LWT - Food Science and Technology 82 (2017) 227e234

Contents lists available at ScienceDirect

LWT - Food Science and Technology

journal homepage: www.elsevier.com/locate/lwt

Production of tofu by lactic acid bacteria isolated from naturally

fermented soy whey and evaluation of its quality
Ce Li, Xin Rui, Yuhui Zhang, Fangyuan Cai, Xiaohong Chen, Mei Jiang*
College of Food Science and Technology, Nanjing Agricultural University, Nanjing, Jiangsu 210095, PR China

a r t i c l e i n f o a b s t r a c t

Article history: Fermented soy whey (FSW) is a traditional tofu coagulant used extensively in China. In this study, 112
Received 1 November 2016 Lactobacillus spp., 18 Streptococcus spp. and 33 yeasts were isolated and identified from naturally fer-
Received in revised form mented soy whey (NFSW). Lactic acid bacteria were found as the major effective strains, and 7 lactic acid
11 April 2017
bacterial strains could produce enough acid to coagulate soy milk. The pH values and acid production
Accepted 17 April 2017
were also monitored during fermentation. The JMC-1 strain showing Gram-positive and rod-shaped was
Available online 18 April 2017
determined to have a high acid-producing activity among the 7 strains. API 50 CHL strip and phylogenetic
analysis of 16S rDNA gene sequence demonstrated that the JMC-1 strain was identified as Lactobacillus
Keywords:
Fermented soy whey
plantarum. Lactic acid and acetic acid were the major organic acids in fermented soy whey, which
Lactic acid bacteria increased during the fermentation. High-performance liquid chromatography revealed that the con-
Identification centrations of these acids increased significantly as fermentation time was prolonged. Compared with
Tofu CaSO4 tofu and MgCl2 tofu, the tofu prepared by L. plantarum JMC-1 was of better quality in terms of
Organic acid sensory evaluation, texture, yield, and retention capacity.
© 2017 Published by Elsevier Ltd.

1. Introduction was used for coagulating tofu. Thus, use of FSW for coagulating tofu
could help to reduce environmental pollution. Bittern tofu prepared
Tofu is a well-known traditional food in many East Asian with magnesium chloride (MgCl2) (Molamma, Conrad, & Suresh,
countries (Rossi, Felis, Martinelli, Calcavecchia, & Torriani, 2016). 2006) is too hard, and MgCl2 is not good for our health. Tofu
Tofu has been distributed worldwide because of high nutrient coagulated by calcium sulfate (CaSO4) has an unpleasant taste,
amounts and several potential health benefits to humans (Li, Qiao, specifically a beany flavor and bitter taste. Compared with these
& Lu, 2012; Yang et al., 2012). Tofu, also known as soybean curd, is a two kinds of tofu, FSW tofu exhibits moderate hardness and better
soft, cheese-like food prepared by curdling fresh hot soymilk with a flavor.
coagulant. Soymilk is coagulated with salts, acids, and enzymes during tofu
Fermented soy whey (FSW) has been used as a traditional tofu production (Kohyama, Sano, & Doi, 1995). Acid- and FSW-
(Mengkebilige, Chen, Wang, Li, & Bao, 2000; Qiao et al., 2010) coagulated tofu have been rarely investigated. Fermented soy
coagulant for more than 600 years in China, especially in rural whey contains a microbial community dominated by lactic acid
areas. In naturally fermented soy whey (NFSW) preparation, soy bacteria (LAB) that play a major fermentative role and thus affect
whey is used as a raw material fermented with various mixtures of the acidity and flavor of tofu products (Qiao et al., 2010). The main
microbes in natural environment. Soy whey is a byproduct of the organic acids in FSW are lactic acid and acetic acid (Qiao et al., 2010;
preparation of soybean products, obtained during the process of €
Ozcelik, €
Kuley, & Ozogul, 2016). However, FSW is produced with
pressing tofu, such as tofu and soy protein isolate, which contain different processing technologies dependent on geographical
high amounts of useful compounds (Xiao et al., 2015; Li, Wu, Wang, location and climate. Consequently, inconsistent flora is produced
& Liu, 2014). Large amounts of soy whey are produced yearly; as a result of different environmental conditions, which makes it
however, this production aggravates environmental pollution. FSW difficult to consistently obtain high quality tofu. The quality of FSW
tofu varies in terms of predominant microorganisms. Moreover, the
fermentation time of NFSW is almost 3 days. Thus, predominant
* Corresponding author. College of Food Science and Technology, Nanjing Agri- microorganisms must be isolated and identified from NFSW for
cultural University, 1 Weigang Road, Nanjing, Jiangsu, PR China.
extensive industrial processing. To the best of our knowledge,
E-mail address: meijiamg9@njau.edu.cn (M. Jiang).

