Study of Quality Characteristics and Efficacy of Extraction Solvent/Technique on the Antioxidant Activity of Bitter Gourd Seed

Momordica charantia Linn belongs to the Cucurbitaceae family and commonly known as Bitter Gourd (BG), bitter melon, karela and grows in the humid and subtropical regions of the world. It is inherent in Asia and now widely cultivated in all parts of the world due to its dietary assessment in its immature or ripened fruits. The BG plant is modified to a wide variation of climates however its best production is in warm areas [1]. It is a curative vegetable which has been used conventionally for diabetes treatment. It is a ground-breaking plant for its adaptability as foodstuff and therapeutic applications. Contemporary scientific proof has proved BG an auspicious antidiabetic, anticancer, antimicrobial, anti-hepatotoxic, antioxidant, antiviral, antiulcerogenic, and larvicidal activities [2].


Introduction
Momordica charantia Linn belongs to the Cucurbitaceae family and commonly known as Bitter Gourd (BG), bitter melon, karela and grows in the humid and subtropical regions of the world. It is inherent in Asia and now widely cultivated in all parts of the world due to its dietary assessment in its immature or ripened fruits. The BG plant is modified to a wide variation of climates however its best production is in warm areas [1]. It is a curative vegetable which has been used conventionally for diabetes treatment. It is a ground-breaking plant for its adaptability as foodstuff and therapeutic applications. Contemporary scientific proof has proved BG an auspicious antidiabetic, anticancer, antimicrobial, anti-hepatotoxic, antioxidant, antiviral, antiulcerogenic, and larvicidal activities [2].
Although the different parts of plant BG have been used as food and drug but the fruit is the most important part [3]. Depending on the maturity stages, BG Seed (BGS) has been found as a rich source of oil (18.1-37.6%) and protein (28-30%). BGS oil is rich in fatty acids like Conjugated Linoleic Acid (CLA). BGS protein is a good source of essential amino acids (Met, Cys, Ile, Phe, Tyr and Lys) and could be a good source of protein for functional ingredients in a food system. BGS also supply important minerals (P, K, Mg, S and Ca), flavonoids and phenolic compounds (catechin and gallic acid) [4]. Antioxidants, especially, the phenolics have gained considerable importance due to their potential health benefits. Consumption of plant foods containing antioxidants is beneficial to health as it lowers down many degenerative processes. Antioxidant compounds can be recovered from plant materials through different extraction techniques taking into account their chemistry and uneven distribution in the plant matrix. Antioxidant compounds most frequently isolated by using solvent extraction technique. However, the extract yields and resulting antioxidant activities of the plant materials are strongly dependent on the nature of extracting solvent, due to the presence of different antioxidant compounds of varied chemical characteristics and polarities that may or may not be soluble in a particular solvent. Polar solvents are frequently employed for the recovery of polyphenols from a plant matrix. The most suitable of these solvents are (hot or cold) aqueous mixtures containing ethanol, methanol, acetone, and ethyl acetate [1].
Therefore huge effort has been made to develop effective selective extraction techniques.
The present study therefore has been conducted with the main objective of investigating the physicochemical composition, antimicrobial and haemolytic activity of BGS oil and to investigate the most effective solvent/technique for extracting potent antioxidant compounds from BGS. Proximate and nutritional compositions of BGS have been reported by few scientist but there is an information gap regarding chemical composition, antioxidant, antimicrobial and hemolytic activities of BG cultivated in Pakistan. This information can add value to the agricultural potential of BG, and can also introduce high-value components in the market towards a nutraceutical potential.

