The addition of lactic acid bacteria in the soybean soaking process of tempeh

Soaking is one of the important steps in tempeh production. Various microorganisms grow during the soaking process, which can affect the nutrition and quality of the tempeh. This study aimed to determine the effect of adding lactic acid bacteria (LAB) to the soaking process and determined the preference levels of the produced tempeh. In this research, Lactobacillus plantarum HPADA3 and Lactobacillus fermentum HPBD2 were added during the soaking of the soybeans. Each LAB had been evaluated for its hemolytic activity and potential virulence factor genes. The soaking duration was much shorter than was commonly done for tempeh production. The resulting tempeh was observed from their appearance and then underwent a sensory evaluation test. It was found that the shorter duration of soaking resulted in compact tempeh. The addition of LAB did not affect the fungal growth in tempeh fermentation. Overall, the bacteria population observed on agar media from the water and tempeh samples were significantly different. All produced tempeh was liked from an evaluation from semi-trained panellists. From all the tempeh variants, the tempeh with the addition of Lactobacillus fermentum HPBD2 was perceived as the most likeable and similar to commercial tempeh.


Introduction
Tempeh is a superior fermented food from Indonesia. Fermented food contains many beneficial compounds that are good for health. It is high in protein, vitamin B 12 , antioxidants, and other bioactive compounds (Keuth and Bisping, 1993;Wu and Hasnah, 2018). Traditionally, tempeh is made from inoculated soybeans that are fermented to form a compact cake. The process of making tempeh includes boiling, dehulling, soaking, rinsing, drying, and fermentation with the inoculum (Mukhoyaroh, 2015).
In tempeh production, one of the most critical phases is the soaking process, where the dehulled soybeans are immersed in water for a length of time. Soaking is important because it improves the moisture content of the beans, supports the activity of microorganisms, and extracts saponin, which are the naturally occurring antimicrobial substances in soybeans. Fermentation takes place while soaking. Lactic acid bacteria (LAB) are one of the microorganisms found in the soaking process. It has the ability to produce acid, which plays a role in the acidification of the soaking process (Nout and Kiers, 2005). The soaking process influences the nutritive value and quality of the tempeh. For example, it supports the growth of beneficial organisms to produce vitamins (Radiarti and Sumarti, 2016) and inhibits the development of pathogenic bacteria (Usman et al., 2018).
Even though the soaking process has a huge impact on the nutritive value of tempeh, the process is also timeconsuming. This research proposes an alternative method by the addition of LAB to the soybeans during the tempeh soaking. LAB have been found to occur naturally in the tempeh making process and has a huge role in tempeh as probiotics (Touw, 2014). It produces acid (lactic acid and acetic acid) that can decrease the growth of spoilage and pathogenic microorganisms. Adding LAB in the soaking process was found to inhibit the growth of Escherichia coli (Winanti et al., 2014). In this research, Lactobacillus plantarum HPADA3 and Lactobacillus fermentum HPBD2 were added to the soaking process of soybeans.
This research aimed to determine the effects of adding LAB in the soaking process of tempeh, and determine the preference level of the tempeh produced.

Production and harvesting of biomass
One loop of each LAB was inoculated into MRSB and incubated at 37ºC until the OD was reached (around 16 hrs). OD was measured to reach biomass that was equal to 10 8 CFU/mL, with OD 600 = 1.4 for L. plantarum HPADA3 and OD 600 = 1.5 for L. fermentum HPBD2, then harvested through centrifugation at 5214 × g for 10 mins at 4°C (Hunaefi et al., 2012;Yadav et al., 2016).

