World Journal
of Microbiology
& Biotechnology
12. 629-635
Selection of starter cultures for the production of kinema, a fermented soybean food of the Himalaya J.P. Tamang”
and S. Nikkuni
Kinema, a traditional fermented soybean food, serves as a low-cost source of protein in the diet of the people of the Himalaya. The traditional method of kinema preparation results in a product with inconsistent quality. Forty five strains of spore-forming bacteria were isolated from nine samples of kinema collected from markets in the Darjeeling hills and Sikkim in India. Of these, 10 strains, identified as Bacillus subtilis (Ehrenberg) Cohn, were selected as possible starter cultures on the basis of enzyme activities and the production of slimy material. Protease activity (U/ml), a-amylase activity (U/ml) and relative viscosity of the selected strains were 6.5 to 81.5, 0.1 to 9.3, and 1.1 to 20.1, respectively. Kinema produced by these strains showed nitrogen contents (expressed as a percentage of the total nitrogen content) of: water-soluble N, 48.4 to 76.5; TCA-soluble N, 16.0 to 27.6: formal N, 5.0 to 12.5; ammonia-N, 4.4 to 7.8%. Reducing sugar was 1.1 to 2.0% of the wet weight and relative viscosity was from 2.0 to 35.2. Correlation matrices of the biochemical parameters and sensory attributes of the kinema produced by these B. subtilis strains were statistically analysed. Strains KK-2:BlO and GK-2:BlO of B. subtilis were the best starter cultures for improved kinema production. Key words: Bacillus subtilis, fermentation,
kinema, soybean, starter culture.
Kinema is an indigenous non-salted fermented soybean food commonly consumed as a low-cost source of plant protein by the people of the Himalayan regions of the Darjeeling hills, Sikkim and north-eastern hills in India, Nepal and Bhutan. During the traditional method of kinema preparation, soybeans are cleaned, soaked overnight, cooked until they can be crushed easily, wrapped in leaves and fermented naturally for I to 3 days at ambient temperature. This product is considered to be of good quality if stickiness occurs when beans are pulled apart and the characteristic flavour has developed (Tamang et al. 1988; Tamang 1995). Kinema is similar to the Japanese natto which is also a non-salted sticky fermented soybean product. Kinema samples collected from markets of the Darjeeling hills and Sikkim showed the presence of Bacillus subtilis,
Enterococcus fuecium, Candida parapsilosis and Geotrichum candidurn (Tamang 1993; Sarkar et al. 19%). The influence of these isolates on monoculture fermentation of kinema was studied and organoleptically it was observed that B. strbfilis was the main fermenting microorganism for kinema preparation and that other organisms have no role in fermentation (Sarkar & Tamang 19%). A considerable variation in the quality of traditionally prepared kinema due to varied fermentation time, temperature and microbial profile results in an inconsistent product. The main objective of this research was to improve the traditional method of kinema preparation, by selecting the most appropriate microorganism for the fermentation.
Materials J.P. Tamang was and S. Nikkuni is with the National Food Research Institute, Ministry of Agriculture. Forestry and Fisheries, 2-l-2 Kannondai, Tsukuba, lbaraki 305. Japan. J.P. Tamang is now with the Microbiology Research Laboratory, Department of Botany, Sikkim Government College, Gangtok, Sikkim 737 102. India; fax: 091-3592-22707. * Corresponding author. @ 1996 Rapid Science
and Methods
Samples Fresh kinema samples were collected from different periodical markets, locally called huat, at Kalimpong, Darjeeling markets in the Darjeeling hills, and Gangtok market in Sikkim, India. Samples
Publishers WorldJournal
of Microbiology 6 tlmtechnology, Vol 12. 1996
629
].P. Tamang and S. Nikkuni were sun dried for 3 days. Dried samples were homogenized liquid nitrogen and stored at - 20°C.
in
orgunisms Bacillus
subfilis (naffo) Miura strain was obtained from Miyagino Natto Co., Sendai, Japan. Type strains B. subfilis JCM 1465 (ATCC 6051) and B. pumilus JCM 2508 (ATCC 7061) were obtained from the Japanese Collection of Microorganisms (JCM), RIKEN, Wako-shi, Japan. Microbial Analysis Homogenized sample (1 g) was mixed with 99 ml of 0.85% (w/v) sterile physiological saline and shaken for 10 min. For spore-forming bacteria, I ml of diluted sample was mixed with 9 ml of sterile physiological saline and heated for 2 min in continuously boiling water. Serial dilutions were prepared in physiological saline, and I ml of appropriate diluted suspension was mixed with molten Trypticase Soy agar (BBL I 1043, Microbiology System, Cockeysville, USA) for total viable count and spore-forming bacteria, and Briggs agar (Nissui Pharmaceutical Co., 05624, Japan) supplemented with I% calcium carbonate and 60 ppm bromocresol purple for lactic acid bacteria. Trypticase agar plates were incubated at 30°C for 48 h and Briggs agar plates were incubated in a Gas-Pak Anaerobic System (BBL) at 3O’C for 72 h. Five colonies of spore-forming bacteria were picked randomly from each of nine kinema samples and maintained on nutrient agar (Difco) at 4’C. Selected strains were deposited in the Applied Microbiology Division of National Food Research Institute, Tsukuba, Japan and the Culture Collection of the Microbiology Research Laboratory, S&m Government College, Gangtok, India. Phenofypic Characterization Colonies of young cultures were Gram-stained (Bartholomew 1962) and cell morphology and motility were examined using a phase contrast microscope (Nikon OPTIPHOT, Japan). To assay for biotin requirement, cultures were grown on sucrose glutamate agar plates containing (g/l): sucrose, SO; monosodium glutamate, 15; potassium dihydrogen phosphate, 2.7; disodium hydrogen phosphate, 4.2; sodium chloride, 0.5; magnesium sulphate, 0.5; agar, 20.0 and 0.1 pg biotin/ml. Plates were examined for growth in the presence and absence of biotin (Hara & Ueda 1982). All other phenotypic characterizations were carried out as described by Gordon et al. (1973). and Claus & Berkeley (1986). DNA Isolation and DNA Base Composifion Chromosomal DNA was isolated and purified according to Saito & Miura (1963). The G + C content of the DNA was determined by HPLC (model 800, JASCO, Japan) using Crest Pak Column (model I BT-5, JASCO, Tokyo, Japan) according to the DNA GC kit (Yamasa Shoyu Co., code no. 450355, Chosi, Japan).
Preliminary Screening Cultures were streaked on surface-dried plates of milk agar and starch agar (Gordon et al. 1973), and incubated at 37°C. After 3 days plates were flooded with iodine solution and the diameters of the clearing zones were used as an assessment of proteolytic (milk agar) and amylolytic (starch agar) activities. Cultures grown on phytone agar (Nagai et al. 1994) at 37°C for 24 h were pulled by touching with an inoculating needle and the stickiness measured by the length of the thread.