Bitter gourd seed (BGS)
Ripened fruits of two BG varities were procured from Ayub Agriculture Research Institute, Department of Oil Seed Crops, Faisalabad, Pakistan. These two varieties have different morphology from each other. The local variety BGS-1 (Guti) has small fruit, deep green color, covered with small triangular tubercles and strongly bitter in taste whereas another varitiety BGS-2 (Goj) has long green fruit, covered with long triangular tubercles and bitter in taste. In Pakistan, BG is commonly known as kerela. The seeds were separated from the fruits and washed several times with water to remove any dust particles and foreign materials. Afterward, the seeds were placed in Pyrex Petri plates (15.00 g in each Petri plate) to dry in an electric oven at 45°C for 24 h until a constant weight was attained. Seeds were stored in polyethylene bags till for further analysis. Duplicate samples of seeds were collected.

Reagents and standards
All chemicals of commercial as well as analytical grade used in this study were purchased from either E. Merck (Darmstadt. Germany) or Sigma-Aldrich (Buchs, Switzerland) unless otherwise noted. Pure standards of tocopherols [DL-α-tocopherol, (+)-δ-tocopherol, and (+)-γ-tocopherol], FAME, Catechin, Triton-X-100 and Gallic acid were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO), microbial media and sensitivity disc for antimicrobial activity were from Oxoid (Hampshire, UK).

Oil extraction
After drying samples in an electric oven, thoroughly mixed seeds were ground by using an electric grinder. Each batch (~75-80.00g) of BGS was fed into Soxhlet extractor fitted with 500 mL round bottom flask and a condenser. The extraction was carried out on water bath (75-80°C) for 8-9 h using 300 mL of n-hexane. After extraction, the solvent was distilled off under vacuum in a rotary evaporator (Eyela Rotary Vacuum Evaporator N.N. Series) at 45°C and the percentage of the oil contents were calculated [5].

Analysis of BGS residue and mineral content determination
After oil extraction, residues contain protein, starch, carbohydrates and different inorganic minerals. Total protein contents of BGS residue were estimated according to AOAC official method [6] using a micro-Kjeldhal apparatus whereas water soluble protein was analyzed by Lowry method using bovine serum albumin as the standard [7]. Starch contents and total carbohydrate were also estimated by modified method as reported by Ali et al [8]. Fiber and ash contents were analyzed according to the International Standard Organization (ISO) method [9].
For mineral analysis, BGS samples were subjected to acid digestion and analyzed by UV-Vis as well as by atomic absorption spectrophotometry (AAS) as described by AOAC, [10].

Analysis of BGS oil
Physical and chemical parameters: Determination of density, refractive index, FFA, PV, iodine value, ester value, saponification value, unsaponifiable matter, Reichert-Meissl Value and p-Anisidine value of the extracted oil was carried out according to standard IUPAC method. Color was estimated by Lovibond tintometer (Tintometer Ltd., Salisbury, United Kingdom) using a 1-in. (2.5 cm) cell. Specific extinction at 232 and 270 nm was determined using Hitachi U-2001 spectrophotometer by following the method of IUPAC [11]. Pour flash, fire and smoke point was analyzed according to the method of AOCS [12].
Tocopherol and fatty acid analysis: Tocopherol analysis was performed by HPLC following the method of Anjum et al. [13] whereas fatty acid composition was determined by instrumental methods like GLC. GLC analysis was performed by preparing fatty acid methyl ester (FAME) according to IUPAC standard method 2.301 [14] and analyzed on a Shimadzu gas chromatograph as reported by Anjum et al [13].

Disc diffusion method:
The antimicrobial activity of both BGS oils was determined by disc diffusion method. Briefly, 100 µL of suspension of tested microorganisms, containing 10 7 colony-forming units (CFU)/mL of bacteria cells and 10 6 spores/mL of fungi spread on NA and PDA medium, respectively. The filter discs (6 mm in diameter) were individually impregnated with 15 µL of extracted oil, placed on the agar plates which had previously been inoculated with the tested microorganisms. Discs without samples were used as a negative control. Amoxycillin (30 µg/dish) (Oxoid, UK) and Flumequine (30 µg/disk) (Oxoid, UK) were used as positive reference for bacteria and fungi, respectively to compare sensitivity of strain/isolate in analyzed microbial species. Plates, after 2 h at 4°C, were incubated at 37°C for 18 h for bacteria and at 28°C for 24 h for fungal strains. Antimicrobial activity was evaluated by measuring the diameter of the growth inhibition zones (zone reader) in millimeters for the organisms and comparing to the controls [15].