Safety assessments of lactic acid bacteria
Safety assessments were divided into hemolytic and virulence factors determination. For the hemolytic test, L. plantarum HPADA3 and L. fermentum HPBD2 were grown in MRSA media with the addition of 0.3% (w/v) CaCO 3 and then streaked onto blood agar plates containing 7% (v/v) sheep's blood (Yadav et al., 2016). The hemolysis activity of both LAB was evaluated after 24 hrs of incubation at 37ºC. The results were recorded by observing the clear zone (β-hemolysis), greenish zone or partial hydrolysis (α-hemolysis), or no reaction.
For virulence factors determination, both genomes from LAB were isolated using a Solg Genomic DNA Prep Kit (SolGent). Genes involved in the virulence factors were amplified using selected primers ( Table 1). The virulence factors were esp for checking surface protein, gelE for gelatinase activity, efaA for surface A antigen, and ace for collagen adhesion (Sanchart et al., 2016).
The reaction conditions (25 μL) were as follows: 3 μL sample DNA, 8 μL PCR master mix (Taq polymerase, MgCl 2 , DNTPs, and PCR buffer), 1 μL Primer F, 1 μL of Primer R, and 12 μL Nuclease Free Water. The PCR protocol was as follows: initial denaturation at 94ºC for 5 mins, denaturation at 94ºC for 1 min, annealing for 1 min (the temperature can be seen in Table 1), elongation at 72ºC for 1 min, final extension at 72ºC for 10 min, and then put on hold at 4ºC. The PCR was done in 30 cycles using a thermal cycler apparatus.
After amplifying, the resulting products were visualized using electrophoresis with 1.5% (w/v) agarose gel in TAE buffer 1×. A 100 bp DNA ladder was inserted in the left well of the gel, followed by each PCR product. Ethidium bromide was used for staining the bands. The bands were visualized with Biorad Universal Hood II Gel Doc System.

Tempeh production
Yellow soybeans (Pasar Modern BSD) were used as the raw materials to make tempeh. Each variation consisted of around 100 g of yellow soybeans. Soybeans were weighed, washed, boiled for 30 to 40 mins using underground water, rinsed, and dehulled by hands. The dehulled soybeans were given treatments as in Table 2. Each LAB in the treatments was added with the amount of 10 8 CFU/mL. After soaking, they were boiled for 30 to 40 mins, rinsed, and left to cool until they reached room temperature. Then, 1% (w/w) tempeh inoculum (Raprima TM ) was mixed into the soybeans (Winarno and Reddy, 1986). The inoculated soybeans were put into a perforated plastic and incubated for 30 hrs at 30ºC. The tempeh was observed from the front, back, and inside. Every variant was made in duplicate. For each soaking variation, water samples were taken before and after the soaking process.

Enumeration of viable bacteria
Plate Count Agar (PCA), Eosin Methylene Blue (EMB), and deMan Rogosa Sharpe Agar (MRS agar) were used to grow the bacteria. PCA was used for growing mesophilic aerobic bacteria, EMB for Enterobacteriaceae, and MRS agar for lactic acid bacteria. For the test, 1 g of each tempeh variant was weighed and diluted with 9 mL sterile physiological saline in test tubes up to 1 mL. The diluted samples were spread on the agar media and observed (Radita et al., 2017). The water used after soaking was also spread on agar media. The results were counted in CFU/mL.

Sensory evaluation
A sensory evaluation test was then performed on the tempeh. It was divided into two parts: a simple overall differences test and a hedonic test. Approximately thirtynine semi-trained panellists were asked to judge the sensory characteristics of the cooked tempeh. The overall differences test compared the control (commercial tempeh) to the tempeh with the addition of LAB, to determine which was the most similar to commercial tempeh. A hedonic test was also performed on the tempeh with LAB addition. All samples were cut to 2×2×1 cm and fried until the colour turned golden brown. The panellists rated the samples on a 7-point hedonic scale (1 = 'Strongly disliked'; 2 = 'Moderately disliked'; 3 = 'Slightly disliked'; 4 = 'Indifferent'; 5 = 'Slightly liked'; 6 = 'Moderately liked', and 7 = 'Strongly liked') for the following attributes: texture, odour, taste, aftertaste, and overall taste. The test was conducted in a sensory evaluation room with the assistance of red light. Water was provided for the panellists between samples. The analysis was done statistically with Statistical Package for the Social Sciences (SPSS) software with 5% probability.