37°C for 72 h on a rotary shaker at 180 rev/min and centrifuged at 13,000 X g for 10 min. The enzyme solution diluted to an appropriate concentration and the substrate solution containing 1% azocasein (Sigma, St Louis, MO, USA) in 0.1 M phosphate buffer @H 6.8) were preincubated separately at 37°C. The enzyme reaction was started by mixing the two solutions and the mixture was incubated at 37°C for 20 min. The reaction was quenched by addition of 2.5 ml of 10% (w/v) trichloroacetic acid. After centrifugation at 13,000 X g for 10 min, 2 ml of supematant was neutralized with an equal volume of 1 N NaOH and the absorbance measured at 450 nm. One unit of protease activity was defined as the quantity required to increase the absorbance by 0.1 under the above conditions. a-Amy/use Activity The blue value method of Fuwa (1954) as modified by Kawaguchi ef al. (1992) was followed for determination of a-amylase activity of the cultures. One unit of cl-amylase activity was defined as the amount of a-amylase which produced a 10% reduction in the intensity of blue colour. Viscosity
Cultures were grown in phytone broth (Nagai et al. 1994) at 37°C for 72 h on rotary shaker at 180 rev/min and centrifuged at 13,000 X g for 10 min. The supematant was taken and viscosity was measured directly by a micro-viscometer (DIN 12879, Haake, Germany) using a 2 mm diameter metal ball at 30°C. For fresh kinema, 30 g was suspended in 30 ml distilled water (about 60”(Z), shaken for 10 min and centrifuged at 13,000 X g for 15 min. Supematant (9 ml) was adjusted to 10 ml with distilled water and viscosity measured as described above. Preparation of Kinema by Sfarfer Culture Samples of the yellow variety of soybean [Glycine max (L.) Merrill] were cleaned and soaked in water overnight. The soaked beans were steamed at 121°C for 40 min and inoculated with selected starter cell suspensions, harvested in nutrient broth (Difco), at 10’ c.f.u./g of cooked beans while the temperature of the beans was above 80°C. Approximately 80 g inoculated beans were put into a polystyrene paper package with dimensions of 10.5 cm X 10.5 cm X 3.0 cm and covered with perforated polyethylene film. The packages were incubated at 40°C with 90% relative humidity for 20 h. Chemical
Analysis
Moisture content was calculated by drying the well-mixed sample at 130 !Z 1°C to constant weight (AOAC 1990). Total nitrogen, water-soluble nitrogen and trichloroacetic acid (TCA)-soluble nitrogen contents were determined according to the Methods of Natto Research (1990) using an auto-Kjeldahl nitrogen analyser (model KN-01, Mitsubishi Kasei Co., Tokyo, Japan). Form01 nitrogen content of the sample was determined by form01 titration (Methods of Natto Research 1990) using auto-Titrator (model TSC-IOA, TOA Co., Tokyo, Japan) and auto-Burette (model TSB-IOA, TOA). Ammonia-nitrogen content of the sample was determined using the ammonia determination kit (Boehringer Mannheim, Germany). Reducing sugar was estimated by the calorimetric method (Somogyi 1945) using glucose as standard solution. Sensory Analysis
Acfivity Protease activity was measured by a modification of the method of Maeda et al. (1993). Cultures were grown in phytone broth at Profease
Kinema samples produced by selected starter cultures were organoleptically evaluated by a panel of 10 judges using a score range of I (bad) to 5 (good) and using natto produced by Bacillus subfilis
Enema starter culture Table 1. markets.
Microbial
load
of
kinema
samples
collected
from
c.f.u.lg
Sample’
Spore-forming bacteria
Total viable count
2.6 x 1Oa
KK DK GK
1.7 x 10’ 3.7 x 10’
’ KK, Kalimpong Kinema. Data source.
Kinema; represent
1.4 x 100 1.7 x 10’ 1.8 x 10”
Lactic acid bacteria 9.3 x 10’ 6.5 x 10’ 2.0 x 10’
DK, Darjeeiing Kinema; GK, Gangtok the means of 3 samples from each
(natto)Miura strain as a control with a score of 3 (moderate). Data obtained were analysed statistically by determining ANOVA using the least square design as described by Snedecore & Cochran (1980) and the SALS software package (version 2.5) at Computer Centre of Agriculture, Forestry and Fisheries Research Secretariate, Tsukuba, Japan.
Results
and Discussion
Microbial Analysis Nine kinema samples collected from different markets were analysed for microbial load. The average total viable count in dried kinema was 2.7 X 10’ c.f.u./g, spore-forming bacteria 1.6 X 10” c.f.u./g, and lactic acid bacteria 5.9 X IO’ c.f.u./g (Table I). Based on the previous finding that B. subtilis was a sole fermenting organism in kinema preparations (Sarkar & Tamang 1994, a total of 45 strains of spore-forming bacteria were isolated from dried kinema samples for the selection of starter cultures for kinema production. Preliminary Screening On the basis of high proteolytic activity ( > 5 mm clearing zones in milk agar plates), amylolytic activity ( > 4 mm clearing zones in starch agar plates), and the ability to produce more stickiness ( > 15 cm) on phytone agar, 10 of the 45 spore-forming strains were selected for further analysis (Figure 1). Idenkification After characterization (Table 2) of the 10 selected strains they were identified as Bacillus subtilis (Ehrenberg) Cohn according to the criteria laid down by Claus & Berkeley (1986). However, with respect to spore position and reduction of nitrate they differed from B. subtilis (natto) Miura and B. subtilis JCM 1465. Another closely related species, B. pumilus, differed from them with regard to spore position, starch hydrolysis and growth at 50°C. The DNA G + C content of the kinema strains illustrated their closeness to type strain B. subtilis. Kinema strains of B. subtilis showed
> 15
0
10
20 Number
30
40
of strains
Figure 1. Preliminary selection of spore-forming bacterial isolated from market kinema on the basis of proteolytic amyloiytic activity and stickiness.
strains activity,
similarity with B. subtilis (natto) Miura strain with regard to the production of stickiness on phytone agar and cooked soybeans, and their requirement of biotin for growth. Similar observations were made on strains of B. subtilis isolated from fermented soybean products from Thailand and China (Hara et al. 1986; Hara et al. 1993).
World Journal
of
Mmbiology
6 Bmfechtwlogy. Vol 12. 1996
631
].l?
and S. Nikkuni
Tamang
Table similar
2. Differential species.
characterlstlcs
of spore-formlng
Characteristic Colony
Kinema strain (n = 10)
shape
Irregular (5) Round (5) 0.7-0.9 2.0-4.0 C/PC + (7) - (3)
Cell width km) Cell length (pm) Spore position Acid from D-xylose Hydrolysis of starch NO, reduction Growth at 50°C Biotin requirement Stickiness on phytone G + C cant (mol %)
agar
bacterial
strains
B. subtilis (natfo) Miura
isolated
from
B. subtilis JCM 1485
klnema
samples,
and
B. pumilus JCM 2508
Round
Irregular
Irregular
0.649 2.c3.0
0.649 2.0-3.0
0.6-0.7 2.0-3.0
C
C
C
+
+
+
+
+
-
-;6,
+
+
-
+ (4) + + + 42.8-t3.2
+
+
+ + 41.2-42.08
-
+ -
43.lb
40.7b
n = number of strains; C/PC, central/pars-central. All isolates were rods, Gram-stained, catalase positive, aerobic, showed motility, produced acid from o-glucose. L-arabinose and o-mannitol but no gas from o-glucose; hydrolysed casein; grew well at pH 5.7 and 6.8, at 7% NaCI, no growth at 55-60X. a Data from lkeda et a/. (1965); b Data from Fahmy eta/. (1985).