Determination of minimum inhibitory concentration (MIC):
For the determination of MIC (which represents the concentration that completely inhibits the growth of microorganisms) a microdilution broth susceptibility assay was used, as recommended by National Committee for Clinical Laboratory Standards [15]. All tests were performed in nutrient broth (NB, Oxoid, UK) for bacterial and sabouraud dextrose broth (SDB, Oxoid, UK) for fungal strains supplemented with Tween 40 detergent to a final concentration of 0.5% (v/v). Bacterial strains were cultured overnight at 37°C in NB and the fungi were cultured overnight at 28°C in SDB. Dilutions series were prepared from 0.03 to 72.0 mg/mL of the compounds in a 96well microtiter plate, 160 µL of NB and SDB for bacteria and fungi, respectively were added onto microplates and 20 µL of tested solution. Then, 20 µL of 5×10 6 CFU/mL of standard microorganism suspension were inoculated on to microplates. Plates were incubated at 37°C for 24 h for bacteria, and at 28°C for 48 h for fungi. The same test was performed simultaneously for the growth control (NB + Tween 40) and sterility control (NB + Tween 40 + test compound). Amoxycillin was used as a reference compound for antibacterial and Flumequine for antifungal activities. The growth was indicated by the presence of a white ''pellet'' on the well bottom.
Hemolytic activity: Hemolytic activity of the oil was studied by the method used by Sharma and Sharma [16]. Three mL freshly obtained heparinized human blood was collected from volunteers after consent and counseling and bovine from the Department of Clinical Medicine and Surgery, University of Agriculture. Blood was centrifuged for 5 min at 1000xg plasma was discarded and cells were washed with three times with 5 mL of chilled (4°C) sterile isotonic Phosphate-buffered saline (PBS) pH 7.4. Erythrocytes were maintained 10 8 cells per mL for each assay. Hundred μL of each oils were mixed with human and bovine erythrocytes (10 8 cells/mL) separately. Samples were incubated for 35 min at 37°C and agitated after 10 min. Immediately after incubation , the samples were placed on ice for 5 min then centrifuged for 5 min at 1000xg. Supernatant 100 μL was taken from each tube and diluted 10 time with chilled (4°C) PBS. Triton X-100 (0.1% v/v) was taken as positive control and phosphate buffer saline (PBS) was taken as negative control and pass through the same process. The absorbance was noted at 576 nm using μQuant (Bioteck, USA). The % RBCs lysis for each sample was calculated.
Preparation of extracts for antioxidant activity: BGS extract was prepared by using 100 g of BGS powder of both types with 1.0L of each of the solvent-water, Absolute ethanol and aqueous ethanol for 3 hours in an orbital shaker (Gallenkamp, UK), or under reflux on a water bath or in ultrasonic bath (1 hour) in separate experiments then filtered with filter paper (whattman filter paper). The residues were reextracted twice. Extract was concentrated and lyophilized. The dried crude concentrated extracts were weighed to calculate the yield and stored at 4°C till for further analysis [17].

Total phenol content (TPC) analysis:
The TPC of both BGS extracts were determined according to the Folin-Ciocalteu method described by Chaovanalikit and Wrolstad [18]. BGS extracts were diluted with 2.5 ml of distilled water, 0.5 ml of the Folin-Ciocalteu stock reagent and 1.0 ml of Na 2 CO 3 reagent (75 g/l) were added to the mixture. The mixture was incubated for 30 min at 30°C and then the absorbance was measured at 765 nm. TPC was expressed in milligrams of Gallic acid equivalents per gram of BGS extract. DPPH radical scavenging activity: DPPH radical scavenging activity was determined by following the modified method of Chang et al. [19]. 1.0 mL of extract containing 25µg/L of dry matter dissolving in methanol, 5.0 mL of freshly prepared solution of 1, 10-diphenyl-2-picrylhydrazyl (DPPH) at concentration of 0.025 g/L was added. Absorbance was measured at 517 nm at the time interval of 0, 0.5, 1, 2, 5 and 10 min. The remaining amounts of DPPH radical were calculated from calibration curve. For comparison of radical scavenging activity of extracts, absorbance was measured at fifth minute.