http://dx.doi.org/10.1016/j.lwt.2017.04.054
0023-6438/© 2017 Published by Elsevier Ltd.
228 C. Li et al. / LWT - Food Science and Technology 82 (2017) 227e234

proximate NFSW composition, bacterial phases, and different bac- Soybeans were washed, soaked overnight in water at room
terial activities on FSW tofu have been seldom analyzed. Changes in temperature, drained, rinsed, and blended with water (soybean to
organic acids in FSW in different fermentation stages and differ- water ratio ¼ 1:6 w/v) in a soymilk grinder (Model MJ-02, Hebei
ences among FSW tofu, bittern tofu, and gypsum tofu have been province, China) (Fig. 1). Subsequently, 40 ml of cooked soybean
rarely evaluated. milk was placed in water bath at 85
C. Then, 20 ml of the FSW was
This work aimed to screen and isolate a high-acid-producing added into the soybean milk. FSW were capable of inducing soy-
bacteria from NFSW for the industrial production of FSW tofu or bean milk coagulation within 5e10 min at 85
C, and then the
other soy products. Changes in pH, acid production, and organic corresponding bacteria were marked.
acid of FSW during fermentation were evaluated. The solidification
characteristics and good flavor of FSW tofu were also investigated. 2.3.3. Determination of pH and the rate of acid production
The FSW of target bacteria were respectively inoculated into the
2. Materials and methods fresh pasteurized soy whey medium at 4% (v/v) and incubated at
37
C up to 24 h for two successive transfers. Then, 10 ml of
2.1. Sample collection fermentation broth was withdrawn after 0, 2, 4, 6, 8, 10, 12, 14, 16,
18, 20, 22, and 24 h of incubation and stirred before measurement
A sample of naturally fermented soy whey was collected from of pH and total acid. The pH of each fermentation broth of pure
Hunan Province, China. The pH value of the sample was 3.95. The cultures was determined using a Schott pH Meter Lab 850 (model,
fermented soy whey (FSW) was made by fermenting soy whey at Lab 850, Mainz, Germany). The acid production (g/100 ml,
28
C for 3 days. expressed as acid degree) of each fermentation broth was accom-
plished according to the method for total acid determination
2.2. Proximate composition analyses (Viljoen & Greyling, 1995). The rate (R) of acid production was
determined in the exponential phase using the equation RM ¼ DM/
Proximate composition analyses were conducted using standard Dt (DM ¼ amount of acid produced [g/100 ml]; Dt ¼ time [h]).
methods (AOVO, 2000). The total protein was determined by micro- Acidity in terms of lactic acid was determined by diluting an aliquot
Kjeldahl method using a protein conversion factor of 6.25, and of the prepared sample with recently boiled distilled water. Then,
crude fat was determined by the Soxhlet extraction method. Ash 2e3 drops of 1% phenolphthalein solution was used as an indicator,
was assayed using the AOAC method (Marmouzi, EI Madani, and titration was conducted with 0.1 mol/L NaOH.
Charrouf, Cherrah, & EI Abbbes Faouzi, 2015). The soluble solid
content was assayed by refractometer after filtration. All the results 2.4. Identification of bacteria
were reported on a wet basis.
2.4.1. Carbohydrate fermentation and assimilation test
2.3. Enumeration and isolation of bacteria API 50 CHL strip (Nigatu, Ahrne , & Molin, 2000) enabled iden-
tification of the isolates to species level. This system conducted 49
2.3.1. Screening of major microorganism in fermented soy whey carbon assimilation or fermentation tests and included a database
Sample (1 ml) of naturally fermented soy whey (NFSW) was with 52 different species. All the procedures for LAB identification
mixed with 9 ml of sterilized 0.85% NaCl (w/v) by vortex mixer. A were conducted in accordance with the manufacturer's in-
series of appropriate dilution of the sample was inoculated into structions. Bacterial suspension was distributed into each of the 50
pour-plates and in triplicate. The de ManeRogosa and Sharp (MRS) couples and then overlaid with sterile paraffin oil. This set of strips
agar was used for the isolation of Lactobacillus spp., and M 17 was was cultured at 37
C for 48 h. During the culture time, the color of
used for Streptococcus spp. and Lactococcus spp. They were incu- some couples changed from violet to black and was taken as a
bated at 37
C for 48e72 h. The Bengal medium for the isolation of positive reaction.
yeast and molds were incubated for 3 days at 28
C. The violet red
bile agar (VRBA) for the isolation of coliforms was incubated for 2.4.2. 16S rDNA sequence analysis
24 h at 30
C. After incubation, the plates with 30e300 colonies 2.4.2.1. DNA extraction and amplification. The JMC-1 strain was
were enumerated and expressed as log10 of colony forming units extracted for sequencing of 16S rDNA using the lysozyme method
per milliliter (cfu/mL) of samples. (Nurgul, Chen, Feng, & Dong, 2009; Liu, Zhang, Yi, Han, & Chi, 2016;
All of the colonies were examined microscopically and checked Handa & Sharma, 2016). The primers pA 50 AGAGTTT-
for catalase reaction before the stocks were prepared. Gram- GATCCTGGCTCAG 30 and pH 50 TACGGCTACCTTGTTACGACTT 30
positive, catalase-negative colonies were identified initially as were used for the amplification of the extracted DNA (1 mL). Then,
Lactobacillus. Colonies showing different appearance (size, shape, 25 mL PCR system containing 18.25 mL deionized water, 0.5 mL of
and color) (Li, Mutuvulla, Chen, Jiang, & Dong, 2012) were selected each 10 mM primer, 2.5 mL Easy Taq buffer, 2 mL of 2.5 mM dNTPs, 5
randomly and further purified by successive streaking using the U of Taq polymerase, and 1 mL bacterial DNA was used. Amplifica-
same medium. The purified bacteria were resuspended and main- tion was implemented in PCR Thermocycler (Bio-Rad Mycycler™)
tained in the same liquid medium containing 50% glycerol programmed as follows: 94
C for 3 min, 35 cycles at 94
C for 30 s,
at 20
C. 59
C for 60 s, 72
C for 2 min, and 72
C for 10 min (final extension).

