Pangium Edule Reinw

Process Biochemistry 35 (1999) 197–204 www.elsevier.com/locate/procbio

Mobilization of primary metabolites and phenolics during natural fermentation in seeds of Pangium edule Reinw.

Nuri Andarwulan a,b, Srikandi Fardiaz b, Anton Apriyantono b, Purwiyatno Hariyadi b, Kalidas Shetty a,*

a Department of Food Science, Uni6ersity of Massachusetts, Chenoweth Laboratory, Box 31410, Amherst, MA 01003, USA b Department of Food Technology and Human Nutrition, Bogor Agricultural Uni6ersity, Bogor, Indonesia

Received 15 December 1998; received in revised form 30 March 1999; accepted 10 April 1999

Abstract

Fermented seeds of the tropical tree Pangium edule Reinw. are a speciality in Indonesia and have been used as spices. The fermentation process of the seeds is a natural spontaneous process, which occurs 40 days following seed maturity and treatment. This study reports some biochemical changes, especially primary metabolites, and antioxidant activity associated with mobilization of lipids and phenolics during seed fermentation. The lipid content increased slightly (46.07–50.95% db) although the dominant fatty acid composition did not change. The dominant fatty acids were oleic acid (C18:1n-9) and linoleic acid (C18:2n-6). During fermentation, the decrease in fatty acid content in lipid coincided with the increasing acid value, which indicated that free fatty acids increased in seeds during fermentation. The dominant tocol in the seed, g-tocotrienol, increased (69.8–123.3 mgg−1 freeze-dried seed) during fermentation. In general, overall protein content and amino acid composition did not change but non-soluble protein increased while soluble protein decreased. The changes in carbohydrate fraction showed that total crude carbobydrate, neutral detergent ﬁbre (NDF, as cellulose, hemicellulose, and lignin) decreased, but reducing sugar increased and starch content did not change. Enzyme assays showed that microorganisms may be involved in the fermentation process. b-glucosidase, an enzyme that can cleave glycosidic bonds of conjugated phenolics and guaiacol peroxidase (GPX) activities increased. The total phenolics content in seeds increased substantially corresponding to the increase in b-glucosidase but antioxidant activity of phenolic extracts did not change. © 1999 Elsevier Science Ltd. All rights reserved.

Keywords: Pangium edule; Fermentation processing; Biochemical changes; Primary metabolises; Phenolics

1. Introduction also a raw material for another product, ‘kecap pangi’ (ketchup, soy sauce like product) and has been used as Pangium edule Reinw. is a tropical tree that grows in a spice in Saparua. An edible oil is also produced from Micronesia, Melanesia, and Southeast Asia, including the seed kernels. Indonesia. The seeds of this tree are poisonous, mostly Keluwak is fermented in a speciﬁc way. The fruits are because of the presence of cyanogenic glucosides [1]. In harvested and placed in the ﬁeld for 10 days until the Indonesia, seed kernels are edible following treatment fruit is tainted. The seeds are then removed, washed, and the removal of cyanogenic glucoside. ‘Dage’ is a and boiled for 3 h. The seeds are then cooled, placed in product from boiled seeds after removal of kernels and a hole in the ground (indoors) and covered by ash. water soaking for 2–3 days. Dage is utilized in West After 40 days, the fermented seeds are cleaned and can Java as a vegetable. Another product is a fermented be used as spices. Previous research on fermented seeds seed material, called ‘keluwak’, which has been used a indicated that the methanol extract of keluwak had spice for soup in Java and South Sulawesi. Keluwak is antioxidant activity [2] and another investigation found that keluwak oil did not contain cyclopentenyl fatty * Corresponding author. Tel.: +1-413-5451022; fax: +1-413- acids, common cyclic fatty acids in the Flacourtiaceae 5451262. g E-mail address: [email protected] (K. Shetty) family, while -tocotrienol is a predominant tocol [3].

