). Los análisis de los residuos (harinas) no mostraron diferencias significativas (P 0,05) en el contenido de fibra y ceniza para los tres métodos de la extracción. Sin embargo, el contenido proteínico de la harina obtenido por métodos enzimáticos y el control sin enzimas fue significativamente menor (P 0,05) que el de la harina obtenida después de la extracción por disolvente. Las diferencias en el índice de yodo (67.1-68.0 g /100 g of oil), densidad en 24 °C (0,865-0,866 g/mL), índice de refracción a 40 °C (1,4622-1,4627) y fracción insaponificable (0,69-0,76 %) no fueron significativamente diferentes para ninguna de las técnicas de extracción. Las extinciones específicas en 232 y 270 nm, el índice de peróxidos, el índice de p-anisidina, la acidez libre y el color de los aceites extraídos con enzimas fueron inferiores a los del aceite obtenido mediante extracción con disolvente. La composición en ácidos grasos de los aceites fue similar en todas los aceites encontrándose sólo pequeñas en los contenidos de ácido esteárico y linoleico. Respecto a los tocoferoles, el contenido en g-tocoferol fue similar en el control y en las extracciones con enzimas y significativamente más elevado (P 0.05) que en el aceite extraído con disolvente, mientras que el contenido en a-tocoferol fue superior en el aceite extraído con disolvente.

The Moringa plant is well-regarded as an important food commodity equipped with high-value nutritious properties (Anwar et al., 2007).Different parts of this plant: root, bark, gum, leaf, fruit (pods), flowers, seed and seed oil have been used for various ailments in the indigenous medicine practices of South Asia, which included the cure of inflammation, cardiovascular, gastrointestinal, hematological and hepatorenal disorders (Singh et al., 1999;Morimitsu et al., 2000;Siddhuraju and Becker, 2003).The seeds of the Moringa are considered to exhibit antipyretic and antimicrobial activities, used for water purifying and are also a good source of non-desiccating oil, commercially known as 'Ben oil' (Anwar et al., 2007).
In Pakistan Moringa is characterized by two species; Moringa oleifera (M.oleifera) and M. concanensis (M.concanensis).The former is widely cultivated across the country; however, the latter is rarely found and only confined to a remote area of Tharparkar, Sindh (Manzoor et al., 2007, Anwar 2003).The species, Moringa concanensis has recently been characterized with regard to its seed oil potential.Its oil is high in oleic acid and resembles in context of fatty acid composition with seed oils of other Moringa species, which includes the M. stenopetala, M. peregrina and M. oleifera (Manzoor et al., 2007).High-oleic oils are gaining importance, especially for replacing polyunsaturated vegetable oils (Corbett, 2003) and are reported to exhibit good oxidative stability during frying (Anwar et al., 2007).
Conventionally, n-hexane is used as a solvent for the extraction of oils from vegetable oil seeds.However, the environmental problems, safety issues and health concerns have prompted the need to find some environmentally-friendly alternatives to hexane (Bhattacharjee et al., 2006).The U.S. Environmental Protection Agency provided incentives for developing alternative methods of edible oil extraction and guidelines for hexane emissions by vegetable oil extraction sources (EPA, 2001).The new tendency to avoid the use of toxic organic solvents in large installations has renewed interest in alternative extraction processes (involving the use of water, alcohol aqueous solutions and supercritical fluids) (Johnson and Lusas, 1983).
A number of aqueous (Rhee et al., 1972;Shi et al., 1998), and aqueous enzymatic extraction (Hanmoungjai et al., 2001;Rosenthal et al., 1996;Rosenthal et al., 2001;Sharma et al., 2002) methods have been developed for the extraction of oil seeds.Shi et al. (1998), reported an aqueous method that resulted in an 80% yield of peanut oil from peanuts.Karlovic et al. (1994) investigated an aqueous enzymatic method for the extraction of corn germs using cellulase which resulted in 80% oil yield.
The main purpose of using the enzymes during aqueous oil extraction is to hydrolyze the structural polysaccharides which form the cell wall of oilseeds or the proteins which form the cell and lipid body membrane.The aqueous enzyme-assisted extraction process removes phospholipids from the oil so that there is no need for degumming, thereby, reducing the overall cost of processing of the oil to the final product (Christensen, 1991).The major disadvantage of using enzymes for oil seed extraction, although safer, is that they are expensive and offer quite lower oil yield as compared to solvent extraction.Nevertheless, the other technical benefits surrounding enzyme-assisted aqueous oil extraction can make this technique viable.
Currently, much effort is being devoted to develop and introduce commercial enzymes for aqueous enzymatic oil extraction.The present work was therefore, an attempt to evaluate the potential of three commercial enzymes for extraction of oil from M. concanensis seeds by applying the aqueous enzymatic technique.The physicochemical characteristics of the aqueous enzymeextracted M. concanensis seed oil were also compared to that of conventional solvent extracted and control oils.