Determination of total flavonoid content (TFC):
The TFC of BGS extracts of both varieties were determined according to the method as described by Zou and Wei [21]. 0.5 ml of sample solution dissolved in 2 mL of distilled water and afterwards with 0.15 mL of 5% NaNO 2 solution. 0.15 mL of 10% AlCl 3 solution was added after 6 min of incubation then allowed to stand for 6 min, followed by adding 2 ml of 4% NaOH solution to the mixture and made volume up to 5 mL, the mixture was mixed thoroughly and left the solution for 15 min. The absorbance was measured 510 nm. TFC was expressed in milligrams of equivalents per gram of BGS extract.
Statistical analysis: Three samples of each BGS variety were analyzed individually in triplicate and the results were reported as mean (n=3×3×1) ± SD (n=3×3×1). Statistical significance of the differences between mean values was assessed by 2 way analysis of variance (ANOVA) using Minitab 2000 Version 13.2 statistical software (Minitab Inc. Pennsylvania, USA). A probability value of p ≤0.05 was considered to denote a statistically significant difference [22].

Proximate analysis
Results of the proximate composition of both BGS varieties are shown in table 1. Significant (P<0.05) differences were observed between BGS-1 and BGS-2 varieties in their contents of moisture (22.91 and 29.32% respectively). In contrast to our results, Ali et al. [8], and Nyam et al. [23], reported significantly lower moisture content (7.62-8.20% and 7.3% respectively) whereas Prashantha et al. [24], and Islam et al. [25], recorded 53-79.5% of moisture which may be attributed to variations in the genetic makeup or due to the collection of sample at different level of fruits maturity.
The oil contents (given as percentage to the dry weight) of both BGS were 31.51 and 40.79% respectively which is appreciably higher than 19.30 and 26.00% average oil as reported by Ali et al. [8], and Nyam et al. [23], respectively. However, the oil content in the BGS-1 was close to the oil content (41-45%) of the four different cultivars of BGS [26]. Our findings are also in good agreement with the findings of Airan and Shah [27], and Prashantha et al. [24], who reported the average yield of 35% and 40.45% of the BGS oil.
The total protein content was found to be 14.92 and 19.17% of which 5.01 and 3.91% were water soluble in BGS-2 and BGS-1 varieties respectively. The present results revealed that BGS is a potential rich source of protein and could be used in various food applications. The protein content estimated by micro-Kjeldahl method showed a considerably higher value than that given by Lowry method. The reason is that Lowry method was applied to a water extract and took into accounts, in this case, the water soluble proteins only. Further, micro-Kjeldahl method takes into account both the protein and non-protein nitrogen. The amount of protein in BGS-1 was in good agreement Values (mean ± SD) are average of three samples of each BGS sample, analyzed individually in triplicate (n = 1x3 x 3), (P < 0.05).  [28], and Ali et al. [8], respectively. Whereas Nyam et al. [23], detected comparatively lower concentration (11.8%) of protein. Islam et al. [25], detected significantly higher amount of protein (27-31.3%). The starch contents were found to be 4.27 and 3.15% indicating the presence of a considerable amount of carbohydrates. An appreciable amount of fiber contents (11.07 and 10.03%) detected in both BGS varieties did not differ significantly (P < 0.05). However, Ali et al. [8], reported comparatively smaller amount of crude fiber (1.02-1.20%). In present research work, the ash contents were found 7.03 and 6.48% in BGS-2 and BGS-1 respectively. Comparatively smaller concentration (2.25-2.73% and 3.60%) of ash was detected in the BGS by Ali et al. [8] and Nyam et al. [23] respectively.  [27], and Prashantha et al. [24], respectively. The refractive index is not a useful property for specifying oils but it is useful purity criteria that can be used as a quality parameter in preliminary testing of the seed oils. The BGS-2 oil was reddish brown in color which may be due to the presence of browning substances that are polar due to active radical [13] whereas BGS-1 oil was brownish yellow oil at room temperature. As given in table 2, the iodine values of the BGS-2 and BGS-1 oils were 109.62 and 124.16 respectively, indicating that BGS-1 oil contained a substantial amount of unsaturated fatty acids. These findings were in agreement with Prashantha et al. [24], and Ali et al. [8], who reported the iodine value of 115.96 and 125.21-131.06.