2.3.2. Soymilk coagulating capacity analysis 2.4.2.2. DNA sequencing and sequence analysis. The amplification
The strains were respectively inoculated into the MRS medium products (5 mL) were extracted for DNA sequencing, and sequence
at inoculums concentration of 4% and incubated at 37
C for 24 h for analysis was conducted by electrophoresis in 0.8% agarose gels.
two successive transfers. The prepared inoculums was then inoc- DNA amplification of the products were performed by Sangon
ulated into the sterile fresh soy whey at 4% (v/v) and incubated at Company (Shanghai, China). The DNA sequences were submitted
37
C for 24 h for two successive transfers. The pH of each FSW was and deposited to The National Center for Biotechnology Informa-
determined using a Schott pH Meter Lab 850 (model, Lab 850, tion (NCBI) to find similar sequences using the basic local alignment
Mainz, Germany). The bacteria and the corresponding FSW with a search tool (BLAST) program. A set of the similar sequences was
pH of less than 4.8 were marked. entered in the MEGA 3.1 to produce the phylogenetic tree.
 C. Li et al. / LWT - Food Science and Technology 82 (2017) 227e234 229

2.5. High performance liquid chromatography (HPLC) analysis of formed were broken and gently transferred to a tofu model
the soy whey during the fermentation (12 cm  12 cm  8 cm depth). The tofu was pressed for 15 min
using special barrel weighing 4 kg. The soy whey was drained off
The organic acids were measured using HPLC (Agilent, USA) naturally. The two other kinds of tofu were coagulated with MgCl2
equipped with a XSelect HSS T3 analysis column and CaSO4. The resulting tofu was stored in a refrigerator at 4
C
(4.6 mm  250 mm, 5 mm particle size, Agilent). Samples were until analysis.
prepared by centrifugation at 4
C and 12,000 rpm for 10 min, and
the supernatant was filtrated with 0.22 mm membrane. The injec-
tion volume was 20 mL. Samples were eluted by 0.01 M (NH4) 2HPO4 2.6.2. Yield and retention ability
at a flow rate of 0.6 ml/min for 20 min. Detection was monitored at Tofu yield was recorded and expressed as weight of tofu ob-
215 nm. The temperature of the column oven was maintained at tained from 100 g dry soybeans (Lee & Kuo, 2011; Wang, Liu, Wang,
30
C. & Chen, 2013). Tofu samples were centrifuged at 3000 rpm and 4
C
for 10 min. Retention ability was calculated as the weight of seep
liquid as a percentage of weight of tofu.
2.6. Tofu quality analysis

2.6.1. Preparation of tofu 2.6.3. Color measurement

Fig. 1 illustrates the preparation of the FSW tofu. Hot soymilk The color values of tofu were measured with a CR-400 chro-
was placed into a stainless steel container and cooled to 85
C. The mameter (Konica Minolta, Japan). The color was expressed in
FSW was slowly poured into the steel container while stirring (for Hunter L (lightness), redegreen (þa or a), and yelloweblue (þb
around 5 min), and FSW addition was stopped when the coagulated or b). A standard white plate with L ¼ 93.57, a ¼ 0.3162, and
soymilk occurred. The coagulated soymilk was allowed to stand b ¼ 0.3326 was applied for calibration. The measurements were
undisturbed for 15 min to ensure a complete coagulation. The tofu made at different locations on each sample and averaged.

Fig. 1. Flow chart of producing FSW tofu. FSW: fermented soy whey.
230 C. Li et al. / LWT - Food Science and Technology 82 (2017) 227e234

Fig. 2. Soymilk with coagulants. A: milk without aggregation B: tofu pudding.

2.6.4. Texture profile analysis of tofu content was 5.61 ± 0.22 mg/ml during soy whey fermentation.
The texture of tofu was determined through texture profile Microbiological analysis by using specific culture medium showed
analysis (TPA) by using a Micro System, model TA.XT Plus (Stable that the naturally fermented soy whey (NFSW) contained
Micro System) with a probe P50. Tofu samples, prior to testing,
were cut into 2.0 cm-thick section, 3.0 cm length and width and
equilibrated to room temperature for half an hour. These samples
were compressed twice to 35% deformation. The test settings were 6.0 JMC-1
JMC-2
set as pre-test speed: 1.00 mm/s, test speed: 5.00 mm/s, and post- JMC-3
test speed: 5.00 mm/s. The textural parameters of hardness, 5.6 JMC-4
JMC-5
cohesiveness, springiness, and chewiness of each tofu sample were JMC-6
determined from the presented TPA curve (Zhu, Wu, Saito, Tatsumi, JMC-7
5.2
& Yin, 2016). Several samples of the four tofu kinds were prepared
for measurement.
4.8
pH