Aqueous enzyme-assisted extraction
For aqueous enzyme extraction, ground seeds (25g) were mixed with distilled water (150 ml) at a ratio of 1:6 w/v for the method followed by Abdulkarim et al. (2005).The mixture was boiled for 5 min and allowed to cool down to room temperature.The pH was then adjusted to the optimal pH for each enzyme with 0.5 N NaOH and 0.5N HCl.Two percent (w/w) E/S (enzyme/substrate) ratio of each of the three enzymemixtures (Kemzyme, Natuzyme, and Feedzyme) were added and the mixture incubated at 45 o C for 24 h with constant shaking at 120 rpm.The oil was recovered after centrifugation (8000 rpm, 20 °C) for 20 minutes using a centrifuge machine (Sigma, 3K 30, Osterode am Harz, Germany) (Abdulkarim et al., 2005).The wet meal was mixed and dried overnight in a vacuum oven (VOC-300 SD; EYELA, Tokyo, Japan) at 85-90 o C. The dry meal was ground and analyzed for ash, fiber, and protein contents.The Control samples were also proceeded under the same set of conditions, except enzyme addition.

Solvent extraction
After the removal of seed hull and other impurities, the seeds (150 g) were crushed and then fed to a Soxhlet extractor fitted with a 1.0 L round-bottom flask and a condenser.The extraction was executed on a water bath for 6 h with 0.5 L of n-hexane.The solvent was distilled off under vacuum in a rotary evaporator (Eyela, N-N Series, Rikakikai Co. Ltd.Tokyo, Japan) at 45 o C. The oil obtained from both extractions was stored under refrigeration (4 o C), until used for further analyses.

Analysis of the oilseed residues/ meals
After oil extraction (both by enzyme and solvent extraction), the meals were analyzed for protein, fiber, and ash contents.Protein content was determined according to the AOAC (1990) method 954.01.Fiber content was estimated according to an ISO (1977) method 5983.A finely ground sample (2.5 g) of meal was weighed and freed from fat by extraction with 15-mL n-hexane.The test portion was boiled with sulfuric acid (0.255 mol L -1 ), followed by separation and washing of the insoluble residue.The residue was then boiled with sodium hydroxide (0.313 mol L -1 ), followed by separation, washing, and drying.The dried residue was weighed and ashed in a muffle furnace (TMF-2100, Eyela, Tokyo, Japan) at 600 o C, and the loss of mass was determined.
Ash content was determined according to an ISO (1977) method 749.Two grams of the test portion were taken and carbonized by heating on a gas flame.The carbonized material was then ashed in an electric muffle furnace (TMF-2100, Eyela, Tokyo, Japan) at 550 o C until a constant mass was achieved.