Physicochemical analysis of oil
The saponification values of BGS oils were in the range of 184.02-186.09 (mg of KOH/g of oil) in agreement with the findings of Ali et al. [8] (184.08-187.01) but lower than 190.70 reported by Prashantha et al. [24]. The percentages of free fatty acids (FFA) (0.59 and 1.06% as oleic acid) were in harmony to the findings (0.86-1.12%) of Ali et al. [8].
The ester values of oils in the both varieties were calculated as 184.93 and 181.86 from the acid and saponification values. Same findings have been reported by Ali et al. [8]. Reichert-Meissl values of both varieties were estimated at 1.84 and 1.79 respectively. The low Reichert-Meissl values thus obtained indicate a low content of lower volatile soluble fatty acids, and this value is also in agreement with the low saponification value obtained. Peroxide value and FFA have frequently been used as two of the most important parameters to monitor the quality of edible oils. These values provide the much needed information for further refining process of seed oils [23]. Both varieties BGS oils contained unsaponifiable matters ranging from 1.07 to 1.79% which is in good agreement with the values 1.12-1.71% mentioned by Ali et al. [8]. The significantly (P < 0.05) lower percentage of unsaponifiable matters as obtained in the BGS-1 oil (1.07%) points to lower amounts of hydrocarbons, higher alcohols and sterols than those in the BGS-2 oil.

Analysis for oxidative stability of oils
Different parameters studied for the oxidative stability of the oils extracted from both BGS are given in table 2. The BGS oils displayed the peroxide values of 3.02-5.97 mEq/kg, lower than that of 6.13-8.50 mEq/kg reported by Ali et al. [8]. Both type of BGS oils contained p-Anisidine value in the range of 3.59-3.89, not significantly (P>0.05) different from each other. The specific extinctions at 232 and 270 nm were 4.32 and 0.52; 5.07 and 1.03 observed in BGS-2 and BGS-1 oils respectively.

Tocopherols analysis
Tocopherols are important biological antioxidants and have been correlated with the reduction of heart diseases, delay of Alzheimer's disease, and preclusion of cancer. Especially, α-tocopherol (vitamin E) prevents oxidation of body lipids including polyunsaturated fatty acids and lipid components of cells and organelle membranes. The concentration of vitamin E (239.81-381.06 mg/g) was found to be highest followed by γ-tocopherol (52.85-61.09 mg/g) and δ-tocopherol (99.74-110.07 mg/g) in BGS oil. A significantly (P<0.05) higher concentration of vitamin E was observed in the BGS-1 oil as compared to the BGS-2 oil (Table 3). Since a rich source of tocopherols BGS oil can be used in food, feed, pharmaceuticals, cosmetics and resins.