Safety of lactic acid bacteria isolates
Lactobacillus plantarum HPADA3 and L. fermentum HPBD2 were both isolated from the tempeh. From the assessments, the blood agar result turned out negative because there were no clear zones. Potential virulence factor genes were determined to be negative except for the efaA gene with a band size of 735 bp (Figure 1). Hemolytic activity was negative indicating they were not pathogenic (Yadav et al., 2016). Potential virulence factor genes had one positive result; the efaA gene was responsible for surface A antigen (Sanchart et al., 2016) or cell wall adhesion which actually might help the LAB to colonize guts and showed its probiotic properties. This gene was also found in potential probiotics from Spanish sheep cheese by another study (Cebrián et al., 2012). Another study also researched commercial probiotics in China that found the presence of potential virulence factors gene (Lei et al., 2015). A study on commonly used probiotics may contain virulence factor genes associated with colonization. The probiotics had been used for 20 years without any detrimental effects on the consumers. This suggested the presence of one or more virulence factors did not arbitrarily make the bacteria pathogenic (Franz et al., 2011).

Tempeh production
Soybeans were soaked for 3 hrs with the addition of LAB then the water used for soaking was measured at the beginning and the end of the soaking period. The soaking duration was shorter than normally done for making tempeh where soybeans could be immersed in water for up to 30 hrs (Gunawan-Puteri et al., 2019). The soaking water experienced a lowering of acidity level by around one degree of acidity (Table 3). pH number was comparable to a study where a soaking duration of 12 to 48 hrs could change pH from 6.5-7.0 to 4.5-5.5 or lowering two degrees of acidity (Mulyowidarso et al., 1989). Another study with 6 hrs of soaking changed the pH from 5 to 4.64. The changes in pH levels were caused by the activity of LAB which produced acid (Lumowa and Nurani, 2014). LAB was found in many stages of the tempeh making process and one of them was during the soaking process (Moreno et al., 2002) All tempeh variants resulted in compact tempeh even though the pH was almost neutral, as seen in Figure 2. This result was similar to former studies. One study showed the effect of pH on Rhizopus oligosporus one of the main microorganisms for tempeh production. Interestingly, it was found that in the study replicating the tempeh fermentation condition, R. oligosporus was able to grow to pH 7.5 at 30ºC depending on the supporting CO 2 and A w, even if the growth was slow   (Sparringa et al., 2002). In addition, Winanti (2014) also found that tempeh from Desa Batang in Banjar Province made with approximately pH 6 of soybeans resulted in compact tempeh.
A study had suggested LAB might affect fungal growth on tempeh, certain LAB strains had been known to inhibit fungal growth. The addition of LAB to aid fermentation in barley tempeh affected fungal growth depending on the types of bacteria and inoculation levels (Feng et al., 2005). Even so, the addition of LAB in this study resulted in tempeh with the same appearance as the CT tempeh. From this result, both LAB did not impact the growth of the inoculum, as the inoculum was able to grow normally and covered the soybeans.

Enumeration of viable bacteria
The water samples and tempeh samples were spread on PCA, EMB, and MRS agar media to determine the population of viable bacteria (Table 4). It was interesting to note that the morphology of bacteria growing from soak water was different on the PCA media. The CT had more variation of bacteria compared to all variants which were seemingly dominated by one type of bacteria. The morphology on the PCA plates appeared to be more uniform than the control, indicating the LAB was able to survive, affect the microbial composition, and dominate the water. A study about adding bacteria in seed soaking water indicated that the bacteria can be absorbed into the seed and increased the chance of survival against the physical condition and lower mesophilic bacteria count (Świeca et al., 2018). On EMB media, there was no metallic green sheen after soaking. The slight increase on EMB plates in the LF variant showed a different pattern than the CT. Even so, there were no metallic green sheen colonies found on EMB plates. There was the possibility that LAB did not inhibit other bacteria. The main mechanism of inhibition by LAB was organic acid but the inhibition was strain-specific and influenced by the amount of organic acid (Belicová et al., 2013).
Most water samples had significantly different bacteria populations than tempeh. Tempeh variants with the addition of LAB had higher bacteria on the MRS agar media compared to the CT. The presence of bacteria in tempeh was influenced by the second cooking process before fungal fermentation (Mulyowidarso et al., 1990) since it could have a lethal effect on bacteria (Nout and Kiers, 2005). Previous studies also had shown the impact of cooking soybeans on the number of bacteria. The number of bacteria was higher in the tempeh without a second cooking process as the population was reduced (Efriwati et al., 2013).
Heat-killed LAB might still play a beneficial role in food even after being inactivated and it may be beneficial to do further study. In one research, heat-killed L. plantarum could enforce the immune system shown by a lowered incidence of upper respiratory tract infection (Hirose et al., 2013). Another study about live and heat-killed Lactobacillus rhamnosus demonstrated that both had the ability to modulate the immune system by inducing cytokines. Even so, it should be noted that   (Jorjão et al., 2015). Tempe has been known to possess good health benefits if consumed. It is able to modulate human gut microbiota, but the effects may vary depending on the microorganisms (Soka et al., 2014).