Physical and chemical parameters of oil
Iodine value, refractive index, density, unsaponifiable matter, peroxide and saponification values of the enzyme extracted and the control M. concanensis seed oils were determined by various standard AOCS (1997) methods.The color of the oil was determined by a Lovibond tintometer (Tintometer Ltd., Salisbury, Wiltshire, United Kingdom) using a 1inch cell.Specific extinctions at 232 and 270 nm were determined using a spectrophotometer (U-2001, Hitachi Instruments, Inc., Tokyo, Japan).Samples of oil were diluted with iso-octane, and spectra were recorded at 232 and 270 nm, and ε 1% 1cm (λ) calculated following the standard IUPAC (1987) method.para-Anisidine values were also determined following an IUPAC (1987) method.

Fatty acid (FA) composition
FAMEs were prepared according to the IUPAC (1987) method 2.301 and were analyzed on a Shimadzu (Kyoto, Japan) gas chromatograph, model 17-A, fitted with a methyl lignocerate coated film thickness 0.20 µm) SP-2330 polar capillary column (30 m ϫ 0.32 mm; Supelco Inc., Supelco Park Bellefonte, PA), and a flame ionization detector (FID).Oxygen-free nitrogen was used as a carrier gas at a flow rate of 3.0 mL min -1 .Other conditions were as follows: initial oven temperature, 180 o C; ramp rate, 5 o C min -1 ; final temperature, 220 o C; injector temperature, 230 o C; detector temperature, 250 o C; and temperature hold, 2 min before and 10 min after the run.A sample volume of 1.5 µL was injected.FAMEs were identified by comparing their relative and absolute retention times to those of authentic standards.A data-handling program, Chromatography Station for Windows (CSW32), was used for quantification.The FA composition was reported as a relative percentage of the total peak area.

Tocopherol content
Tocopherols (α, γ, and δ) were analyzed using an HPLC following the CPFA (Current Protocols in Food Analytical Chemistry) (Wrolstad, 2003) methods.0.1 g oil sample and 0.05 g ascorbic acid were weighed accurately into a 16 ϫ 125-mm test tube.Five milliliters of 90.2% ethanol and 0.5 ml of 80% aqueous KOH solution were added to the test tube and vortexed for 30 sec.The test tube was flushed with nitrogen, capped and incubated in a water bath (70 o C) for 30 min with periodical vortexing.The tubes were placed in an ice bath for 5 min then 3 ml deionized water and 5 ml n-hexane were added and vortexed for 30 sec followed by centrifugation for 10 min at 1000 ϫ g (room temperature).The upper hexane layer was transferred to another test tube.The aqueous layer and the residue were re-extracted by repeating the same procedure.The upper hexane layers from both the extractions were combined and evaporated to dryness under nitrogen streaming.One milliliter of mobile phase was added to the tube and vortexed 30 sec to re-dissolve the extract and then transferred to an HPLC sample vial.A 20-µL sample was injected into a Supelcosil LC-Si column (250 ϫ 4.6 mm, Supelco Inc., Supelco Park, Bellefonte, USA).A mobile phase of ethyl acetate/acetic acid/hexane (1:1:198, v/v/v) was used at the rate of 1.5 mL min -1 .Detection was monitored at 295 nm.Tocopherols were identified by comparing their retention times with those of pure standards of α-, γ-, and δtocopherols, and were quantified on the basis of peak area of the unknowns with those of QUALITY ASSESSMENT OF MORINGA CONCANENSIS SEED OIL EXTRACTED THROUGH SOLVENT… pure standards (Sigma Chemical Co.).Quantification was based on an external standard method.A D-2500 Hitachi Instruments, Inc., Tokyo, Japan Chromatointegrator model with a built-in computer program for data handling was used for quantification.

Statistical analysis
Three M. concanensis seed samples for each of the three enzyme treatments were analyzed individually in triplicate.Data were reported as means Ϯ SD (n ϭ 3 ϫ 3).One way ANOVA was used to determine significant differences between groups, considering a level of significance of less than 5% (P Ͻ 0.05) by using the statistical software STATISTICA 5.5 (Stat Soft Inc, Tulsa, Oklahoma, USA).