Table strains strain.
3. Enzyme isolated
activities and relative from market klnema,
Strain KK-2:B6 KK-2:BlO KK-3:B14 DK-l:B2 DK-2:B9 DK-2:BlO GK-l:B5 GK-2:B9 GK-2:BlO GK-3:B13 6. subtitis 8. subtilis
Protease (U/ml)
(natto) Miura JCM 1465
Data represent
the means
10.5 27.4 81.5 6.5 17.0 16.4 17.5 16.0 23.0 21.4 48.0 2.6 of duplicate
viscosity of B. subtilis natto straln and type
a-amylase (U/ml) 2.0 3.0 9.3 3.0 0.7 3.8 1.1 3.1 0.1 0.2 0.1 1.0
Relative viscosity 1.1 1.4 20.1 1.1 1.7 1.1 1.1 1.1 1.3 1.1 3.0 1.0
experiments.
Biochemical Analysis Table 3 shows the protease activity, a-amylase activity and relative viscosity of B. subtilis strains. The KK-3:B14 strain showed higher protease activity (U/ml) and relative viscosity than the natto strain. However, kinema strains showed high a-amylase activities (U/ml) than that of B. strbtilis (natfo) Miura strain. Kinema was prepared using starter cell suspensions of the B. subtilis strains. Kinema produced by strains KK-2:BIO and KK-3:B14 had a higher proportion of their total nitrogen as water-soluble, TCA-soluble and formol forms while kinema produced by strains KK-2:BIO, KK-3:B14, GK-2:Bo
and GK3:BIO showed high viscosity (Table 4). There was significant correlation at the 5% level between protease activity of the strains and the water-soluble nitrogen content of kinema produced and also between protease activity of strains and viscosity of the product and water-soluble nitrogen and formol nitrogen contents at the 1% level (Table 5). These data suggest that water-soluble nitrogen increases with high protease activity of B. stlbtilis strains in kinema fermentation. Sensory Analysis Table 6 shows the sensory analysis of kinema produced by 10 selected strains of B. subtilis. There was no significant difference (P < 0.05) in colour and aroma attributes among the IO samples. However, a significant difference (P c 0.05) in stickiness, texture, taste and general acceptability attributes was observed among some samples. There was no significant difference (P < 0.05) in general acceptability of kinema produced by strains KK-2:B10 and GK-2:BIO and that of natto produced by a starter culture of B. subtilis (natto) Miura strain. Although strain KK-3:B14 showed high protease activity and viscosity, organoleptically the general acceptability score of the kinema produced by it differed significantly (P < 0.05) from kinema produced by the KK2:BlO and GK-2:BlO strains. Since strain KK-3:B14 had a wrinkled dull, rough colony, the kinema produced by it did not have a good appearance. The appearance of B. sttbtilis is an important attribute for the stickiness score. There was a significant correlation at the 0.1% level between general acceptability and stickiness of kinema, and between general acceptability and taste of the product
Kinema starter culture Table
4. Chemical
analysis
and viscosity
of kinema
produced
Strain
KK-2:B6 KK-2:BlO KK-3:Bl4 DK-l:B2 DK-2:B9 DK-2:BlO GK-I:85 GK-2:B9 GK-2:BlO GK-3:Bl3 Ef. subtilis 6. subtilis
% Total
(natto) Miura JCM 1465
WSN, water-soluble ammonia nitrogen; ments.
Table 5. Correlation matrlx viscosity of klnema produced from market klnema. Protease Protease WSN/TN FN/TN Viscosity
nitrogen; RS, reducing
TCA-N
66.0 76.5 75.6 48.4 54.6 66.3 65.7 56.5 62.5 64.4 84.6 36.1
24.5 27.6 21.4 16.0 19.6 27.0 23.5 20.7 19.5 20.4 30.6 10.0
TCA-N, sugar;
trichloroacetic RV, relative
cultures
of 6. subtilis
strains.
nitrogen
acid viscosity.
FN
NH3-N
RS%
RV
9.5 12.5 10.5 7.2 5.0 10.6 9.0 a.3 7.0 a.5 12.6 1.5
6.2 6.6 7.1 6.8 4.4 7.8 5.8 5.1 4.4 6.8 6.8 0.7
1.5 2.0 1.6 1.1 1.4 1.5 1.2 1.2 1.1 1.1 2.0 1.7
3.0 30.0 35.2 2.0 5.0 2.4 2.1 24.7 30.5 2.6 33.7 1.2
soluble nitrogen; Data represent
FN, formol nitrogen; the means of duplicate
NHa-N, experi-
3
of protease, nitrogen contents and by 10 stralns of 6. subtilis Isolated
. 2 1
WSN/TN
1.000 0.632* 0.363 0.655
n = 10. ‘Significance water-soluble nitrogen to total nitrogen.
WSN
by starter
1.000 0.824’* 0.476 at 5%. to total
FN/TN
.*
vlscoslty
1 .ooo
0.311
** Significance at 1%. nitrogen; FN/TN. formol
1.000 WSN/TN, nitrogen
CoelIicient . 1
(Table 7). The quality of kinema is mainly judged by its stickiness and flavour (Sarkar & Tamang 1994, which are almost the same attributes as in natto (Akimoto et al. 1990). The data also showed a significant correlation at the I% level between the taste and stickiness attributes of kinema. This datum helps to formulate the sensory quality of kinema, by considering stickiness and taste as the most important attribute indices for general acceptability of the product. There was a significant correlation at the 5% level between relative viscosity of kinema and general acceptability scores (Figure 2) This indicates that increased viscosity of kinema scored high for sensory evaluation. A similar correlation between sensory attributes and stickiness was observed for natto (Akimoto et al. 1990). On the basis of biochemical analysis and sensory evaluation of kinema produced separately by 10 selected strains of B. subfilis, strains KK-2:B10 and GK-2:BlO were considered to be the best starter cultures for the production of kinema.
. .
of correlation=0.652*
I* 2
Sensory
3
4
5
scores
Figure 2. Correlation between relative viscosity and sensory evaluation (general acceptability) of kinema produced by 10 strains of B. subtilis isolated from market kinema. 1, KK-2:BlO; 2, GK-2:B10; 3, KK-3:Bl4. n = 10. * Significance at 5%.
Acknowledgement The present work forms part of the post-doctoral research of Dr J.P. Tamang at the National Food Research Institute, Tsukuba, sponsored by the United Nations University-Kirin Brewery Co. fellowship (1994-95). The authors are grateful for the financial support of the UNU-Kirin Brewery of Japan.
References AOAC
1990 Official Methods of Analysis 15th edn. Virginia: of Official Analytical Chemists. Akimoto, T., Yamada, S. & Matsumoto, I. 1990 The relation Association
World ~oumal of Mzcrobiology 6 Biotechndogy, Vol 12, 1996
633
J.P. Tamang Table
and S. Nikkuni
6. Sensory
evaluation
Strain
oi kinema
produced
Coiour
by starter
cultures
Aroma
of 8. subtilis
strains.
