STUDIES ON THE PRODUCTION OF PHASEOLUS VULGARIS TEMPE IN DISCONTINUOUSLY ROTATING BIOREACTOR

Hanna Miszkiewicz¹, Marcin Bizukojc², Anita Rozwandowicz³, Stanisław Bielecki^1
1 Institute of Technical Biochemistry, Lodz University of Technology, Poland
² Department of Bioprocess Engineering, Technical University of Lodz
³ Institute of Technical Biochemistry, Technical University of Lodz

ABSTRACT

Phaseolus vulgaris fermentation was conducted in a discontinuously rotating bioreactor (DRB). Effects of rotation frequency and the bean mass thickness on Rhizopus oligosporus growth and biosynthesis of selected hydrolases were examined. Morphology and physiology of the fungus were estimated by using the digital image analysis. The results were compared with that of the reference stationary process. The fastest growth of R. oligosporus and the most efficient synthesis of selected hydrolases were achieved for the bed (hydrated, acidified, dehulled, and autoclaved cotyledons of P. vulgaris) thickness of 6 cm, and the frequency of agitation of 1 min every 4 h. Solid substrate fermentation in DRB was shown to be superior to that conducted under stationary culture conditions because R. oligosporus hyphae displayed higher physiological activity and the activity of extracellular hydrolases was 20% larger. In consequence, tempe produced in DRB under optimized conditions contained twice more reducing sugars and 25% more soluble proteins as compared to the reference sample.

Key words: solid-substrate fermentation, discontinuously rotating bioreactor, Phaseolus vulgaris, tempe, Rhizopus oligosporus, digital image analysis.

INTRODUCTION

Tempe is a traditional fermented food manufactured in Eastern Asia with different strains of Rhizopus grown on dehulled and cooked cotyledons of legumes, mainly soybean, but also bean, pea and mixtures of legume and corn seeds [6]. The quality tempe is a compact, sliceable mass, penetrated by Rhizopus mycelium and has pleasant flavor and taste, different from that of raw material. This high value added, healthy and nutritious product is rich in easily digestible proteins and vitamins, and almost free of anti-nutritive compounds.

The principal role of Rhizopus mycelium during the fermentation is synthesis of enzymes which digest the raw material and provide the desired texture and sensory attributes. Concomitantly, anti-nutritive substances are decomposed what enhances nutritious value of tempe [ 12].

Currently, the traditional method of tempe manufacturing has been modified, i.e. cotyledons of soybean or other seeds, properly prepared and inoculated with Rhizopus spores, are not wrapped in banana leaves to be fermented for 48 h, but they are put into plastic flat boxes or trays with perforated bottom, allowing for efficient gas exchange [5].

Our studies aimed at production of a novel, healthy, fermented food, characterized by a low content of anti-nutritive ingredients, which can be consumed without further processing or used for production of other foodstuffs, like snacks, etc. For this purpose we used cotyledons of P. vulgaris inoculated with spores of R. oligosporus and fermented in a discontinuously rotating bioreactor. Our investigations focused on optimization of bed thickness, and agitation frequency as well as on determination of dynamics of hydrolases synthesis by R. oligosporus.

MATERIALS AND METHODS

Raw material
Commercially available Phaseouls vulgaris Lens Aura seeds were purchased from the company delivering seeds for horticulture and forestry in Ożarów Mazowiecki (Poland). Prior to inoculation with R. oligosporus spores the beans were soaked in 0.85% aqueous lactic acid solution, dehulled, crumbled to 0.5 x 0.25 cm pieces, and autoclaved at 121°C for 15 min.

Microorganism
R. oligosporus strain from the pure culture collection of the Institute of Technical Biochemistry of the Technical University of Łódź was maintained on dextrose agar slants at 4°C. Samples (100 g) of autoclaved P. vulgaris cotyledons were inoculated with 1 ml of a suspension R. oligosporus spores (10⁶ cfu ml^-1).

Fermentation on Petri dishes
Perti dishes (diameter of 5 cm) with tempe (1cm thick layer) were incubated at 37°C for 72 h.