Recently, we have also shown that the dominant fatty The Soxhlet oil extracts of all freeze-dried fermented acids in seeds during germination were oleic and seeds were analyzed in duplicate for fatty acid profiles. linoleic acid, and the antioxidant g-tocotrienol in germi- Methyl ester derivatives were prepared according to nated seed was dominant in early stage and changed to AOCS standard method No. Ce 2-66 and IUPAC be a tocotrienol during germination [4]. standard method No. 2.301 with modification. Approx- Prior to this study, it was not known whether fer- imately 1 mg of sample oil (diluted using hexane as the mentation or post-harvest ripening occurred during solvent) was placed in a small vial with a teflon cap. A keluwak processing. We suspected that fermentation 0.5 to 0.7 ml of 2 N NaOH (in methanol) was added to was the process because of the possibility growth of each sample. After homogenization, the vial was placed microorganism inside and on the surface of the seed. in a heating block at 80°C for 10 min. The vial was

The physical properties of seeds such as texture, colour removed and 1 ml of BF3-methanol reagent (Sigma, St. and flavour changed during post-harvest processing. Louis, MO) was added. Subsequently, following ho- The potentially fermented seeds during post-harvest mogenization, the vial was placed in a heating block at processing have soft texture and dark colour (dark red 80°C for 10 min and homogenized every 3 min. Half a to dark brown). These changes are thought to be due to millilitre of hexane was added to the reaction mixture biochemical reactions linked to the enzyme activity of after it cooled. Following homogenization, saturated microorganisms. The objective of this research was NaCl solution was added to the mixture and this which primarily to investigate the mobilization of lipids, was followed by centrifugation at 3000 rpm for 1–2 protein, carbohydrate, free phenolics and the antioxi- min. The upper phase (hexane phase) was removed and dant activity associated with seeds during post-harvest placed in a vial, which contained anhydrous Na2SO4. processing and natural fermentation. A secondary ob- The hexane phase, which contained fatty acid deriva- jective was to determine the activity of key enzymes, tives, was stored at 4°C in amber crimp vials wrapped b-glucosidase, which is an enzyme for breakdown of in aluminum foil and was analyzed within a week. glycosidic bond of conjugated phenolics and guaiacol Standards of methyl ester derivatives were obtained peroxidase, which may be potentially involved in conju- from Sigma. gation of phenolic aglycones. Fatty acid derivatives were analyzed with a Varian Model 3700 gas chromatograph with flame-ionization detector and equipped with an integrator (SP 4270). 2. Materials and methods The column was a Supelco™10 fused silica capillary column with dimension of 30 m×0.20 mm and 20 mm 2.1. Plant material and fermentation process film thickness. The initial column temperature of 150°C was increased at a rate of 3°C min−1 to a final temper- Pangium edule Reinw. seeds were obtained from ature of 240°C which was held for 10 min. The injector Bogor, Indonesia. The fruits were harvested in Novem- and detector temperatures were 250 and 300°C, respec- ber 30, 1997 and placed in the field for 10 days until the tively. The fatty acid content was expressed as percent- fruit was tainted. The seeds were removed, washed, and age of total fatty acids. The total fatty acid content was boiled for 3 h. Following this, the seeds were cooled, calculated using margaric acid (C17:0) as internal stan- placed in a hole in the ground (indoor) and covered by dard and expressed as mg fatty acid g−1 lipid. The acid ash. The post-harvest fermentation process began on value assay was described in AOAC methods [5]. December 4, 1997 and proceeded until January 14, 1998 2.3. Tocol analysis (40 days). Three groups of fermented seeds were used in this study according to the time of fermentation: 0 day The tocopherols (T) and tocotrienols (T3) were ex- refers to the boiled seed following cooling, 20 and 40 tracted in duplicate from the freeze-dried fermented days refer to fermented seeds following boiling and seed powder in minimal light [6,7]. 1 g seed was homog- cooling. The fermented seeds after removal of kernels enized for 1 min in HPLC grade methanol (20 ml) and were freeze-dried and ground. The freeze-dried fer- filtered through Whatman c42 media. The superna- mented seeds were stored at −20°C until analysed. tant was removed and placed in a 25 ml glass vial and evaporated under nitrogen. The residue was resus- 2.2. Lipid content and fatty acids composition pended in 15 ml HPLC grade methanol and the homogenization and filtration steps were repeated. The Lipid or oil content was measured gravimetrically supernatant was removed and added to the first extract after the freeze-dried fermented seed powder was defat- and dried under nitrogen. The dried extract was dis- ted using a Soxhlet extraction method for 6 h using solved in 2 ml of HPLC grade hexane, mixed briefly in hexane as the solvent. Oil was obtained after the sol- a vortex mixer and centrifuged at 13 000 rpm for 5 min, vent was evaporated under reduced pressure using the placed ina2mlamber crimp vial and immediately AOAC official method 963.15. analyzed. N. Andarwulan et al. / Process Biochemistry 35 (1999) 197–204 199