Stickiness
KK-2:B6
3.0
f
o.ooa
2.4 f
0.97a
1.1. f
KK-2:BlO KK-3:B14
3.0 2.9
f 0.00" + 0.338
2.5 f 2.5 f
0.97a 0.71a
3.6 f 1.5 f
DK-l:B2 DK-2:B9 DK-2:BlO
2.6 f 0.73" 2.9 f 0.32a 2.8 + 0.428
2.4 + 0.73a 2.3 f 0.68a 2.1 + 0.998
GK-l:B5 GK-2:B9
2.9 f 0.33a 2.9 f 0.32a
2.7 f 2.9 f
GK-2:BlO GK-3:B13
2.9 f 3.0 f
Texture
2.1 f
0.99a
1.8 f
0.63=
3.0 2.2
f f
0.67b 0.79=
3.1 f 1.9 f
0.74bd 0.74=
1.3 zk 0.4a= 1.7 k 0.68-' 1.9 k 0.3249
2.4 f 0.97de 2.5 IL 0.8tTd'g 2.9 + 0.57ad
1.7 f 2.2 f 2.2 f
0.82= 0.79= 0.92*-
1.6 f 0.848 1.8 k 0.63= 2.1 f 0.74==
0.67a 0.74a
2.1 + 0.32d 2.5 k 0.53e
2.6 2 0.52ad 3.1 f 0.32=
2.1 2.9
f 0.32a + 0.74b
2.1 2 0.32= 2.6 f 0.70bC
0.32= o.ooa
3.0 f. 0.828 2.8 f 0.86a
4.2 + 0.79' 1.9 If: 0.57Q
2.5 f 2.8 f
2.6 f 2.1 f
Natto produced by 6. subtilis k SD (n = 10). Means having
(natto) Miura was the same superscripts
Table
7. Correlation
matrix
of sensory
atiributes
Coiour Coiour Aroma Stickiness Texture Taste General acceptability n = 10.
l
* Significance
between
qualities
protease, of natto.
0.84b 0.53c
of kinema
produced
Aroma
0.9Vd
0.71d'g 0.63-
Texture
1.000
0.277 0.278
0.565 0.120
1.000 0.272
1.000
0.491 0.410
0.476 0.593
0.832'* 0.970***
0.461 0.401
at 1%. '"'Significance
and y-glutamyltranspeptidase Nippon Shikuhin Kogyo
0.7gbC 0.88a
5, good. Values (P -C 0.05).
of B. subtilis
1.000 0.314
isolated
represent
from
market
Taste
3.2 + 0.92d 2.2 f 0.63= the
mean
scores
kinema. General acceptability
1.000 0.906”*
1.000
atO.l%.
activities Gakkaishi 37
and
(II),
155.
D.
by 10 strains
Stickiness
J.W. 1962 Variables influencing results and the precise definition of the steps in Gram staining as a means of standardizing the results obtained. Stain Technology 37, UP-
& Berkeley, R.C.W. 1986 Genus Bacillus Cohn 1872, Manual of Syslematic Bacteriology Volume 2, eds Sneath, P.H.A., Mair, N.S., Sharpe, M.E. & HoIt, J.G. pp 1X051139. Baltimore: Williams & Wilkins. Fahmy, F., Floossdorf, J. & Claus, D. 1985 The DNA base composition of the type strains of the genus Bucillw. Systematic and Applied Microbiology 6, 6(r65. Fuwa, H. 1954 A new method for microdetermination of amylase activity by the use of amylose as the substrate. Journal of Biochemistry 41, 583-603. Gordon, R.E., Haynes, W.C. & Pang, C.H.-N 1973 The genus Bacillus, Handbook No. 427. Washington: United States Department of Agriculture. Hara, T. & Ueda, S. 1982 Regulation of polyglutamate production in Batiks srrbtilis (n&o): transformation of high PGA productivity. Agricuhurul and Biological Chemistry 46, 2275-2281. Hara, T., Chetanachit, C., Fujio, Y. & Ueda, S. 1986 Distribution of plasmids in polyglutamate-producing Bacillus strains isolated from ‘natto’-like fermented soybeans, ‘thua-nao’, in Thailand. ]ournu~ of General and Applied Microbiology 32, 241-249. Hara, T., Ogata, S. & Ueda, S. 1993 Plasmid distribution in y-polyglutamate-producing Bacillus strains isolated from 174. In Bergey’s
f
as a control; score 1, bad; score in each row do not differ significantly
872-877. Bartholomew,
Claus,
2.7
General acceptability
2.9 _+ 0.32=' 2.0 f 20.67be
used
0.32a
Taste
‘dan-do&i’, in China.
a ‘natto’-like non-salty ]ortrnul of General and
fermented soybean Applied Microbiology
food 39,
75-82.
Ikeda, Y., Saito, H., Miura, K., Takagi, J. & Aoki, H. 1965 DNA base composition, susceptibility to bacteriophages, and interspecific transformation as criteria for classification in the genus Bacillus. ]oumul of General and Applied Microbiology 11, 181-190.
Kawaguchi, T., Nagae, H., Murao. S. & Arai, M. 1992 Purification and some properties of a haim-sensitive cc-amylase from newly isolated Bacillus sp. No. 195. Bioscience, Biotechnology, and Biochemistry 56, 1792-1796. Maeda, T., Takagi, M. & Imanaka, T. 1993 Purification and characterization of a new metal protease which hydrolyzes the cyclic decapeptide, gramicidin S. ]ournul of Fermentation and Bioengineering 75, 173-l 77. Methods of Natto Research. 1990 Society for study of natto. Tsukuba: National Food Research Institute. Nagai, T., Nishimura, K., Suzuki, H., Banba, Y., Sasaki, H. & Kiuchi, K. 1994 Isolation and characterization of Bucillw subfilis strain producing natto with strong Umami-taste and high viscosity. Nippon Shokuhin Kogyo Gakkuishi 41, 123-128. Saito, H. & Miura, K. 1963 Preparation of transforming deoxyribonucleic acid by phenol treatment. Biochimu ef Biophysics Acta 72,619-629.
Sarkar, and Food Sarkar,
P.K. & Tamang, J.P. 1994 The influence of process variables inoculum composition on the sensory quality of kinema. Microbiology 111, 317-325. P.K., Tamang, J.P., Cook, P.E. & Owens, J.D. 1994 Kinema
Kinema
-
a traditional soybean fermented food: proximate composition and microflora. Food Microbiology 11, 47-55. Snedecor, G.W. & Cochran, W.G. 1980 Statistical Methods 7th edn. Ames: Iowa State University Press. Somogyi, M. 1945 A new reagent for the determination of sugars. ]oumal of Biologica! Chemistry 160, 61-62. Tamang, J.P. 1993 Studies on the microflora of some truditiona~ fermented foods of Durjeeling hills and Sikkim. PhD. Thesis. Dajeeling: North Bengal University. Tamang, J.P. 19% Study of traditional fermented foods production in
starter
culture
Da+&ng hills and Sikkim, with emphasis on Kinema. Postdoctoral Report. Tokyo: United Nations University. Tamang, J.P., Sarkar, P.K. & Hesseltine, C.W. 1988 Traditional fermented foods and beverages of Dajeeling hills and S&m - a review. Journal of Science of Food and Agriculture 44, 3i’~385.