Fermentation in DRB
The main unit of the discontinuously rotating bioreactor (Fig.1) was 320 cm³ glass jar with perforated lid. The diameter of holes in the lid was 2 mm, and the distance between them was 1.5 cm. The jar was equipped with a helical stirrer, and was fastened on a rotary shaker. Rotation of the stirrer and the shaker was controlled by an electronic programmer. The series of cultures with different agitation frequency 1 min every (2, 3, 4, 5 or 6 h) was executed. The number of rpm was 10 and 12 for the stirrer and the shaker, respectively. All cultures in DRB were carried out at 37°C for 72 h.

Analytical methods
Extracts of tempe
1 g of tempe was ground in a mortar with 10 ml of distilled water at 4°C for 20 min. The pellet obtained by centrifuging (11,000 rpm, 4°C) was discarded, and the supernatant was diluted with buffers or distilled water prior to analyses.

Solid substance estimation
Solid substance content was determined in duplicate by drying of 1 g samples of tempe at 105°C to constant weight.

Reducing sugars content
Reducing sugars were assayed according to Somogyi and Nelson [11] and expressed in mg of glucose per g of tempe.

Soluble proteins content
Soluble proteins concentration was determined according to Lowry et al. [4] using bovine serum albumin as a standard.

Determination of enzymatic activity
Enzymatic activities were assayed by relevant standard methods and APIzym test (Biomerieux), and expressed in standard activity units, i.e. micromoles of product released from substrate in 1 min per 1 g of tempe. Estimation of activity by using APIzym test was carried out according to Biomerieux instruction. Products released by the examined enzymes were as follows: fatty acids (lipases, sunflower oil), maltose (amylases, soluble starch), glucose (cellulases, carboxymethylcellulose (CMC) and filter paper), and tyrosine (proteases, casein). All assays were made in triplicate, at optimum pH and temperature for each of the enzymes.

Microscopic preparations
1 mm slices prepared from samples of tempe were successively stained with 0.01% crystalline violet (5-10 min), 1% eosine (30 min) and differential mixture composed of cedar oil, terpentine and phenol (1:2:2, v/v/v), placed on microscope glasses and observed under fluorescent microscope. The images of hyphal entities were stored in a digital form what allowed for determination of their length by using the relevant computer program.

Digital image analysis
1 g of tempe was suspended in 10 ml of 0.2% aqueous solution of Tween 20, homogenized for1 min, 5 fold diluted in distilled water, and filtered through 1.2 mm filter (diameter of 47 mm) to obtain R. oligosporus hyphae free from P. vulgaris cotyledons and other medium components [3]. The hyphae collected from filter was suspended in 5 ml of 0.8% Tween 20, pH of 7.0, stained with methylene blue (hyphae suspension: methylene blue solution, 1:1.5), and observed under OLYMPUS BX-40 light microscope under a magnification of 200 with phase contrast. The microscope was coupled with a camera and a computer equipped with Micro Image 4.0 software. Different respiratory activities of hyphal cells were visualized by their different colors. Apical and growing cells with high respiratory activity (zone B) were white, cells with moderate activity (zone A) were violet, those with low activity (zone C) - violet, and almost or completely inactive cells (zone D) were black. Images of approximately 25 objects, which were made during the process were processed by median and high-pass Gauss filter, and segmented to obtain four distinct color zones. To calculate the zone fraction, the ratio of each zone area versus total hyphal area was calculated [1].

RESULTS AND DISCUSSION

Optimization of agitation conditions
In the first step, a series of R. oligosporus cultures on P. vulgaris cotyledons (bed thickness of 6 cm) was executed. The sole variable parameter was the frequency of 1 min agitation (every 2, 3, 4, 5 or 6 h). For reasons of comparison also the stationary culture was made. Analyses of tempe derived from each batch revealed that the most beneficial changes in fermented beans (i.e. the smallest drop in solid substance as well as the highest rise in reducing sugars and soluble proteins) occurred when the bed was agitated every 4 h (Fig.2).