Sensory characteristics and evaluation of produced tempeh
Sensory evaluation was conducted by 39 semitrained panellists. For the first differentiation test, LF tempeh was perceived to be the most similar to commercial tempeh, followed by LFLP tempeh and LP tempeh. 14 panellists chose LF tempeh (35.90%), while 13 panellists chose LFLP tempeh (33.33%), and 12 panellists chose LP tempeh (30.77%). For further analysis, the likeability of all tempeh was determined through a hedonic test with the score showing that the texture and odour score between the tempeh was not significantly different. Taste, after taste, and overall testing were significantly different. The taste score of LF tempeh was significantly different compared to the other tempeh, receiving the highest score. The same pattern was also found on the aftertaste score. For overall testing, LF tempeh was significantly different to LP tempeh but not that different to LFLP tempeh. LFLP tempeh was also not too different from LP tempeh. From all the samples, LP tempeh received the least score on overall testing. The comparison of attributes can be seen in Figure 3.
A sensory evaluation test was done using commercial tempeh and all tempeh variants. Of all the tempeh, the highest similarity with commercial tempeh was perceived on LF tempeh. LF tempeh was preferred significantly in terms of taste and after taste compared to LFLP and LP tempeh. Overall, it was significantly preferred over LP tempeh. In this study, the soybeans, water, processing method, and mould used were the same except for the use of bacteria in the soaking process. Differences in taste and after taste between the tempeh may be caused by different bacteria in the soaking stage. A similar case had been documented where bacteria affected a tempeh's flavour. Tempeh with a higher Bacillus sp. count had a more pronounced bitter taste compared to other tempeh (Barus et al., 2008).
In general, all the tempeh had fulfilled the quality requirement of texture, odour and colour from the National Standardization Agency of Indonesia (SNI 3144:2015) (BSN 2015). Odour and colour were considered normal. The attributes (odour, colour, and taste) were liked with an overall score ranging from 4.95 (LP tempeh) to 5.65 (LF tempeh). LF tempeh received positive remarks from panellists with a score between 5 and 6 (5.81) in taste, was slightly liked with a score of 5.41 in the aftertaste, and a score of 5.65 (almost 6moderately liked). Food can be claimed as probiotic food if it contained a minimum of 10 6 CFU/g viable bacteria (FAO/WHO, 2002). All the tempeh could be called probiotic tempeh since they all met this requirement. Although the number of LAB decreased after heat treatment, the killed bacteria might still have the potential to give health benefits to the immune system and should be studied (Hirose et al., 2013). All tempeh had a shorter soaking duration compared to conventional tempeh and LF tempeh was found to be the best variant. Tempeh as a superfood from Indonesia has increasingly gained a reputation in the worldwide community and has the potential to diversify food products and support local businesses (Astawan et al., 2017).

Conclusion
Lactobacillus fermentum HPBD2 and L. plantarum HPADA3 added to the soaking water grew without affecting the fungal growth. The addition Lactobacillus isolates increased the number of lactic acid bacteria in tempeh compared to control. All tempeh enriched with Lactobacillus in this study met the national standard requirement in texture, odour, and color. Tempeh added with Lactobacillus fermentum HPBD2 received better preference than tempeh added with L. plantarum HPADA3. The addition of LAB influenced the tempeh bacteria composition that might affect the health benefits of tempeh, while cell debris and metabolites produced may still have beneficial effects on health and was recommended for further study. Since all tempeh contained a minimum of 10 6 CFU/mL of viable bacteria on MRS agar, it can also be called probiotic tempeh.