RESULTS AND DISCUSSION
The oil content (23.54-27.46%) of aqueous enzyme-assisted M. concanensis seed was significantly (P Ͻ 0.05) higher than that of the control (15.41%), but lower than that extracted with hexane (38.40%) (Table 1).The oil concentration was highest (27.46%) in the Kemzyme treated seed samples, whereas the seeds treated with Feedzyme afforded the lowest oil content (23.54%).The increase in oil yield in the present aqueous enzyme-assisted extraction of M. concanensis seed as compared to the control might be attributed to the enzyme activity.The multiple activity complexes and enzyme mixtures are especially effective for degradation of seed cell walls, due to their synergistic action thus making more oil available for extraction (Sineiro et al., 1998).The analysis of the meals revealed the protein content (19.06-20.17%) of the M. concanensis seed treated with enzymes to be significantly (P Ͻ 0.05) lower than that of the control and solvent-extracted seeds.This decrease may be due to the extraction of some protein in the aqueous and the emulsion phases.Fiber and ash contents in the enzyme treated oil seeds ranged from 5.94 to 6.12, and 6.98 to 7.17%, respectively and were almost comparable to those of the control and solvent-extracted and showed no significant (P Ͼ 0.05) variations.
Table 2 shows the data for various physicochemical parameters of the enzyme-, solventextracted and the control-M.concanensis seed oils.There were no significant (P Ͼ 0.05) differences in the iodine value (67.1-68.0g of iodine/100 g of oil), density at 24 o C (0.865-0.866 g mL Ϫ1 ), refractive Index at 40 o C (1.4622-1.4627)and unsaponifiable matter (0.69-0.71%) of the M. concanensis seed oils extracted through three extraction methods.The saponification (173-175 mg of KOH/g), of the aqueous enzyme-extracted M. concanensis seed oil was found to be comparable to the control but S. LATIF AND F. ANWAR  Mean values in the same row followed by the same superscript letters are not significantly different (P Ͼ 0.05).
slightly lower than that of the solvent-extracted oils.
The contents of free fatty acids (% as oleic acid) (0.17-0.20) and color (1.2-1.3 r ϩ 12-14 y) of the enzyme-extracted M. concanensis seed oils were found to be well in line with that of the control (0.18, and 1.2 r ϩ 14 y respectively) but significantly lower (P Ͻ 0.05) than that of the solvent-extracted oil (0.32, and 2.1 r ϩ 19 y respectively).
The results of different oxidation parameters of the control, enzyme-, and solvent-extracted M. concanensis seed oils are shown in Table 3.The specific extinctions at 232 and 270 nm, which revealed the oxidative deterioration and purity of the oils (Anwar et al., 2006a), of enzyme-extracted M. concanensis seed oils ranged from 2.88 to 2.94 and 0.57 to 0.62, respectively.The content of conjugated diene and triene of the enzyme-extracted oils was found to be comparable to that of the control (2.95, 0.60), but significantly (P Ͻ 0.05) lower than that of solvent-extracted oil (3.17, 0.67).The peroxide and p-anisidine values of enzyme-extracted M. concanensis seed oil (1.08-1.22meq/kg and 1.61-1.65 respectively) were comparable to that of the control (1.24 meq/kg and 1.59 respectively), however, these were significantly (P Ͻ 0.05) lower than that of the solvent-extracted oil (1.73 meq/kg and 1.82, respectively).A noticeably lower level of the values of different oxidation parameters and thus good oxidative stability exhibited in the present analysis of enzyme-extracted M. concanensis seed oil as compared to the solvent-extracted oil might be attributed to the extraction technique applied.The conventional vegetable oil seed extraction process is performed by means of organic solvents, where the high operational temperature might affect the oil quality and its composition, particularly, with regard to the oxidation state.No previously reported data on the oxidation parameters of enzyme-extracted M. concanensis seed oil is available with which to compare the results of our present analysis.
Table 5 shows the contents of different tocopherols in the control, enzyme-, and solventextracted M. concanensis seed oils.et al., 2006a), in the Kemzyme-extracted oil, was found to be higher, whereas, Feedzyme-, and Natuzye-extracted oils showed significantly (P Ͻ 0.05) lower values than that of solvent-extracted oil.
The concentration of δ-tocopherol, with potent antioxidant activity (Anwar et al., 2006a), in the Natuzyme-, and Kemzyme-extracted oils (30.50 and 29.27 mg kg Ϫ1 , respectively) was in close agreement with the control, and solvent-extracted oil (30.78 and 29.61 mg kg Ϫ1 , respectively).However, a significantly (P Ͻ 0.05) higher content of δ-tocopherol was observed in Feedzyme-extracted oil (35.49mg kg Ϫ1 ).The level of γ-tocopherol in enzyme-extracted oils (16.93-21.22mg kg Ϫ1 ), was significantly higher than that of solvent-extracted (11.38 mg kg Ϫ1 ) oil.In context of δ-tocopherol concentration, Kemzyme-extracted (16.93 mg kg Ϫ1 ) oil was noted to have close resemblance to that of the control (17.56 mg kg Ϫ1 ).No previous reports on the tocopherol contents of enzyme-extracted M. concanensis seed oils are available in the literature which with the data of present analysis could be compared.
It could be concluded from the results of the present investigation that aqueous enzymatic treatment enhanced the oil yield of M. concanensis seeds with respect to the control although the yield of solvent extraction was not reached.The oil extracted through this approach revealed improved oxidative stability without exhibiting any considerable change in the contents of major fatty acids.Furthermore, the aqueous phase obtained through this extraction method seems to be a potential source of edible protein, and may be utilized for value-addition in beverages and other food commodities.