Independently on agitation frequency, the peaks of lipases were detected after 24 h (18.3 U/g) and 60 h (13.3 U/g), amylases and cellulases - after 36 h (1.5 and 8.0 U/g), and acidic proteases (optimum pH of these enzymes was 3.0 and 5.5, respectively) - after 72 h (3.7 and 1.8 U/g). However, activities of all aforementioned enzymes were the highest in tempe samples agitated every 4 h (Fig. 3). Therefore this variant of agitation conditions was chosen for further experiments.

Optimization of bed thickness
In this step of investigations, the variable parameter was bed thickness (3, 6 and 9 cm). The thickness of 6 cm was shown to be superior to the two other because it provided the highest amount of free glucose (10.5 mg/g, after 36 h) resulting from relatively high activity of amylases and cellulases (1.5 and 8.0 U/g, respectively), and the largest content of soluble proteins after 72 h (22.7 mg/g, 5.2 fold more than in raw beans) whereas a decline in solid substance was only 6.1% (Fig. 4). These beneficial changes resulted from the highest yields of extracellular enzymes of R. oligosporus as compared to two other bed volumes (Fig.5).

Morphology and physiology of R. oligosporus
The growth of R. oligosporus was examined under optimum fermentation conditions (tempe thickness of 6 cm, agitation every 4 h). Average hyphal length (average length of 5 different hyphal entities) was determined every 12 h (Tab. 1). Up to 36 h the fungus grew intensively (the average length of 877 µm after 36 h). Within the next 12 h it grew slower (the average length of 917 µm after 48 h of fermentation), and in the last stage - the septa within the hyphae, with a minimum length of 254 µm (after 72 h) were visible.

Activities of R. oligosporus enzymes were also assayed by APIzym test (Tab. 2) and expressed in units estimated according to Biomerieux chart. The tests revealed that the fungus efficiently synthesized alkaline and acidic phosphatases, naphtalene-AS-BI-phosphohydrolase and b-glucosidase throughout the whole process (40, 40, 30 and 30 nanomoles, respectively). Activities of C8 and C14 lipases peaked on the first day (20 and 10 nanomoles, respectively) while activities of C4 esterase (20 nanomoles), a-galactosidase (20 nanomoles), and leucine arylamidase (30 nanomoles) were the highest on the second and third day. Two other enzymes, valine arylamidase (30 nanomoles) and N-acetyl-b-glucoseaminidase (10 nanomoles) were most efficiently secreted on the third day.

Comparison of fermentations under stationary and agitated conditions
Traditional tempe fermentation is conducted under stationary conditions. It was represented by R. oligosporus culture on Petri dishes. Comparison of changes in pH, reducing sugars, soluble proteins and activities of extracellular enzymes for stationary and agitated cultures (under optimized conditions) is shown in Fig. 6-8.

During fermentation in DRB, pH of tempe rose gradually from 5.49 to 7.02 after 36 h and further declined below its initial value (to 5.25 after 72 h). This decrease in pH was caused by an increase in the yield of lipase biosynthesis (the second peak after 60 h). The latter enzyme released enough fatty acids to significantly reduce pH. In contrast, pH of tempe fermented on Petri dishes rose from 5.97 to 7.26 (after 36 h) and was elevated (more slowly) to 7.53 after 72 h. This increase was brought about by oxidation of amino acids to CO₂, H₂O and NH₃. The similar alteration in pH was observed by Ruiz-Teran who cultured R. oligosporus on soybeans [9].

Maximum concentration of reducing sugars in DRB (14.9 mg/g after 36 h), twice larger than on Petri dishes, was detected 12 h later than under stationary conditions. The enhanced content of shorter sugars resulted from more efficient synthesis of amylo- and cellulolytic enzymes by R. oligosporus under agitated culture conditions. These hydrolases were secreted with the highest yield after 36 h of the process run in DRB whereas the fungus grown on Petri dishes most efficiently synthesized cellulases on the first day (6.0 U/g), and concurrently the amount of reducing groups released by these enzymes achieved a peak on that day (Fig. 6-8).