Extracts were analyzed directly by high perfor- reagent (1:3 dilution of stock solution which con- mance liquid chromatography (HPLC) using a tained 50 mg orthophtaldehyde, 4 ml methanol, Hewlett Packard model HP1090 equipped with a No- 0.0025 ml mercaptoethanol, 0.05 ml of 30% Brij-30 vapak C18 column (3.9 mm×150 mm) with guard and 1 ml of 1 M borate buffer pH 10.4) was added cartridge and diode array detector at 298 nm. The and the mixture stored for 1 min. The sample was eluent was HPLC grade acetonitrile–methanol (85:15) then ready for HPLC assay. Amino acids were ana- at a flowrate of 0.8 ml min−1 (Table 1). A sample lyzed by HPLC (Shimadzu) equipped with a Ultra volume of 10 ml was used. Identification of to- Techspere ODS 3 column (HPLC Technology) (4.6 cotrienols (a-, and g-tocotrienols) were compared with mm×75 mm), a fluorescence detector, and an inte- the retention time and derivative spectrum of each grator (Shimadzu C-R6A). The eluent was a gradient compound peak [8]. Tocopherol standards (a-, g-, and of buffer A (0.025 M, acetate buffer pH 6.5, contain- d-tocopherols) were obtained from Sigma (Sigma). ing of 0.025 M Na-acetate pH 6.5, 0.05% Na-EDTA, The tocols content was expressed in mgg−1 freeze- 9% methanol and 1% THF) and B (95% methanol) dried seed using d-tocopherol as external standard. (Table 1) at a flow rate of 1 ml min−1. Amino acids standard for calibration and calculation were pur- 2.4. Protein content and amino acids composition chased from Sigma.