2.6.5. Sensory evaluation

Fresh tofu was subjected to sensory evaluation based on the 4.4
method proposed by Fasoyiro (2014). Nine semi-trained panelists
who participated in previous related projects evaluated the sensory 4.0
attributes of the fresh tofu under natural light in a room. The
panelists were not given a time limit for evaluation, although most
0 4 8 12 16 20 24
of them completed the rating of all samples within 10e15 min. Tofu
was cut into cubes and placed on a plastic plate with a random Time (h)
number (Cai & Chang, 1998). The attributes selected as the indices
Fig. 3. The pH values of fermented soy whey with seven strains (n ¼ 3).
for sensory evaluation were color, flavor, texture, taste, and overall
acceptability. The results were recorded with a 5-point scale
(1 ¼ dislike, 2 ¼ neither like nor dislike, 3 ¼ like slightly, 4 ¼ like 0.08 JMC-1
moderately, and 5 ¼ like very much) for each sample. JMC-2
JMC-3
JMC-4
2.7. Data analysis 0.06 JMC-5
JMC-6
Rate of acid production

JMC-7
Experiments were performed in triplicate, and data were sub-
jected to ANOVA in Statistical Analysis System (SAS) software 0.04
package. Different tofu groups were examined through Duncan's
multiple ranges in SAS. Differences among group means were 0.02
considered significant at P < 0.05.

3. Results and discussion 0.00

Soy whey is rich in carbohydrates, lipids and proteins. A sig- 0 4 8 12 16 20 24

nificant decrease in protein from 6.85 ± 0.41 mg/ml to Time (h)
4.59 ± 0.27 mg/ml was observed during soy whey fermentation.
Furthermore, the fat content was 4.20 ± 0.13 mg/ml and the ash Fig. 4. The acid production rate of fermented soy whey with seven strains (n ¼ 3).
 C. Li et al. / LWT - Food Science and Technology 82 (2017) 227e234 231

Table 1 Lactobacillus spp. [11.34 ± 0.02 log(cfu/ml)] and yeasts [4.41 ± 0.03
Physiological and biochemical characteristics of strain JMC-1. log(cfu/ml)]. Escherichia coli and molds were not detected.
Carbohydrate Resultsa Carbohydrate Results

Control e Esculin þ
3.1. Enumeration and isolation of bacteria
Glycerol e Salicin þ
Erythritol e Cellobiose þ
D-Arabionose e Maltose þ A total of 163 strains were isolated from the naturally fermented
L- Arabionose þ Lactose þ soy whey. FSW of seven strains had the capacity to induce soybean
Ribose þ Melibiose þ milk coagulation respectively (Fig. 2), thus were selected for further
D eXylose e Saccharose þ
analysis. All the seven strains were also identified as Lactobacillus
L -Xylose e Trehalose þ
Adonitol e Inulin e spp. owing to being Gram-positive, catalase-negative, and rod-
b-Methyl-xyloside e Melezitose þ shaped. The colonies were all round and bulging, but the color,
D eGalactose þ D eRaffinose þ transparency, and size in diameter were different. These results are
D eGlucose þ Amidon e
in agreement with many authors who observed that lactic acid
D eFructose þ Glycogen e
D eMannose þ Xylitol e
bacteria were the main microorganisms in Chinese traditional fer-
L-Sorvose e b-Gentiobiose þ mented products (Xiong, Li, Guan, Peng, and Xie, 2014; Xiong et al.,
Rhamnose e D eTuranose þ 2014; Liu, Han, & Zhou, 2011; Chao, Wu, Watanabe, & Tsai, 2009).
Dulcitol e D eLyxose e
Inositol e D eTagatose e
Mannitol þ D-Fucose e 3.1.1. Determination of pH and rate of acid production
Sorbitol e L-Fucose e
Tofu coagulated with fermented soy whey (FSW) is known as
a-Methyl-D-mannoside þ D-Arabitol e
a-Methyl-D-glucosamine e L-Arabitol e acid-coagulated tofu because of the large amount of acid produced
N-Acety1 glucosamine þ Gluconate þ by Lactobacillus spp. in FSW (Kohyama et al., 1995). The added acid
Amygdalin þ 2-Keto-gluconate e changes the double-layered structure on the surface of the protein
Arbutin þ 5-Keto-gluconate e
and enables protein aggregation to form a gel (Grygorczyk &
a
Symbols denote positive (±) and negative () in sugar utilization patterns of API Corredig, 2013). The acid-producing Lactobacillus sp. was
50 CHL kit. screened according to this characteristic.
Figs. 3 and 4 show the alterations of pH and acid production rate
during the soy whey fermentation by the seven selected strains.
During fermentation, pH decreased with time compared with the
acid production rate, which increased before 8 h and then
decreased with time until 24 h of fermentation. The pH of the seven
strains rapidly decreased in the first 8 h then slowly decreased up to
24 h. The seven strains could be divided into three categories based
on the pH reduction trend. The soy whey fermented with JMC-1
strain had the lowest pH value, followed by JMC-4, JMC-3, JMC-6
strain, and then JMC-2, JMC-5 and JMC-7 strains. However, FSW
with JMC-1 strain was the first to reach a pH value less than 4.0
before 16 h, had a higher acid production rate than the other strains
before 18 h, and the highest rate of approximately 0.66 at 8 h,
corresponding to the changes of pH value (Fig. 3). The acid pro-
duction of JMC-1 strain was approximately 0.7 g/100 ml. The JMC-1
strain showed the highest acid formation, and soy whey fermented
for 16 h was more suitable as the tofu coagulant. Qiao et al. (2010)
reported that the pH value of 3.8e4.0 was the best FSW acidity for
making tofu and FSW of 18 h fermentation and NFSW of 72 h
fermentation was generally used as coagulant in the production.
Fig. 5. PCR products of amplified 16S rDNA gene coding regions from different LAB
The fermentation time (16 h) was shorter compared to that re-
using primers for JMC-1. Lanes: M molecular weight marker, 0 blank, 1 JMC-1. ported in Qiao's research. Compared with NFSW, FSW with JMC-1
strain saved much fermentation time and contained none of the
spoilage bacteria.