Fatty acid (FA) analysis
FA compositions of BGS oils were determined by GLC and are presented in table 3. Except myristic and oleic acids, significantly differences (P<0.05) were observed in the remaining FAs in both BGS oils. BGS-2 and BGS-1 oils contain highest amount of eleostearic acid  (48.14 and 44.05%) followed by stearic acid (25.44 and 29.14%), oleic (16.10 and 15.87%) and linoleic (6.87 and 5.41%) acids respectively.
Similar trend in the FA composition findings were also reported in the Chinese BGS oil [26]. However, a higher concentration of eleostearic acid was observed in the Chinese BGS oil as compared to our results. Contrary to our results, Nyam et al. [23], recorded mainly stearic (32.4%), oleic (1.5%) and linoleic acid (2.6%) in the same fruit. The fatty acid profiles (except stearic and eleostearic acid) evaluated in this study are in close agreement with the values reported by Ali et al. [8]. This variation in the fatty acid composition may be due to the variety, soil, and climatic conditions.

Antimicrobial activity
The antimicrobial activity of BGS oil of both varieties were evaluated against a set of microbes like Gram positive and Gram negative bacteria and selected fungal strains ( Table 4). The oil obtained from both BGS exhibited different degree of antimicrobial activity against the all microorganism tested. Results obtained from disc diffusion method, followed by measurement of minimum inhibitory concentration (MIC), showed the following sensitivity order:

Bacillus subtilis > Staphylococcus aureus> Pasturella multocida> Escherichia coli
The results indicated that Bacillus subtilis and Staphylococcus aureus were the most sensitive bacteria among selected bacterial strain tested by both BGS. In a previous study, BGS oil showed antibacterial activity against methicilin-resistant Staphylococcus aureus [29]. But this may be the first report regarding the antimicrobial activity of both BGS oil. A slightly higher antimicrobial activity was recorded for the oil extracted from BGS-2 than BGS-1. Similar pattern was also observed for MIC values (Table 4). A comparatively higher activity of both BGS oil was exhibited against Gram positive bacteria than Gram negative strains. The antibacterial activity of both varieties BGS oil was comparable with the standard drug, Amoxicillin.
The oil of both BGS varieties exhibited antifungal activity but less than standard drug (Flumequinene). The following sensitivity order of selected fungal strains was observed:

Microsporum canis > Fusarium solani > Candida albicans > Aspergillus flavus
Our findings were similar to the results depicted by Braca et al. [30]. It can be inferred that the BGS oil has stronger and broader spectrum of antimicrobial activity. However, BGS oil is more potent against bacterial as compared to fungal strains.

Hemolytic activity
The BGS oil of both varieties was screened using a rapid assay against human and bovine erythrocytes. The results are summarized in table 5. No toxicity was observed by BGS-2 oil but very low hemolytic activity was recorded by BGS-1 oil (Table 5). Shu-Jing and Ng [20] reported the hemolytic activity of steryl glycoside fraction from tubers of Momordica cochinchinensis. However, no previous report was available about the toxicity of Momordica Charantia.