Total crude protein was performed using the micro 2.5. Carbohydrate content Kjeldahl procedure. Non-soluble protein was the protein that was precipitated in TCA solution and Crude or total carbohydrate was determined by dif- was also determined using the micro Kjeldahl proce- ference from the proximate analysis (100%−water dure. Prior to the nitrogen assay for non-soluble content%−ash content%−protein content%−lipid protein, the freeze-dried fermented seed was extracted content%). Neutral detergent fibre (NDF, as cellulose, using water and then centrifuged. 10% trichloro acetic hemicellulose, and lignin) was estimated by the van acid (TCA) was added to the supernatant and after Soest method [9]. Reducing sugar (D-glucose) was es- centrifugation, protein in the residue (non-soluble timated by the dinitrosalicylic acid (DNSA) method protein) was determined. The soluble protein was de- [10]. The starch content was determined after acid termined by the difference (Total crude protein %− (HC1) hydrolysis and sugar analysis by an anthrone non-soluble protein %). method. The amino acid composition of fermented seeds was determined using a high performance liquid chro- 2.6. i-glucosidase acti6ity assay matography (HPLC) method and carried out as fol- lows: protein in freeze-dried fermented seed powder The enzyme was extracted from freeze dried fer- were hydrolyzed using 6 N HCl in stoppered vials at mented seed powder in buffer at a ratio of 1:5 and 100°C for 24 h, and then freeze-dried. Five millitres 5% PVP-4 was added. The extraction buffer consisted of 0.01 N HCl was added to the freeze dried material of 150 mM bis-tris propane, 2 mM EDTA, 3 mM and the mixture was filtered through a 0.45 mm filter DTT at pH 7.5. The mixture was centrifuged at 4000 (Millipore) followed by addition of 1 M K-borate rpm, 4°C for 30 min and the supernatant utilized for buffer, pH 10.4 at a ratio of 1:1. Ten microlitres of enzyme and protein assays. The protein content of sample solution was placed in a vial and 25 ml OPA each enzyme extract was determined using a Bio-Rad Protein Assay™ (Bio-Rad Laboratories, Hercules, Table 1 CA) [11]. Gradient program of eluent for amino acid assay using HPLC b-Glucosidase in the fermented seed extract was method quantified in terms of its specific activity. 25 ml40 Time (min) % B (45% methanol) mM PNP-b-D-glucopyranoside and 450 ml of 0.1 M phosphate buffer (pH 6.3) were mixed in a reaction 00test tube and 25 ml of supernatant added. After 2 h 10incubation at 30°C the reaction was stopped by 2 15 m 515adding 800 lof1MNa2CO3. After the mixture was 13 42 homogenized, absorption was measured at 400 nm us- 15 42 ing water as reference and the blank was directly in- 20 70 activated enzyme and substrate. Enzyme activity was 22 100 calculated from PNP (p-nitrophenol) produced using 26 100 m −1 28 0 an extinction coefficient ( ) of PNP=10 500 l mol −1 38 0 per cm. Enzyme activity is reported as mmols min per g protein. 200 N. Andarwulan et al. / Process Biochemistry 35 (1999) 197–204