98 JMC-1
36 Lactobacillus plantarum strain IMAU80102 16Sribosomal RNA gene partial sequence (GU125524.1)
Lactobacillus pentosus strain IHBB6854 16S ribosomal RNA gene partial sequence (KF668473.1)
52 Lactobacillus plantarum subsp. Plantarum gene for 16S rRNA partial sequence strain: AA11 (AB603681.1)
Lactobacillus plantarum strainKCC-116S ribosomal RNA gene complete sequence (KC422316.1)
Lactobacillussp. KLDS1.0717 16S ribosomal RNA gene partial sequence (EU600920.1)
Lactobacillus plantarum strain KLDS1.0728 16S ribosomal RNA gene partial sequence (EU626013.1)
47

0.5

Fig. 6. Phylogenetic tree based on 16S rDNA sequence of JMC-1 strain. The number on the branches indicates the support proportion of each branch.
232 C. Li et al. / LWT - Food Science and Technology 82 (2017) 227e234

3.2. API and 16S rDNA

Note: at the fermented same time, means without same lowercase superscripts differ significantly (P < 0.05); at the organic acid, means without same capital superscripts differ significantly (P < 0.05). - indicated the acid was
10486.95 þ 610.16aA
1894.96 þ 115.05bA
153.22 þ 13.70cdAB
198.81 þ 19.44cdA

359.85 þ 30.59cdC

411.96 þ 37.32cG
3.50 þ 0.13dCDE
29.03 þ 1.17cdC
38.29 þ 2.72cdE
Previous results showed that JMC-1 strain was a strong acid-
producing lactobacillus, and the colonies were white with a

13576.57
smooth and opaque surface and diameter of 0.8e1.2 mm. The cells
of the JMC-1 strain were Gram-positive rods within single cell or

24

e
two cells, with no spore.

10322.20 þ 351.13aA
1679.52 þ 106.42bB
159.28 ± 14.44deAB
200.85 þ 16.86deA
The JMC-1 strain was examined for fermentation of different

32.99 þ 2.81eABC
368.56 þ 25.44dC

530.98 þ 47.34cF
43.78 þ 3.92eDE

3.38 þ 0.15eDE
carbohydrates using the API 50 CHL medium. Based on API 50 CHL
kit results, the JMC-1 strain positive for L-arabinose, ribose, D-

13341.54
galactose, D-glucose, D-fructose, D-mannose, methyl-a-D-manno-

18
side, N-acetyl glucosamine, amygdalin, arbutin, esculin, salicin,

e
10248.78 þ 438.23aA
cellobiose, maltose, lactose, melibiose, saccharose, trehalose,

155.82 þ 10.70deAB

1629.64 þ 97.41bBC
198.73 þ 10.88deA

374.89 þ 31.75cdC

618.87 þ 52.48cEF
3.61 þ 0.21eBCDE
melezitose, D-raffinose, b-gentiobiose, D-turanose, and gluconate

45.91 þ 4.19eDE

35.19 þ 2.95eAB
was identified as L. plantarum (Table 1). The 16S rDNA fragment of

13311.44
strain JMC-1 was amplified by two selected primers resulting in an
amplicon of approximately 1472 bp in size (Fig. 5). The 16S rDNA

16

e
sequencing data in BLAST showed that JMC-1 strain was highly

10051.86 þ 476.82aA
1521.76 þ 89.88bCD
similar to L. plantarum strain IMAU80102, and the partial rDNA

199.26 þ 15.53deA

714.93 þ 38.56cDE
380.43 þ 17.91dC
146.15 þ 8.79deB

51.49 þ 2.55eCD

37.43 þ 3.667eA

3.94 þ 0.10eAB
sequence of the strain matched 99% (Fig. 6).

13107.25
3.3. HPLC analysis of the soy whey during the fermentation

14

e
Table 2 shows the HPLC analysis results of the acid slurry. Formic

1389.76 þ 84.90bDE
155.77 þ 10.25deAB

8948.95 þ 500.47aB

802.52 þ 59.47cCD
196.60 þ 5.12deA

374.77 þ 37.33dC
acid and succinic acid were not detected. Ten types of organic acids,

undetected and content of it was subtle or zero. Values were the average of 3 replicates independent experiment, with standard deviation.
58.54 þ 2.92eC

28.12 þ 2.35eC

4.08 þ 0.17eA
namely, oxalic acid, DL-tartaric acid, pyruvic acid, malic acid, a-

11959.11
ketoglutaric acid, lactic acid, acetic acid, maleic acid, citric acid, and
fumaric acid, were separated and detected effectively during the

12

e
fermentation of JMC-1 strain. These organic acids are intermediates

1399.66 þ 65.90bDE
6898.89 þ 370.73aC
179.09 þ 21.96deA
195.60 þ 17.86deA

872.39 þ 52.33cBC
and metabolites of various biochemical processes that occur during

372.56 þ 32.11dC
30.96 þ 2.98eBC
67.72 þ 6.81eB
fermentation and are important secondary carbon sources for