Table 1 Proximate composition of Moringa concanensis seeds A Enzyme-extracted
SD, calculated as percentage on dry seed weight basis for three Moringa concanensis seed samples for each enzyme analyzed individually in triplicate.Mean values in the same row followed by the same superscript letters are not significantly different (P Ͼ 0.05).
A Values are means Ϯ

Table 2 Physico-chemical properties of Moringa concanensis seed oils A
A Values are means Ϯ SD, of three Moringa concanensis seed oils analyzed individually in triplicate.

Table 3 Determination of the oxidative state of Moringa concanensis seed oils A Enzyme-extracted Constituent Solvent-extracted Natuzyme Kemzyme Feedzyme Control
Values are means Ϯ SD, of three Moringa concanensis seed oils analyzed individually in triplicate.Mean values in the same row followed by the same superscript letters are not significantly different (P Ͼ 0.05). A

Table 4 Fatty acid composition (g/100g) of Moringa concanensis seed oils A
A Values are means Ϯ SD, of three Moringa concanensis seed oils analyzed individually in triplicate.Mean values in the same row followed by the same superscript letters are not significantly different (P Ͼ 0.05).
Abdulkarim SM, Long K, Lai OM, Muhammad SKS, Ghazali HM. 2005.Some physico-chemical properties of Moringa oleifera seed oil extracted using solvent and aqueous enzymatic methods.Food Chem.93, 253-263.AOAC.1990.Official Methods of Analysis of the Association of Official Analytical Chemists,

Table 5 Comparison of tocopherol in Moringa concanensis seed oils A
Values are means Ϯ SD, of three Moringa concanensis seed oils analyzed individually in triplicate.Mean values in the same row followed by the same superscript letters are not significantly different (P Ͼ 0.05) QUALITY ASSESSMENT OF MORINGA CONCANENSIS SEED OIL EXTRACTED THROUGH SOLVENT… A