2.7. Guaiacol peroxidase acti6ity assay

The enzyme was extracted from the freeze dried fermented seed powder in buffer, under cold conditions. A sample (50 mg) was homogenized with a glass mortar in 2.5 ml extraction buffer in an ice bath. The extraction buffer consisted of 0.1 M potassium phosphate (Fisher Scientific) buffer, pH 7.5, containing 2 mM EDTA (Fisher Scientific) and 1% PVP-40 (Fisher Scien- tific). The homogenate was centrifuged at 13 000 rpm for 10 min and the supernatant utilized for the enzyme and protein assays. The protein content of each enzyme extract was determined using a Bio-Rad Protein As- say™ (Bio-Rad Laboratories, Hercules, CA) [11]. Fig. 1. Changes in lipid content of Pangium edule Reinw. seed during fermentation. GPX in the fermented seed extract was quantified in terms of its specific activity. 360 ml of 0.056 M guaiacol 6 m 2.9. Antioxidant acti ity test (Acros Organics, NJ), 40 lof50mMH2O2 (Fisher Scientific, Fair Lawn, NJ), and 600 ml of the 0.1 M The antioxidant activity of the phenolic extract was Phosphate buffer (pH 6.8) were mixed in a reaction test evaluated using a modification of the b-carotene– tube. This gavea1mlreaction mixture containing 50 linoleate model system described by Miller [17,18]. A mM potassium phosphate buffer (pH 6.8), 2 mM of solution of b-carotene (Sigma) was prepared by dissolv- H O and 20 mM of guaiacol. At zero time, 50 mlof 2 2 ing 2.0 mg of b-carotene in 10 ml of chloroform. One supernatant was transferred to the reaction tube and milliliter of this solution was then pipetted into a mixed. The oxidation of guaiacol by GPX was followed roundbottom flask. After chloroform was removed un- by monitoring the increase in absorbance (u 470 nm). = der vacuum, using a rotary evaporator at 40°C, 20 mg The rate of change of absorbance per minute was used of purified linoleic acid, 200 mg of Tween 40 emulsifier to quantify the enzyme in the mixture using an extinc- (Aldrich, Milwaukee, WI), and 50 ml of aerated dis- tion coefficient (m) of the oxidized product (tetra guaia- tilled water were added to the flask with vigorous col) of 26.6 mM−1cm−1. Enzyme activity is reported as shaking. Aliquots (5 ml) of this prepared emulsion were mmoles min−1 per g protein [12,13]. transferred into a series of tubes containing 2 mg dry 2.8. Total phenolic content weight of extract. As soon as the emulsion was added to each tube, the zero time absorbance was read at 470 Phenolics were extracted from defatted freeze dried nm. Subsequent absorbance readings were recorded at fermented seed powder. Sample (6 g) were homogenized 30 min intervals, the samples being incubated in a water in 100 ml of ACS grade OmniSolv® methanol (EM bath at 50°C. The protection factor (PF) used to ex- Science, Inc., Gibbstown, NJ) for 3 h using shaker. press antioxidant activity was determined as the ratio of After the mixture was placed in a water bath at 70°C absorbance of the sample at 30 min to that of the for 1 h, it was filtered through Whatman c42 media. control. The residue w as rinsed using methanol and the super- natants combined. The solvent was removed in a rotary evaporator at 40°C under reduced pressure. Total phenolics of the extracts were determined from modified 3. Results and discussion assay [14], which is similar to the original method [15] and as used in this laboratory [16]. Approximately 1 ml The lipid content in Pangium edule Reinw. seed in- of 5 mg of DW extracts ml−1 were taken and placed in creased slightly (46.07–50.95% db) (Fig. 1). It was a test tube to which 1 ml of 95% ethanol (ACS grade) suspected that more non-polar compounds were re- and 5 ml of filtered/deionized water added. Folin-Cio- leased and/or synthesized during fermentation. The calteu reagent (50%, 0.5 ml; Sigma Chemical Co.) was colour of the seed oil from the fermentation process added to each sample. After 5 min, 1 ml of 5% Na2CO3 was darker following Soxhlet extraction. This dark (Fisher Scientific) was added mixed with a vortex colour could be from browning reaction products and/ mixer, and allowed to stand for 60 min in darkness. or lignin degradation products. The browining reaction Samples were again homogenized with a vortex mixer or non-enzymic browning, called the Maillard reaction, and absorbance was measured at 725 nm. A standard might have occurred when seeds were boiled and during curve was prepared using gallic acid (Fisher Scientific) fermentation. The seeds contained sufficient levels of in 95% ethanol. The total phenolic content was ex- protein, lipid and carbohydrate, which also contained pressed as mg g−1 of DW of phenolic extracts. reducing sugars. Heat treatment of seeds has generally N. Andarwulan et al. / Process Biochemistry 35 (1999) 197–204 201

Table 3 Tocotrienol content in Pangium edule Reinw. seed during fermentationa

Fermentation g-T3 (ug g−1 fw a-T3 (ug g−1 fw (day)freeze dried sample) freeze dried sample)

0 69.8 Trace 20 97.0 ND 40 123.3 ND

a Each value is a mean of two independent assays.