3.25 þ 0.14eE

0.42 þ 0.01eE
numerous microbial genera that proliferate during food

10020.54
fermentation.
10

As shown in Table 2, lactic and acetic acids had the highest

6311.19 þ 332.91aCD
contents and all increased in the process of fermentation, which
170.13 þ 13.90eAB

1289.62 þ 86.90bE

876.20 þ 58.94cBC
386.06 þ 35.60dC
196.26 þ 23.28eA

3.86 þ 0.21eABC
30.85 þ 2.72eBC
could possibly be attributed to a significantly greater supply of
66.24 þ 4.03eB

2.31 þ 0.14eD
available reducing sugars. Lactic acid, which increased fourfold
from 2413.45 mg/L to 10486.63 mg/L and accounted for about 77%

9332.72
of the total acid after 16e24 h, was the major organic acid. This was
€
8

in accordance with an earlier study reported by Ozcelik, Kuley and

5942.95 þ 309.88aDE

€
1252.77 þ 77.28bEF

Ozogul (2016). Acetic acid increased onefold from 922.2 mg/L to

912.73 þ 62.72cAB
400.33 þ 38.34dC
198.26 þ 22.27eA

3.76 þ 0.07eABC
32.91 þ 2.62eBC
145.69 þ 5.44eB

85.71 þ 6.98eA

1894.78 mg/L and accounted for about 12% of the total acid after
5.72 þ 0.23eC

16 h. The changes in lactic and acetic acids also indicated that the
8980.83

JMC-1 strain was heterofermentative. The contents of malic, keto-

glutaric, maleic, citric, and fumaric acids decreased after 24 h of
6

1135.87 þ 65.02bFG

fermentation. The contents of malic and citric acids were higher in

950.48 þ 41.95cAB
5379.16 þ 73.24aE
167.09 þ 4.85eAB

518.06 þ 42.58dB
199.26 þ 20.21eA

the fermented soy whey, and these acids were the important flavor
30.59 þ 2.86fBC
29.81 þ 2.86fF

6.82 þ 0.17fB
HPLC analysis of JMC-1 strain fermented for different periods.

3.29 þ 0.37fE

substances. Malic acid decreased almost in half, and citric acid

decreased most from 1010.91 mg/L to 411.88 mg/L. There were
8420.43

traces of ketoglutaric, maleic, and fumaric acids in the FSW. The

4

contents of oxalic and DL-tartaric acids were essentially invariant in

1020.86 þ 48.92bGH
2563.55 þ 196.21aF

24 h of fermentation. Pyruvic acid showed a trend of elevation and

174.60 þ 21.36dA
199.60 þ 13.03dA

988.33 þ 78.15bA
651.01 þ 46.10cA

7.26 þ 0.54eAB
29.28 þ 2.03eC

decrease. However, the content of total acid was also increased by

3.31 þ 0.15eE

1.49 times. Most acids tended to be stable after 16e24 h fermen-

5637.8

tation, which further indicated that the soy whey fermented for
16 h was more suitable as a tofu coagulant.
e
2
Acid (mg/L) Fermentation time (h)

159.64 þ 12.97deAB

2413.47 þ 188.33aF

1010.87 þ 80.35bA
32.98 þ 3.96efABC

922.21 þ 59.23bH
198.93 þ 18.03dA

708.66 þ 72.86cA

3.71 þ 0.25fBCD

3.4. Tofu quality analysis

7.68 þ 0.63fA
5458.15

3.4.1. Yield and retention ability of tofu

Table 3 shows that the yields of NFSW tofu and FSW JMC-1 tofu
e
0

were higher than traditional bittern (MgCl2) tofu and gypsum

Ketoglutaric

Total acid

(CaSO4) tofu under the same tofu-making process. From 100 g of

Fumaric
Tartaric
Pyruvic

Maleic
Table 2

Oxalic

Acetic
Lactic
Malic

Citric

dried soybean, 196.87 ± 9.32 g tofu was obtained by using JMC-1-

fermented soy whey as coagulant. This was consistent with yield
 C. Li et al. / LWT - Food Science and Technology 82 (2017) 227e234 233