(Received
in revised
form
28 April
1996;
accepted
4 May
1996)
of Microbiology
& Biotechnology
12. 629-635
Selection of starter cultures for the production of kinema, a fermented soybean food of the Himalaya J.P. Tamang”
and S. Nikkuni
Kinema, a traditional fermented soybean food, serves as a low-cost source of protein in the diet of the people of the Himalaya. The traditional method of kinema preparation results in a product with inconsistent quality. Forty five strains of spore-forming bacteria were isolated from nine samples of kinema collected from markets in the Darjeeling hills and Sikkim in India. Of these, 10 strains, identified as Bacillus subtilis (Ehrenberg) Cohn, were selected as possible starter cultures on the basis of enzyme activities and the production of slimy material. Protease activity (U/ml), a-amylase activity (U/ml) and relative viscosity of the selected strains were 6.5 to 81.5, 0.1 to 9.3, and 1.1 to 20.1, respectively. Kinema produced by these strains showed nitrogen contents (expressed as a percentage of the total nitrogen content) of: water-soluble N, 48.4 to 76.5; TCA-soluble N, 16.0 to 27.6: formal N, 5.0 to 12.5; ammonia-N, 4.4 to 7.8%. Reducing sugar was 1.1 to 2.0% of the wet weight and relative viscosity was from 2.0 to 35.2. Correlation matrices of the biochemical parameters and sensory attributes of the kinema produced by these B. subtilis strains were statistically analysed. Strains KK-2:BlO and GK-2:BlO of B. subtilis were the best starter cultures for improved kinema production. Key words: Bacillus subtilis, fermentation,
kinema, soybean, starter culture.
Kinema is an indigenous non-salted fermented soybean food commonly consumed as a low-cost source of plant protein by the people of the Himalayan regions of the Darjeeling hills, Sikkim and north-eastern hills in India, Nepal and Bhutan. During the traditional method of kinema preparation, soybeans are cleaned, soaked overnight, cooked until they can be crushed easily, wrapped in leaves and fermented naturally for I to 3 days at ambient temperature. This product is considered to be of good quality if stickiness occurs when beans are pulled apart and the characteristic flavour has developed (Tamang et al. 1988; Tamang 1995). Kinema is similar to the Japanese natto which is also a non-salted sticky fermented soybean product. Kinema samples collected from markets of the Darjeeling hills and Sikkim showed the presence of Bacillus subtilis,
Enterococcus fuecium, Candida parapsilosis and Geotrichum candidurn (Tamang 1993; Sarkar et al. 19%). The influence of these isolates on monoculture fermentation of kinema was studied and organoleptically it was observed that B. strbfilis was the main fermenting microorganism for kinema preparation and that other organisms have no role in fermentation (Sarkar & Tamang 19%). A considerable variation in the quality of traditionally prepared kinema due to varied fermentation time, temperature and microbial profile results in an inconsistent product. The main objective of this research was to improve the traditional method of kinema preparation, by selecting the most appropriate microorganism for the fermentation.
Materials J.P. Tamang was and S. Nikkuni is with the National Food Research Institute, Ministry of Agriculture. Forestry and Fisheries, 2-l-2 Kannondai, Tsukuba, lbaraki 305. Japan. J.P. Tamang is now with the Microbiology Research Laboratory, Department of Botany, Sikkim Government College, Gangtok, Sikkim 737 102. India; fax: 091-3592-22707. * Corresponding author. @ 1996 Rapid Science
and Methods
Samples Fresh kinema samples were collected from different periodical markets, locally called huat, at Kalimpong, Darjeeling markets in the Darjeeling hills, and Gangtok market in Sikkim, India. Samples
Publishers WorldJournal
of Microbiology 6 tlmtechnology, Vol 12. 1996
629
].P. Tamang and S. Nikkuni were sun dried for 3 days. Dried samples were homogenized liquid nitrogen and stored at - 20°C.
in
orgunisms Bacillus
subfilis (naffo) Miura strain was obtained from Miyagino Natto Co., Sendai, Japan. Type strains B. subfilis JCM 1465 (ATCC 6051) and B. pumilus JCM 2508 (ATCC 7061) were obtained from the Japanese Collection of Microorganisms (JCM), RIKEN, Wako-shi, Japan. Microbial Analysis Homogenized sample (1 g) was mixed with 99 ml of 0.85% (w/v) sterile physiological saline and shaken for 10 min. For spore-forming bacteria, I ml of diluted sample was mixed with 9 ml of sterile physiological saline and heated for 2 min in continuously boiling water. Serial dilutions were prepared in physiological saline, and I ml of appropriate diluted suspension was mixed with molten Trypticase Soy agar (BBL I 1043, Microbiology System, Cockeysville, USA) for total viable count and spore-forming bacteria, and Briggs agar (Nissui Pharmaceutical Co., 05624, Japan) supplemented with I% calcium carbonate and 60 ppm bromocresol purple for lactic acid bacteria. Trypticase agar plates were incubated at 30°C for 48 h and Briggs agar plates were incubated in a Gas-Pak Anaerobic System (BBL) at 3O’C for 72 h. Five colonies of spore-forming bacteria were picked randomly from each of nine kinema samples and maintained on nutrient agar (Difco) at 4’C. Selected strains were deposited in the Applied Microbiology Division of National Food Research Institute, Tsukuba, Japan and the Culture Collection of the Microbiology Research Laboratory, S&m Government College, Gangtok, India. Phenofypic Characterization Colonies of young cultures were Gram-stained (Bartholomew 1962) and cell morphology and motility were examined using a phase contrast microscope (Nikon OPTIPHOT, Japan). To assay for biotin requirement, cultures were grown on sucrose glutamate agar plates containing (g/l): sucrose, SO; monosodium glutamate, 15; potassium dihydrogen phosphate, 2.7; disodium hydrogen phosphate, 4.2; sodium chloride, 0.5; magnesium sulphate, 0.5; agar, 20.0 and 0.1 pg biotin/ml. Plates were examined for growth in the presence and absence of biotin (Hara & Ueda 1982). All other phenotypic characterizations were carried out as described by Gordon et al. (1973). and Claus & Berkeley (1986). DNA Isolation and DNA Base Composifion Chromosomal DNA was isolated and purified according to Saito & Miura (1963). The G + C content of the DNA was determined by HPLC (model 800, JASCO, Japan) using Crest Pak Column (model I BT-5, JASCO, Tokyo, Japan) according to the DNA GC kit (Yamasa Shoyu Co., code no. 450355, Chosi, Japan).
Preliminary Screening Cultures were streaked on surface-dried plates of milk agar and starch agar (Gordon et al. 1973), and incubated at 37°C. After 3 days plates were flooded with iodine solution and the diameters of the clearing zones were used as an assessment of proteolytic (milk agar) and amylolytic (starch agar) activities. Cultures grown on phytone agar (Nagai et al. 1994) at 37°C for 24 h were pulled by touching with an inoculating needle and the stickiness measured by the length of the thread.
