World Journal
of Microbiology
& Biotechnology
12. 53-56
Microbiological cassava-starch
and technological fermentation
aspects
of
J.L. Parada,* E. Zapata, S.V. de Fabrizio and A. Martinez were Strepfococcus, Bacillus, The major genera found in the microflora of fermented, sour, cassava-starch Lactobacillus and Saccharomyces with amylase activity. Lactic acid bacteria predominated whereas the presence of moulds was not significant. No coliforms were detected. Electron microscopy showed bacteria and yeasts in contact with the starch granules and signs of erosion on the granule surface. Lactic acid was the main metabolite; no oligosaccharides, maltose or glucose were detected, indicating their rapid utilization. The degree of acidification, which correlated with the decrease in viscosity and the final quality of the product, was influenced by the variable microbial ecology. Key words:
Cassava, fermentation,
starch.
Cassava (Muniho/ sculentu Crantz) is a root grown in large regions of the world and is an important source of food for many people of tropical and subtropical countries. In Latin America, it is either consumed like potatoes in Europe or it is processed to obtain cassava flour and starch, which are used in different cooked products. Sour cassava starch is obtained by a natural fermentation, which lasts 20 to 30 days. This product, much appreciated in several countries (Campbell-Platt 1994), is known as ‘almidbn agrio’ in Colombia and as ‘polvillo azedo’ in Brazil. Its acidity, which is mainly due to the lactic acid produced in the fermentation vat (Seeley & Dain 1960; Steinkraus 1992), is responsible for its preservation (Parada 1984). Sour starch is distinct from the unfermented, ‘sweet’ starch, having lower pH and improved functional and rheological properties. It is irreplaceable for manufacturing many traditional and nutritious bakery products. The fermentation of cassava starch takes place in rural or home-scale industries, where the composition of the natural inoculum and the evolution of the microflora are not well understood. Thus, the fermentation time is variable and the quality of the product uneven. J.L. Parada and S.V. de Fabrizio are with the Laboratorio de Microbiologia de Alimentos. Oepartamento de Quimica Organica, Facultad de Ciencias Exactas y Naturales. Pabell6n II, 3er Piso. Ciudad Universitaria (1426), Buenos Aires, Argentina; fax: 54 1 782 0529. E. Zap&a and A. Martinez are with the lnstituto de lnvestigaciones Tecnol6gicas (IIT), Colombia. ‘Corresponding author. (Q 1996 Rapid Science
Publishers
The aim of the present study was to improve our understanding of some of the technological and microbiological aspects associated with sour-cassava-starch production.
Materials Description
and Methods of
the Traditional
fermentationProcess
Cassava roots were harvested, washed and peeled manually. After grating, the starch was extracted with running water to give a ‘starchy’ liquid of pH 6.0 to 6.5. This liquid was transferred to wooden or glass vats, where the fermentation occurred in 3 or 4 weeks at room temperature (around 30°C). During this process, a thick dark layer formed on the surface of the liquid and this was separated from the sour starch, dried and considered a subproduct. By the end of the fermentation, the pH had dropped to 3.5 to 4.0.
Finally, the sour starch was removed and sun-dried for several days to give a snow-white cial distribution.
product
ready
for storage
and commer-
Sampling in the Production Centres Fermentation of sour cassava starch was studied in the same season in two consecutive years at two small production units located in a rural village near Cali, Colombia. During the fermentation, samples were taken, IO cm below the surface, with a sterile spoon. They were transferred to sterile flasks or to flasks containing the recovery medium of Elliker (Elliker et al. 1956) without any sugar and with the corresponding selective agent, pimaricin or oxitetracycline, added.
JL.
Parada et al.