Effects of solvent/technique on the extracts yields
Amounts (g/100g of dried BGS) of the antioxidant extract using three different solvents (water, ethanol and aqueous ethanol (ethanol: water, 80:20 v/v) and three extracting techniques: shaker, reflux and ultrasonic waves are shown in table 6.
The vigor of the extraction procedure mainly affected the amount of the antioxidant components extracted from BGS of both varieties which may probably vary from sample to sample. Amongst other contributing factors, efficiency of the extracting solvent to dissolve endogenous compounds might also be very important. For the effectiveness of extracting technique, the results showed that yields of the extract were better when extraction was done under ultrasound treatment followed by reflux and then shaking, regardless of the plant material and solvent used. This indicates that ultrasonic waves not only reduce the particle Values (mean ± SD) are average of three samples of each BGS sample, analyzed individually in triplicate (n = 1x3 x 3), (P < 0.05); A Amoxiclin (30 µg/disk) for bacterial and Flumequine (30 µg/disk) for fungal strains. B DD, Diameter of inhibition zone (mm) including disc diameter of 6 mm. C MIC, minimum inhibitory concentration (µg/mL). Values (mean ± SD) are average of three samples of each BGS sample, analyzed individually in triplicate (n = 1x3 x 3), (P < 0.05). size but also increase the rate of dissolution of antioxidant components using hot solvent systems, thus offering higher extract yields which is also comparable with the yield under reflux. This is in agreement with the findings of Shon et al. [31], who investigated that alcohol and hot water are more efficient to extract antioxidant compounds from Phellinus baumii. Table 6a shows the Total Flavonoid Content (TFC) and Total Phenol Content (TPC) of BGS extracts of both varieties under different conditions; shaking, reflux and ultrasound using three solvents: water, absolute and aqueous ethanol. Flavonoids and phenolic compounds are the main antioxidative compounds of fruits and vegetables [17] Under shaking treatment, among the different solvents, aqueous ethanol extracts of BGS-2 and BGS-1 found to have high TPC (59.42 and 70.31) followed by ethanol (55.11 and 68.11) and water extracts (44.17 and 40.59 mg/g respectively). Results of the present study showed that among all the solvent extracts; the aqueous ethanol extracts had the highest TPC. This may be due to the fact that phenolics are often extracted in higher amounts in more polar organic solvents as compared with absolute ethanol and water. These values approached to maximum when treated under ultrasonic treatment as ultrasound induces cell destruction, cell size reduction and propagation of ultrasonic waves through solid material, leading to a larger solid-liquid interface and better contact of the solvent with the components. Same findings has been reported by Mladenovic et al. [32], who extracted ultrasonically the higher content of phenols (69.46 mg GAE/g dry extract) and flavonoids (33.53 mg CE/g dry extract) in the ethanolic extract of leek stem. In contrast to the trends noted for extraction yields, the TPC of all medicinal plant materials extracted using the reflux technique decreased, regardless of the nature of the extracting solvent used.

Effects of solvent/technique on total phenolic contents
The decrease in the amounts of TP of these plant material extracts, prepared under reflux might have been due to the thermal decomposition of some phenolic antioxidants at the higher temperatures used for reflux extraction. It has been reported that thermal processing conditions might result in the loss of natural antioxidants because heat may accelerate their oxidation and other degenerative reactions. Thus, heating temperature is of much consideration during processing. An accelerated shelf-life test at 80°C for 4 days resulted in 20-40% decrease of the antioxidant activity of the apple juice Van der Sluis et al. [33] and Cheng et al. [34], reported that antioxidant activity of wheat bran decreased up to 61% by heating at 100°C for 9 days. On the other hand, Dutra et al. [35] reported that among different extraction techniques (reflux, maceration, ultrasound, heating plate), extraction made under reflux using ethanol/water (70:30, v/v) offered the highest polyphenol levels in Vogel seeds. This might be attributed to an effective extraction under reflux conditions leading to higher release of some bound phenolics [34].

Effects of solvent/technique on the total flavonoids
Total flavonoid contents (TFC) of BGS, extracted with three solvent systems, using shaker, reflux and ultrasonic waves extracting techniques, are given in table 6a. TFC were determined as catechin equivalents (CE). Aqueous ethanol extracts of BGS-2 and BGS-1 varieties showed highest level of TFC (69.01 and 82.01 mg/g) followed by aqueous (66.03 and 77.48) and absolute ethanol (41.25 and 37.52). Amount of TFC in all solvents extracts generally decreased when reflux technique employed for their preparation whereas abruptly increased when extraction was made by ultrasound treatment and maximum TFC was offered by aqueous ethanol (74.11 and 85.17) followed by water (68.31 and 81.11) and absolute ethanol (50.38 and 44.21). Shu-Jing and Ng [20] also reported similar trend for TFC (44.00 mg/g) and TPC (68.80 mg/g) while using ethanol as an extracting solvent whereas 62.00 and 51.60 mg/g was estimated for TFC and TPC in the extract obtained in aqueous ethanol.