37°C for 72 h on a rotary shaker at 180 rev/min and centrifuged at 13,000 X g for 10 min. The enzyme solution diluted to an appropriate concentration and the substrate solution containing 1% azocasein (Sigma, St Louis, MO, USA) in 0.1 M phosphate buffer @H 6.8) were preincubated separately at 37°C. The enzyme reaction was started by mixing the two solutions and the mixture was incubated at 37°C for 20 min. The reaction was quenched by addition of 2.5 ml of 10% (w/v) trichloroacetic acid. After centrifugation at 13,000 X g for 10 min, 2 ml of supematant was neutralized with an equal volume of 1 N NaOH and the absorbance measured at 450 nm. One unit of protease activity was defined as the quantity required to increase the absorbance by 0.1 under the above conditions. a-Amy/use Activity The blue value method of Fuwa (1954) as modified by Kawaguchi ef al. (1992) was followed for determination of a-amylase activity of the cultures. One unit of cl-amylase activity was defined as the amount of a-amylase which produced a 10% reduction in the intensity of blue colour. Viscosity
Cultures were grown in phytone broth (Nagai et al. 1994) at 37°C for 72 h on rotary shaker at 180 rev/min and centrifuged at 13,000 X g for 10 min. The supematant was taken and viscosity was measured directly by a micro-viscometer (DIN 12879, Haake, Germany) using a 2 mm diameter metal ball at 30°C. For fresh kinema, 30 g was suspended in 30 ml distilled water (about 60”(Z), shaken for 10 min and centrifuged at 13,000 X g for 15 min. Supematant (9 ml) was adjusted to 10 ml with distilled water and viscosity measured as described above. Preparation of Kinema by Sfarfer Culture Samples of the yellow variety of soybean [Glycine max (L.) Merrill] were cleaned and soaked in water overnight. The soaked beans were steamed at 121°C for 40 min and inoculated with selected starter cell suspensions, harvested in nutrient broth (Difco), at 10’ c.f.u./g of cooked beans while the temperature of the beans was above 80°C. Approximately 80 g inoculated beans were put into a polystyrene paper package with dimensions of 10.5 cm X 10.5 cm X 3.0 cm and covered with perforated polyethylene film. The packages were incubated at 40°C with 90% relative humidity for 20 h. Chemical
Analysis
Moisture content was calculated by drying the well-mixed sample at 130 !Z 1°C to constant weight (AOAC 1990). Total nitrogen, water-soluble nitrogen and trichloroacetic acid (TCA)-soluble nitrogen contents were determined according to the Methods of Natto Research (1990) using an auto-Kjeldahl nitrogen analyser (model KN-01, Mitsubishi Kasei Co., Tokyo, Japan). Form01 nitrogen content of the sample was determined by form01 titration (Methods of Natto Research 1990) using auto-Titrator (model TSC-IOA, TOA Co., Tokyo, Japan) and auto-Burette (model TSB-IOA, TOA). Ammonia-nitrogen content of the sample was determined using the ammonia determination kit (Boehringer Mannheim, Germany). Reducing sugar was estimated by the calorimetric method (Somogyi 1945) using glucose as standard solution. Sensory Analysis
Acfivity Protease activity was measured by a modification of the method of Maeda et al. (1993). Cultures were grown in phytone broth at Profease
Kinema samples produced by selected starter cultures were organoleptically evaluated by a panel of 10 judges using a score range of I (bad) to 5 (good) and using natto produced by Bacillus subfilis
Enema starter culture Table 1. markets.
Microbial
load
of
kinema
samples
collected
from
c.f.u.lg
Sample’
Spore-forming bacteria
Total viable count
2.6 x 1Oa
KK DK GK
1.7 x 10’ 3.7 x 10’
’ KK, Kalimpong Kinema. Data source.
Kinema; represent
1.4 x 100 1.7 x 10’ 1.8 x 10”
Lactic acid bacteria 9.3 x 10’ 6.5 x 10’ 2.0 x 10’
DK, Darjeeiing Kinema; GK, Gangtok the means of 3 samples from each
(natto)Miura strain as a control with a score of 3 (moderate). Data obtained were analysed statistically by determining ANOVA using the least square design as described by Snedecore & Cochran (1980) and the SALS software package (version 2.5) at Computer Centre of Agriculture, Forestry and Fisheries Research Secretariate, Tsukuba, Japan.
Results
and Discussion
Microbial Analysis Nine kinema samples collected from different markets were analysed for microbial load. The average total viable count in dried kinema was 2.7 X 10’ c.f.u./g, spore-forming bacteria 1.6 X 10” c.f.u./g, and lactic acid bacteria 5.9 X IO’ c.f.u./g (Table I). Based on the previous finding that B. subtilis was a sole fermenting organism in kinema preparations (Sarkar & Tamang 1994, a total of 45 strains of spore-forming bacteria were isolated from dried kinema samples for the selection of starter cultures for kinema production. Preliminary Screening On the basis of high proteolytic activity ( > 5 mm clearing zones in milk agar plates), amylolytic activity ( > 4 mm clearing zones in starch agar plates), and the ability to produce more stickiness ( > 15 cm) on phytone agar, 10 of the 45 spore-forming strains were selected for further analysis (Figure 1). Idenkification After characterization (Table 2) of the 10 selected strains they were identified as Bacillus subtilis (Ehrenberg) Cohn according to the criteria laid down by Claus & Berkeley (1986). However, with respect to spore position and reduction of nitrate they differed from B. subtilis (natto) Miura and B. subtilis JCM 1465. Another closely related species, B. pumilus, differed from them with regard to spore position, starch hydrolysis and growth at 50°C. The DNA G + C content of the kinema strains illustrated their closeness to type strain B. subtilis. Kinema strains of B. subtilis showed
> 15
0
10
20 Number
30
40
of strains
Figure 1. Preliminary selection of spore-forming bacterial isolated from market kinema on the basis of proteolytic amyloiytic activity and stickiness.
strains activity,
similarity with B. subtilis (natto) Miura strain with regard to the production of stickiness on phytone agar and cooked soybeans, and their requirement of biotin for growth. Similar observations were made on strains of B. subtilis isolated from fermented soybean products from Thailand and China (Hara et al. 1986; Hara et al. 1993).
World Journal
of
Mmbiology
6 Bmfechtwlogy. Vol 12. 1996
631
].l?
and S. Nikkuni
Tamang
Table similar
2. Differential species.
characterlstlcs
of spore-formlng
Characteristic Colony
Kinema strain (n = 10)
shape
Irregular (5) Round (5) 0.7-0.9 2.0-4.0 C/PC + (7) - (3)
Cell width km) Cell length (pm) Spore position Acid from D-xylose Hydrolysis of starch NO, reduction Growth at 50°C Biotin requirement Stickiness on phytone G + C cant (mol %)
agar
bacterial
strains
B. subtilis (natfo) Miura
isolated
from
B. subtilis JCM 1485
klnema
samples,
and
B. pumilus JCM 2508
Round
Irregular
Irregular
0.649 2.c3.0
0.649 2.0-3.0
0.6-0.7 2.0-3.0
C
C
C
+
+
+
+
+
-
-;6,
+
+
-
+ (4) + + + 42.8-t3.2
+
+
+ + 41.2-42.08
-
+ -
43.lb
40.7b
n = number of strains; C/PC, central/pars-central. All isolates were rods, Gram-stained, catalase positive, aerobic, showed motility, produced acid from o-glucose. L-arabinose and o-mannitol but no gas from o-glucose; hydrolysed casein; grew well at pH 5.7 and 6.8, at 7% NaCI, no growth at 55-60X. a Data from lkeda et a/. (1965); b Data from Fahmy eta/. (1985).