C&m Media and lncubafion Conditions Moulds and yeasts were isolated in Sabouraud agar (Janke 1961) containing 40 & oxitetracycline/ml and acidified to pH 5.6 with 2M HCI. Bacteria were isolated in modified Elliker starch/agar (Elliker et al. 1956) containing 20 ,ug pimaricin/ml, The re-isolation of the bacteria was performed on Rogosa agar (Rogosa et al. 1951) and Elliker (5% v/v) milk/agar, both with pimaricin. The selection of amylolytic and acid-producing bacteria was carried out on Lee starchjagar (Hobbs & Green 1984). The utilization of different sugars was tested in the corresponding fermentation media (Hobbs & Green 1984). The presence of coliform bacteria was investigated on Violet Red/bilis/lactose/ agar (Mossel et al. 1962) and eosin/methylene blue/agar. The cultures were preserved using all-purpose/Tween-SO/agar (Evans & Niven 1951) with lactose as sole source of energy. The bacteria were incubated aerobically at 35°C for 24 to 48 h, or under reduced pressure of 0, for the micro-aerophilic organisms. Moulds and yeasts were also incubated aerobically at 30°C for 3 to 5 days. Densities of mesophilic bacteria (c.f.u./g) were determined on plate-count agar at 32°C for 48 h. Microscopy Gram staining of the fermented starch permitted the starch granules and the predominant morphologies of the microflora to be observed by light microscopy during the fermentation process. Samples of sour cassava starch were also dehydrated, goldcoated and observed with a scanning electron microscope. Identification Tests Standard tests, as given by Sneath (1986), were used for the preliminary identification of the bacteria. The characterization of moulds and yeasts was based on the criteria proposed by Pitt 8s Hocking (1985) and Lodder & Kreger-van Rij (1984) respectively.
(bacteria) or Sabouraud starchiagar plates (yeasts), after incubation for 72 to 96 h. The plates were exposed to I, vapour or flooded with a solution of I% I, in 2% KI. Clear zones around colonies, on a blue background, indicated a-amylase activity. Some of the hydrolytic colonies were picked up and studied. a-Amylase activity was assayed in the supematants of cultures, incubated at 35°C for 72 h, in Lee starch broth without CaCO,. All the cultures were adjusted to an A,,, of 0.8, for comparative purposes. Then, 6-ml samples were centrifuged at 5000 x g and the supernatants used to measure the enzyme activity of exogenous amylase. Phosphate buffer (0.1 M) containing 0.15 M NaCl and 0.5 mg soluble starch substrate/ml at pH 7.0 was used, following the iodometric modified method of Smith & Roe (1949). The pellets containing the cells were washed, centrifuged and each resuspended in 3 ml sterile distilled water, to investigate amylase activity in the whole cells. One unit of activity was defined as the amount of enzyme present in the supematant (or cell suspension) which was able to hydrolyse 10 mg starch in 30 min at 37°C.
Results Evolution
a-Amylase Activity Starch hydrolysis
54
World]olrmal
was preliminarily
products Lee starch medium were analysed Kieselguhr G in 0.02 M sodium was 65 ml ethyl acetate and 35 The plates were stained with heated at 100°C for 5 to 10 min. and other simple sugars were
observed
on Lee starchlagar
ofh4icrobiology & Biofcchmlogy, Vo’of12. 1996
of
the Process
The natural
fermentation
by a mixed
microflora
the
crude,
final
pH
in the
Acidity, Water Confent and Viscosity Total acidity, determined by titration of the fermentation supematants with 0.5 M NaOH, using phenolphthalein as indicator, was estimated as g lactic acid/l00 g (g%) (Anon. 1976). Contents of lactic, acetic and butyric acids were estimated by titration of distilled supematants (Anon. 1976). pH was measured potentiometrically during the fermentation. Acid production was evaluated by inoculating Lee starch broth without CaCO, and incubating at the appropriate temperature for 40 h. The cultures were then centrifuged at 5000 x g. Acid in the supematant in excess of that in the fresh medium was taken to be the acid produced during the incubation. Water content was determined at 105°C according to Anon. (1990). The viscosity of a 5.5% starch suspension was measured with a Brabender’s visco-amylograph after adjusting the pH to 3.5 (Cardenas & Buckle 1980). Ckromatog7apky of the Fermentation Supematants from the cultures in by TLC using plates spread with acetate buffer. The solvent system ml propanol/water (2:1, v/v). anisaldehyde/sulphuric acid and Controls of glucose, maltose included.
and Discussion
first
7
sweet of 3.5 stage
3
of cassava which
starch, to 3.7.
from The
main
6
0
drop
12
was
increasing
an initial
of the fermentation
Fermentation
starch
produced
pH
was
(Figure
I).
16
16
out
acidity
6.0
in pH
carried to 6.5
in to
a
observed
21
time (days)
FlQure 1. Changes in the densities of mesophilic bacteria (Oyear 1; A-year 2) and yeasts and moulds (O-year 1; l year 2), acidity (0) and pH (m) during cassava-starch fermentation. Samples came from two traditional producers in the same season of each of two consecutive years. are mean results for triplicate experiments microbial densities, however, are shown products in year 1 (7) and one producer’s (A).