Effects of solvent/technique on DPPH and metal chelation
DPPH scavenging activity of different plant materials as affected by extracting methods is depicted in table 6b. Absorbance in this assay was recorded at 0.5 to 10 min time intervals from initiation of the reaction. Observed scavenging activity was similar at the beginning Values (mean ± SD) are average of three samples of each BGS sample, analyzed individually in triplicate (n = 1x3 x 3), (P < 0.05) ND: Not detected. Table 5: Hemolytic activity of bitter gourd seed (BGS) against human and bovine erythrocytes.  68.10 a a ± 2.99 Values (mean ± SD) are average of three samples of each BGS sample, analyzed individually in triplicate (n = 1x3 x 3), (P < 0.05); Superscript letters within the same row indicate significant (P< 0.05) differences of means within the extracting solvent; Subscript letters within the same column indicate significant (P< 0.05) differences of means among the extracting technique. of the reaction and changed with increase in the reaction time until it stabilized by the 10th min. Significant (p<0.05) differences of DPPH scavenging capacities among extracts were observed at 5th minute of the reaction. The DPPH scavenging ability of the sample extracts was reported as mg/mL of DPPH scavenged (mg/mL DPPH. scavenging). As expected, a higher percent of DPPH scavenging is correlated to a higher antioxidant activity [36].

BGS Extraction by shaker
In Ultrasonic waves extraction technique exhibited better scavenging activity than their corresponding extracts, obtained by reflux as well as shaking technique whereas reflux technique has no any significant effect on scavenging activity. In contrast to present findings, Wu et al. [20], reported more free radical scavenging activity of BGS-H 2 O extract than BGS-EtOH. These results are also contradictory to the previous findings of Dutra et al. [35]  411.03 b a ± 4.21 Values (mean ± SD) are average of three samples of each BGS sample, analyzed individually in triplicate (n = 1x3 x 3), (P < 0.05); Superscript letters within the same row indicate significant (P< 0.05) differences of means within the extracting solvent; Subscript letters within the same column indicate significant (P< 0.05) differences of means among the extracting technique. highest DPPH scavenging activity. This reduction in the radical scavenging activity of the extracts, obtained by the reflux technique might be ascribed to the thermal decomposition of phenolics [35].

Elemental analysis
With regard to minerals, both BGS varieties contain relatively large quantities of calcium (425.08 and 399.81 μg/g) ( Table 7) which is in close agreement with those (383.45-440.96 μg/g) reported by Ali et al. [8]. The amount of copper (3.51 and 2.89 μg/g) was higher when compared to the values described by Assubaie and El-Garawany [3]. Similar results for copper (2.85-3.52 μg/g) have been reported by Ali et al. [8]. In both BGS varieties, the iron (43.06 and 49.21 μg/g) contents were in good agreement with the findings (41.10-45.03 μg/g) of Ali et al. [8] and were appreciably higher than those for BGS (12.35-32.33 μg/g) [3]. The contents of zinc in both BGS varieties (10.88 and 12.91 μg/g) were slightly higher than those in BGS (7.76-10.50 μg/ g) studied by Assubaie and El-Garawany [3]. An appreciably higher amount of calcium and phosphorus was observed in the BGS-2 as compared to BGS-1.

Conclusion
The findings imply that both BGS can be used as a potentially source of lipid, protein, unsaturated fatty acids (especially eleostearic acid), minerals, tocopherols and antioxidants such as flavonoids and phenols. BGS-1 was found to have appreciably higher contents of lipid and protein as against the BGS-2. BGS oil of both varieties exhibited notable antimicrobial activity against selected bacteria and fungi strains, assessed by disc diffusion and measurement of minimum inhibitory concentration by micro dilution method. Aqueous ethanol exhibited better potential for antioxidant extraction under ultrasound effect. However, BGS-1 showed remarkably higher level of TPC and TFC as compared to BGS-2 under ultrasonic treatment. These antimicrobial and antioxidant activities of BGS may contribute to its therapeutic applications.