The profile of changes in soluble proteins content was similar for both types of P. vulgaris fermentation, however, due to more efficient synthesis of acidic proteases by the fungus grown in DRB (22.7 U/g versus 16.8 U/g) the final concentration of these proteins was 5.9 mg g^-1 higher under agitated culture conditions. Because the discontinuous digging through the tempe facilitated exchange of gases and heat, R. oligosporus produced all examined enzymes simultaneously and with elevated yield.

To find correlation between the respiratory activity of hyphal cells and productivity of synthesis of the aforementioned enzymes, R. oligosporus hyphae was stained with methylene blue and subjected to microscopic observation. Digital image analysis revealed four distinct zones with different respiratory activity (Fig.9). During the first two days of fermentation conducted in the bioreactor the most physiologically active zone B (white) predominated and the content of inactive zone D (black) was negligible. It provided evidence of a balanced fungus growth. On the third day the apparent degeneration of the hyphae began because the zone B fraction declined to 0.75 and concomitantly the zone D fraction rose to 0.185. Because of the autolysis of R. oligosporus mycelium (maximum from 60 to 72 h of fermentation), the proteases were secreted to the medium what caused an increase in soluble proteins content in tempe. Zone A and C fractions (violet and blue fragments of hyphae) was minor during the whole process.

Profiles of variation in zone fractions are different for fermentations under stationary and discontinuously agitated conditions (Fig. 9 and 10). In DRB, the R. oligosporus hyphae retained its satisfactory respiratory activity for 24 h longer, because a drop in zone B fraction and an increase in zone D fraction were observed after 60 h of tempe production. Maximum values of zone B fraction (0.8-0.98) were achieved from 24 to 60 h of the process. In contrast, degradation of the mycelium grown on Petri dishes began after 36 h. In this case, maximum values of zone B fraction were lower (0.62-0.76) and were observed from12 to 36 h of the culture. Maximum zone D fraction was 0.185 after 72 h of R. oligosporus growth in DRB, and 0.4 after 48 h growth on Petri dishes.

Because of limited availability of oxygen in SSF, microbial biomass forms a wet film on bed´s surface and only slightly penetrates beneath it. Penetrating the substrate, fungal filaments secrete enzymes, mainly hydrolases, which digest polymers occurring in deeper parts. Assimilation of degradation products by the growing mycelium is accompanied by the concomitant change in the texture of bed´s particles, which are composed mainly of starch and other polysaccharides. Hydrolysis of these polymers results in a drop of particle´s size and alteration in bed´s density [7].

The most important problems encountered in SSF include an excessive metabolic heat, and impaired control and scale enlargement [8]. Application of rotating, swing or stirred fermentors for SSF may increase the homogeneity of the overall system, unachievable in the stationary conditions. However, certain SSF products, like tempe, are directly consumed and their texture, attainable only in a stationary culture, is their captivating attribute. Tempe fermentation, conducted in agitated conditions, in one of the aforementioned fermentors, would bring about disruption of R. oligosporus hyphae and impede formation of secondary metabolites.

Temperature of tempe determines the rate of R. oligosporus growth in a traditional, stationary tempe fermentation. This temperature is usually increased by 10-16°C and can reach 44°C. At the latter, maximum temperature, the growth of the fungus is strongly reduced and the heat is slowly dissipated what causes a significant drop in bed´s temperature and further growth of the fungus after 36 h. A rapid rise in temperature (by 3°C per 1 cm of bed), intensive consumption of oxygen (depletion to 2%), and up to 21% increase in CO₂ volume were reported for tempe fermentation in a fermentor with bed´s thickness of 6.5 cm. The upper bed layer shrunk during the later phase of fermentation and empty gaps between the bed and fermentor wall were observed. These gaps facilitated diffusion of gases and promoted growth of the fungus in deeper parts of bed. Because carbohydrates from the upper layers were depleted, the further growth of the fungus was supported by polysaccharides from the bottom part [7].