The fatty acid composition appeared to be stable indi- Fig. 2. Changes in fatty acid content and acid value of lipid fraction cating that oxidative protection inside the seed was of Pangium edule Reinw. seed during fermentation. present during fermentation. The dominant tocol, g-tocotrienol, doubled (69.78– resulted in colour or flavour modification because of 123.25 mgg−1 freeze-dried seed) during fermentation Maillard reaction [19]. The reaction products vary from (Table 3). The synthesis of g-tocotrienol during fermen- low to high molecular weight compounds which in- tation may result from the activities of microorganism. cludes compound like melanoidin (dark red to brown g-tocotrienol in seed may have antioxidant activity colour [20]). The water content of seeds during fermen- which protected lipid and/or fatty acid from oxidation. tation was 62.65–65.10%, and it was predicted that Rancid or off flavour was not present in fermented water activity (Aw) supported the Maillard reaction. seed. Lignin degradation products were phenolic compounds Total protein content and amino acid composition which also varied, from low to high molecular weight (Fig. 3 and Table 4) did not change, but soluble protein and several have dark colours [21]. Lignin could be decreased while non-soluble protein increased (Fig. 3). degraded by microorganisms because of lignocellulolitic The increasing non-soluble protein could be an indica- enzyme activity. tor of microbial growth [22]. In this case, microorgan- Fatty acid content decreased (836.1–786.5 mg g−1 isms might utilize soluble protein for growth. Glutamic lipid) and free fatty acid, which is expressed as acid acid was the dominant amino acid in the fermented value, increased (0.22–2.68 mg KOH g−1 lipid) (Fig. seed, and this was one of flavour sources for spices 2). These results did not correlate with lipid content. (Table 2). The decrease of fatty acid content in the lipid fraction Microbial growth could also be monitored through and the increase in free fatty acid in seeds indicated that degradation of carbohydrate [22]. During fermentation, during fermentation hydrolysis reactions due to lipase total crude carbohydrate content and neutral detergent activity of microorganisms may have occurred. fibre (NDF as cellulose, hemicellulose, and lignin) de- Minor fatty acids in the lipid fraction (Table 2) creased while reducing sugar increased and starch con- slightly decreased, but the dominant fatty acids (oleic tent did not change (Fig. 4). The increase in reducing and linoleic acids) did not change during fermentation. sugar was not equal to the decrease of total carbohy- Table 2 drate content or NDF. It may be that total carbohy- Fatty acids composition of Pangium edule Reinw. seed during drate, including cellulose was partially degraded and fermentationa

Fatty acid (%) Fermentation (day)

02040

C14:0 NDND ND

C16:0 7.58 8.11 8.08 C16:1 0.93 0.15 0.14

C18:0 3.52 3.14 3.09

C18:1 43.71 42.13 42.55

C18:2 41.1542.81 42.61

C18:3 3.15 2.88 2.84 C20:0 0.22 0.24 0.21 C20:1 0.40 0.38 0.37 C22:0 0.17 0.11 0.13 Fig. 3. Changes in protein content of Pangium edule Reinw. seed a Each value is a mean of two independent assays. during fermentation. 202 N. Andarwulan et al. / Process Biochemistry 35 (1999) 197–204

Table 4 Amino acids composition of Pangium edule Reinw. seed during fermentationa

Amino acid (% db) Fermentation (day)

0 20 40

Aspartic acid2.6 2.5 2.2 Glutamic acid 4.14.4 4.3 Serime1.0 1.1 1.1 Histidine 0.40.3 0.1 Glysine0.9 0.8 0.8 Threonine 0.60.9 0.8 Arginine2.6 2.4 2.4 Tyrosine 0.20.2 0.2 Methionine0.1 0.1 0.2 Fig. 5. b-glucosidase and guaiacol peroxidase activity in Pangium Valine0.9 0.9 0.9 edule Reinw. seed during fermentation. Phenylalanine 1.31.4 1.4 Isoleusine0.6 0.7 0.7 correlated with the increase in reducing sugar during Leucine 1.41.5 1.5 fermentation. Since Pangium edule seed is known to be Lysine0.6 0.5 0.5 Total (% db) 19.3 19.8 19.4 toxic because of the presence of cyanogenic glucosides; this enzyme may also facilitate the degradation of these a Each value is a mean of two independent assays. toxins in the fermented seed [23]. The enzyme present in the seed may be from microbial activity. Fungi such as became soluble but did not completely release all the Aspergillus sydowi and Fusarium equiseti contain lina- reducing sugars. Another possibility was that the reduc- marase for cyanogenic glucoside detoxiﬁcation [24]. ing sugars reacted with amino compounds in Maillard b-glucosidase also has polygalacturonase (pectinase) ac- reactions during seed fermentation. Reducing sugars tivity [25], and this activity may soften the texture seed may also be released from conjugated phenolics. during fermentation. The speciﬁc enzyme activity could also be used for In addition to the Maillard reaction, the increasing monitoring microbial growth. The activity of b-glucosi- activity of peroxidase might also be correlated with dase, an enzyme that breaks glycosidic bond of conju- darkening of fermented seeds and growth of microor- gated phenolics, and potential phenolic condensation ganisms. Guaiacol peroxidase in plant has the potential enzyme, peroxidase increased during fermentation (Fig. to convert free phenolics to polymerized derivatives like 5). The increase in activity of b-glucosidase may be lignans and lignins. This enzyme, from microbial fer-