of NFSW tofu which was 198.93 ± 11.26 g (based on 100 g dried JMC-1 tofu was lower than NFSW tofu and higher than bittern tofu
beans). The retention ability of FSW JMC-1 tofu (87.57%) was closed and gypsum tofu. So in this study, the JMC-1 coagulant could make
to NFSW tofu (86.97%) and they were lower than bittern tofu a good quality tofu.
(93.34%) and gypsum tofu (89.42%). These results could be attrib-
uted to organic acids in FSW being weak acids with lower ioniza-
tion capacity. The stability of protein molecules in soymilk can be 3.4.3. Sensory evaluation
attributed to the net negative charge and steric repulsion. With acid Sensory evaluation was performed on four kinds of fresh tofu,
addition, the slow release of hydrogen ions neutralizes the negative and the results are presented in Fig. 7. JMC-1 retained more flavor,
surface charges on soy proteins particles and causes the protein to texture, mouthfeel, and overall acceptability for tofu than MgCl2
form a network structure which hold together by noncovalent and CaSO4 but had a low surface value. Gypsum tofu was whiter
bonds. The network structure can hold water molecules constantly, than other tofu, suggesting it had a higher surface value. The sen-
in the process of forming the network structure, thereby increasing sory evaluation value of FSW JMC-1 tofu was similar to the NFSW
the tofu yield. But the water was easier to flow out than other two tofu. In addition, FSW JMC-1 tofu had a better mouth feel than
kinds of tofu. This also indicated that coagulating protein had a NFSW tofu. The dominant lactic acid, acetic acid, and other organic
good effect on tofu prepared with JMC-1 strain. acids with low content and lower ionization capacity in the FSW
imparted the special flavor, slightly sweet taste, and good texture
properties to the tofu.
3.4.2. Texture profile analysis and color of tofu
TPA of tofu was a very useful method for determining tofu
quality and consumer acceptability. As shown in Table 3, the four 4. Conclusions
kinds of tofu varied depending on the coagulant and full production
in the lab to exclude the effect of beans. The bittern (MgCl2) tofu The FSW with isolated JMC-1 could coagulate soybean milk and
had the greatest hardness value (2295.40 g), which could possibly yield the highest acid production rate (0.66). Gram-positive and
be attributed to the fast coagulation rate of soymilk. The gypsum catalase-negative isolates were initially identified as Lactobacillus
(CaSO4) tofu was also harder. FSW JMC-1 tofu and NFSW tofu had spp. API 50 CHL strip and 16S rDNA sequence analysis further
an intermediate hardness. The springiness of FSW JMC-1 tofu and revealed that JMC-1 strain was a L. plantarum strain. Compared
bittern tofu were slightly better than others. The difference in the with the bittern (MgCl2) tofu and gypsum (CaSO4) tofu, the organic
chewiness values of the four kinds of tofu was the same as in the acids produced by L. plantarum strain slowly released hydrogen
hardness values. The NFSW tofu and FSW JMC-1 tofu were easy to ions made the tofu with good flavor and texture. FSW could be
chew, suitable for old man and children. In addition, all of the tofu more suitable for industrial applications in the presence of certain
products were light yellow, showing that the tofu quality was good bacteria and under specific processing conditions. Further studies
because a good-quality tofu is supposed to be light yellow or white on optimum fermentation conditions and FSW applications for
in color (Li et al., 2015). However, the b value (yellowish) of FSW environmental pollution reduction are underway.

Table 3
Textural properties of tofu made with different coagulants.
A
Coagulant Yield Hardness Springiner Chewiness Color determination

L a b

NFSW 198.93 ± 11.26a 1148.48 ± 31.84c 0.92 ± 0.01a 859.17 ± 31.04c 87.03 ± 0.26b 1.29 ± 0.14b 20.38 ± 0.34a
JMC-1 196.87 ± 9.32a 1196.40 ± 39.54c 0.94 ± 0.02a 913.60 ± 19.47c 86.73 ± 0.07b 1.38 ± 0.12b 19.32 ± 0.45b
MgCl2 164.57 ± 10.02b 2295.50 ± 55.46a 0.94 ± 0.02a 1800.00 ± 48.24a 87.87 ± 0.59a 0.81 ± 0.07a 17.83 ± 0.04c
CaSO4 182.29 ± 10.41a 1905.79 ± 66.49b 0.93 ± 0.04a 1531.87 ± 68.19b 87.50 ± 0.46ab 0.86 ± 0.05a 18.81 ± 0.82b

NFSW: fresh soy whey fermented in natural environment for 2 d pH 4.0 at 28
C; JMC-1: sterile soy whey fermented with strain JMC-1 for 16 h pH 4.0 at 37
C; Tofu made with
NFSW and JMC-1 were acid slurry tofu. Tofu made with MgCl2 was Bittern tofu; Tofu made with CaSO4 was Gypsum tofu. A: Color value of upper surface of tofu; B: Values were
average of triplicates; (aec)in column indicated significant differences (P < 0.05) among tofu.

Fig. 7. Sensory evaluation of fresh tofu coagulated with 4 kinds of coagulants (1 ¼ dislike, 5 ¼ like very much).
234 C. Li et al. / LWT - Food Science and Technology 82 (2017) 227e234

Acknowledgments Engineering, 142, 201e209.