In contrast to the traditional tempe fermentation, process conducted in agitated conditions in fermentors (rotating, swing or stirred) yielded the granulated product. Accumulation of heat was less intensive and the mass was more uniform [8i10].

Studies on wheat bran fermentation with R. oligosporus in a rotating drum either equipped with baffles or not, revealed that the baffles improved the process because of easier aeration, heat dissipation and exchange of humidity between the bed and gaseous phase [2]. That resulted in faster metabolism and shorter fermentation time. Optimization of the number of baffles, aeration intensity and rotation velocity improved the product quality.

Observation of maize grits fermentation by R. oligosporus NRRL 2710 in a swing fermentor (1 swing per 5 min) showed that sufficient bed aeration, its cooling by evaporation and humidity optimization beneficially affected process yield.

In case of P. vulgaris fermentation in DRB, the maximum generation of metabolic heat occurred at optimum temperature for R. oligosporus growth (approximately 37°C) and the temperature inside the bed was reduced due to its temporary rotation. In consequence, the metabolic heat was secreted for up to 70 h, and the substrate was more strongly modified as compared to the stationary conditions [8].

CONCLUSIONS

The discontinuously rotating bioreactor provided better conditions for R. oligosporus growth. Its mycelium displayed higher respiratory activity and 20% more efficiently synthesized extracellular hydrolases. That yielded the quality tempe, which contained twice more reducing sugars and 25% more soluble proteins as compared to that produced under stationary conditions.

REFERENCES

1. Bizukojc M., Ledakowicz S., 2003. Morphologically structured model for growth and citric acid accumulation by Aspergillus niger. Enz. Microbiol.Technol. 32: 268-281.

2. Fung C.J., Mitchell D.A. 1995. Baffles increase prformance of solid-state fermentation in rotating drum bioreactors. Biotechnology Techniques Vol9.No.4: 295-298.

3. Kamiński P., Hedger J., Williams J., Bucke C., Swadling I., 2000. Dielectric permittivity as a method for the real time monitoring of fungal growth during a solid substrate food fermentation of Quinoa grains. Food Biotechnology: Progress in Biotechnology. Elsevier Science. Vol. 17: 393-398.

4. Lowry O. H., Rosebrough N. I., Farr A. L., Randall R. I., 1951. J. Biol. Chem. 193: 265.

5. Nout M.J.R., Rombouts, F.M., 1990. A Review: Recent developments in tempe research. J. Appl. Bacteriol., 69:609-633.

6. Nowak J., Szebiotko K., 1992. Some biochemical changes during soybean and pea tempeh fermentation. Food Mikrobiol. 9: 37-43.

7. Rathbun B.L., Shuler M. L. 1983. Heat and mass transfer effects in static solid-substrate fermentation: Design of fermentation chambers. Biotechnol. And Bioeng. Vol. XXV: 929-938.

8. de Reu J. C., Zwietering M.H., Rombouts F.M., Nout M.J.R. 1993. Temperature control in solid substrate fermentation through discontinuous rotation. Appl. Microbiol. Biotechnol. 40:261-265.

9. Ruiz-Teran F., Owens J.D., 1996. Chemical and enzymic changes during the fermentation of bacteria-free soya bean tempe. J. Sci. Food Agric. 71: 523-530.

10. Sargantanis J., Karim M.N., Murphy V.G., Ryoo D. 1993. Effect of operating coditions on solid substrate fermentation. Biotechnol and Bioeng. Vol. 42: 149-158.

11. Somogyi M., Nelson N., 1944. J. Biol. Chem. 153: 375.

12. Varzakas T., 1998. Rhizopus oligosporus mycelial penetration and enzyme diffusion in soya bean tempe. Process Biochemistry Vol. 33. No 7: 741-747.

The work was supported by the grant no. AR 73/25/PBZ/021/P06/25/2001 from the Polish Committee of Scientific Researches.

Responses to this article, comments are invited and should be submitted within three months of the publication of the article. If accepted for publication, they will be published in the chapter headed 'Discussions' and hyperlinked to the article.