Most values shown with both products; for both producers’ product in year 2
Cassava-starch fermenfafion Table 1. a-Amylase activity and isolates from 40-h, cassava-starch
acid production fermentations.*
a-Amylase activity in supernatantst
Isolate
of bacterlal
Total acfdlty (lactic acid g%)
1987). One of the strains selected because of its high amylolytic activity, was a Lacfobacillus sp. No coliform bacteria were detected, indicating that the manufacture of the starch was relatively safe. The number of c.f.u./g decreased in the final stages of the fermentation, when only the acid-tolerant microorganisms survived.
(u/l) Streptococcus sp. C51 S.bovis
C281
Lactobacillus sp. Lb30 Bacillus sp. B6 13. subtilis BCl *The values tNo activity
100 390
0.038 0.019
1340 900 2580
0.039 0.059 0.095
are mean results of triplicate was detected in the washed
experiments. cells.
The titrated final acidity, 0.40 to 0.53 g%, was comparable with that of the commercial powdered product (0.40 to 0.80 8%). Lactic acid was the major fermentation product (60% to 62%), followed by acetic acid (5% to 10%) and butyric acid (0% to 3%). Other organic acids were also present in small and variable amounts. The duration of the fermentative process (20 to 30 days) depended on the environmental temperature and the differences in the producers’ practice. When the production season began, the fermentation time was usually longer, due to the initial adaptation of the microflora. The water content of the fresh, wet samples ranged from 41% to 52%, but this fell to 10% to 12% after sun-drying, greatly increasing the product’s stability. The microbial composition and densities in the samples taken in the first year of sampling from two different producers were quite similar (Figure 1). The density of yeasts in the following year was lower, although most of those present were amylolytic. As producers did not use a defined inoculum, fermentations were not reproducible and the final quality of the sour starch was therefore variable. Since the process occurred in open vats, the numbers and types of microorganisms present depended on ecological and environmental conditions. Fermentation Microflora The microflora mainly consisted of aerobic and microaerophilic bacteria and yeasts, with very few fungi. Grampositive microorganisms and yeasts were present on the starch granules. The five, representative, amylolytic yeasts isolated, all had the ovoidal morphology characteristic of Saccharomyces. Moulds were classified as Penicillium and Aspergillus. Gram-positive cocci and bacilli, sporeformers and nonsporeformers, were the main components of the flora. Two strains were identified as Bacillus spp. two coccal isolates belonged to the genus Streptococcus and a third to the genus Enrerucucctrs, according to the new taxonomy (Schleifer
a-Amylase Acfivifv and Acid Production a-Amylase activity was assayed on culture supematants and washed cells to determine if it was extracellular or celllinked. One strain of Bacillus, BCl, had higher a-amylase activity than the streptococci, lactobacilli or the isolated yeasts, as might be expected (Fogarty 1983). La&bacillus Lb30 showed remarkable activity, even higher than that of Bacillus B6 (Table 1). Although five bacterial strains and the isolated yeasts were amylolytic, no degradation products of starch (i.e. maltose or glucose) were found. As no activity was detected in the washed cells, all a-amylase activity appeared to be extracellular, where it can interact with the surface of the starch granules. Bacilli were better acid producers than cocci or lactobacilli (Table I). However, no correlation between a-amylase activity and total acidity was observed, indicating that acid production depended, at least in part, on other metabolic features. Microbial Acfion on the Sfarch The appearance of the sour-cassava-starch granules was modified by the combined action of amylolytic enzymes, lactic acid and other, microbial, acidic metabolites. Previous studies (Cardenas & Buckle 1980; Cereda 1987) have demonstrated that the addition of low concentrations of HCl modifies the surfaces and texture of such granules. The amylolytic microorganisms in close contact with the starch granules (Figure 2) facilitated enzymatic attack. Although the Gram-positive amylolytic bacteria and yeasts presumably hydrolysed the starch to more simple carbohydrates, for their own assimilation, such products could also be utilized by other microorganisms, in a cascade sequence. The visco-amylographic curves for sour cassava starch were similar to those obtained for sweet cassava starch, but with lower viscosity values (data not shown). These results are in agreement with those of Cardenas & Buckle (1980) and indicate an improvement in rheological properties and baking performance following fermentation of the starch. However, no significative differences in viscosity were observed between samples taken after 8 days and those obtained at the end of the fermentation process. Since at least 20 days’ fermentation were necessary to achieve the appropriate baking characteristics, it is suggested that factors other than viscosity contribute to the desired properties. In order to reduce the fermentation time, improve the reproducibility of the fermentation and improve the quality
World Jownal
of Minobiology
6 Biotechnology. Vol T2. 1996
55
J.L. Parada et al. Chemists. St Paul, MN: American Association of Cereal Chemists. Anon. 1990 Official Methods of Analysis. Washington DC: Association of Official Agricultural Chemists. Campbell-Platt, G. I994 Fermented foods--a world perspective. Food Research International 2 7, 253-257. Cardenas, O.S. & Buckle, T.S. 1980 Sour cassava starch production: a preliminary study. Journal of Food Science 45, 15091512. Cereda, M.P. 1987 Tecnologia e qualidade do polvillo azedo. Informe Agropecuario (BeloHorizonte) 13, 63-70.