Table strains strain.
3. Enzyme isolated
activities and relative from market klnema,
Strain KK-2:B6 KK-2:BlO KK-3:B14 DK-l:B2 DK-2:B9 DK-2:BlO GK-l:B5 GK-2:B9 GK-2:BlO GK-3:B13 6. subtitis 8. subtilis
Protease (U/ml)
(natto) Miura JCM 1465
Data represent
the means
10.5 27.4 81.5 6.5 17.0 16.4 17.5 16.0 23.0 21.4 48.0 2.6 of duplicate
viscosity of B. subtilis natto straln and type
a-amylase (U/ml) 2.0 3.0 9.3 3.0 0.7 3.8 1.1 3.1 0.1 0.2 0.1 1.0
Relative viscosity 1.1 1.4 20.1 1.1 1.7 1.1 1.1 1.1 1.3 1.1 3.0 1.0
experiments.
Biochemical Analysis Table 3 shows the protease activity, a-amylase activity and relative viscosity of B. subtilis strains. The KK-3:B14 strain showed higher protease activity (U/ml) and relative viscosity than the natto strain. However, kinema strains showed high a-amylase activities (U/ml) than that of B. strbtilis (natfo) Miura strain. Kinema was prepared using starter cell suspensions of the B. subtilis strains. Kinema produced by strains KK-2:BIO and KK-3:B14 had a higher proportion of their total nitrogen as water-soluble, TCA-soluble and formol forms while kinema produced by strains KK-2:BIO, KK-3:B14, GK-2:Bo
and GK3:BIO showed high viscosity (Table 4). There was significant correlation at the 5% level between protease activity of the strains and the water-soluble nitrogen content of kinema produced and also between protease activity of strains and viscosity of the product and water-soluble nitrogen and formol nitrogen contents at the 1% level (Table 5). These data suggest that water-soluble nitrogen increases with high protease activity of B. stlbtilis strains in kinema fermentation. Sensory Analysis Table 6 shows the sensory analysis of kinema produced by 10 selected strains of B. subtilis. There was no significant difference (P < 0.05) in colour and aroma attributes among the IO samples. However, a significant difference (P c 0.05) in stickiness, texture, taste and general acceptability attributes was observed among some samples. There was no significant difference (P < 0.05) in general acceptability of kinema produced by strains KK-2:B10 and GK-2:BIO and that of natto produced by a starter culture of B. subtilis (natto) Miura strain. Although strain KK-3:B14 showed high protease activity and viscosity, organoleptically the general acceptability score of the kinema produced by it differed significantly (P < 0.05) from kinema produced by the KK2:BlO and GK-2:BlO strains. Since strain KK-3:B14 had a wrinkled dull, rough colony, the kinema produced by it did not have a good appearance. The appearance of B. sttbtilis is an important attribute for the stickiness score. There was a significant correlation at the 0.1% level between general acceptability and stickiness of kinema, and between general acceptability and taste of the product
Kinema starter culture Table
4. Chemical
analysis
and viscosity
of kinema
produced
Strain
KK-2:B6 KK-2:BlO KK-3:Bl4 DK-l:B2 DK-2:B9 DK-2:BlO GK-I:85 GK-2:B9 GK-2:BlO GK-3:Bl3 Ef. subtilis 6. subtilis
% Total
(natto) Miura JCM 1465
WSN, water-soluble ammonia nitrogen; ments.
Table 5. Correlation matrlx viscosity of klnema produced from market klnema. Protease Protease WSN/TN FN/TN Viscosity
nitrogen; RS, reducing
TCA-N
66.0 76.5 75.6 48.4 54.6 66.3 65.7 56.5 62.5 64.4 84.6 36.1
24.5 27.6 21.4 16.0 19.6 27.0 23.5 20.7 19.5 20.4 30.6 10.0
TCA-N, sugar;
trichloroacetic RV, relative
cultures
of 6. subtilis
strains.
nitrogen
acid viscosity.
FN
NH3-N
RS%
RV
9.5 12.5 10.5 7.2 5.0 10.6 9.0 a.3 7.0 a.5 12.6 1.5
6.2 6.6 7.1 6.8 4.4 7.8 5.8 5.1 4.4 6.8 6.8 0.7
1.5 2.0 1.6 1.1 1.4 1.5 1.2 1.2 1.1 1.1 2.0 1.7
3.0 30.0 35.2 2.0 5.0 2.4 2.1 24.7 30.5 2.6 33.7 1.2
soluble nitrogen; Data represent
FN, formol nitrogen; the means of duplicate
NHa-N, experi-
3
of protease, nitrogen contents and by 10 stralns of 6. subtilis Isolated
. 2 1
WSN/TN
1.000 0.632* 0.363 0.655
n = 10. ‘Significance water-soluble nitrogen to total nitrogen.
WSN
by starter
1.000 0.824’* 0.476 at 5%. to total
FN/TN
.*
vlscoslty
1 .ooo
0.311
** Significance at 1%. nitrogen; FN/TN. formol
1.000 WSN/TN, nitrogen
CoelIicient . 1
(Table 7). The quality of kinema is mainly judged by its stickiness and flavour (Sarkar & Tamang 1994, which are almost the same attributes as in natto (Akimoto et al. 1990). The data also showed a significant correlation at the I% level between the taste and stickiness attributes of kinema. This datum helps to formulate the sensory quality of kinema, by considering stickiness and taste as the most important attribute indices for general acceptability of the product. There was a significant correlation at the 5% level between relative viscosity of kinema and general acceptability scores (Figure 2) This indicates that increased viscosity of kinema scored high for sensory evaluation. A similar correlation between sensory attributes and stickiness was observed for natto (Akimoto et al. 1990). On the basis of biochemical analysis and sensory evaluation of kinema produced separately by 10 selected strains of B. subfilis, strains KK-2:B10 and GK-2:BlO were considered to be the best starter cultures for the production of kinema.
. .
of correlation=0.652*
I* 2
Sensory
3
4
5
scores
Figure 2. Correlation between relative viscosity and sensory evaluation (general acceptability) of kinema produced by 10 strains of B. subtilis isolated from market kinema. 1, KK-2:BlO; 2, GK-2:B10; 3, KK-3:Bl4. n = 10. * Significance at 5%.
Acknowledgement The present work forms part of the post-doctoral research of Dr J.P. Tamang at the National Food Research Institute, Tsukuba, sponsored by the United Nations University-Kirin Brewery Co. fellowship (1994-95). The authors are grateful for the financial support of the UNU-Kirin Brewery of Japan.
References AOAC
1990 Official Methods of Analysis 15th edn. Virginia: of Official Analytical Chemists. Akimoto, T., Yamada, S. & Matsumoto, I. 1990 The relation Association
World ~oumal of Mzcrobiology 6 Biotechndogy, Vol 12, 1996
633
J.P. Tamang Table
and S. Nikkuni
6. Sensory
evaluation
Strain
oi kinema
produced
Coiour
by starter
cultures
Aroma
of 8. subtilis
strains.