Fig. 4. Changes in carbohydrate content of Pangium edule Reinw. seed during fermentation. N. Andarwulan et al. / Process Biochemistry 35 (1999) 197–204 203 mentation, could catalyze the polymerization of pheno- hydrates were partially hydrolysed and may have con- lic aglycones released by b-glucosidase activity as well tributed to the fermentation process and microbial as from lignin degradation products. growth. Total phenolics substantially increased without The total phenolics in fermented seeds increased sub- concurrent increase in antioxidant activity. Phenolics stantially during fermentation but antioxidant activity may contribute partially to oxidation stability, flavour of the phenolic extracts did not change (Fig. 6). The and to some extent antimicrobial activity. This will be increasing total phenolics due to b-glucosidase activity further explored in on-going studies. (Fig. 5) and the antioxidant activity indicated that antioxidant protection might not be linked only to phenolics but also to other non-phenolic metabolises. Acknowledgements Seeds were treated at high temperature by boiling for 3 h prior to processing and therefore an enzymic re- We would like to thank Indonesian Government sponse from these seeds were unlikely. Likewise during (URGE) exchange program, the Indonesian Cultural post-harvest processing there was no seed germination Foundation, and Directorate General of High Educa- and therefore it is unlikely that the increase in b-glu- tion, Department of Education and Culture, the Re- cosidase, total phenolics and guaiacol peroxidase was public of Indonesia for grants supporting N.A. due to the endogenous activity from the seeds. Enzymic activity may therefore be due to natural fermentation References by microorganisms. Preliminary studies in this regard have identified certain bacteria and fungi that may be [1] Burkill IH. A dictionary of the economic products of the Malay associated with increased enzyme activities and total Peninsula, 1935;2:1653–1654. phenolics (data not shown). [2] Fardiaz D, Romlah S. Antioxidant activity of picung (Pangium edule Reinw.) seed. In: Development of Food Science and Tech- nology in ASEAN. Proceedings of the Fourth ASEAN Food Conference, February 17–21, Jakarta, 1992:814–819. 4. Conclusion [3] Puspitasari-Nienaber NL, Aitzetmuller K, Werner G. Analytical investigation on Pangium edule seed oil, Langfassung Pub- likation Stand 20.12.94 (Poster unter Nr. 98025), 1994. Natural microbial fermentation of Pangium edule [4] Andarwulan N, Fardiaz D, Wattimena GA, Shetty K. Antioxi- Reinw. seeds following postharvest heat treatment was dant activity associated with lipid and phenolic mobilization suspected based on enhanced activity of selected en- during seed germination of Pangium edule Reinw. J. Agric. Food zymes linked phenolic mobilization. During fermenta- Chem. 1999 (in press). tion processing metabolites including lipids, protein, [5] AOAC Official Methods of Analysis, 11th ed, AOAC, Washing- ton, DC, 1975. and carbohydrates were mobilized. Among the fatty [6] Budin JT, Breene WM, Putnam DH. Some compositional prop- acids present in lipids, oleic and linoleic acids did not erties of camelina (Camelina sati6a L. Crantz) seeds and oils. J change but the concentration of the antioxidant, g-to- Am Oil Chem Soc 1995;72:309–15. cotrienol increased. Glutamic acid was the major amino [7] Oomah BD, Kenaschuk EO, Mazza G. Tocopherols in flaxseed. acid that was hydrolyzed from protein, which may J Agric Food Chem 1997;45:2076–80. [8] Brzuskiewicz LM, Analysis of tocopherols and tocotrienols using contribute to flavour. Among other metabolites, carbo- HPLC with UV-Vis photodiode-array detection. MS Thesis, Department of Biochemistry at the University of Massachusetts, Amherst, 1989 [9] Roberstson JB, van Soest PJ, The detergent system of analysis and its application to human food. In The analysis of Dietary Fiber in Food. In: James WPT, Theander O. Marcel Dekker, New York, 1981:123–56. [10] Miller GL. Use of dinitrosalicylic acid reagent for determination of reducing sugar. Anal Chem 1959;31:426–8. [11] Bradford M. A rapid and sensitive method for quantification of microgram quantities of protein utilizing the principle of protein- dye binding. Anal Biochem 1976;72:248–54. [12] Laloue H, Weber-Lofti F, Lucau-Danila A, Guillemat P. Iden- tification of ascorbate and guaiacol peroxidase in needle chloro- plasts of spruce trees. Plant Physiol Biochem 1997;35:341–6. [13] George P. Intermediate compound formation with peroxidase and strong oxidizing agents. J Biol Chem 1953;201:413–51. [14] Chandler SF, Dodds JH. The effect of phosphate, nitrogen, and sucrose on the production of phenolics, and solasodine in callus cultures of Solanum laciniatum. Plant Cell Rep 1983;2:105–8. [15] Singleton VL, Rossi JA Jr. Colorimetry of total phenolics with Fig. 6. Changes in total phenolics and antioxidant activity of phenolic phosphomolybdic-phosphotungstic acid reagents. Am J Enol extracts of Pangium edule Reinw. seed during fermentation. Vitic 1965;16:14458. 204 N. Andarwulan et al. / Process Biochemistry 35 (1999) 197–204