The authors acknowledge the financial support of the Food Technology Program of the Organization of American States (OEA), CONICET (Argentina) and the CIRAD/ SAR(France)-CEE.
Elliker, P.R., Anderson, A.W. & Hannesson, G. 1956 An agar culture medium for lactic acid streptococci and lactobaciUi. ]ournal of Dairy Science 39, 1611-1612. Evans, J.B. & Niven, C.F. 195I Nutrition of the heterofermentative lactobacilli that cause greening of cured meat products. Journal of Bacteriology 62, 599--603. Fogarty, W.M. 1983 Microbial amylases. In Microbial Enzymes and Biotechnology, ed Fogarty, W.M. pp. 1-92. London: Applied Science. Hobbs, W.E. & Green, V.W. 1984 Acid producing microorganisms. In Compendium of Methods for Microbiological Examination of Foods, ed Speck, L.M. pp. I84-195. Washington DC: American Public Health Association. Janke, D. 1961 Pilznfihrboden nach Sabouraud, modifiziert Merck, ein neuer Trockenn/ihrboden zur Z/ichtung yon Dermatophyten. Zeitschrift der Haut- und Geschlechts-Krankheiten 15, 188-193. Lodder, J. & Kreger-van Rij, NJ.W. (eds) 1984 The Yeast. A Taxonomic Study. Amsterdam: Elsevier Science. Mossel, D., Mengerink, W. & Scholts, H. 1902 Use of a modified MacConkey agar medium for the selective growth and enumeration of all Enterobacteriaceae. ]ournal of Bacteriology 84, 381386. Parada, J.L. 1984 Bacterias l$cticas y el mejoramiento de microorganismos de uso industrial. La Alimentaci6n Latinoamericana 146, 93-96. Pitt, J.I. & Hocking, A.D. 1985 Fungi and Food Spoilage. Sidney: Academic Press. Rogosa, M., Mitchell, J. & Wiseman, R. 1951 A selective medium for the isolation of oral and faecal lactobacilli. Journal of Bacteriology 62, 132-133. Schleifer, K.H. 1987 Recent changes in the taxonomy of lactic acid bacteria. FEMS Microbiology Reviews 46, 201-203. Seeley, H.W. & Dain, J.A. 1960 Starch hydrolysing streptococci. Applied and Environmental Microbiology 79, 230--233. Smith, B.W. & Roe, J.H. 1949 A photometric method for the determination of a-amylase in blood and urine, with use of the starch-iodine color. Journal of Biological Chemistry 179, 53-59. Sheath, P.H.A. (ed), 1986 Bergey'sManual of Systematic Bacteriology, Vol, 2. pp. 999-1003, 1208-1260. Baltimore: Williams & Wilkins. Steinkraus, K.H. 1992 Lactic acid fermentations. In Applications of Biotechnology to Traditional FermentedFoods, pp. 43-51. Washington DC: National Academy Press.
References
(Received in revised form 6 October I995; accepted 9 October
Anon. 1976 Approved Methods of the American Association of Cereal
~995)
Figure 2. Scanning electron micrographs of microorganisms on the surface of starch granules taken from a fermentation. Scale bars = 1 nm.
of the product, selected amylolytic starters should be used to produce sour cassava starch.
Acknowledgements
~
World]oumal of Microbiology~@ Biotechnology,Vol 12, 1996