Marmouzi, I., EI Madani, N., Charrouf, Z., Cherrah, Y., & EI Abbbes Faouzi, M. Y.
(2015). Proximate analysis, fatty acids and mineral composition of processed
This work was supported by the National Natural Science Moroccan Chenopodium quinoa Willd. and antioxidant properties according to
Foundation of China (No. 3150100970) and Key Research and the polarity. SpringerPlus, 13, 110e117.
Development Projects of Jiangsu Province (BE2015359). Mengkebilige, Chen, P., Wang, L. Y., Li, S. Y., & Bao, S. J. (2000). Studies on Biological
properties of Lactic acid bacteria from acidic slurry. Journal of Inner Mongolia
Agricultural University, 3, 90e93 (in Chinese).
References Molamma, P. P., Conrad, O. P., & Suresh, V. (2006). Effect of different coagulants on
the isoflavones levels and physical properties of prepared firm tofu. Food
AOVO. (2000). Official methods of analysis. Arlington: Association of Official Chemistry, 99, 492e499.
Analytical Chemists. Affiliated with Ethiopian Health and Nutrition Research Institute, P.O. Box 80928,
Cai, T. D., & Chang, K. C. (1998). Characteristic of production-scale tofu as affected by Addis Ababa, Ethiopia Nigatu, A., Ahrne , S., & Molin, G. (2000). Temperature-
soymilk coagulation method: Propeller blade size, mixing time and coagulation dependent variation in API 50 CH fermentation profiles of lactobacillus species.
concentration. Food Research International, 31, 289e295. Current Microbiology, 41, 21e26.
Chao, S.-H., Wu, R.-J., Watanabe, K., & Tsai, Y.-C. (2009). Diversity of lactic acid Nurgul, R., Chen, X., Feng, M., & Dong, M. (2009). Characterization of the dominant
bacteria in suan-tsai, traditional fermented mustard products of Taiwan. In- microflora in naturally fermented camel milk shubat. World Journal Microbi-
ternational Journal of Food Microbiology, 135, 203e210. ology Biotechnology, 25, 1941e1946.
€
Ozcelik, €
S., Kuley, E., & Ozogul, F. (2016). Formation of lactic, acetic, succinic, pro-
Fasoyiro, S. B. (2014). Physical, chemical and sensory qualities of roselle water
extract-coagulated tofu compared with tofu from two natural coagulants. pionic, formic and butyric acid by lactic acid bacteria. LWT-Food Science and
Nigerian Food Journal, 32, 97e102. Technology, 73, 536e542.
Grygorczyk, A., & Corredig, M. (2013). Acid induced gelation of soymilk, comparison Qiao, Z., Chen, X., Cheng, Y., Liu, H., Liu, Y., & Li, L. (2010). Microbiological and
between gels prepared with lactic acid bacteria and glucono-d-lactone. Food chemical changes during the production of acidic slurry, a traditional Chinese
Chemistry, 141, 1716e1721. tofu-coagulant. International Journal of Food Properties, 13, 90e104.
Handa, S., & Sharma, N. (2016). In vitro study of probiotic properties of Lactobacillus Rossi, F., Felis, G. E., Martinelli, A., Calcavecchia, B., & Torriani, S. (2016). Microbio-
plantarum F22 isolated from chhang e a traditional fermented beverage of logical characteristics of fresh tofu produced in small industrial scale and
Himachal Pradesh, India. Journal of Genetic Engineering and Biotechnology, 14(1), identification of specific spoiling microorganisms (SSO). LWT-Food Science and
91e97. Technology, 70, 280e285.
Kohyama, T., Sano, Y., & Doi, E. (1995). Rheological characterization and gelation Viljoen, B. C., & Greyling, T. (1995). Yeasts associated with Cheddar and Gouda
mechanism of tofu (soybean curd). Journal of Agricultural and Food Chemistry, making. International. Journal of Food Microbiology, 28(1), 79e88.
43, 1808e1812. Wang, J., Liu, Q., Wang, Y., & Chen, X. (2013). Research on soybean curd coagulated
Lee, C. Y., & Kuo, M. (2011). Effect of g-polyglutamate on the rheological properties by lactic bacteria. SpringerPlus, 2, 250.
and microstructure of tofu. Food Hydrocolloids, 25, 1034e1040. Xiao, Y., Wang, L., Rui, X., Li, W., Chen, X., Jiang, M., et al. (2015). Enhancement of the
Li, M., Chen, F., Yang, B., Lai, S., Yang, H., Liu, K., et al. (2015). Preparation of organic antioxidant capacity of soy whey by fermentation with Lactobacillus plantarum
compatible magnesium chloride incorporated with polysaccharide coagulants. B1-6. Journal of Functional Foods, 12, 33e44.
Food Chemistry, 167, 168e174. Xiong, T., Li, X., Guan, Q., Peng, F., & Xie, M. (2014). Starter culture fermentation of
Li, W., Mutuvulla, M., Chen, X., Jiang, M., & Dong, M. (2012). Isolation and identi- Chinese sauerkraut: Growth, acidification and metabolic analyses. Food Control,
fication of high viscosity-producing lactic acid bacteria from a traditional fer- 41, 122e127.
mented milk in Xinjiang and its role in fermentation process. European Food Xiong, T., Peng, F., Liu, Y., Deng, Y., Wang, X., & Xie, M. (2014). Fermentation of
Research Technology, 235, 497e505. Chinese sauerkraut in pure culture and binary co-culture with Leuconostoc
Li, B., Qiao, M., & Lu, F. (2012b). Composition, nutrition, and utilization of okara mesenteroides and Lactobacillus plantarum. LWT e Food Science and Technology,
(soybean residue). Food Reviews International, 28(3), 231e252. 59, 713e717.
Liu, S., Han, Y., & Zhou, Z. (2011). Lactic acid bacteria in traditional fermented food. Yang, H. Y., Lin, H. S., Chao, J. C. J., Chien, Y. W., Peng, H. C., & Chen, J. R. (2012). Effects
Food Research International, 44, 643e651. of soy protein on alcoholic liver disease in rats undergoing ethanol withdrawal.
Liu, H., Zhang, L., Yi, H., Han, X., & Chi, C. (2016). Identification and characterization The Journal of Nutritional Biochemistry, 23(6), 679e684.
of plantaricin Q7, a novel plantaricin produced by Lactobacillus plantarum Q7. Zhu, Q., Wu, F., Saito, M., Tatsumi, E., & Yin, L. (2016). Effect of magnesium salt
LWT-Food Science and Technology, 71, 386e390. concentration in water-in-oil emulsion on the physical properties and micro-
Li, R., Wu, Z., Wang, Y., & Liu, W. (2014). Pilot study of recovery of whey soy proteins structure of tofu. Food Chemistry, 201, 197e204.
from soy whey wastewater using batch foam fractionation. Journal of Food