[16] Shetty K, Curtis OF, Levin RE, Witkowsky R, Ang W. Pre- Maillard Reaction in Chemistry, Food, and Health, The Royal vention of vitriﬁcation associated with in vitro shoot culture of Society of Chemistry, Cambridge, 1994:11–19. oregano (Oreganum 6ulgare)byPseudomonas spp. J Plant [21] Crawford RL. Lignin Biodegradation and Transformation. Physiol 1995;147:447–51. Toronto: John Wiley and Sons, 1991. [17] Hammerschmidt PA, Pratt DE. Phenolic antioxidants of dried [22] Carvalheiro F, Roseiro JC, Collaco MAT. Biological conver- soybeans. Food Sci 1978;43:556–9. sion of tomato pomace by pure and mixed fungal cultures. [18] Wanasundara U, Amarowicz R, Shahidi F. Isolation and iden- Process Biochem 1994;29:601–5. tiﬁcation of an antioxidative component in canola meal. J [23] Ikediobi CO, Onyike E. The use of linamarase in gari produc- Agric Food Chem 1994;42:1285–90. tion. Process Biochem 1982;17:2–8. [19] De Bry L. Anthropology of the Maillard reaction. In: [24] Ikediobi CO, Ogundu EC, Ukoha AI. Production of linama- Labuza TP, Reineccius GA, Monnier VM, O’Brien J, Baynes rase by Aspergillus sydowi and Fusarium equiseti. Process JW editors. Maillard Reaction in Chemistry, Food, and Biochem 1985;20:99–102. Health, The Royal Society of Chemistry, Cambridge, [25] Brimer L, Cicalini AR, Federici F, Petruccioli M. Beta-glucosi- 1994:2836. dase as a side activity in commercial pectinase preparations of [20] Rizzi G, The Maillard reaction in foods. In: Labuza TP, fungal origin. The hydrolysis of cyanogenic glycosides. Ital J: Reineccius GA, Monnier VM, O’Brien J, Baynes JW, editors. Food Sci 1995;4:387–94.