Stickiness
KK-2:B6
3.0
f
o.ooa
2.4 f
0.97a
1.1. f
KK-2:BlO KK-3:B14
3.0 2.9
f 0.00" + 0.338
2.5 f 2.5 f
0.97a 0.71a
3.6 f 1.5 f
DK-l:B2 DK-2:B9 DK-2:BlO
2.6 f 0.73" 2.9 f 0.32a 2.8 + 0.428
2.4 + 0.73a 2.3 f 0.68a 2.1 + 0.998
GK-l:B5 GK-2:B9
2.9 f 0.33a 2.9 f 0.32a
2.7 f 2.9 f
GK-2:BlO GK-3:B13
2.9 f 3.0 f
Texture
2.1 f
0.99a
1.8 f
0.63=
3.0 2.2
f f
0.67b 0.79=
3.1 f 1.9 f
0.74bd 0.74=
1.3 zk 0.4a= 1.7 k 0.68-' 1.9 k 0.3249
2.4 f 0.97de 2.5 IL 0.8tTd'g 2.9 + 0.57ad
1.7 f 2.2 f 2.2 f
0.82= 0.79= 0.92*-
1.6 f 0.848 1.8 k 0.63= 2.1 f 0.74==
0.67a 0.74a
2.1 + 0.32d 2.5 k 0.53e
2.6 2 0.52ad 3.1 f 0.32=
2.1 2.9
f 0.32a + 0.74b
2.1 2 0.32= 2.6 f 0.70bC
0.32= o.ooa
3.0 f. 0.828 2.8 f 0.86a
4.2 + 0.79' 1.9 If: 0.57Q
2.5 f 2.8 f
2.6 f 2.1 f
Natto produced by 6. subtilis k SD (n = 10). Means having
(natto) Miura was the same superscripts
Table
7. Correlation
matrix
of sensory
atiributes
Coiour Coiour Aroma Stickiness Texture Taste General acceptability n = 10.
l
* Significance
between
qualities
protease, of natto.
0.84b 0.53c
of kinema
produced
Aroma
0.9Vd
0.71d'g 0.63-
Texture
1.000
0.277 0.278
0.565 0.120
1.000 0.272
1.000
0.491 0.410
0.476 0.593
0.832'* 0.970***
0.461 0.401
at 1%. '"'Significance
and y-glutamyltranspeptidase Nippon Shikuhin Kogyo
0.7gbC 0.88a
5, good. Values (P -C 0.05).
of B. subtilis
1.000 0.314
isolated
represent
from
market
Taste
3.2 + 0.92d 2.2 f 0.63= the
mean
scores
kinema. General acceptability
1.000 0.906”*
1.000
atO.l%.
activities Gakkaishi 37
and
(II),
155.
D.
by 10 strains
Stickiness
J.W. 1962 Variables influencing results and the precise definition of the steps in Gram staining as a means of standardizing the results obtained. Stain Technology 37, UP-
& Berkeley, R.C.W. 1986 Genus Bacillus Cohn 1872, Manual of Syslematic Bacteriology Volume 2, eds Sneath, P.H.A., Mair, N.S., Sharpe, M.E. & HoIt, J.G. pp 1X051139. Baltimore: Williams & Wilkins. Fahmy, F., Floossdorf, J. & Claus, D. 1985 The DNA base composition of the type strains of the genus Bucillw. Systematic and Applied Microbiology 6, 6(r65. Fuwa, H. 1954 A new method for microdetermination of amylase activity by the use of amylose as the substrate. Journal of Biochemistry 41, 583-603. Gordon, R.E., Haynes, W.C. & Pang, C.H.-N 1973 The genus Bacillus, Handbook No. 427. Washington: United States Department of Agriculture. Hara, T. & Ueda, S. 1982 Regulation of polyglutamate production in Batiks srrbtilis (n&o): transformation of high PGA productivity. Agricuhurul and Biological Chemistry 46, 2275-2281. Hara, T., Chetanachit, C., Fujio, Y. & Ueda, S. 1986 Distribution of plasmids in polyglutamate-producing Bacillus strains isolated from ‘natto’-like fermented soybeans, ‘thua-nao’, in Thailand. ]ournu~ of General and Applied Microbiology 32, 241-249. Hara, T., Ogata, S. & Ueda, S. 1993 Plasmid distribution in y-polyglutamate-producing Bacillus strains isolated from 174. In Bergey’s
f
as a control; score 1, bad; score in each row do not differ significantly
872-877. Bartholomew,
Claus,
2.7
General acceptability
2.9 _+ 0.32=' 2.0 f 20.67be
used
0.32a
Taste
‘dan-do&i’, in China.
a ‘natto’-like non-salty ]ortrnul of General and
fermented soybean Applied Microbiology
food 39,
75-82.
Ikeda, Y., Saito, H., Miura, K., Takagi, J. & Aoki, H. 1965 DNA base composition, susceptibility to bacteriophages, and interspecific transformation as criteria for classification in the genus Bacillus. ]oumul of General and Applied Microbiology 11, 181-190.
Kawaguchi, T., Nagae, H., Murao. S. & Arai, M. 1992 Purification and some properties of a haim-sensitive cc-amylase from newly isolated Bacillus sp. No. 195. Bioscience, Biotechnology, and Biochemistry 56, 1792-1796. Maeda, T., Takagi, M. & Imanaka, T. 1993 Purification and characterization of a new metal protease which hydrolyzes the cyclic decapeptide, gramicidin S. ]ournul of Fermentation and Bioengineering 75, 173-l 77. Methods of Natto Research. 1990 Society for study of natto. Tsukuba: National Food Research Institute. Nagai, T., Nishimura, K., Suzuki, H., Banba, Y., Sasaki, H. & Kiuchi, K. 1994 Isolation and characterization of Bucillw subfilis strain producing natto with strong Umami-taste and high viscosity. Nippon Shokuhin Kogyo Gakkuishi 41, 123-128. Saito, H. & Miura, K. 1963 Preparation of transforming deoxyribonucleic acid by phenol treatment. Biochimu ef Biophysics Acta 72,619-629.
Sarkar, and Food Sarkar,
P.K. & Tamang, J.P. 1994 The influence of process variables inoculum composition on the sensory quality of kinema. Microbiology 111, 317-325. P.K., Tamang, J.P., Cook, P.E. & Owens, J.D. 1994 Kinema
Kinema
-
a traditional soybean fermented food: proximate composition and microflora. Food Microbiology 11, 47-55. Snedecor, G.W. & Cochran, W.G. 1980 Statistical Methods 7th edn. Ames: Iowa State University Press. Somogyi, M. 1945 A new reagent for the determination of sugars. ]oumal of Biologica! Chemistry 160, 61-62. Tamang, J.P. 1993 Studies on the microflora of some truditiona~ fermented foods of Durjeeling hills and Sikkim. PhD. Thesis. Dajeeling: North Bengal University. Tamang, J.P. 19% Study of traditional fermented foods production in
starter
culture
Da+&ng hills and Sikkim, with emphasis on Kinema. Postdoctoral Report. Tokyo: United Nations University. Tamang, J.P., Sarkar, P.K. & Hesseltine, C.W. 1988 Traditional fermented foods and beverages of Dajeeling hills and S&m - a review. Journal of Science of Food and Agriculture 44, 3i’~385.
(Received
in revised
form
28 April
1996;
accepted
4 May
1996)
